NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Roundtable on Evidence-Based Medicine; Yong PL, Saunders RS, Olsen LA, editors. The Healthcare Imperative: Lowering Costs and Improving Outcomes: Workshop Series Summary. Washington (DC): National Academies Press (US); 2010.

Cover of The Healthcare Imperative

The Healthcare Imperative: Lowering Costs and Improving Outcomes: Workshop Series Summary.

Show details

Appendix AWorkshop Discussion Background Paper

Presentations and Related Literature Summary of the Estimates

Prepared for

The Healthcare Imperative:

Lowering Costs and Improving Outcomes

Workshop Series

May, July, September 2009

Institute of Medicine

Washington, DC

INTRODUCTION

The presentations throughout the first two workshops in the Institute of Medicine (IOM) Roundtable on Value & Science-Driven Health Care’s series The Healthcare Imperative: Lowering Costs and Improving Outcomes, provided a vast survey of the impact of waste and inefficiency on national healthcare expenditures and the potential cost-saving strategies available for implementation now. To supplement this information, a working paper was commissioned, which placed the presenters’ estimates in the context of similar national estimates published in the peer-reviewed literature and by think tanks and government agencies.

Health reform in the United States has long focused on the means to expand health insurance coverage to the growing numbers of uninsured. In the current debates, significant attention has also been drawn to the necessity to simultaneously address our rapidly escalating national health expenditures, which fully consume one-sixth of our economy.

To more fully explore the drivers and solutions to controlling our healthcare spending, the IOM Roundtable on Value & Science-Driven Health Care, with the support of the Peter G. Peterson Foundation, engaged in a three-part workshop series titled The Healthcare Imperative: Lowering Costs and Improving Outcomes.

The goals of the series were threefold: (1) to identify, characterize, and discuss the major causes of excess healthcare spending, waste, and inefficiency in the United States; (2) to consider strategies that might reduce per capita health spending in the United States while improving health outcomes; and (3) to explore policy options relevant to those strategies.

The presentations at the first two workshops in the series offered many estimates on the costs of inefficiency and the potential savings that could be realized through application of much discussed cost-control strategies. This working paper aims to provide brief summaries of estimates provided during those two workshops, including the methods of calculation and any limitations as noted by the presenters. In addition, these estimates are placed in the context of similar national estimates published in the peer-reviewed literature and by think tanks and government agencies. By doing so, a broader sense of the range of costs and savings available throughout the healthcare system will emerge.

Several observations noted in the course of completing this work are discussed in the following sections.

Varying sources of presentation estimates The estimates presented throughout the workshop series were calculated by varying methods, including original peer-reviewed research by the presenter and the presenter’s synthesis of the published literature. In the case of the latter, few additional national estimates were found that were not referenced by the presenter.

Differences in underlying methodologies Variation in the estimates within each category often stemmed from differing methodologies, sources of data, study time periods, and scope of work, often making direct comparisons between estimates extremely difficult.

Variations in number of available comparison estimates The number of national estimates identified within each category varied significantly, with several well-studied categories containing multiple estimates while other topics containing few or zero comparisons.

Limited focus to national estimates While estimates existed for several topics detailing potential costs and/or savings at an institutional or statewide level, this paper focused on national estimates (if they could be identified).

As this paper focused on the estimates provided throughout the IOM workshops, our preliminary literature survey focused primarily on comparable national estimates on waste, inefficiency, and cost-savings strategies as applied to the healthcare delivery system. In the course of the work, two notable observations arose and are discussed in the following sections.

Range of estimates varied For those estimates in which multiple comparisons existed, some estimates, such as those for tort reform and telehealth, grouped closely with those in the literature while others lay amidst a large range of estimates, such as those for tertiary prevention and health information technology. These variations often stemmed from differing methodologies, study time periods, sources of data, and scope of work, and made direct comparisons between estimates extremely difficult.

Need for additional research As the number of national estimates identified within each category varied significantly, with several well-studied categories containing multiple estimates while other topics containing few or zero comparisons, those with few comparisons, such as transparency and retail clinics, indicate areas in need of additional research to calculate national impacts and could build on the studies of smaller scope noted throughout the report. In addition, in areas with large ranges in estimates, further rigorous research would be beneficial in resolving the differences.

The next sections contain brief summaries highlighting the workshop estimates as well as identified literature estimates. A table summarizing the estimates discussed throughout the paper is included as an appendix. Also included in the appendixes is a summary of the lower-bound estimates developed by the staff of the IOM Roundtable on Value & Science-Driven Health Care based on the information cited throughout the background paper.

OVERVIEW OF THE WORKSHOP SERIES

In 2009, the IOM Roundtable on Value & Science-Driven Health Care, with the support of the Peter G. Peterson Foundation, engaged in a three-part workshop series titled The Healthcare Imperative: Lowering Costs and Improving Outcomes.

The goal of the series was three-fold:

  • Identify, characterize, and discuss the major causes of excess health-care spending, waste, and inefficiency in the United States.
  • Consider strategies that might reduce health spending in the United States while improving health outcomes.
  • Explore policy options relevant to those strategies.

Through the efforts of a planning committee consisting of leaders representing the various stakeholders throughout the healthcare sector, a series of three workshops were defined:

  • The first workshop, titled Understanding the Targets and convened May 21–22, explored the major drivers of healthcare spending growth, focusing on five broad categories: unnecessary services; inefficiently delivered services; excess administrative costs; prices that are too high; and missed prevention opportunities.
  • The second workshop, titled Strategies That Work and held July 16–17, focused on the potential of various strategies to lower health-care spending while improving outcomes, including knowledge enhancement-based strategies; care culture and system redesign-based strategies; transparency of cost and performance; payment- and payer-based strategies; community-based and transitional care strategies; and entrepreneurial strategies and potential changes in the state of play.
  • The final workshop in the series, titled The Policy Agenda and held September 9–10, delved into the policy options relevant to implementation and adoption of the strategies discussed in July in ways that maximize their impact on controlling the drivers of healthcare spending.

UNDERSTANDING THE TARGETS

The initial workshop focused on the identification of categories of waste and inefficiency in the healthcare system and their respective order of magnitude as a percentage of U.S. care spending, including:

  • Unnecessary services;
  • Inefficiently delivered services;
  • Excess administrative costs;
  • Prices that are too high; and
  • Missed prevention opportunities.

Session 1: Unnecessary Services

In a climate of growing concerns about how much the United States spends on health care, it has been estimated that as much as 30 percent of spending could be saved without compromising outcomes (Fisher et al., 2003a, 2003b). Indeed, existing studies find no relationship between higher levels of spending and the quality of care received by patients (Baicker and Chandra, 2004; Yasaitis et al., 2009).

The presenters in this session on the provision of unnecessary services focused on

  • Overuse of services beyond evidence-established benchmarks;
  • Use of services beyond benchmarks where evidence is not established; and
  • Choice of higher-cost services over evidence-established equivalents.

Overuse of Services Beyond Evidence-Established Benchmarks

Several studies examining the drivers of excess spending have focused on overuse of services and testing that may not bring clinical benefits to patients, highlighting excessive use of antibiotics, imaging and diagnostic tests, avoidable emergency department (ED) use, and surgical procedures (Bentley et al., 2008; Chassin et al., 1987; Merenstein et al., 2006; Winslow et al., 1988).

This section presents analyses presented by Amitabh Chandra that examined the degree to which costs and mortality could be simultaneously reduced. Subsequently, comparable estimates are presented, and the authors’ findings are placed in the context of the existing empirical literature.

Savings from reducing overuse of services Chandra (2009) made the argument that healthcare reform could save both money and lives. Chandra estimated that improving hospital performance to the level of the highest-performing hospitals (based on mortality and cost data) could result in 8 percent reductions in both cost and mortality for three high-mortality conditions (acute myocardial infarction, hip fraction, and colon cancer), saving over $1 billion annually and enabling more than 11,500 patients to live at least 1 more year. Chandra also found evidence suggesting that greater use of bundled payments within Medicare is a viable option for restraining cost growth.

In this analysis, the authors extended their prior work demonstrating a lack of association between spending and quality (Yasaitis et al., 2009). Using mortality as a quality measure and actual Medicare spending per beneficiary as the expenditure measure, they failed to find an association between spending and outcomes but rather found high-quality providers at each level of spending. To quantify the savings that might be achieved by improving performance, they first assigned each hospital to one of five categories, ranging from highest to lowest performance, based on spending and quality. Those in the highest performance category had both low mortality and costs; those in the lowest performance category had both high mortality and costs. The authors then simulated what would happen if lower-rated hospitals could perform like those in the higher-rated groups to arrive at the reductions noted above.

The authors also found that half of the variation in spending could be explained by the use of Part B services. Given that Part A payments are bundled and Part B payments are not, this finding suggested that combining reimbursements for inpatient, outpatient, and home health into a single payment might achieve savings.

The authors noted two main limitations to their study. First, the validity of the authors’ findings relies on the accuracy of their risk adjustment measure (the International Statistical Classification of Diseases and Related Health Problems [ICD]-9 diagnoses codes from Part A claims records), as survival is substantially more sensitive to risk adjustment than quality measures such as those used in Yasaitis and colleagues (2009). Second, as with all other work that relies on benchmarking methods, their study cannot speak about what policy levers could be used to achieve their estimated cost and mortality improvements. Hence, it is not certain how their estimated savings could be realized.

Additional estimates Chandra and colleagues’ analysis was one of the first to examine the relationship between hospital-level mortality and spending. A subsequent literature review found that Yasaitis and colleagues (2009), as referenced above, was the study closest to Chandra (2009). There is a sizeable empirical literature that uses more technical methods (and makes more restrictive assumptions) to estimate hospital inefficiency holding quality constant, including stochastic frontier analyses and data envelopment analysis. Studies analyzing national hospital data using stochastic frontier analyses estimate uniformly higher cost inefficiencies, in the range of 10.8 to 25.5 percent.

As mentioned above, Bentley and colleagues (2008) estimated that spending on eight selected wasteful services—excessive antibiotic use, avoidable ED use, and overuse of noninvasive diagnostic imaging, among others—might be as much as $65.1 billion, the equivalent of 3.4 percent of U.S. healthcare spending. Merenstein and colleagues (2006) found that urinalyses, electrocardiograms, and X-rays were frequently performed despite evidence and guidelines recommending against their use in asymptomatic patients at an estimated annual direct medical cost of up to $194 million. It has been estimated that the cost of excess medical and surgical services, including coronary artery bypass surgery and percutaneous coronary interventions is $600 billion (Delaune and Everett, 2008). Avoidable ED use has been estimated to cost $21.4 billion nationally, and the overuse of antibiotics has been estimated to cost $1.1 billion annually (Delaune and Everett, 2008). Kaplan (2009) discussed analyses indicating that $5.1 billion annually could be saved from a 50 percent decline in unnecessary visits for common conditions—headaches, back pain, and benign breast conditions. Additionally, the same author estimated $6.5 billion in annual savings from reducing unnecessary MRI testing for back pain and headaches, extrapolating from their institution’s experience after implementation of an evidence-based protocol. Others have calculated $300 million in annual spending on unnecessary MRI scans for back pain (Delaune and Everett, 2008). While focusing on duplicative and redundant testing, Jha (2009) found that costs amounted to $8.2 billion in 2004.

Estimates comparison As above, the finding by Chandra (2009) that hospital-level mortality and spending are uncorrelated in their data is consistent with the findings in Yasaitis and colleagues (2009). That being said, Chandra and colleagues’ (2009) percentage cost savings estimate appears to fall within a reasonable range. The dozens of data envelopment analysis studies of U.S. hospitals cited by Bruce Hollingsworth (2003) have not yet been surveyed. However, Chirikos and Sear (2000) compared the inefficiency estimates generated by these different empirical strategies using data from hospitals in Florida from 1982 to 1983 and found that the data yielded convergent evidence about hospital efficiency at the industry level. This is suggestive, if weak, evidence for the notion that the data envelopment analysis and stochastic frontier analyses estimates for national savings would roughly be of the same magnitude.

Although the costs of overuse of clinical services cannot be directly compared given the inclusion of different services in each estimate, it is worth noting that the estimates of Bentley and colleagues (2008) cover the broadest range of services in their analyses, including excessive antibiotic use for viral upper respiratory infections and otitis media, avoidable ED use, avoidable hospitalizations of nursing home patients, overuse of cytology for cervical cancer screening, inappropriate hysterectomies, unnecessary hospital admissions in ED triage of patients with chest pain, overuse of noninvasive radiologic imaging, and inappropriate spinal fusion surgeries. Although the estimates of Bentley and colleagues (2008) of $18.2 million to $33.3 million in 2004 dollars (1 to 1.8 percent of U.S. healthcare spending) for overuse of noninvasive radiologic imaging far exceeded that of Mecklenburg and Kaplan (2009), the latter included only MRIs while the former included use of other imaging modalities in their calculations.

Use of Services Beyond Benchmarks Where Evidence Is Not Established

A number of studies have found that the amount of spending across regions of the United States can vary twofold or greater (CBO, 2008; Fisher et al., 2003a); yet low-spending regions arguably deliver equal or higher quality care than high-spending regions (Baicker et al., 2004; Fisher et al., 2003a). The variation in spending appears to be driven by the use of discretionary medical services (Fisher et al., 2003b; Sirovich et al., 2008). This suggests that interregional comparisons might provide insights into the savings that could be achieved from coaxing better performance out of existing medical institutions.

This section reviews estimates presented by Elliot S. Fisher that calculated the potential annual savings that could be achieved within Medicare by eliminating excess use of discretionary services. Comparable estimates are presented and compared.

Savings from reducing use of services beyond benchmarks Exploiting this interregional variation in spending, Fisher and Bronner (2009) estimated that annual savings in the area of $50 billion (an 18 to 20 percent reduction) could be achieved within Medicare.

By ranking U.S. hospital referral regions according to the intensity of care provided, estimates of potential savings could be calculated by shifting use rates in high-use regions to patterns seen in low-use regions. In particular, they compared regions against benchmarks defined by hospital referral regions ranked in the best decile and quintile.

Drawing from sources such as the Dartmouth Atlas of Health Care, Fisher and Bronner found the potential reductions in use rates for a number of services could be substantial. For example, inpatient days could be reduced by up to 21.3 percent and medical specialist visits could be reduced by up to 44.1 percent. In fact, they find large potential reductions across all five services they considered (see Table A-1 below), and the decrease in these use rates would result in an expenditure reduction of $47.8 billion to $53.9 billion, when moving to the top quintile and top decile benchmarks, respectively.

TABLE A-1. Percentage Reduction in Discretionary Services by Benchmark.

TABLE A-1

Percentage Reduction in Discretionary Services by Benchmark.

There are two main limitations to Fisher and Bronner’s approach. First, benchmarking by hospital referral region unavoidably ignores the substantial variation in cost and quality within each region. For example, the gains from improving administrative efficiency or reducing defensive medicine practices through tort reform do not enter into the authors’ calculations. Along the same lines, possible expenditure reductions from reforming the payment system or implementing greater integration and coordination of care are also excluded. Therefore, the authors may actually be underestimating the potential gains to healthcare reform. Second, benchmarking methods in general are silent on how the predicted benefits might actually be achieved.

Even if the authors’ analysis suggests that savings of $50 billion or more in Medicare are achievable in principle, it does not say by what mechanism these savings can be manifested nor does it account for the costs of improving performance to the benchmarked regions.

Additional estimates Based on a similar type of benchmarking analysis, Wennberg and colleagues (Wennberg et al., 2002) estimated that $40 billion, or 28.9 percent of spending, could have been saved in 1996 if Medicare spending levels were reduced to the lowest spending decile nationally. Reviews in recent reports from the Council of Economic Advisers (Romer, 2009) and the Congressional Budget Office (CBO) (2008) relied very heavily on this paper’s findings, and subsequent searches identified few other estimates in the literature.

Estimates comparison Although the absolute savings of approximately $50 billion presented by Fisher and Bronner (2009) is larger than the literature estimate of $40 billion by Wennberg and colleagues (2002), the latter estimate represents a 10 percentage point difference in total spending. While the reasons underlying the difference remain unclear, perhaps factors other than discretionary services, such as the burden of chronic illness or the efficiency of delivery of clinical services, may have become relatively more significant drivers of Medicare spending over time. Also, as Fisher and Bronner (2009) analyzed disaggregated data from a more recent time period, their estimate may be more relevant to the current policy debate than prior estimates.

Choice of Higher-Cost Services Over Evidence-Established Benchmarks

Roughly one-third of all medical decisions require choosing between or among two or more treatment options (Center for the Evaluative Clinical Sciences, 2005). These “preference-sensitive” care decisions drive approximately one-fourth of all Medicare expenditures (Wennberg et al., 2009). Treatment options often range from conservative to aggressive and range in costs as well, but recent studies have found that patients exposed to decision aids were more likely to choose conservative treatment (O’Connor et al., 1999, 2003). These findings suggest that preference-sensitive care may present a significant opportunity to reduce costs without affecting outcomes.

In this section, analyses by David Wennberg are presented. The author estimated the potential savings from increased use of shared decision making (SDM). A comparison to other estimates is also presented.

Savings from reduced choice of higher-cost services Shared decision-making programs are designed to assist patients confronted with two or more treatment options in making informed decisions. Often facilitated with decision aids, SDM aims to provide unbiased estimates of the risks and benefits for each treatment option available to the patient. By fostering communication and collaboration between patients and their providers, patients become empowered to make informed choices. Patients using SDM often choose more conservative (and less expensive) treatment after carefully weighing the trade-offs. After reviewing the literature, the author concluded that a 1 to 1.5 percent reduction in net health spending could be achieved with systematic use of SDM, while the combination of SDM with changes in provider incentives and benefit design could lead to a greater than 5 percent reduction in net health costs.

The author expressed three caveats. First, no other healthcare system could provide a counterfactual system on which he could base his estimate as SDM has not been systematically applied in any other healthcare system. Second, existing provider interventions have not occurred on a large enough scale for analysts to produce credible estimates of the effect of provider-based SDM on total expenditures. Finally, the hypothesized effects of provider reimbursement and benefit design have not yet been subjected to any test.

Additional estimates As Wennberg discussed in his presentation, evidence from semiquantitative studies presented in the Dartmouth Atlas suggested a 10 to 20 percent reduction in costs might be possible (O’Connor et al., 2004), while another investigation found that health coaching combined with decision aids reduced total population costs by 3.6 percent (Wennberg, 2007). However, a recently published systematic literature review by Leatherman and Warrick (2008) on SDM found that “few studies provide assessment of impact on health outcomes, quality of care, utilization, or costs” (p. 79S). Further searches did not identify any comparable national estimates of savings. Recent events across the United States also suggested that there might not even be a state-level estimate against which to judge the findings in Wennberg (2009). Legislation passed in Washington State in 2007 officially recognized SDM “as a high standard of informed consent” (Kuehn, 2009), and required a demonstration project to assess the cost-effectiveness of SDM (currently under way). Four other states (as well as the federal government) are currently considering legislation mandating similar pilot programs (Kuehn, 2009).

Estimates comparison The estimate offered by Wennberg (2009) appears to be unique in its national scope. However, results from published literature indicate the significant potential for SDM to improve the quality of patient decision making while simultaneously lowering expenditures.

Session 2: Inefficiently Delivered Services

Concerns about waste in U.S. healthcare spending have not just focused on the provision of unnecessary services, but also on operational waste (Bentley et al., 2008). Operational waste is concerned with the resources necessary to provide those services and the efficiency (or inefficiency) with which they are used.

The presenters in this session on inefficiently delivered services addressed:

  • Costs from mistakes and duplicative tests;
  • Costs from care fragmentation;
  • Costs from inefficient use of higher-cost providers; and
  • Costs from inefficiencies in physician offices and hospitals.

Costs from Mistakes (Medical Errors, Preventable Complications) and Duplicative Tests

The landmark IOM report To Err Is Human: Building a Safer Health System estimated that between 44,000 and 98,000 deaths are caused by medical errors every year (IOM, 2000). When placed among the leading causes of death in the United States, medical errors rank above deaths from motor vehicle accidents, AIDS, and breast cancer (IOM, 2000). Such adverse events, which are defined as “medical errors resulting in injury” (IOM, 2000), not only increase direct costs to the healthcare system, but also represent dollars spent on additional care and increased insurance premiums that could have been better spent elsewhere (IOM, 2000).

This section summarizes the results presented by Ashish Jha that examined the costs of the top 10 preventable adverse events and duplicative testing in U.S. hospitals. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Savings from preventable medical errors Jha (2009) estimated the annual direct medical costs associated with preventable adverse events to be $16.6 billion (2004 dollars) in U.S. hospitals; when including redundant tests, the estimate increased to $24.8 billion.

The costs of medical errors were limited to 10 adverse events chosen via an intensive literature review. The analysis used data from the National Inpatient Sample to calculate the proportion of the population that was at risk for a particular adverse event. To determine the number of adverse events that occurred, the author multiplied the at-risk population by the incidence rate for the adverse event, taking into account variation by using a range of incidences from the literature. Finally, the number of adverse events was multiplied by the percent that were considered preventable. Both the number of adverse events and the proportion that were preventable were then multiplied by the direct medical costs associated with each event to determine the overall national annual cost of each event. To account for variation in the data available, Jha used a Monte Carlo statistical simulation. A similar approach was used to determine the costs associated with redundant tests.

Results show that there were an estimated 5.7 million adverse events in 2004, of which 2.2 million were adverse drug events (589,000 of which were preventable) and 1.7 million were hospital-acquired infections (1.4 million preventable). For all adverse events and redundant tests that were considered preventable with currently available approaches, Jha estimated an avoidable cost of $16.6 billion and $8.2 billion, respectively (Table A-2). This represents 8.2 percent of all inpatient costs in the United States.

TABLE A-2. Total and Avoidable Costs Due to Adverse Events and Redundant Tests.

TABLE A-2

Total and Avoidable Costs Due to Adverse Events and Redundant Tests.

As this analysis depended on estimates from the literature that are several years old, it may therefore not accurately represent the incidence of medical errors today (Jha, 2009). There were also important patient populations—such as women admitted for labor and delivery—for whom no reliable estimates could be incorporated into the analysis. The omission of these hospitalizations likely led to an undercount of the number of adverse events and their associated costs. Also, the analyses only addresses direct medical costs resulting from medical errors and redundant testing. Finally, the costs associated with implementing strategies to reduce the incidence of preventable adverse events were not taken into account in the study; thus, the net savings may be lower (Jha et al., 2009).

Additional estimates Few studies have been conducted that provide national estimates of the costs associated with adverse events, and only one cost estimate was found related to the reduction of redundant radiology tests. The IOM report estimated that total costs (direct and indirect medical costs, such as lost wages and disability, among others) associated with preventable adverse events range between $17 billion and $29 billion (IOM, 2000). According to the study by Thomas and colleagues (1999), on which the IOM cost estimates are based, more than half of these costs are attributable to direct medical costs. The study by Thomas and colleagues (1999) examined nearly 15,000 medical records from hospitals in Utah and Colorado to determine the incidence of adverse events.

The IOM report also cited a more targeted study by Bates and colleagues (1997), which involved a case-control study using self-reported hospital data and chart reviews to determine the incidence of adverse drug events in two hospitals. The authors used regression analysis to compare differences in resource use and length of stay between the cases and the controls in the study. By extrapolating from the results, the authors estimated annual costs for adverse drug events to be $4 billion; for preventable adverse drug events they estimated annual costs of $2 billion. However, the ability to extrapolate these results may be very limited as there were only two hospitals in the study. A second study by Classen and colleagues (1997), conducted over a 4-year period starting in 1990 at a single hospital in Utah, estimated nationwide hospital costs for adverse drug events to be $1.6 billion annually. This result is lower than that achieved by Bates and colleagues (1997) and is likely due to a lower assumed rate of adverse drug events.

More recently, Zhan and Miller (2003) conducted a study using the Nationwide Inpatient Sample to identify medical errors and calculate excess hospital costs due to injuries. The authors used the Agency for Healthcare Research and Quality (AHRQ) Patient Safety Indicators to determine incidence of medical errors, and conducted a case-control analysis to determine the differences in lengths of stay and charges. By extrapolating from the results using a 0.5 cost-to-charge ratio, the authors estimated total national healthcare costs for the 18 medical injuries included in the study to be $4.6 billion (2000 dollars) (Zhan and Miller, 2003). The primary limitation of this study lies in its reliance on adequate and accurate coding in administrative data to determine the incidence of medical errors; the validity of the results varies to the extent that the codes do not accurately reflect the diagnosis, to the extent that different hospitals code differently, and to the extent that errors are under- or overcoded in administrative claims records (Zhan and Miller, 2003).

Estimates comparison The estimates of the costs associated with medical errors appear to vary due to specific differences in study design. These differences include whether or not both direct and indirect costs were included, the incidence rate of adverse events found in each study, the year in which the study was completed, and the number of adverse events included in the study. Differences in the incidence rate of adverse events may be attributable to differences in the methods used to define an adverse event. As a result of these factors, it can be difficult to directly compare the estimates.

However, it can be noted that the overall estimates of the potential savings associated with adverse events are relatively comparable. The estimate by Jha and colleagues (2009) of $16.6 billion (in 2004 dollars) in direct medical expenses is comparable to the range of costs presented by the IOM ($17 billion to $29 billion) when taking into account that less than half of the costs estimated by the IOM are due to direct medical costs and the fact that the cost estimates are for a time period approximately a decade earlier. The estimate by Zhan and Miller (2003) of $4.6 billion in healthcare costs appears to be low not only compared to the IOM report, but also when taking into account the fact that Jha and colleagues (2009) studied fewer adverse events. As Zhan and Miller (2003) based their estimates on coding for medical errors in administrative data, which likely underestimated the incidence of errors, and Jha and colleagues (2009) used incidence rates and costs from studies generally relying on comprehensive chart review-based data, the latter’s estimates may be more accurate. It is also worth noting that while Thomas and colleagues (1999) found that over half of adverse events occur outside the hospital setting, the estimates discussed here focused only on the inpatient setting.

Costs from Care Fragmentation (Including Duplicate Services and Treatment Delays)

Uncoordinated or fragmented health care can lead to a number of adverse consequences for patients. Patients with chronic illnesses may see many different physicians, and these physicians may unknowingly prescribe contraindicated or conflicting medications, advice, or treatments (Peikes et al., 2009). In addition, patients may not have received sufficient advice on how to care for their condition, and may not be able to adequately adhere to prescribed treatment regimens (Peikes et al., 2009). At least partly as a result of uncoordinated care, chronically ill patients treated for a number of different conditions represent a disproportionate share of Medicare expenditures (Thorpe and Howard, 2006). These expenditures are largely driven by spending on hospital stays and readmissions to the hospital (MedPAC, 2008).

This section presents estimates from Mary Kay Owens that examined the costs of uncoordinated care in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

The cost of uncoordinated care Owens (2009) estimated that a program designed to identify patients with the most extreme uncoordinated care and reduce their uncoordinated care could result in an average of $240.1 billion (8.8 percent) in annual national savings.

This estimate was derived from an analysis of claims data from more than 9 million Medicaid and Medicaid/Medicare dually eligible patients in five states determining the extent to which uncoordinated care contributed to higher than expected costs. Specific claim-level events were identified, such as excessive numbers of prescriptions; therapeutically duplicative drugs; frequently changing drug therapies; using multiple treating providers, multiple prescribers, and multiple pharmacies concurrently and in random patterns; accessing the ED frequently for nonemergent or preventable care; and numerous other care patterns indicative of uncoordinated care. These events were then evaluated using various statistical methods, and then criteria-driven algorithms defining combinations of markers standardized across the study populations were used to determine the incidence and magnitude of uncoordinated care in each state’s study population. Matched comparison groups were also created in order to estimate potential savings that could be achieved via adoption of care coordination for the most extreme group of uncoordinated care patients. Statistical analysis of variables indicated that those variables that predicted higher than expected costs were also correlated with episodes of uncoordinated care.

Results showed that the small percentage of patients (10 percent) considered to have experienced “extreme uncoordinated care” were associated with an average of 36 percent of total costs. Extending the work, Owens (2009) estimated that interventions designed to reduce episodes of care fragmentation, including coordination of care between providers, can, on average, save 35 percent of costs for the most extreme group of uncoordinated care patients. Extrapolating these findings nationally, overall estimated national savings from a program with enabled care coordination, assuming 3 years to phase in, were an average of $240.1 billion per year, or 8.8 percent of annual national projected costs.

Although a standard definition of uncoordinated care was applied across all the state populations, some limitations of the analysis may include the lack of uniform marker values applied across all the state populations to identify those with extreme uncoordinated care. However, the values of the markers were allowed to vary based on statistical definitions specific to each state’s population and subpopulation (Medicaid vs. Medicaid/Medicare duals) to adjust for differences in demographic and disease characteristics. While markers may be a plausible proxy for the measurement of uncoordinated care, the sensitivity and specificity of these markers for identifying those patients lacking care coordination is limited to the populations studied. In addition, the analysis is based on patients enrolled in public healthcare plans, and as a result the magnitude of savings attributed to extreme uncoordinated care may be less in employer-sponsored plans. Finally, it is important to note that the estimate does not reflect net savings and that those people identified as receiving uncoordinated care may vary over time as individual insurance status, medical, and social circumstances change.

Additional estimates A review of the literature related to coordination of care and fragmented care found few studies that addressed the costs of uncoordinated care. Berenson and colleagues (2009) estimated that chronic care management and care coordination for dually eligible Medicare and Medicaid beneficiaries could result in 10-year (2010–2019) savings of $200 billion, assuming that care coordination could yield 5 percent savings per year. Berenson and colleagues (2009) suggested that the estimate might be conservative as it only applies to a very small percentage of the population at risk.

The Medicare Coordinated Care demonstration explored the potential cost, hospitalization, and quality impacts of a care coordination program at 15 sites across the United States (Peikes et al., 2009). The authors measured outcomes using Medicare claims data and patient surveys of volunteer participants who were randomly assigned to the care coordination program or usual care. Results from the April 2002 to June 2005 study indicated that none of the programs generated net savings, although three program sites had monthly expenditures lower than the usual care group. The study was primarily constrained by having limited power to detect whether reductions in standard Medicare expenditures would be sufficient to offset any program fees.

Another study examined outcomes associated with a care coordination program for four conditions at Permanente Medical Group in northern California (Fireman et al., 2004). Annual cost measures were obtained from the health plan’s cost management system and were compared to the average costs for adult patients without these conditions. Results indicated that costs substantially increased and that the predicted savings (mainly from reduced days in the hospital) were not observed. However, while costs did not decrease, the trends in quality indicators were favorable. Limitations included the fact that a randomized controlled trial comparing the effect of treatment for those with the same condition was not possible (Fireman et al., 2004).

As failure to coordinate care in the transition from inpatient to out-patient care has been identified as a significant factor contributing to the 17.6 percent of hospital Medicare admissions resulting in readmissions within 30 days of discharge (accounting for $15 billion in spending), care coordination has been suggested as a method of reducing the incidence of avoidable readmissions (MedPAC, 2007).

Estimates comparisons It is extremely difficult to compare the various cost estimates for coordinated care from the literature. Owens (2009) presented the only extrapolated national cost estimate that included both the publically and privately insured. Even were the estimates to be compared on a study group level, the different array of patient groups studied (ranging from patients enrolled in public programs to those in a private health maintenance organization [HMO]) make the estimates difficult to compare. However, with that being said, it can be noted that the estimate by Berenson and colleagues (2009) of $200.5 billion savings over 10 years for dually eligible beneficiaries is considered an underestimate by the authors of the potential total savings that could be achieved by a national effort to improve care coordination because of their focus on dually eligibles.

Also of note, two of the studies in the literature, one of which was based on a randomized controlled trial, found no net savings from the implementation of care coordination programs. This is a very different outcome from the other estimates suggesting that significant savings are possible. It may be that the particular application of the care coordination program was unsuccessful, such that a change in the design of the program might improve the likelihood of realizing savings, or that savings from care coordination may require a significant time to realize.

Costs from Inefficient Use of Higher-Cost Providers

A significant amount of the cost of producing health care is due to the cost of labor. As a result, ensuring that the labor inputs to health care are used in an efficient manner has the potential to reduce healthcare use and thereby healthcare costs as a whole. Systems changes that encourage shifts to more efficient care provision have applied business models of operation, such as the Toyota Production System and Six Sigma, to the healthcare market. These models seek to remove the waste present in the system and help improve quality through process standardization (Klein and McCarthy, 2009).

This section presents a discussion by Gary S. Kaplan of the results he gained from examining the potential costs from inefficient use of higher-cost providers in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Savings from the efficient use of caregivers Based on results from a targeted intervention designed to reduce unnecessary caregiver visits at Virginia Mason Medical Center (VMMC), Mecklenburg and Kaplan (2009) estimated that national cost savings from the reduction of unnecessary outpatient visits to be $5.1 billion annually and savings from eliminating unnecessary visits for imaging procedures to be $6.5 billion annually. In addition, they estimated savings of $8.3 billion when factoring in increased use of lower-cost providers, such as advanced registered nurse practitioners (ARNPs) and physician assistants (PAs). A further $2.3 billion could be saved by substituting low-cost telephone or computer-based visits for conventional visits for chronic conditions. Savings from these four independent categories were estimated to total $22.2 billion annually.

In 2002, VMMC implemented methods used by Toyota to improve safety and remove waste in the system. In 2004, the health system established a collaborative with employers and implemented a model designed to lower costs while providing quick access to care (Mecklenburg and Kaplan, 2009). The primary outcome measured was service use, which was matched with 2009 data on reimbursements and cost in order to determine the savings achieved as a result of the new model.

Results indicated that unnecessary visits for common conditions—including headaches, back pain, and benign breast conditions—declined by 50 percent after the model was implemented. Assuming that outpatient visits for these three common conditions comprise 8.8 percent (based on data from VMMC) of all such visits nationally, the authors estimated that a 50 percent reduction suggests that 48.4 million outpatient visits per year could be eliminated via adoption of this care model. National cost savings are estimated to be $5.1 billion annually. The study hospital also experienced a 30 percent reduction in imaging visits, which, when extrapolated to the national level, yielded an estimated $6.5 billion in annual savings. Assuming that, on a national level, half of visits for uncomplicated conditions could be handled capably by an ARNP or PA rather than by a physician, additional savings could equal $8.3 billion. Use of telephone or computer-based visits would save an estimated $2.3 billion per year (Mecklenburg and Kaplan, 2009).

Some limitations of this study include potential questions related to the generalizability of the findings at VMMC to the general U.S. healthcare system. Variations in labor and supply costs may influence the amount of savings achievable at individual institutions. However, it is important to note that over 75 percent of the savings detailed are from simple categories of improvements that are commonly (yet not consistently) used in general practice.

Additional estimates Multiple studies have concluded that use of physician extenders is a cost-effective practice, but none offer national estimates of potential savings. Adjusted for patient case mix, it has been found that practices that more extensively used PAs and ARNPs in care delivery had lower average practitioner labor costs and total labor costs per visit (Roblin et al., 2004). However, this same study found that, because pediatric visits were more costly than internal medicine visits on average, the savings in pediatric visits was smaller. A systematic review of the literature on nurse midwives concluded that, compared to other models of care for pregnant women, use of nurse midwives led to lower use of multiple interventions (e.g., antenatal hospitalization and episiotomy) and improved outcomes, with evidence supporting lower costs as well (Hatem et al., 2008). Another study found that the total cost per episode seen by a PA was less than a similar episode managed by a physician, regardless of the patient’s age, gender, and health status (Hooker, 2002). Few differences emerged between physicians and PAs in the use of resources and the rate of return visits in the same study. A recent study concluded that expanded use of ARNPs and PAs in the delivery of primary care could save $4.2 billion to $8.4 billion over the next decade in the Commonwealth of Massachusetts (Eibner et al., 2009).

A related body of literature examines the volume of visits across the various physician specialties. Farrell and colleagues (2008) reported that, between 2003 and 2006, while total physician office visits remained stable at approximately 900 million annually, visits to primary care doctors decreased by 0.5 percent per year while specialist visits increased by 1.6 percent annually. Fisher and colleagues (2009) also discussed evidence suggesting that the volume of specialist visits could decrease by approximately 40 percent without harming the quality of care.

Estimates comparison As no other national estimates for use of nonphysician providers were found, no comparison was undertaken. As described above, physician extenders have not been shown to harm clinical outcomes—and, in fact, may improve outcomes—and physician extenders may lower costs if used for appropriate medical conditions. A recent analysis concluded significant cost savings from increased use of PAs and ARNPs in the delivery of primary care services in Massachusetts. However, it has been suggested that extrapolations of savings must be done cautiously as the degree of savings depends on the magnitude of salary differential between physicians and nurses, and may be offset by potential lower productivity of nurses compared to doctors and lack of changes in physician workloads if the additional labor allows expansion of care to meet previously unmet needs (Laurant et al., 2005). Additionally, with a trend toward decreasing primary care visits and increasing visits to specialists and simultaneous evidence that the number of discretionary visits to specialists could decrease by approximately 40 percent without harming quality of care, a substantial opportunity potentially exists for cost savings from redistribution of visits between primary care and specialists.

Costs from Inefficiencies in Physician Offices and Hospitals

Operational waste can be seen in both inpatient and outpatient settings. Ensuring that hospitals and physician offices operate in the most efficient manner possible may serve to reduce excess healthcare costs (Bentley et al., 2008).

This section presents the results from two presentations: (1) William F. Jessee, who examined the potential savings associated with increasing efficiency in physician offices; and (2) Arnold Milstein, who summarized analyses conducted by the Medicare Payment Advisory Commission (Med-PAC) examining the potential savings offered by high-performing hospitals. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

The cost of delivery inefficiencies in physician offices Jessee (2009) estimated that about $6.4 billion to $25.5 billion (2007 dollars) could be saved annually by reducing costs in physician offices. The estimate was based on an effort to estimate the clinical and administrative waste in physician office practices via an annual survey of practice costs and revenues. The survey was sent to 10,586 physician offices, and the response rate was 14 percent. The author calculated the distribution of costs by relative value unit (RVU), both including and excluding physician compensation, as a measure of the cost of production. The distribution of costs was skewed in a similar manner both when including and excluding physician compensation, indicating that differences in practice efficiency could be attributable at least in part to the higher costs of production. The author obtained an estimate of the waste in physician practices by normalizing the total cost by RVU distribution curve and comparing it to the observed curve.

Assuming the differences between the two curves are a measure of waste, $25.5 billion (2007 dollars) may be saved via reducing costs in physician practices. Jessee noted, however, that most differences between the curves are attributable to physician compensation, which implies that the majority of the difference is not directly attributable to efficiency differences across practices. When excluding physician compensation, an (arbitrary) estimate of the potential savings attributable to increased efficiency is about $6.4 billion annually (25 percent of the total estimated cost savings), or approximately 0.2 percent of total U.S. healthcare costs.

Some limitations of the study include potential bias in the survey results owing to the low response rate from invited participants. In addition, the cost estimates may understate total physician practice costs, as the costs for inpatient care provided are not included in the analysis (Jessee, 2009). Finally, it is unclear whether inclusion of physician compensation in the RVU provides a good measure of inefficiency.

Savings from paying for high performance Milstein (2009) presented results from a MedPAC analysis that estimated that overall U.S. healthcare spending would decrease by almost 2 percent if all hospitals were to achieve the same performance as the top 12 percent of hospitals (MedPAC, 2009). MedPAC’s March 2009 report to Congress used Medicare data to estimate the potential reduction in Medicare spending if all hospitals provided care at the same cost as hospitals that provide low-cost, high-quality care. MedPAC selected the 12 percent of hospitals considered to be “relatively efficient” based on risk-adjusted cost and quality measures. Costs per case were then standardized, and a composite mortality rate was calculated for eight common conditions. Each mortality rate was then weighted by the share of discharges in that hospital.

Results indicated that, if all hospitals were to achieve the same performance as the top 12 percent, mortality, readmission rates, and the cost of inpatient care would all decline. In terms of cost, hospital inpatient spending would decrease by approximately 10 percent. If these savings were passed on to consumers, overall U.S. healthcare spending would decrease by almost 2 percent. In addition, results from other data indicate that lower hospital costs are also associated with payers other than Medicare being able to negotiate lower average prices per case (Milstein, 2009).

Some limitations of the analysis include the fact that the savings may not be generalizable to other types of providers, such as physician practices. In addition, the lack of more specific data on hospital structures, processes, and outcomes precludes a more complete understanding of the factors that contribute to hospital performance, and as a result the potential changes that could be implemented to improve performance. In addition, because the analysis used diagnosis-related groups (DRGs), which are only a measure of payment, as the unit of cost measurement, the actual resource costs of the hospitals are not taken into account. As a result, it may be that hospitals that exhibit high quality with low cost also simply have lower resource costs and thus are able to deliver lower-cost care in addition to charging lower-cost DRGs.

Additional estimates A literature search for studies examining the potential savings from increased efficiency and performance in physician offices resulted in no additional studies estimating the cost savings associated with such improvements in efficiency. However, one study was found that supported these general findings. An analysis by Andes and colleagues (2002) measured the efficiency of physician practices using a linear programming technique called data envelopment analysis. This technique combined a number of different measures in order to compute one measure of efficiency. Results indicated that there was a range of efficiency levels across the 115 physician practices included in the analysis. Of these practices, only 7 were considered to be relatively efficient (Andes et al., 2002). In addition, the authors found that the practices found to be most efficient were those that did not have the highest charges, but instead were able to achieve their high efficiency through more efficient use of resources (Andes et al., 2002).

A review of the literature associated with the Toyota model, Six Sigma, and “lean” health care yielded a number of analyses related to the effects of such models on quality, cost, and outcomes. A potential Medicare savings of $400 billion over 10 years could be realized if U.S. hospitals reduced their inpatient costs to the level of ThedaCare, a hospital and clinic system in Wisconsin that has implemented efficiency improvements based on manufacturing methods (Toussaint, 2009). By extension, private payers (non-Medicare) could save an estimated $1.3 trillion over 10 years. Based on an annualized average hospital savings of $3.4 million, application of lean production systems to all U.S. hospitals could save an estimated $19.4 billion annually from elimination of non-value-added activities (Hafer, 2009). In addition, a case study (Klein and McCarthy, 2009) of the Gundersen Lutheran Health System in Wisconsin, Iowa, and Minnesota examined the care coordination techniques employed by the health system as an attempt to become a more efficient provider of health care. Results from the health system indicated that after a year in the coordinated care program, charges per patient fell an average of $7,300, and there was a reduction in the number of hospitalizations for patients in the program (Klein and McCarthy, 2009).

Estimates comparison The literature suggests that opportunities to increase the efficiencies within physician practices exist, however, as no other national estimates were found, a comparison was not undertaken. While the estimates of potential savings from increasing hospital efficiency from Milstein (2009) and Hafer (2009) differ in magnitude, these differences might reflect the various methodologies undertaken in calculating each estimate. While the former focused on savings achieved from an analysis benchmarked to the top 12 percent of hospitals in terms of cost and quality, the latter considered the average annual savings achieved from application of lean production methods. Also, the Milstein (2009) estimates focused on Medicare spending, while those of Hafer (2009) included all hospital spending. The estimates offered by Toussaint (2009) are difficult to compare directly to those previously discussed given their varying time frames; however, even a rough annual savings estimate (which may either under- or overestimate savings achieved in any single year during the 10-year time frame) far exceeds the estimates of Milstein (2009) and Hafer (2009).

Session 3: Excess Administrative Costs

Administrative costs in the U.S. healthcare system are significant, and reflect the complexity of a multipayer system and the costs of safety and quality assessments. Given the concern regarding the costs of health care, interest in estimating the amount of expenditures consumed by administrative activity has been long-standing. Henry J. Aaron (2003) wrote an overview of papers dating back to 1986 estimating this potential source of waste. More recently, economists such as Paul Krugman (2009) and Greg Mankiw (2009) have taken up the issue. As the issue is still not settled, hopes remain that reducing excess administrative costs could generate tremendous cost savings in the U.S. healthcare system.

In this section, we present analyses on administrative costs discussed by James G. Kahn, Lawrence P. Casalino, James L. Heffernan, Andrew L. Naugle, and Peter K. Smith. Additional estimates are also presented and compared.

Estimates of excess administrative costs Kahn (2009), Casalino and colleagues (2009b), and Heffernan and colleagues (2009) provided estimates of excess administrative costs at the provider level, as well as for the entire U.S. healthcare system.

Kahn used the results of studies on billing and insurance-related (BIR) costs and applied them to U.S. national health expenditures to determine total and excess administrative costs. Casalino and colleagues (2009b) applied the results of a U.S. survey of providers to national health expenditures to estimate the administrative costs for physician offices. Meanwhile, Heffernan and colleagues described using data from the Massachusetts General Physicians Organization to estimate excess administrative complexity attributable to billing and payment activities and the time costs of physicians and staff associated with paperwork needed to file for reimbursements.

Results are presented in Table A-3. Excess spending is defined as the amount spent above a given benchmark comparison. Based on the analyses, provider-specific excess spending ranges anywhere from $26 billion (Blanchfield et al., 2009) to $75 billion (Kahn, 2009). For physician offices, estimated excess administrative costs for BIR ranged from $26 billion (Blanchfield et al., 2009) to $32 billion (Casalino et al., 2009b) annually. A synthesis of the estimates conducted by all these authors identified a total spending excess, based on a ratio of U.S. to Canada administrative costs, of between $168 billion and $183 billion per year (Kahn, 2009).

TABLE A-3. Synthesis of Estimates from Presentations on Excess Administrative Costs.

TABLE A-3

Synthesis of Estimates from Presentations on Excess Administrative Costs.

Some limitations of these analyses include the focus of Heffernan and colleagues (2009) on a single physician group office. The fact that there is some evidence indicating that the studied office is more efficient than other offices indicates that the excess costs may be underestimated (Blanchfield et al., 2009). In addition, there are varying levels of uncertainty in the estimates of BIR, with the most complete knowledge being about physician offices and the less certain estimates relating to hospitals and other providers (Kahn, 2009). The benchmarks used for the comparisons were also not definitive; as a result, the estimates of excess BIR costs may be lower than if other benchmarks were used. Finally, excess BIR costs may be associated with excessive clinical services, which, if independently reduced, would reduce the associated BIR costs by some amount (Kahn, 2009).

Enhancing clinical data as a knowledge utility Smith (2009) estimated that by reducing documentation requirements of nurses, $87.9 billion could be saved annually. Medical documentation requirements currently result in a vast data set that is not relevant to patient-specific needs. In addition, current documentation considers important clinical elements relevant to a patient’s specific problem to be secondary to the necessity of supporting payment requirements and ensuring the ability to defend against medical liability claims (Smith, 2009). In particular, payment requirements result in additional data elements that are not valuable to the patient experience.

Currently, a three-level patient evaluation requires a total of 90 minutes of physician time, with significant amounts of clinical data produced. Nurses are also required to document additional data elements, requiring further documentation designed to support payment and legal defenses. An analysis indicated that surgical nurses spend the greatest proportion of their time (36 percent) on documentation, compared to 19 percent on patient care activities and 21 percent on care coordination. Applying this proportion to the national health expenditure estimates, Smith estimated that nursing documentation costs an estimated $146.5 billion per year; reducing this documentation by 60 percent could yield $87.9 billion in savings, representing 4 percent of total national expenditures (Smith, 2009).

Potential reduction in administrative expenses Naugle (2009) quantified the total savings opportunity in administrative costs potentially available to commercial payers. Employing a benchmarking method, the author found that if administrative expenses for fully insured (the insurance company takes the financial risk on the claims cost) commercial products were reduced to the best-practice administrative expense of 7.5 percent of premiums, total savings of approximately $13.9 billion could be achieved. Furthermore, he found that additional savings of $6.2 billion to $9.1 billion could be realized for payers in the self-insured (purchaser takes the financial risk on the claims cost) market.

To calculate potential savings associated with fully insured commercial products, the author estimated the total savings that could be generated if the best-practice level of administrative expenses were adopted by all commercial payers. These estimates were based on data from a variety of sources such as the Milliman Healthcare Reform Database, the Medical Expenditure Panel Survey, and privately held data on commercial premiums. Naugle then calculated the potential savings for the self-insured market as a percentage of the savings for the fully insured commercial products.

There are a number of limitations to this study. First, the savings estimates apply only to payers. Though secondary savings may also come to providers, purchasers, and patients, these quantities are not included in the analysis. Second, only commercial products are considered. It is possible that additional savings might be achieved in other settings (e.g., Medicare and Medicaid), but this quantity is also excluded from the analysis. These two factors suggest that the author may have underestimated the potential savings from reducing administrative expenses. Third, as is common to all benchmarking analyses, the method is silent on what interventions could allow administrative payments to approach the best-practice level. In particular, benchmarking cannot address what is possible for a specific plan or group of plans, and it may not be possible for all payers to achieve the best-practice benchmark. This analysis also does not account for the costs of changing current practice, suggesting that the net savings realized may be lower than the estimate provided.

Additional estimates Most studies providing estimates comparable to the national savings estimates provided in the papers above rely on cross-national comparisons. The work of Woolhandler and colleagues (2003) is an often-cited example of work comparing administrative costs in the United States to those in Canada. The authors estimated excess annual administrative costs to be $209 billion in 1999 dollars ($415 billion in 2009 dollars, if growing as fast as health expenditures).

Existing studies using microlevel data have also focused on BIR activities spending. They have estimated spending as percent of physician, hospital, and private insurer revenue. In fact, the findings from these studies, such as Casalino and colleagues (2009b) and Sakowski and colleagues (2009), were used as inputs in the workshop synthesis calculation. Hence, we do not compare the estimates from these existing studies to workshop synthesis estimates. Relevant to the estimate of Naugle (2009), published analyses claim that the administrative expenses of commercial products, excluding profits, are 9.2 percent of premiums (Sherlock, 2009). Also, Russo (2009) found that requiring health insurers to spend a set amount (85 percent) of premium revenues on medical care would increase insurer efficiency and could save roughly $100 billion over 10 years (as a rough estimate). Finally, comparable national estimates for the potential savings in Smith (2009) were not found.

In terms of the administrative costs of health services regulation, it has been estimated that the total costs exceed $339.2 billion, which include regulation of health facilities, health professionals, health insurance, drugs and medical devices, and the medical tort system, including the costs of defensive medicine. After subtracting $170.1 billion in benefits, the net burden of health services regulation still amounts to $169.1 billion annually (Conover, 2004).

Estimates comparison There have been conceptual and methodological objections to the research based on single-country comparisons, as summarized by Aaron (2003). Thus, the research presented in this workshop provides a useful check on their macroapproach. The workshop synthesis (Kahn, 2009) estimate of excess annual spending of $188 billion to $203 billion is lower than the Woolhandler and colleagues (2003) estimate of $209 billion, if adjusted for the different time frame. The synthesis estimate is highly sensitive to a preliminary estimate for the United States: Canada BIR ratio for physicians, and will change to the extent this ratio deviates from 2:1 (Casalino et al., 2009a). Although the strength of the synthesis estimate stems from its inclusion of multiple analyses, the determination of administrative costs as a percentage of revenue and the chosen benchmark of a single-payer system may represent an upper bound that is difficult to attain given current reform directions, if nonetheless a valuable point of reference. In particular, the comparison to a single-payer system may indicate an unrealizable amount of savings given that a national U.S. single-payer system appears politically untenable.

The workshop synthesis estimate of $63 billion to $75 billion in potential savings to insurers is three times higher than the $20.1 billion to $23.0 billion estimate in Naugle (2009). The difference is explained mainly by the definition of administrative costs used by each analysis: Naugle (2009) compared current administrative levels to a best-practice benchmark in the current system. Jensen (2009) compared entire countries, in a regression model, finding that the U.S. private insurers contribute $63 billion more to costs than if the United States had (mostly nonprofit) private insurance as is the situation in member countries of the Organisation of Economic Co-operation and Development (OECD) countries. Kahn (2009) obtained an estimate similar to Jensen by comparing current private overhead from the national health expenditures to public program overhead rates.

Session 4: Prices That Are Too High

The prices of medical services and products also have been identified as an area of potential waste in the U.S. healthcare system. The presentations in this session focused on:

  • Hospital service prices;
  • Prices of medications;
  • Prices of durable medical equipment; and
  • Prices of medical devices.

Hospital Service Prices

Hospital consolidations may help reduce operating costs by increasing efficiency; however, consolidations may also result in increased prices. Because public programs such as Medicare and Medicaid reimburse hospitals based on set fee schedules, private payers are affected by increases in prices caused by consolidation (Capps, 2009). These price increases, in turn, can drive up the cost of insurance for those with private coverage (Capps and Dranove, 2004). A review of 87 papers on hospital consolidation and its impact on costs, quality, and pricing indicated there were small cost savings brought about by most mergers and acquisitions, nil or negative effects on quality, and the potential for substantial price increases, particularly when hospital mergers occur in a geographically narrow area (Vogt and Town, 2006).

This section reviews the results presented by Cory S. Capps from his examination of the magnitude by which hospital consolidations have increased prices, and by extension national health expenditures. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Price increases attributable to hospital consolidation Capps (2009) estimated, based on conservative assumptions, that hospital consolidations have caused an increase of approximately $10 billion to $12 billion in annual national healthcare expenditures. To reach this conclusion, Capps identified the 94 metropolitan statistical areas (MSAs) that satisfied two conditions: (1) each had a population large enough to support multiple independent hospitals, and (2) each was concentrated. The author calculated the predicted price change in these MSAs if the market concentration were reduced from the actual level to the “relatively unconcentrated” level.

A comprehensive survey of the literature on concentration and hospital pricing conducted by Vogt and Town (2006) concluded that hospital prices increased by 1 percent on average for every 160-point increase in the Herfindahl-Hirschman Index (HHI) of concentration (the HHI is a widely used measure of concentration in antitrust analysis).

Results from Capps’ analysis indicated that private payers’ payments to hospitals are about 3 percent higher nationwide than they would have been without the market consolidation. By extension, this means that national healthcare expenditures are an estimated 0.4 percent to 0.5 percent higher (a total of $10 billion to $12 billion) on an annual basis than they would have been absent the extensive consolidation of hospital ownership that began in the mid-1990s.

Some limitations of the analysis include the assumption that inpatient and outpatient prices move in the same manner; this may be inaccurate as outpatient competitive conditions may be different from those in the inpatient market. The analysis also only identified the direct price effect when there may also be other types of effects attributable to consolidation. For example, larger hospital systems with market power may be able to resist payer attempts to control use; reduced hospital competition may also increase hospitals’ incentive to operate efficiently. Both could increase costs of hospital care to public as well as private payers. Finally, the analysis only provides general trends and averages, and may not reflect a specific market’s price experience due to consolidation.

Additional estimates A review of the literature on hospital consolidations indicated that, in general, studies found evidence of price increases after hospital mergers (Capps and Dranove, 2004; Krishnan, 2001; Krishnan and Krishnan, 2003; Vogt and Town, 2006). However, no studies extrapolated their results to the national level.

A review by Vogt and Town (2006) of different types of hospital consolidation studies found that most studies found evidence of large merger-induced price increases. For example, a review of event studies found that hospital prices typically increased by at least 10 percent after a merger. More specifically, out of 13 studies, 10 found increases of at least 2 percent (Vogt and Town, 2006).

Capps and Dranove (2004) examined the effect of hospital consolidations on negotiated prices with preferred provider organizations based on data from hospital contracts. They conducted a multivariate regression analysis designed to estimate the effect of mergers on the negotiated price. For 9 of the 12 hospitals that experienced an increase in market power sufficient to potentially trigger antitrust scrutiny, prices increased by significantly more than the median price increase. A cross-sectional analysis of four markets in which consolidations occurred also found that prices increased for hospitals that merged in three of the four markets.

Finally, studies by Krishnan and Krishnan (2003) and Krishnan (2001) found that prices increased more for hospitals that experienced a merger compared to those that did not. Krishnan and Krishnan (2003) analyzed data from 113 hospitals in California, of which 20 experienced an acquisition between 1995 and 1996. The authors found that acquired hospitals had increased revenue per patient but did not have lower operating costs attributable to the merger. One limitation of the study was that it examined prices for only 1 year after the merger, thus it cannot be determined whether the higher prices remained beyond that time horizon (Krishnan and Krishnan, 2003).

Estimates comparison Estimates from other papers support the analysis presented by Capps (2009), which relies on studies showing that hospital consolidations lead to price increases. In addition, the magnitude of the price increases used by Capps to calculate his estimate (3 percent) is similar to the price increases found in other peer-reviewed literature. However, a comparison of the impact of higher prices on the nation as a whole cannot be made, as other papers did not extrapolate their findings to the national level.

Although not directly relevant to hospital service pricing, a related literature survey examines the income and salaries of physicians. Farrell and colleagues (2008) compared physician incomes in the United States to those in other OECD countries, concluding that U.S. generalists make 4.1 times per capita gross domestic product (GDP), compared with 2.8 times per capita GDP in other OECD countries, while specialists make 6.5 times per capita GDP, compared with an OECD average of 3.9 times. Farrell and colleagues additionally found that, across all U.S. physicians, higher earnings add $64 billion in costs to the U.S. system, the sum of $49 billion more for specialists and $15 billion more for generalists. Although physician salaries may not be growing after adjustment for inflation, recent analyses indicated that primary care physicians in the United States have seen more significant negative impacts than specialists (Litzau, 2009; Tu and Ginsburg, 2006). In their review, the CBO (2008) additionally reported that physicians typically increase the volume of their services in response to reductions in payment rates so as to offset between 20 percent and 40 percent of the rate cut’s impact on their total payments.

Prices of Medications

A number of factors affect prescription drug prices in the United States. These factors include whether the drug is a brand-name or generic drug, and who pays for the drug. Government purchasers have access to price ceilings and mandated rebates, among other mechanisms, and as a result generally pay lower prices than private purchasers (CBO, 2005). Further complicating the question of whether prescription drug prices are too high is the fact that the supply and payment chains move differently from each other. Purchasers generally do not take possession of the prescription drug, and as a result pay the supplier (pharmacies) and negotiate rebates separately with the manufacturer (Hoadley, 2009).

Recent trends in drug spending and pricing show some changes over prior years. Drug spending growth hit a 45-year low of 4.9 percent in 2007 (Hartman et al., 2009; Sisko et al., 2009). This was likely attributable to lower price growth generally, safety concerns, and the recession. The average growth rate, however, hides significant differences in price trends for brand-name, generic, and specialty drugs; for example, brand-name drug prices increased much faster than overall drug prices (8.7 percent vs. 4.5 percent) in 2008 (Purvis, 2009). Another factor affecting trends is the increasing use of generic drugs as opposed to brand-name drugs. Because generics are much less expensive than brand names, substituting a generic for a brand name is generally less costly for both the purchaser and the patient (Hoadley, 2009). Looking internationally, studies have found that U.S. prices for brand-name drugs are about twice those of four other developed countries, but generic drug prices are much lower in the United States than in those other countries (Paris and Docteur, 2006).

This section presents a review by Jack Hoadley of prescription drug-pricing trends and savings estimates in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Prescription drug prices Hoadley (2009) reviewed a variety of estimates that have been released indicating that reductions in prescription drug prices could save significant amounts. One estimate by the CBO found that the government could save $10 billion annually (between 2010 and 2019) by requiring manufacturers to pay a 15 percent rebate on Medicare Part D drug purchases (CBO, 2005). In general, a broad estimate looking at the effect of a 5 percent across-the-board price reduction (excluding government purchasers that already receive significant discounts) found that total savings for the health system could be about $9 billion annually (Hoadley, 2009).

A study comparing drug prices across different countries found that, while brand-name drug prices in the United States are roughly twice those in Australia, Canada, France, and the United Kingdom, generic drug prices in the United States are between 10 and 65 percent below prices in those countries (Paris and Docteur, 2006). By taking advantage of the low generic drug prices, an industry estimate indicated that increasing the generic dispensing rate by 3 percent annually could save $10.5 billion (Genetic Pharmaceutical Association, 2009). Finally, the CBO (2008) estimated that allowing manufacturers to create follow-on biologics (generic versions of biologic medications) could save $13 billion over 10 years.

Some limitations to discussions about prescription drug pricing include a lack of standards for establishing the “optimal” price, and the fact that lowering prices in one commercial market may increase prices in another, thereby reducing or eliminating the potential savings. In addition, lower drug prices may potentially reduce funds available for investment in research and development. Finally, studies assessing the potential share of drug use amenable to switching to generics are lacking; as a result the above estimate of a 3 percent shift is solely an example.

Additional estimates A review of the literature related to prescription drug pricing and the potential savings associated with different policies yielded one national estimate and two other papers related to the comparison of U.S. prices with those of other countries. Gellad and colleagues (2008) used Medical Expenditure Panel Survey drug use and spending data from 2003–2004 to estimate the savings for Medicare Part D beneficiaries if Medicare drug prices in Part D were reduced to Federal Supply Schedule (FSS) prices. The authors estimated annual savings to the Medicare program of $21.9 billion for the top 200 drugs used by beneficiaries after inflating the drug costs to 2006 dollars (Gellad et al., 2008). Of note, this estimate may overstate the true potential savings; the lower-end sensitivity estimate of $11 billion may be more reasonable. In addition, the comparison of FSS prices to retail transaction prices may be inappropriate, as retail prices do not take into account manufacturer rebates. As this study used data prior to the initiation of Part D, it does not reflect any changes in use associated with the start of the program.

Schoen and colleagues (2007) estimated the potential savings from allowing the Secretary of the Department of Health and Human Services to negotiate prices for Medicare Part D. The estimate is based on a three-tiered policy approach, by which Part D would pay Medicaid prices for dual eligibles, prices would be set for unique drugs, and the Secretary would establish a purchasing collaborative comprising all government payers (with voluntary private-sector participation). Schoen and colleagues (2007) estimate that this policy change would yield net savings of $15.8 billion over 5 years ($43.4 billion over 10 years).

Two analyses compared pharmaceutical prices in the United States to those of other countries, using index measures related to wealth. Farrell and colleagues (2008) found that prices in the United States were 50 percent higher compared to other countries; however, the prices varied depending on the type of drug. Brand-name drug prices were 77 percent higher in the United States while generic drugs were 11 percent lower in the United States (Farrell et al., 2008). Danzon and Furukawa (2008) compared drug prices in the United States to those in 11 other countries. When taking income into account, the authors found that most countries’ price indices were relatively close (within 10 percent) to the United States except for the three Latin American countries examined and Japan. Danzon and Furukawa (2008) also found that generic drugs are less expensive in the United States, with other countries price indices being anywhere from 8 percent to 111 percent higher. CBO analyses also found that importation of medications from a broad set of industrialized countries could reduce drug spending by approximately $40 billion over 10 years (CBO, 2004b).

Additionally, a recent analysis found that Medicare Part D pays on average 30 percent more for drugs than does Medicaid (U.S. House of Representatives Committee on Oversight and Government Reform, 2008). Prior to the implementation of Medicare Part D, those dually eligible for Medicare and Medicaid received their prescription drug coverage through Medicaid, which is legally allowed to negotiate drug discounts with manufacturers. However, now all dually eligible beneficiaries receive their prescription drug coverage through Medicare Part D, which is offered through private insurers who do not have the ability to negotiate drug prices with manufacturers. Since the medications the dually eligibles receive through Medicare Part D are, on average, 30 percent more expensive than those previously received through Medicaid, it was estimated that drug spending for this population increased by over $3.7 billion in the first 2 years of the Medicare Part D program. It was also estimated that if Medicare Part D paid the same price as Medicaid for all drug purchases, the total savings over the next 10 years could be as much as $156 billion.

Estimates comparison Among the multiple analyses of pharmaceutical pricing, it can be noted that all international comparisons found similar trends in pricing across countries. More specifically, various sources found that while brand-name drug prices were higher in the United States than in other countries, generic drug prices are lower in the United States. It is more difficult to compare estimated savings attributable to the various proposals, given that the estimates focus on different policy options to lower medication expenditures. For example, the CBO estimate (2008) of the potential savings created by requiring a 15 percent rebate for Medicare Part D targets pharmaceutical manufacturers, while Gellad and colleagues (2008) estimate of the savings from requiring FSS pricing would likely require mandating changes in the current supply and payment systems in order to achieve such a price (Hoadley, 2009). In addition, both the estimates provided by Hoadley and CBO will vary depending on the magnitude of the price reduction used in the respective calculations.

Prices of Durable Medical Equipment

Durable medical equipment (DME) prices may be too high as a result of two factors. First, patients often have insurance coverage for such equipment, and second, patients often have no choice as to whether they need to purchase the equipment (Hoerger, 2009). In 2007, the United States spent a total of $24.5 billion for DME (CMS, 2009). Currently, Medicare DME payments are based on a set fee schedule as opposed to bids.

This section presents a review by Thomas J. Hoerger (2009) and Mark E. Wynn (2009) of DME pricing and savings estimates for the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Prices for durable medical equipment Hoerger (2009) estimated that a reduction in Medicare reimbursements, fraud, and waste for DME could save the program $2.8 billion annually (0.1 percent of total national health expenditures in 2007). There is a significant body of evidence indicating that Medicare pays too much for DME, including an Office of the Inspector General (OIG) report that found that Medicare fees in 2003 exceeded Web site prices by 37 percent (Department of Health and Human Services, 2004). The OIG also found that Medicare payments for oxygen concentrators were almost $7,000 higher than the supplier purchase cost (Department of Health and Human Services, 2006). As a result, reducing spending on DME may result in savings to the Medicare program.

Hoerger (2009) and Wynn (2009) examined the results of Medicare demonstration projects designed to determine whether Medicare could achieve a lower price via alternative methods of determining payment levels. Two demonstration projects found that by implementing a program of competitive bidding, Medicare could save between 19 and 20 percent off the fee schedule. In addition, the costs of operating the bidding program were lower than the savings achieved, indicating a potential net savings to the government overall. In 2003, Congress established a program of competitive bidding for DME, and initial bids were 26 percent lower than the fee schedules. The program has not yet gone into effect, however, as Congress has delayed it for 18 months and required that bids be submitted again (Wynn, 2009).

In addition to competitive bidding, Hoerger (2009) suggested that fraud and waste contributes to some of the excessive payments for DME in the Medicare program. By using estimated Medicare overpayments of 10 percent (2006) as a proxy for fraud and waste, and combining that estimate with the demonstration findings of a possible 20 percent reduction in prices, Hoerger (2009) estimated that Medicare could save $2.8 billion (annually) on DME. This represents 11.5 percent of the total national spending on DME.

Some limitations to the above estimates include the fact that the 20 percent reduction in fees may no longer be possible owing to subsequent changes in the fee schedule and in the market. However, even though reductions in the fee schedule occurred prior to the demonstrations, they still yielded bids that were 20 percent lower (Hoerger, 2009). Another consideration is that the estimated savings would only accrue to the Medicare program, and the fee reduction may not have much effect on DME use. Given that use is the primary factor in expenditures for DME, this may reduce the potential savings available (Hoerger, 2009).

Additional estimates A review of the literature found no other published studies related to the potential savings achievable by Medicare besides those already discussed. However, one analysis by Farrell and colleagues (2008) found that DME spending in the United States is actually $19 billion less than expected, relative to wealth. Farrell and colleagues (2008) attribute this finding to the general lack of health insurance coverage in the United States for DME and because of the slow growth rate of DME spending over 4 years (2003–2006).

Estimates comparison The estimates presented by Hoerger (2009) and Wynn (2009) are not comparable to the findings by Farrell and colleagues (2008) for a variety of reasons. First, Hoerger (2009) and Wynn (2009) were examining the potential savings attributable to changes in the reimbursement structure for the Medicare program, while Farrell and colleagues (2008) examined overall expected spending relative to the wealth of the entire United States. Thus, Farrell and colleagues were including data from all payers in the country, not just the Medicare program. As Hoerger and Wynn focused on Medicare spending for DME, there may be potential for additional savings in the private health insurance market for DME.

Prices of Medical Devices

The medical device market in the United States is characterized by differentiated products and strong influence by medical staff on the purchasing decisions for these products. In addition, the prices of such devices are often confidential, reducing the ability of hospitals to bargain effectively with the device manufacturer (Pauly and Burns, 2008).

This section presents an analysis by Jeffrey C. Lerner of the potential savings possible by negotiating lower medical device prices in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Prices for medical devices Lerner (2009) estimated that hospitals could have saved approximately $4.7 billion (2008 dollars) if they had negotiated to the average price paid for medical devices.

To calculate the estimate, the author evaluated data from 123 hospitals, which incorporated information on supplies purchased over a 4-month period, as well as data from 1,500 hospitals and health systems on the prices offered for 5 types of capital equipment over a 12-month period. Results from the 123 hospitals indicate that, for medical supplies, if hospitals were to negotiate to the average price paid for each device, hospitals could have saved approximately 3.1 percent ($4.7 billion in 2008) off their costs. For capital equipment purchases, wide variation in the prices offered to hospitals was seen; the average discount achieved across all technologies was 29.6 percent. Of note, smaller hospitals sometimes received lower price offers than large teaching hospitals.

Some limitations of the analysis include the fact that the price information was for supplies and not for all medical devices; this could bias the results depending on the variation in prices for devices not included in the data. In addition, the analysis focused on hospital spending; other providers also purchase devices, and their potential savings could not be disambiguated for the analysis. Also, it may be that some buyers are unable to negotiate prices effectively. In this case, the estimated savings would not be plausible. Finally, the savings indicated represent simply a transfer of resources and do not represent a reduction in unnecessary use. This, however, could lead to restraining payment increases in some DRGs if Medicare feels less pressure to increase payment levels because the technology costs less (Lerner, 2009).

Additional estimates A review of the literature related to medical devices found no other studies estimating the potential savings from lower medical device prices. However, Pauly and Burns (2008) suggest that increased transparency of medical device prices, allowing for a range of prices to be made publicly available, should increase the ability of hospitals to work with physicians to negotiate with manufacturers. In turn, this increased bargaining power could result in lower prices for medical devices.

Estimates comparison A review of the literature related to medical device pricing found no other studies estimating either national or local cost savings from negotiating lower medical device purchase prices. As a result, a comparison of the cost estimates cannot be conducted for this section.

Session 5: Missed Prevention Opportunities

Almost 40 percent of deaths every year are attributable to modifiable behavioral risk factors, including tobacco, poor diet, physical inactivity, and alcohol consumption (Mokdad et al., 2004). Recently, preventive services have received increased attention from policy makers. In January 2000, the Department of Health and Human Services launched Healthy People 2010, with the aim of promoting health and encouraging disease prevention across the United States (National Center for Health Statistics, 2009). More recently, Michigan appointed a surgeon general to address health promotion and disease prevention, while Vermont integrated prevention into health reform, including community workshops on healthy lifestyles (Wilson, 2009). Although the health benefits of increased prevention seem clear, the possibility that it might also lower spending by preventing the occurrence of future disease is an enticing one. During the presidential campaign, it was claimed that “[g]uaranteeing access to preventive services will improve health and, in many cases, save money” (Cutler and DeLong, 2008).

In this section, two estimates of the potential impact of increasing the delivery of preventive services on healthcare costs in the United States are discussed. Thomas J. Flottemesch considered the role increased primary and secondary preventive services could play, while Michael P. Pignone performed a similar investigation for tertiary preventive services. Other comparable studies—when they exist—are also presented, and the estimates from these studies are compared.

Treatment Costs from Missed Prevention Opportunities

Savings from increased primary and secondary prevention Flottemesch (2009) estimated the effect of increasing primary and secondary preventive clinical services on national healthcare expenditures. In particular, they modeled the impact of increasing the use rate of preventive services to 90 percent on 2006 national expenditures. The author found that increasing the target use rate to 90 percent for all recommended preventive services would have led to a decrease in net expenditures of $3.7 billion (0.2 percent of U.S. personal healthcare spending in 2006), with primary preventive services alone yielding an estimated net savings of $7.0 billion (0.4 percent of 2006 U.S. healthcare spending).

Flottemesch examined preventive services recommended for the general population by the U.S. Preventive Services Task Force and the Advisory Committee on Immunization Practices. Each service is classified as a primary (meant to prevent the occurrence of a medical condition) or secondary (meant to identify medical conditions in an asymptomatic state) or, in some cases, both. The data for the calculations were culled from literature reviews, and the estimates were generated using models developed in support of the work of the National Commission on Prevention Priorities, which are carefully designed so as to allow consistent comparison among and between clinical preventive services.

Table A-4 shows how the projected impact on medical expenditures varies by preventive service. For example, increasing the 2006 delivery level of tobacco screening from 28 percent to 90 percent would have decreased net expenditures by $5.6 billion, and increasing the delivery of discussing daily aspirin use from 33 percent to 90 percent would have decreased net expenditures by $3.3 billion. On the other hand, increasing delivery of cholesterol screening from 79 percent to 90 percent would have increased net expenditures by $1.5 billion. Therefore, Flottemesch’s calculations suggested that lumping prevention into one large undifferentiated group may be counterproductive, and that investing in an evidence-based package of preventive services could produce net cost savings.

TABLE A-4. Projected Impact on Medical Expenditures by Preventive Service.

TABLE A-4

Projected Impact on Medical Expenditures by Preventive Service.

The authors noted a number of limitations to their study. First, measurement error is a serious concern given that the analyses drew from a wide variety of sources. Second, costs may have been omitted or counted twice. Furthermore, properly modeling the effect of multiple risk factors is a priori unclear and perhaps leads to overstatement or understatement of the effect on net expenditures. Importantly, costs needed to achieve increased use, such as outreach to patients and delivery system changes to improve clinicians’ ability to offer these services, were not included in this analysis. Finally, since indirect and transitional expenditures—such as productivity gains and losses—are excluded from the analysis, it is possible that costs and savings are understated across the board. The author concluded that it was most prudent to interpret their findings as, at best, net expenditure neutral.

Savings from increased tertiary prevention In a complementary analysis, Pignone (2009) attempted to estimate the effect of increasing the use of tertiary prevention. This type of prevention focuses on patients with established health conditions, particularly chronic conditions, with the goals of preventing additional morbidity, improving quality of life, and reducing disability. The author estimated that annual savings of $45 billion could be achieved through enhanced tertiary prevention.

Pignone examined examples of effective interventions in areas such as disease management, discharge coaching aimed at reducing rehospitalization, and palliative care. With respect to discharge coaching, for instance, Coleman and colleagues (2006) examined one such program and found that mean costs in the noncoached group were $2,546; costs were $2,058 for the coached group, a 19 percent reduction. Based on his survey, Pignone suggested that currently available interventions could, conservatively speaking, produce 10 percent spending reductions on average. If 30 percent of the $1.5 trillion currently spent on patients with chronic conditions could be affected by enhanced tertiary prevention, this 10 percent change on the spending base would yield an estimate of $45 billion in savings.

Three main limitations to this savings estimate were noted. First, because the proportion of real-world spending amenable to tertiary prevention is difficult to estimate, this estimate is far from certain. Second, external validity may not hold: the effectiveness of a successful intervention may not be replicable elsewhere, especially when that intervention is implemented on a wide scale. Consider that the original programs in which the interventions were implemented often have highly experienced and specially trained staff with high levels of enthusiasm. Limitations in skills or training and lower degrees of enthusiasm may produce more modest results. Current administrative arrangements may also preclude the establishment and maintenance of multidisciplinary, patient-centered teams. Finally, it is not clear that the proper incentives are in place for successful tertiary preventive measures to be widely implemented. In the current fee-for-service payment system, many payers have no means of compensating providers for more efficient, nontraditional means of service delivery.

Additional estimates The CBO reported that “although different types of preventive care have different effects on spending, the evidence suggests that for most preventive services, expanded utilization leads to higher, not lower, medical spending overall” at the federal level (Elmendorf, 2009). Russell (2009) reviewed nearly 600 cost-effectiveness studies from 2000 to 2005 and found that less than 20 percent of the preventive services were found to be cost saving. Russell also noted that studies over the past 4 decades have “shown that prevention usually adds to medical spending” (p. 45). Heavily cited reviews published in previous years, such as Coffield and colleagues (2001) and Stone and colleagues (2000), have also presented findings along similar lines. Other reports from outside the peer-reviewed literature have examined particular interventions and come to qualitatively different conclusions. The Commonwealth Fund (2009) concluded that substantial savings could be achieved from reducing the use of tobacco (a net cumulative reduction in national health expenditures of $255 billion over 11 years) and the incidence of obesity ($406 billion savings over the same time period). Berenson and colleagues (2009b) analyzed the cost-saving potential of interventions aimed at preventing diabetes among those at highest risk. Not only could such a program decrease the incidence of diabetes by half, the authors estimated net savings of 0.6 percent of personal healthcare expenditures over 10 years. The total 10-year savings would be $191 billion, of which 75 percent ($142.9 billion) would constitute savings to Medicare and Medicaid. PriceWaterhouseCoopers’ (2009) Health Research Institute estimated annual excess costs attributable to smoking and conditions related to obesity at $567 billion to $161 billion and $200 billion, respectively; the costs of poorly controlled diabetes were $22 billion, while nonadherence cost another $100 billion.

As of this writing, the publication that is perhaps closest to Pignone’s analysis (Pignone, 2009) is a Milken Institute report published in 2007 (DeVol et al., 2007). The authors estimated the impact increased prevention and early intervention for seven common chronic diseases—cancer, diabetes, hypertension, stroke, heart disease, pulmonary conditions, and mental disorders—could have on medical expenditures on the national level. Assuming “reasonable improvements in health-related behavior and treatment,” they found that “the cumulative avoidable treatment costs from now to 2023 would total a whopping $1.6 trillion” and the single-year savings in 2023 would be $217 billion in their most optimistic modeling scenario. Underuse of appropriate medications for chronic conditions has been cited as a large factor contributing to waste in disease management and tertiary prevention, with the underuse of generic antihypertensives and controller medications in pediatric asthma estimated to cost over $5.5 billion annually (Delaune and Everett, 2008). However, several reviews of the disease management literature have found mixed evidence and have been cautious in projecting cost savings (Delaune and Everett, 2008; Goetzel et al., 2005).

Estimates comparisons The result in Flottemesch (2009) is generally consistent with the findings of the peer-reviewed literature. Both Flottemesch (2009) and the peer-reviewed literature provide much lower estimates than those presented in the nonpeer-reviewed reports mentioned above.

Pignone (2009) and DeVol and colleagues (2007) are not directly comparable, primarily because the former estimated a single year’s savings while the latter provided estimates projected more than a decade into the future. The estimate fromPriceWaterhouseCoopers’ Health Research Institute (2009) is also not directly comparable given the difference in focus between the report and IOM conference paper. It is worth noting that surveys of studies with less comprehensive estimates come to more guarded conclusions than those of Pignone (2009), DeVol and colleagues (2007), and PriceWaterhouseCoopers’ Health Research Institute (2009). a The CBO (2004a) surveyed peer-reviewed evaluations of disease management programs for (primarily) diabetes, coronary artery disease, and congestive heart failure, and found that there was not enough evidence to support the claim that these programs generally reduce federal spending. Another review by researchers at RAND (Mattke et al., 2007) and at the New England Health-care Institute (Delaune and Everett, 2008) considered a broader range of diseases and also found that the evidence could not conclusively determine if these programs reduced costs. On the other hand, Kim R. Pittenger presented results in the July workshop that supports the estimate of Pignone (2009). Although these data were nonexperimental and the generalizability of the estimate may be limited, the finding is indeed provocative.

Ultimately, Steven H. Woolf argued at the May workshop that asking “how much can we save” is the wrong question. Rather, the focus should be shifted from cost savings to value, as lack of savings does not mean lack of cost-effectiveness. He asserted that “the first priority in bending the curve to slow growth in spending is less about searching for the handful of services that produce net savings and more about shifting spending from low-value to high-value services” (Woolf, 2009).

STRATEGIES THAT WORK

Building on the discussions in the first workshop, the July workshop explored methods of decreasing inefficiency and waste and their likely net yield, including:

  • Knowledge enhancement-based strategies;
  • Care culture and system redesign-based strategies;
  • Transparency of cost and performance;
  • Payment- and payer-based strategies;
  • Community-based and transitional care strategies; and
  • Entrepreneurial strategies and potential changes in the state of play.

Session 1: Knowledge Enhancement-Based Strategies

The ability to transform the delivery of care at the level of patient-provider interactions will certainly depend on the ability to generate and apply knowledge at the point of care. Amid the dialogue of reform, the American Recovery and Reinvestment Act of 2009 allocated $1.1 billion to comparative effectiveness research (CER), a key tool in optimizing efficient use of healthcare resources. Use of evidence-based protocols has been employed to improve quality and efficiency in the delivery of patient care. Significant attention has also focused on the ability of health information technology to provide clinical decision support and facilitate care coordination.

The presentations in this session focused on strategies to enhance the knowledge base, including:

  • Comparative effectiveness research;
  • Evidence-based clinical protocols; and
  • Electronic health records with decision support.

Comparative Effectiveness Research

Health care provided in the United States is not always based on evidence supporting the effectiveness of a particular intervention. Complicating this, in cases where more than one treatment option exists, there may not be comparative evidence showing the relative effectiveness of each treatment (Deloitte & Touche LLP, 2009). As a result, the implementation of care models that incorporate evidence about the effectiveness of specific interventions could help lower costs and improve quality in the healthcare system.

This section presents a discussion by Carolyn M. Clancy (2009) that examined the current state of comparative effectiveness research in the United States. Cost estimates related to the use of comparative effectiveness research from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

The comparative effectiveness research agenda Clancy discussed AHRQ’s agenda and efforts to undertake research on comparative effectiveness. Comparative effectiveness analyses should be conducted and organized in a manner that provides those who are making decisions about health care access to the most recent evidence-based information related to the options for treatment (Clancy, 2009).

Clancy also discussed the latest funding provided by Congress for comparative effectiveness research, via the American Recovery and Reinvestment Act (ARRA) of 2009. The Act provided $1.1 billion for research, which is split among AHRQ, the National Institutes of Health (via AHRQ), and the Office of the Secretary for Health and Human Services. The ARRA legislation also required the IOM to develop priorities for CER funding. In addition, funding priorities will consider the definition of comparative effectiveness research offered by a newly established Federal Coordinating Council for CER, which incorporates comparisons of interventions and decision making that is tied to the individual needs of patients (Clancy, 2009).

Finally, Clancy addressed some issues to consider in the area of comparative effectiveness research. First, comparative effectiveness, while a useful tool, is not sufficient by itself to change the delivery of care. Results from such analyses do not tell doctors how to practice medicine, do not make health decisions, and are not related to decisions as to whether to pay for care. Rather, CER is primarily useful in that it presents evidence in a manner that enables decision makers to make the best possible decisions given the evidence.

Additional estimates A review of the literature related to comparative effectiveness research returned no peer-reviewed papers that estimated the total system savings associated with comparative effectiveness. However, several other reports were found that addressed the potential savings.

One estimate assumed that a Center for Comparative Effectiveness would be created that would fund research on CER and make copayment and pricing recommendations based on this research (The Commonwealth Fund, 2009). If these recommendations were adopted by public and private payers into benefits design and payment and pricing policies, the authors estimated that national savings could be $480 billion between 2010 and 2019. Berenson and colleagues (2009b) described the potential uses for comparative effectiveness research, but they declined to provide an estimate because of the uncertainty associated with the methods by which the research would be applied and whether payers would in fact be able to limit coverage of technologies that were shown to be less effective. The CBO cited uncertainty on the impact of CER on expenditures given the difficulties in predicting adoption and use; however, they detailed the potential of CER to reduce healthcare costs over the long term—possibly by substantial amounts if CER were rigorously performed and if the results were ultimately tied to changes in financial incentives for providers and consumers (CBO, 2007).

Estimates comparison Because only one paper presented an estimate of the potential savings attributable to comparative effectiveness research, a comparison cannot be conducted. However, savings from comparative effectiveness research will depend on the ability of payers and government to change other aspects of the current healthcare system to realign incentives to encourage the use of more effective treatments.

Evidence-Based Clinical Protocols

As discussed in the section on CER, healthcare decisions often do not take into account the evidence associated with the effectiveness of a particular treatment, and comparisons of the effectiveness of multiple treatment options are often not available (Deloitte & Touche LLP, 2009). Over the past several years there has been a growing call for the development of processes and procedures by which evidence can be incorporated in care delivery. Evidence-based health care is characterized by a focus on the evidence on the effectiveness of a particular treatment, as opposed to treatment based on clinical observation and experience (EBM Working Group, 1992).

This section presents the results from an analysis by Lucy A. Savitz that examines the potential savings associated with the implementation of a targeted evidence-based care model in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Cost savings from evidence-based care models An analysis presented by Savitz estimated $2 billion in annual savings from a targeted evidence-based clinical protocol designed to improve quality of care and reduce unnecessary admissions for febrile infants. This estimate was based on extrapolation from savings estimated from implementation of an evidence-based care process model at a large healthcare system in Utah.

Evidence-based care models provide clinicians with guidance on care management by presenting them with state-of-the-art knowledge. These models provide information based on accessible references and guidelines and can often improve on the clarity of prior guidelines, as well as providing timely support for decisions related to a patient’s condition. Evidence-based care models at the Utah system were designed to target cost drivers, including length of stay, readmissions, and ED visits. To apply these models across the United States, however, care coordination across currently uncoordinated and nonintegrated systems is needed, and there is some concern as to the degree to which savings realized in the Intermountain Healthcare delivery system could be realized and sustained in other settings.

Savitz presented some caveats associated with adoption of evidence-based health care. First, facilitating diffusion of these model requires outside intervention (Dopson and Fitzgerald, 2005), and thus concrete steps must be taken in order to encourage adoption. Second, savings achievable by other clinics and health systems may vary owing to differences in the costs of adoption in each system. Finally, the sustainability of cost savings after initial implementation of the model remains unclear. However, it does appear that, by focusing efforts, improvements will occur (Wachter and Pronovost, 2006).

Additional estimates A review of the literature resulted in a number of articles that discussed the savings realized from implementation of evidence-based care. A number of savings estimates from implementation of specific evidence-based models have been completed. First, PriceWaterhouseCoopers (2009) has estimated that $1 billion in wasteful healthcare spending is caused by the overprescribing of antibiotics. Establishing clinical protocols designed to reduce such overprescribing could yield some savings. In addition, Stuart and colleagues (1997) estimated over $500,000 per year in health system savings from an evidence-based model designed to manage patients with symptoms of acute dysuria. The savings were primarily attributable to reduced visits, lab tests, and prescriptions (Stuart et al., 1997). Another estimate by Wagner and colleagues (2001) found that an evidence-based care model applied to diabetes patients could save $400 to $4,000 per patient over a 3-year time period. Finally, results from the application of new clinical guidelines for the treatment of high blood pressure in elderly patients could result in $20.5 million in savings for the Medicaid program (Fischer and Avorn, 2004).

UnitedHealth Group (2009a) estimated the potential savings to the Medicare program from implementing changes that included the application of evidence-based clinical guidelines. These changes, referred to as an “integrated medical management program,” also included annual care assessments, changes in the benefit design and reimbursements, and assistance in patient decision making. UnitedHealth Group compared hospital admissions from their Medicare Advantage plans to those of fee-for-service Medicare, and estimated potential savings from such a program to be $102 billion over 10 years (2010–2019). Additionally, UnitedHealth Group modeled the application of evidence-based standards to reimbursement policies, including radiology benefit management and prospective claims review, estimating an additional $75 billion in potential federal savings over the next decade. Using evidence-based clinical guidelines, Mecklenburg and Kaplan (2009) estimated a potential $6.5 billion in annual savings from reducing unnecessary MRI testing for back pain and migraines.

Estimates comparison Although the cost estimates are not directly comparable because they address different clinical problems and protocols, the evidence suggests that evidence-based care protocols have the potential to improve quality and provide cost savings. However, while Savitz’s national estimate is derived from one specific care model for febrile infants and cannot be directly compared to the estimate of savings presented by UnitedHealth Group, they are likely complementary as they address separate clinical problems.

Of note, some policy options discussed by Savitz (2009) include requiring inclusion of evidence-based care models in research on comparative effectiveness; creation of a clearinghouse where systems can access previously created evidence-based care models; and elimination of flaws in reimbursement that lead to perverse incentives to increase care. The last of these is incorporated in the UnitedHealth Group (2009a) estimate of savings.

Electronic Health Records with Decision Support

Most medical records in the United States are still stored on paper in physician offices, making coordination of care with other healthcare providers, quality measurement, and reduction of medical errors extremely difficult (Hillestad et al., 2005). With the ability to facilitate improved care coordination and reduction in medical errors, the adoption of electronic health records (EHRs) can result in savings to the healthcare system (Hillestad et al., 2005). In addition to the patient medical record, EHRs can allow providers to write prescriptions electronically (e-prescribing), request tests and treatments via computer (computerized physician order entry [CPOE]), and obtain decision support via computerized systems.

This section presents a discussion by Rainu Kaushal that examined the potential savings associated with the implementation of EHRs with decision support in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Savings from implementation of EHRs Kaushal (2009) discussed some cost saving estimates associated with implementation of various types of EHRs. A study by Walker and colleagues (2005) estimated that adoption of nationwide electronic information exchange and interoperability could save $77.8 billion annually. When CPOE is adopted in the inpatient scenario, savings estimates range from $1 million to $3 million annually per hospital after an initial investment (Massachusetts Technology Collaborative & New England Healthcare Institute, 2009). In addition, savings from adoption of EHRs in the ambulatory setting are estimated to be $86,400 per provider over 5 years (Wang et al., 2003). The CBO (1998) estimated savings from switching to generic drugs to be $8 billion to $10 billion per year, which could be facilitated by e-prescribing.

Key characteristics of EHRs include improvement of access to information, an increase in timely feedback, increased accuracy in coding and billing, and an overall change in healthcare delivery. Costs that can be reduced via EHRs include preventive care, chronic care, transitions (from one provider setting to another), and medications. In addition, Kaushal discussed the various efficiency effects of EHRs, including reduction of transcription costs, billing errors, and office visits, and reduction of redundant tests. An increase in quality has also been seen, including improved guideline adherence and improvement on performance. Kaushal (2009) indicated that the estimates of improvements in quality and cost savings depend on the specific type and definition of EHRs being studied.

Kaushal also discussed some caveats associated with implementation of EHRs. First and foremost, implementation is very difficult. It requires significant financial investment, adjustments in workflow design, and requires support staff for technical issues. Difficulties also arise when considering expanding EHRs nationwide and making them interoperable. Finally, evaluation of the value associated with EHRs is even more difficult than implementation. Not only is it expensive, but it requires extremely focused research efforts.

Additional estimates A review of both peer-reviewed literature and other reports yielded a significant number of papers that sought to estimate the costs and savings associated with implementation of EHRs.Hillestad and colleagues (2005) used the Healthcare Information and Management Systems Society survey and a literature review to estimate the rates and costs of adoption of EHRs in various provider settings. Based on their analyses, they estimated that EHRs could eventually save $81 billion annually. When combining efficiency savings with the savings associated with increased safety, the authors estimated net savings in hospital systems to be $371 billion over 15 years; in physician offices the savings could be $142 billion over the same period (Hillestad et al., 2005).

A similar analysis by Wang and colleagues (2003) used data from the authors’ institution and literature reviews to estimate the costs and benefits associated with adoption of EHRs in physician offices, compared to the traditional paper method of keeping records. Results indicated that the net benefit over 5 years could be $86,400 per provider. These savings result mainly from reduced drug expenditures and better use of radiology tests, among other factors. One important limitation of this study is the fact that the effectiveness of EHRs in physician offices has not yet been firmly established, thus the results may be somewhat uncertain (Wang et al., 2003).

Girosi and colleagues (2005) estimated that savings from 80 percent implementation of EHRs in the United States could reach $80 billion; however, this study has been criticized by the CBO and others because of its sole focus on literature showing positive results from EHRs (Berenson et al., 2009). Of note, this estimate did not take into account the effects of current payment systems on EHRs, which could reduce the effectiveness of EHRs (Berenson et al., 2009). The Commonwealth Fund released a study that estimated investment in health information technology (IT) could result in savings of $261 billion over 10 years (The Commonwealth Fund, 2009). Expanding on this estimate, Russo (2009) estimated that the spillover effects from adoption of EHRs could lead to savings of $800 billion, owing to coordinated care and disease management savings. Berenson and colleagues (2009b) estimated that net 10-year savings from adoption of EHRs could be $97 billion. Finally, PriceWaterhouseCoopers (2009) estimated that ineffective use of information technology has resulted in $81 billion to $88 billion in waste. However, the methods by which this estimate was derived are not clear and thus the validity of the estimate cannot be assessed.

Other studies have focused more specifically on e-prescribing and CPOE. PriceWaterhouseCoopers (2009) found that the use of paper prescriptions has resulted in $4 billion in wasteful spending in the healthcare system. However, the methods by which this estimate was derived are not clear and thus the validity of the estimate cannot be assessed. Chaudhry and colleagues (2006) conducted a systematic literature review, finding that 8 of 10 studies examining the effects of EHRs on healthcare use found decreased rates of use that primarily resulted from the use of CPOE. Fischer and colleagues (2008) examined prescribing behavior changes and savings resulting from the use of e-prescribing, using administrative data in a pre- and poststudy with controls. Results indicated that e-prescribing led to a 3.3 percent increase in prescribing for tier 1 (less expensive) medication. Based on the average cost of prescriptions for insurers, the authors estimated that e-prescribing could lead to savings of as much as $4 billion per 100,000 patients each year (for full adoption) (Fischer et al., 2008).

Finally, Hillestad and colleagues (2005) used results from other literature to estimate the potential reduction in adverse drug events (ADEs) from using CPOE. Results indicated that full use of CPOE could lead to the elimination of 200,000 adverse drug events in the hospital setting nationwide, leading to about $1 billion in savings per year. In the outpatient setting, an estimated 2 million ADEs could be avoided, leading to savings of $3.5 billion per year (Hillestad et al., 2005).

In its analysis of these studies, the CBO concluded that, in certain circumstances, health IT has reduced costs and improved outcomes. However, in general, health IT appeared to be necessary but not sufficient on its own to generate cost savings (CBO, 2008).

Estimates comparison Significant variation appears to exist in the estimates of the savings associated with the adoption of EHRs. This is likely the result of a number of factors, including the time horizon presented, the type of technology being examined, and the extent to which the authors assume the technology will be adopted. In addition, it is not always clear whether the savings are net of costs. For savings over longer time horizons (10 to 15 years), estimates appear to range between $77 billion and $800 billion, though there is some question as to the savings from other effects on the delivery system that are included in each analysis.

It is important to note that a number of papers have called the estimated savings from EHRs into question. Others have questioned the assumptions behind the savings estimates (Himmelstein and Woolhandler, 2005) and have suggested that the limitations of EHRs in ambulatory care are not fully addressed in the cost estimates (Sidorov, 2006). In addition, the CBO has suggested that while savings are possible in EHRs, the majority of estimates overstate the potential for EHRs to result in such savings (CBO, 2008).

Session 2: Care Culture and System Redesign-Based Strategies

Strategies to lower healthcare costs and improve outcomes depend not only on knowledge enhancement, but also on changes in the delivery system infrastructure and care culture. Such changes as care site integration and medical liability reform will likely have both individual effects as well as synergistic effects with the strategies discussed in other sessions.

The presenters in this session discussed care culture and system redesign-based strategies, including:

  • Caregiver profile, efficiency, and team care;
  • Care site efficiency and productivity initiatives and incentives;
  • Care site integration initiatives;
  • Antitrust interventions;
  • Promoting information technology interoperability and connectivity;
  • Service capacity restrictions; and
  • Medical liability reform.

Caregiver Profile, Efficiency, and Team Care

Researchers such as Smedley and Stith (2003) have shown that socioeconomic and racial health disparities are persistent. There is also a mismatch between physician training and supply and patients needing chronic disease management. Social factors such as the availability of transportation have also operated as barriers to obtaining care (Arcury et al., 2005; Baker et al., 2008).

To address these problems, it might be informative to draw lessons from efforts to decentralize the care giving process. Early in the 1900s, the large fixed costs of acquiring state-of-the-art medical equipment and a limited supply of qualified labor (i.e., physicians) contributed to the rise of hospitals as the sole source of solutions to complex medical problems. However, advances in education and technology have led to increased provision of medical care outside the hospital. Use of mid-level practitioners and the development of retail clinics staffed by nurse practitioners provide salient examples. Community health workers, important contributors in the healthcare systems of South America (Hwang, 2009), may be another resource for care delivery (AHRQ, 2009). The appearance of these lower-cost providers suggests that they might play a role in reducing the growth in healthcare costs (Hwang, 2009).

In this section, we describe the presentations of Michelle J. Lyn and Jason Hwang on potential savings from new models of care. A discussion of other estimates is also included.

Potential savings from improved team careLyn and colleagues (2009) made the case for developing new models of care through community engagement and provided a few suggestive examples of how this might be done. They described Just for Us (JFU), an in-home care program for the low-income frail elderly and disabled in Durham, North Carolina, as one example of this approach. JHU is a collaborative effort between Duke University, local government, and one of the area’s federally qualified health centers. This effort deploys interdisciplinary teams to provide care to patients in their homes. Yaggy and colleagues (2006) analyzed expenditures for Medicaid beneficiaries enrolled in JFU and reported that ambulance costs decreased by 49 percent, ED costs decreased by 41 percent, and in-patient costs decreased by 68 percent, while prescription costs increased 25 percent and home health costs increased 52 percent. Another ongoing study found improvement in hypertension control among enrollees over the course of 1 year. The authors looked to Community Care of North Carolina (CCNC) as another example of an interinstitutional collaboration deploying multidisciplinary teams. Estimated savings for 2006 were between $1.5 billion to $1.7 billion (Mercer, 2007). The authors concluded that the choice of appropriate performance measures was still an open question, and evaluation of these programs remains a difficult, complicated, and important issue to resolve.

Citing a recent report by AHRQ (2009), Hwang indicated that there is currently a paucity of evidence on the effect of community health workers on health and costs due to small sample sizes, an inability to perform randomized controlled trials, and the difficulty of identifying and accounting for confounders.

However, as a complementary strategy, the Commonwealth Fund (2009) estimated the potential savings that revising the Medicare fee schedule for primary care could have on U.S. healthcare spending. By changing relative value weights and applying differential updates such that primary care would be emphasized, as well as revising payments for overvalued services, the authors found that this package of policies would “reduce national health spending, relative to currently projected levels, by an estimated $71 billion through the year 2020.”

Additional estimates A literature search found no comparable national estimates for use of community health workers, which is consistent with the lack of data suggested by the author and the findings of AHRQ (2009). However, in the July workshop, James G. Kaplan estimated potential national savings of $8.3 billion if half of outpatient visits for uncomplicated conditions could be handled capably by an ARNP or PA rather than by a physician (Mecklenburg and Kaplan, 2009). Similarly, a recent study concluded that expanded use of ARNPs and PAs in the delivery of primary care could save $4.2 billion to $8.4 billion over the next decade in Massachusetts (Eibner et al., 2009).

It may also be informative to consult other papers presented at this conference given that they present savings from similar efforts focusing on community-level interventions. For example, Levi (2009) reported $16 billion in potential savings from the expansion of community-based wellness programs, and Thygeson (2009) found potential savings in the range of $2 billion to $7.5 billion from increased use of retail clinics, which primarily employ mid-level practitioners.

Estimates comparison As no competing estimates were identified, a direct comparison cannot be made. However, the literature suggests that use of alternative caregivers has the potential to yield significant cost savings to the healthcare system.

Care Site Efficiency and Productivity Initiatives and Incentives

As noted previously in the discussion of Mecklenburg and Kaplan (2009), a significant amount of the cost of producing health care is attributable to the cost of labor. Deploying labor more efficiently could reduce healthcare costs by lowering the costs of production.

Kim R. Pittenger reported on the efficiency gains at VMMC that were brought about by the implementation of a new production model based on Toyota methods, and he estimated the potential savings if this new production method were adopted across the United States. His results are summarized below. Other estimates are also presented and compared.

Savings from increased care site efficiency Pittenger (2009) described the results of VMMC’s move in 2002 to a production system based on the Toyota Production System. Under the Virginia Mason Production System (VMPS) work is done in small batches (“flow production”) in order to decrease waits, delays, errors, and higher costs; mistake-proof devices and practices are used to reduce errors at all levels; and medical care is explicitly standardized to improve performance. Working from data collected by VMMC, Pittenger estimated $58 billion in savings could be achieved with widespread implementation of the VMPS.

Savings opportunities were classified into three main categories: operational, clinical, and patient safety. Examples of operational savings at VMMC included a decrease in liability and malpractice premiums by more than 35 percent over 2 years, and a greater than 10 percent reduction in cost per RVU for primary care owing to the implementation of flow production in result reporting, incoming phone calls, and refills. Assuming a 10 percent reduction in cost per RVU for the 302 million preventive care and 351 million chronic condition visits nationally (CDC, 2007) and a 30 percent savings from the current $10.7 billion spent on liability premiums (A.M. Best, 2009), widespread implementation of the VMPS could result in $7.5 billion in annual savings nationwide.

In the clinical category, the author restricted attention to potential savings from standardization in diabetes care. An outpatient initiative between Boeing and VMMC showed a 35 percent cost reduction. Assuming a 30 percent reduction in the $116 billion spent in 2007 (American Diabetes Association, 2009) for diabetes treatment, VMPS-related savings could be $40.6 billion. Notably, the author endorsed the similar chronic care savings estimate provided by Pignone during the May workshop. With respect to patient safety, VMMC has seen significant declines in the rate of ventilator-associated pneumonias, surgical-site infections, and central-line infections. Assuming national rates approach VMMC-observed rates with adoption of the VMPS, total savings from increased patient safety could be $4.1 billion. Additional savings from improved care processes for MRI imaging related to lower back pain and headaches were also calculated to be in the amount of $1.3 billion.

Additional estimates As in the discussion of Milstein (2009) above, our literature review found several savings estimates directly related to efficiency initiatives. While the primary literature examining the Toyota model, Six Sigma, and lean paradigms in health care have focused on improvement in outcomes, it has been estimated that $19.4 billion in annual savings could be realized from application of lean production systems to all U.S. hospitals by eliminating nonvalue-added activities (Hafer, 2009). Please refer to the discussion of Milstein (Milstein, 2009) for further details.

Estimates comparison Although these estimates differ by a factor of three, several differences exist between the data calculations, including type of reengineering method—the Virginia Mason Production System (Pittenger, 2009) compared to lean production (Hafer, 2009), and the scope of services—inpatient and outpatient compared to just the former, respectively. Additionally, the estimates offered by Pittenger were extrapolated from a savings seen in a single medical center where the VMPS was implemented in 2002 while the extrapolations of Hafer were based on annualized average hospital savings across 75 institutions at various stages of implementation.

Case Site Integration Initiatives

As noted in previous sections, improved care coordination could reduce medical expenditures. In this section, a report from Timothy G. Ferris (2009) about a care coordination project is discussed. The findings in this paper are compared to other existing estimates as well.

Potential savings from case site integration initiatives Ferris (2009) described a 3-year care coordination demonstration project for the Centers for Medicare & Medicaid Services (CMS) instituted in the Boston area for Medicare beneficiaries with high illness burdens. If similar preliminary estimates of savings could be realized nationally, he estimated that between $0.6 billion and $1.5 billion could be saved for Medicare over a 2-year period.

While this project is currently ongoing and data analyses are incomplete, Ferris reported that, relative to the matched control group, patients in the intervention group had lower costs, fewer admissions, lower mortality, and greater use of hospice. If the program is able to achieve the 5 percent cost savings target set by CMS and the effect of this intervention is externally valid over the entire population, a 1 to 2 percent savings could be achieved. The core of the care coordination program was what he deemed “mass customization.” For example, when any individual in the intervention group registered in a local ED, his or her primary care provider and case manager would receive pages notifying them of the ED visit. The provider and manager would then proceed to the emergency room and help ensure the patient received appropriate care. Hence, patients who would otherwise have been admitted to inpatient care unnecessarily would instead be taken care of in, and released from, the ED because of the primary care provider and case manager’s detailed knowledge of the medical and social histories of the patients.

To provide national savings estimates, the authors used a model based on 1.6 percent target population savings and 45 million Medicare beneficiaries with an average annual cost of $7,000. Estimating the size of the Medicare population receiving care within an integrated delivery system as between 40 and 60 percent and the proportion of those integrated delivery systems that have the necessary information technology infrastructure as being between 30 and 50 percent, the authors calculated savings of $0.6 billion and $1.5 billion for Medicare over a 2-year period from implementation of this care delivery model targeting the highest-risk patients.

Ferris noted that the program relied crucially on information technology for care coordination. First, the use of EHRs allowed real-time communication of changes in patient status or care plans. Second, administrative systems allowed physicians and care managers to track patients, manage workflow, and—as described above—know when an enrolled patient arrived at an ED. Third, analysis of the data from care management and administrative systems allowed the program to track trends in use.

Additional estimates There are few studies investigating the potential cost savings of improved case site integration. UnitedHealth Group (2009a) estimated that 10-year savings from improved institutional preadmission policies, transitional care management from inpatient to outpatient settings, and advanced illness programs including palliative care services, as well as disease and integrated medical management, would yield savings of approximately $367 billion to the federal government. The Commonwealth Fund (2009) estimated patient-centered medical homes could save $175 billion over 10 years. Berenson and colleagues (2009) estimated that chronic care management and care coordination for dually eligible Medicare and Medicaid beneficiaries could result in 10-year (2010–2019) savings of $201 billion, assuming that care coordination could yield 5 percent savings per year. The authors suggested that the estimate might be conservative as it only applies to a very small percentage of the population at risk.

Given the high costs of readmissions (MedPAC, 2007), care coordination has been discussed as a method of reducing avoidable readmissions. Berenson and colleagues (2009) suggested that reducing payment for potentially preventable readmissions within 15 days of discharge to 60 percent of the usual payment would provide incentives to reduce preventable readmissions and potentially save Medicare and Medicaid $15 billion over the next decade. A multifaceted program to improve the hospital discharge process through focused patient education and enhanced attention to communication between inpatient and outpatient providers has been demonstrated to lower rates of postdischarge readmissions and ED visits by 30 percent and save nearly $400 per patient (Jack et al., 2009).

In comparison, estimates on the use of medical homes have been less optimistic about near-term savings. Although Berenson and colleagues (2009) believed that a commitment to increased reliance on primary care and medical homes would be a wise investment for the long term, they did not believe it would produce cost savings within the next decade. The CBO (2008) reported that more evidence on the effect of medical homes is needed before further extrapolations to the Medicare program can be completed. Improving care could reduce spending among some patients by eliminating duplicated services, increasing appropriate use of specialists, and averting serious complications from chronic illnesses through better medical management, but it could also result in increases in spending for chronically ill patients who are not receiving all recommended care.

Estimates comparison The potential annual savings estimate by Ferris (2009) is in the lower range of the estimates presented above on care coordination. However, the estimates are difficult to compare as Ferris (2009) used savings from preliminary findings in a demonstration project and focused on a target Medicare population with a high illness burden. In comparison, UnitedHealth Group extrapolated nationally from savings realized among their current beneficiaries from their current disease management programs and initiatives. It is worth noting that, regardless of the approach taken, the authors all endorse the concept of care coordination as a potential method of improving health and care coordination. For additional discussion of the potential for improved care coordination to reduce costs, please refer to the prior section on Owens (2009).

Antitrust Interventions

In a previous session in May, Capps (2009) described the relationship between hospital consolidation and prices in the market for health care. In the session described below, Roger Feldman discussed the role of competition policy in restraining these prices.

Potential savings from antitrust interventions Feldman (2009) provided an overview of the role of antitrust regulation in ensuring efficiency in the provision of health care and outlined suggestions for its improvement. He first reminded us of the legal foundations of antitrust regulation in the United States: the Sherman Act of 1890 and the Clayton Act of 1914. Of relevance to current regulators is the Hart-Scott-Rodino Act of 1976, which requires that parties of mergers meeting certain criteria both notify the Federal Trade Commission (FTC) and the Department of Justice (DOJ) in advance of their merger and delay completion of the merger until one of these agencies has evaluated the merger’s effect on competition. Although this legislation may significantly affect companies in other industries, many mergers in health care are of too low a dollar value (i.e., below $130.3 million in 2009 dollars) to trigger the premerger review.

Feldman presented a detailed discussion of recent developments in horizontal (i.e., two competing hospitals) merger policy. In sum, horizontal merger activity proceeded unchecked in the 1990s. The decrease in competition resulted in higher prices, especially for minorities and lower-income communities (Town et al., 2007). The FTC and DOJ did attempt to challenge these mergers, but for various reasons the federal courts decided to reject their claims. More recently, in 2004, the FTC successfully challenged two hospital mergers by using its internal administrative processes instead of appealing to the courts. Feldman interpreted this development as evidence for the view that the trend of unchecked merger activity may be shifting. He also addressed vertical (i.e., a hospital and a physician group) mergers. As the economics and law of vertical mergers are not settled, the FTC and DOJ have to exercise much more discretion in bringing antitrust action, and it is still unclear whether their current policy is successful at protecting quality and pricing of health services.

Finally, Feldman proposed a few measures that might improve regulatory policy in the United States. First, he suggested lowering the Hart-Scott-Rodino financial trigger so that more healthcare mergers would trigger premerger reviews. Second, better coordination between federal and state antitrust agencies would perhaps foster more effective regulation. Third, the FTC and DOJ should start to challenge physician mergers, given the abundance of anecdotal evidence that these groups do exercise market power (Strunk et al., 2001). Fourth, the FTC and DOJ should be prepared to insist on divestiture as a remedy. Fifth, the FTC and DOJ should no longer accept hospital community payments (e.g., promises to provide more charity care) as just compensation for the loss of competition given that these promises are very difficult to enforce.

Additional estimates As this presentation paper focused on the history of antitrust regulation and the lessons for future policy, the discussion of Capps’ (2009) estimates that hospital consolidations have caused an increase of approximately $10 billion to $12 billion in annual national healthcare expenditures in Section II, Session 4 is highly relevant.

Estimates comparison As above, no comparison of estimates will be made. The discussion of Capps in Section II, Session 4 provides much relevant discussion. Furthermore, the view of the FTC is also of interest (FTC, 2008). The FTC describes its role in regulating practices that will either likely increase costs or limit competition. The FTC also focuses on spurring innovation through antitrust enforcement, particularly in the areas of healthcare provider clinical integration, healthcare mergers, and pharmacy benefit management services.

Promoting Information Technology Interoperability and Connectivity

The U.S. healthcare system’s heavy reliance on a paper-based system has tremendous implications for cost and access. Though the cost of a single transaction is negligible, it becomes substantial over billions of transactions. In fact, it was estimated that 90 percent of the 30 billion transactions in the U.S. healthcare system were paper based (Menduno, 1999). Because information stored on paper may not be easily accessible across physicians or institutions, a paper-based system can cause physicians to perform redundant tests or lead to unnecessary hospitalizations. For example, suppose a test result from a prior examination would be sufficient to inform a physician in a future clinical encounter. If that physician will not have access to the result because the file is in another hospital, he might need to recollect that data and order a redundant test.

In this section, the potential cost savings from improved information technology interoperability described by Ashish Jha are discussed.

Potential savings from improved information technology interoperability Jha (2009) examined the prospects for a simpler, more integrated way to exchange clinical and administrative data. Jha first summarized two of the most prominent papers in the literature.Richard Hillestad and colleagues (2005) presented the most comprehensive estimate of potential national effects of improved EHRs systems interoperability. These authors found potential savings of approximately $81 billion through improvements to safety and efficiency. He also highlighted the work ofJan Walker and colleagues (2005). These authors found potential savings of $337 billion during a 10-year implementation period and annual savings of nearly $78 billion in each subsequent year (amounts measured in 2003 dollars). However, he noted that both studies were substantively vulnerable to methodological critiques. Hillestad and colleagues (2005) depended on what can perhaps be characterized as a best-case scenario. Their estimate is only plausible if pivotal delivery system changes actually occur; the authors also overestimated the then-current penetration of EHRs at the hospital level by at least a factor of two. Walker and colleagues (2005) relied heavily on expert consensus and likely underestimated administrative costs.

Additional estimatesHillestad and colleagues (2005) and Walker and colleagues (2005) are reviewed earlier in this working paper in Section III, Session 1 discussing the findings in Kaushal (2009). Please refer to that section for further details.

Estimates comparison As above, please refer to the discussion of Kaushal (2009) for further details.

Service Capacity Restrictions

Hospital competition combined with widespread health insurance coverage, physician preference for a high quantity and quality of care, and retrospective cost reimbursement can lead to an increase in the cost of care. Taken together, these forces can foster nonprice competition among hospitals such that they invest in facilities and services to compete for patients. This phenomenon is known as the “medical arms race.”

In this section, comments by Frank A. Sloan on future policy options to combat this arms race are summarized. Other estimates are also presented and discussed.

Potential savings from service capacity restrictions Sloan (2009) provided an overview of policies aimed at combating the medical arms race. He described the National Health Planning and Resources Development Act of 1974 and the certificate of need (CON) requirement that attempted to curtail cost growth through regulating capital investment. He then described the roughly contemporaneous introduction of selective contracting and prospective payment, and discussed their joint role in perhaps preventing CON from achieving cost containment. He concluded his review with the 1983 repeal of the CON requirement.

The final section of his presentation considered the conditions under which CON-type regulation could constitute good policy. Supposing market competition remains the mechanism by which we expect to contain costs, he suggested that expenditure regulation and capacity reduction would not be relevant cost-containment tools. Supposing the government effectively implements price controls, it is possible that expenditure regulation and capacity reduction could be of use in restraining cost growth. Sloan returned to the CON requirement and wondered what role it could play in future policy reforms. The empirical studies show (Salkever, 2000) that CON programs have not succeeded in cost containment, and it is not altogether clear what effect they have had on access to and quality of care. The first reason for this may be that “need” has not been well defined and has not given policy makers much guidance in their oversight. Second, CON programs do not have capital budgets, which allow them to be affected by pressure from stakeholders. Third, CON programs grant a de facto franchise to incumbents. Therefore, if CON-type programs are to be implemented in the presence of price controls, policy will need to address the shortcomings above. To this end, Sloan provided some straightforward solutions to these problems and suggested that capital expenditure regulation may still be a feasible option going forward.

Additional estimates As the author believes that the effectiveness of capacity restrictions depends on other future policy decisions on cost containment, no savings estimate was provided. Note that in addition to the review cited by the author, recent work such as Vivian Ho (2007) and Grabowski and colleagues (2003) further support the notion that CON programs have not succeeded in cost containment.

Estimates comparison As no savings estimate was provided, no comparisons could be undertaken.

Medical Liability Reform

Tort reform has long been a concern of practicing physicians in the United States. Evidence for this concern can be found in recent articles and editorials from the American Medical Association (American Medical Association, 2008; Sorrel, 2008). Tort reform has also recently surfaced as an important issue in the current debate over healthcare reform (Garber, 2009).

In this section, we present the analyses of Randall R. Bovbjerg and his discussion of the potential savings that might be achieved by medical liability reform. Other estimates are also presented and discussed.

Potential savings from medical liability reform Bovbjerg (2009) assessed the evidence on malpractice reform’s role in reducing healthcare spending over the next 10 years. Drawing upon published work (Berenson et al., 2009b), he found that conventional tort reform could be expected to reduce total spending by 0.9 percent per annum, saving almost $20 billion in 2010 and almost $260 billion over a full decade.

Bovbjerg pointed out that savings from malpractice reform could come from three sources. First, a policy change affecting malpractice payouts—such as a cap on the total award or on the nonmonetary component of the award—could lead to lower liability premiums.a The CBO (2004a) summarized prior literature and estimated that a $250,000 cap on noneconomic damage awards would reduce these premiums by an average of 25 to 30 percent. This estimate implied savings of $7 billion to $9 billion (0.3 percent to 0.4 percent of national health spending) in 2007 had such a reform been implemented.

Second, savings could also come by reducing the incidence of defensive medicine. Estimates of potential savings in this arena vary. Most are between zero and 0.3 percent of spending (CBO, 2004a; Currie and MacLeod, 2008; Dubay et al., 1999; Sloan and Shadle, 2009; Sloan et al., 1997). The highest peer-reviewed estimate is 4 percent of total spending (Hellinger and Encinosa, 2006; Kessler and McClellan, 1996). Sloan (2009) stated that he considered potential savings equal to or higher than those from liability premiums a reasonable view. In particular, he suggested that savings could be perhaps 0.5 percent of total health spending.

Finally, Bovbjerg explained that savings could come through synergistic interaction with other reforms. For instance, the spread of evidence-based medicine could increase the effect of malpractice reform on spending. Conversely, tort reform could soften provider resistance to use oversight owing to reduced liability concerns. However, for purposes of his national savings estimate from tort reform alone, the author conservatively excluded such additional savings. The implication was that interactive savings need estimation apart from any single component of reform.

The author concluded with the suggestion that making tort reform part of a broad health reform package could have other positive effects. Patients as a class would benefit if changing tort law could help build coalitions to enact comprehensive health system reform, as suggested by Bill Bradley (2009). One benefit is that near-universal coverage would probably ensure that those who were permanently injured during medical care would not have to rely on a liability award.

Additional estimates The maximum savings estimate that could be supported by the quantitative literature would be $90 billion, assuming the Kessler and McClellan (1996) result holds. Even larger estimates of the national costs of defensive medicine exist outside the scientific peer-reviewed literature (for example, PriceWaterhouseCoopers, 2009), but frequently do not provide details on the methods of calculation and may not specify the quantitative impact of policy interventions.

Estimates comparison Additional review of the econometric evidence and the policy literature suggests that the estimate presented in Bovbjerg (2009) is reasonable. Higher estimates can be based on such findings as Kessler and McClellan (1996), but that is a minority finding. Moreover, Patricia Danzon (2000) noted in her extensive literature review some uncertainty about the validity of that result, mainly from confounding caused by the growth of managed care in California, which was not accounted for in the original paper. Subsequent work by the CBO (2004a) was unable to replicate the finding of Kessler and McClellan (1996), and a recent extension of the latter paper’s methods that also included physician spending found no impact of direct reform. Hence, it would be difficult at present to justify substantially higher savings absent further developments. One such development would be interventions combining tort reform with other initiatives as discussed by Bovbjerg (2009) and others (Gabel, 2009).

Session 3: Transparency of Cost and Performance

Transparency has been valued as a tool for quality improvement. However, transparency has also been touted as a potential means of enhancing competition and lowering costs. The presentation in this session discussed the potential impact of transparency on costs and outcomes, including

  • Transparency in prices;
  • Transparency in comparative value of treatment options;
  • Transparency in comparative value of providers;
  • Transparency in comparative value of hospitals and integrated systems; and
  • Transparency in comparative value of health plans.

Transparency in Prices

Economic theory suggests that the ability of consumers to compare products based on price leads to choice of higher-value providers, lower prices, and better quality (Ginsburg, 2007). In the healthcare market, consumers have often not had access to price information, or cared little about the price because of the presence of insurance coverage. However, there has been increased interest in the release of pricing and other information as part of an effort to lower costs and increase quality in health care.

This section summarizes a discussion by John Santa regarding the potential for increased price transparency in healthcare services in the United States.

The potential for price transparency Santa (2009) discussed the rationale behind increasing transparency in pricing of healthcare services. In the current U.S. system, rising costs have resulted in increased cost sharing and increases in bankruptcy cases among consumers specifically attributable to high healthcare costs. These changes further suggest that consumers should have access to information on the comparative effects, prices, and costs of products and services. Santa pointed out, however, that the presence of third-party payers changes the dynamic of the purchase of healthcare services by influencing the setting of reimbursements (public programs) and establishing cost sharing (employers).

In addition to a system in which the purchasing decision is separated from the consumer and prices are not easily accessible, consumers must rely on physicians to make decisions on their care. Physicians, however, may not always make a decision that is consistent with guidelines, and may have financial relationships with companies that are in conflict with patient needs. Santa (2009) suggested comparative effectiveness research as a tool to provide meaningful comparisons of different healthcare alternatives. In addition, transparent provision of price, effectiveness, and adverse events, among others, should be a goal.

Additional estimates A search for literature related to increased transparency of healthcare prices in the United States returned no studies attempting to analyze the cost savings associated with publishing price data.

Estimates comparison A comparison of cost estimates cannot be conducted for this section as there are no estimates of the potential savings from such a policy change. However, a review by the Congressional Research Service (2007) suggested that there is potential for increased price transparency to improve outcomes and lower prices. This conclusion was based on a review of empirical evidence from other markets as to the effects of increased transparency. In general, sites on the Internet that provide comparison pricing appear to have lowered prices for products, and lifting restrictions on advertising for products such as eye care, which are by nature complicated products, has also led to lower prices. Online data provision by some states and insurers of hospital costs, on the other hand, have so far showed little pricing effects (Congressional Research Service, 2007).

Transparency in Comparative Value of Treatment Options

As an extension of CER, cost-effectiveness analyses can also provide important information as to the relative effectiveness of treatments while also taking cost into account. By providing such information to decision makers, including patients and payers, the United States can move toward a system where resources are allocated in an optimal manner (Gazelle, 2009).

This section presents a discussion by G. Scott Gazelle regarding the potential for using the results of cost-effectiveness analyses in the United States.

The potential for cost-effectiveness research Gazelle (2009) discussed the use of cost-effectiveness research as a means by which effectiveness and price information can be provided in a transparent manner. Neither comparative effectiveness research, from which cost information is excluded, nor price transparency as separate policies will enable optimal allocation of resources in the healthcare system. To reduce costs in a systematic manner while preserving health, services that are more cost-effective should receive more resources than those that are less cost-effective.

Gazelle discussed some challenges present in the movement toward use of cost-effectiveness analyses in influencing coverage and reimbursement policy. These challenges include an evidence base that is currently very small, varying quality in the analyses that have already been completed, and a limited pool of researchers who are currently able to conduct such analyses. Finally, prior attempts to apply coverage decisions based on cost-effectiveness that have not been successful may “bias against the feasibility and acceptability of such an approach” (Gazelle, 2009).

In choosing to use cost-effectiveness results, there are a number of different policy options available. The most aggressive approach would be to define a set threshold for the cost-effectiveness ratio, and approve or deny coverage based on services and products meeting or exceeding that threshold. A less aggressive approach would align incentives, such as copayments and tiering, to the relative cost-effectiveness of services. Another approach would establish standards for cost-effectiveness analyses, encourage (and fund) its development, and allow the market to determine how to use the results. Finally, another option would be to focus exclusively on the effectiveness of different treatments without consideration of cost.

Additional estimates A search of the literature related to cost-effectiveness yielded no studies that present cost estimates related to the use of cost-effectiveness research. The current policy debates appear to focus more on comparative effectiveness research (i.e., not incorporating cost), and as a result there are some estimates of savings due to comparative effectiveness, but not cost-effectiveness. The estimates related to comparative effectiveness research are presented in Session 2 of this section.

Estimates comparison As there are not any cost estimates related to the potential savings attributable to the use of cost-effectiveness research, a comparison cannot be completed at this time.

Transparency in Comparative Value of Providers

Quality and outcomes can vary significantly across providers within and between different healthcare markets. In many cases physicians that perform poorly relative to their peers are not aware that they achieve poorer outcomes (UnitedHealth Group, 2009a). As a result, providing report cards to physicians and to consumers can have two potential effects. First, payers and consumers can select physicians that have higher quality ratings; and second, physicians may be given an incentive to compete on quality (Werner and Asch, 2005).

This section summarizes a discussion by Paul B. Ginsburg regarding the release of comparative quality information about providers in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

The potential for quality reporting Ginsburg (2009) discussed the opportunities for providing comparative quality information about physicians, as well as increasing transparency of price data for consumers. Two trends are currently converging, making the provision of quality and price information more plausible. These include an increasing belief in accountability and transparency and a growing consumerism movement in the healthcare market (Ginsburg, 2009).

There currently exists potential for increased quality in the healthcare system, which can also lead to increased efficiency, via encouraging consumers to make wiser choices of providers and pressuring providers who have lower ratings to improve. Unfortunately, there is currently a lack of consumer interest in information related to quality; this is likely because of a lack of awareness of the significant variations in quality across providers and the challenges in determining how best to make such information accessible and useful to consumers. However, there has been a large provider response to quality information, mostly related to professionalism (Ginsburg, 2009). Ginsburg (2009) suggested three steps for generating and providing effective quality information: (1) develop measures that take into account provider input, (2) audit the reported data, and (3) take into account the different audiences that will see the data when analyzing it.

A series of options for making pricing data meaningful may include changing pricing so that it is quoted on a per episode basis as opposed to an individual service basis. This will make prices more understandable to consumers. Also, customizing pricing data for the different insurance companies, recognizing the insurer as an important player in the health-care system, will ensure consumers will know what prices are applicable to their own experience. Finally, creating a benefit structure that distinguishes provider choice as an important metric will ensure applicability of quality measures to pricing differences (Ginsburg, 2009).

Finally, Ginsburg discussed the role of governments in the provision of quality and pricing data. He recommended that governments should require the collection of quality data, convene stakeholders to encourage agreement on quality and pricing measures, and pool information on different providers. Governments should also encourage the adoption of IT systems (Ginsburg, 2009).

Additional estimates A review of the literature returned few cost estimates related to public reporting of provider quality data. One estimate by UnitedHealth Group (2009b) explored the potential savings associated with providing quality and efficiency measures for specialist physicians solely among physicians. Based on results from UnitedHealth Group’s own data-sharing programs, an estimated $14.5 billion (2010–2019) could be saved from the sharing of such information in the Medicare program.

Estimates comparison A comparison of the cost estimates cannot be conducted, given that only one estimate was found in the literature. However, it is important to note that even though savings were estimated via one type of reporting initiative, it may be that quality reporting may lead to adverse results as well. These may include unintended incentives for providers to avoid treating sick patients so as to keep their quality score high; placing little emphasis on patient preferences and clinical judgment in favor of meeting the quality score; and having providers attempt to achieve only the benchmark rate for healthcare interventions (Werner and Asch, 2005).

Transparency in Comparative Value of Hospitals and Integrated Systems

As discussed above, physician quality reporting can create positive effects in the market. These effects include payers and consumers being able to select physicians that have higher quality ratings and physicians having incentives to compete on quality (Werner and Asch, 2005).

This section presents an analysis by Peter K. Lindenauer estimating the potential cost savings associated with the release of hospital quality data in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Potential savings from release of hospital quality data Lindenauer (2009) estimated that specific public reporting requirements on hospital readmissions, complications, and healthcare-associated infection rates could result in as much as $2.5 billion to $5 billion per year in savings. The savings estimate was extrapolated from the effectiveness of a New York State reporting system, combined with data from government sources on the costs and preventability of the complications listed above (MedPAC, 2009; Scott, 2009; Zhan and Miller, 2003). The savings estimate was based on the assumption that transparency could result in 10 to 20 percent reductions in adverse events, including readmissions and complication rates, given that public reporting in New York resulted in a 14 percent reduction in mortality following bypass surgery. In addition, savings is assumed to accumulate to payers (Lindenauer, 2009).

Lindenauer discussed two possible pathways through which increased transparency on price and quality of hospitals could improve the value of health care. The first such pathway is the selection pathway, through which patients and other stakeholders use information about performance to make a choice about their care. The second pathway is the change pathway, through which performance data is used to stimulate improvement efforts at the hospital. However, transparency effects of quality information may not be significant, owing to the complexity of care and lack of ability to choose the hospital under a number of different circumstances, such as emergencies (Lindenauer, 2009). Price transparency effects are also uncertain. Although release of information could reduce price discrimination and price dispersion, there could be unintended consequences on average prices (Austin and Gravelle, 2007).

Some caveats to this analysis include the fact that the current evidence on the benefits of transparency is weak, and that reporting systems provide the catalyst for change but do not improve care directly and could result in double-counting of the savings (Lindenauer, 2009).

In the short term, achievement of the benefits of transparency involves broadening and strengthening current reporting requirements and ensuring future reporting initiatives make a concerted effort to reach out to patients and encourage their use of the data. Over the long term, measures should be created that represent greater value to patients, and the data collection requirements should be made more efficient for providers. In addition, hospital payment systems must be changed in order to implement further goals, including combination of quality and cost information and extending requirements beyond current data windows (Lindenauer, 2009).

Additional estimates A review of the literature related to estimates of savings from public reporting initiatives yielded few results. Most results were related to findings of changes in quality of care associated with reporting. One study by Dranove and colleagues (2003) explored the effect of report cards in New York and Pennsylvania. Using Medicare claims data and information on U.S. hospital characteristics, the authors compared outcomes for patients with acute myocardial infarction or elective coronary artery bypass grafts in hospitals both in states with reporting requirements and in states without requirements.

The authors found that report cards were associated with higher Medicare expenditures (although these were not statistically significant when compared to the control states), as well as higher rates of adverse outcomes (Dranove et al., 2003). However, Dranove and colleagues (2003) cautioned that report cards should not be considered as generally harmful. Some limitations of the analysis included the fact that it only measured short-run outcomes, and thus longer-term effects were not represented.

Estimates comparison There is only one cost estimate for this section, and it appears that there is no consensus on the relative savings available as a result of implementing public reporting initiatives. Lindenauer (2009) estimated significant savings as a result of public reporting, while Dranove and colleagues (2003) showed an increase in Medicare expenditures as a result of reporting initiatives for one type of surgical intervention.

Transparency in the Comparative Value of Health Plans

As previously discussed, economic theory suggests that the ability of consumers to compare products based on price leads to choice of higher-value providers, lower prices, and better quality (Ginsburg, 2007). The same is likely true for health plans operating in the U.S. healthcare system: transparency of quality data should lead to selection of higher-quality plans.

This section presents a discussion by Margaret E. O’Kane regarding the advantages of releasing health plan quality information in the United States.

The potential for releasing health plan quality information O’Kane (2009) discussed the usefulness of quality and satisfaction measures in informing different stakeholders about the performance of the healthcare system. To date, transparency has not had much effect, if any, on insurance cost trends and plan performance. This is likely due to restrictions on the ability of plans to establish networks based on value of providers because of monopsony1 providers and access requirements. Additionally, there is ambivalence about use management, resulting in plans having limited ability to deny coverage based on value measures. In addition, consumer concerns that increasing transparency will lead to trade-offs of cost for quality results in a lack of desire to pursue either course (O’Kane, 2009).

O’Kane asserted that the perfect market as defined in economics does not exist in health care. Market conditions that are essentially monopsonies exist alongside third-party payers that separate consumers from the true cost of care. In addition, the product provided is not the same across all areas and can be very difficult to define. Information related to the quality of the product is very limited and can be difficult to comprehend (O’Kane, 2009).

According to O’Kane, an agenda that promotes value in the healthcare system should incorporate the following. Health plans operating under public programs should be required to report quality data. Hospital payments should be adjusted to align with performance across all payers; these changes should incorporate payments that reward high performance and deny payment for events that should never occur. Payments to physicians should also be reformed in a manner that rewards coordination of care and allows for incentives to develop integrated systems. Finally, consumer incentives to reward patients who use value networks, medical homes, and encourage the use of high-value treatments, should be established (O’Kane, 2009).

Additional estimates A review of the literature resulted in no studies estimating the costs or savings associated with releasing quality information about health plans. However, some of the issues referenced in O’Kane’s discussion are relevant to other sections, for example hospital payment changes and value-based insurance. Readers are referred to these other sections for a more complete discussion.

Estimates comparison As there are no specific cost estimates related to health plan quality information, a comparison cannot be completed at this point.

Session 4: Payment- and Payer-Based Strategies

Strategies targeting payment models and payers have also received significant attention as a means of lowering costs and incentivizing patient-centered care. Ranging from bundled payments to value-based insurance design, ongoing efforts to employ these strategies have occurred in both the private and public sector.

Presentations in this session discussed payment-and payer-based strategies, including:

  • Value-based payments such as bundled and fee-for-episode payments;
  • Managed competition;
  • Value-based insurance design; and
  • Administrative simplification.

Value-Based Payments: Bundled and Fee-for-Episode Payments

Fee-for-service payment systems encourage the overuse of services and do not provide incentives for care coordination or for care delivery efficiency (Schoen et al., 2007). Alternate payment systems have been proposed to better align provider incentives, thereby improving health care and reducing overall costs. One potential value-based system is an episode-of-care payment system, which provides payment for all services provided during a single episode of care. This type of bundled payment system creates incentives for providers to use higher-value treatments, and enables stakeholders to see the full cost of treating a patient and to compare the provider costs (Miller, 2009). Although an episode of care payment model presents some advantages relative to other payment reforms, a disadvantage is that the system does not provide incentives to reduce unnecessary episodes of care. Alternative payment methods such as comprehensive care payments set a fixed amount to cover all services for a given condition during a set period of time, thereby creating further accountability for the use of resources. (Miller, 2009).

This section discusses analyses presented by Amita Rastogi that estimated the potential savings from elimination of potentially avoidable complications via changes in reimbursement models in the United States. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

Cost savings from a fee-for-episode payment system Rastogi (2009) estimated that use of a bundled payment model for 13 specific conditions, including heart attacks, diabetes, asthma, and congestive heart failure, would induce a significant reduction in potentially avoidable complications (PACs) and could save $165 billion for the 200 million commercially insured patients in the United States. By extension, completely eliminating PACs could save $355 billion for the same population.

Rastogi (2009) based her estimates on work engaged with the Prometheus payment project, which bases payments on a complete episode of care. Evidence-informed case rates (ECRs) form the basis for the payment, and include costs for “necessary care for a given condition across the care continuum for a predefined period of time” (Rastogi, 2009). Each ECR also includes a built-in payment for PACs; if complications arise, care is provided and paid for out of the additional PAC allowance. To the extent that complications are avoided, physicians are able to keep the PAC allowance as a bonus. In addition, ECRs encompass a quality scorecard, from which payments are made depending on the scores achieved by the providers and their counterparts.

The estimate of potential savings from use of bundled payments was derived from data from a large national employer, which found that 15 percent of the total $45 billion in annual costs of care were due to PACs. Of the 15 percent, Rastogi (2009) applied best practices from the literature to estimate that PACs could be reduced by 50 percent (de Brantes et al., 2009). These results were then extrapolated to the U.S. population to derive the above estimate.

Additional estimates There has been relatively limited experience with such payment methods, though projects that have been completed indicated that payers achieved savings ranging from 10 percent to 40 percent without negative impacts on quality (Cromwell et al., 1997; Edmonds and Hallman, 1995; Johnson and Becker, 1994). An estimate completed by Schoen and colleagues (2007) for the Commonwealth Fund estimated the potential savings to Medicare from changing to a system of payments based on episodes of care (for acute care episodes) to be $96 billion over 5 years and $229 billion over 10 years. The authors assumed that providers would seek to shift costs to other payers in response to reductions in Medicare payments. However, they also assumed that other payers would not change their methods of payment. If other payers were to follow Medicare’s lead in changing payment systems, the savings could be higher (Schoen et al., 2007).

The Medicare program conducted a demonstration project in the 1990s that examined the effects of paying for heart bypass care based on bundled payments. Results indicated that most participating sites lowered their operating costs and lowered Medicare spending, while quality remained high (Cromwell and McCall, 1998). Finally, the Geisinger Health System implemented a program designed to pay a flat rate for coronary artery bypass graft surgery and for care related to the surgery for a period of 90 days after the surgery (Mechanic and Altman, 2009). An evaluation of the program found that during the first year there were fewer adverse events and lower hospital charges compared to the control group (Casale et al., 2007).

Estimates comparison There are few estimates as to the overall potential national cost savings associated with changing to a payment system based on episodes of care. However, evidence suggests that shifting to a payment system based on episodes of care can save costs. While the national savings estimates of Rastogi (2009) and Schoen and colleagues (2007) are similar in magnitude, it is difficult to compare them for two reasons: (1) the estimate by Rastogi focused on payments for 13 specific conditions while the estimates of Schoen and colleagues focused on payments for acute care episodes; and (2) the estimates focus on different populations (i.e., private compared to public insurance beneficiaries).

Managed Competition

Health insurance exchanges provide individuals, households, and small employers with the ability to purchase insurance that may be more available or more affordable than if it were not provided via the exchange. Such exchanges provide a number of functions, including collecting and providing information about the health plan options, promoting risk pooling, establishing the benefit packages, and negotiating premiums (Frank and Zeckhauser, 2009). Exchanges are able to provide consumers that often do not have sufficient information about the available plans with the information they need to better evaluate their insurance options (Frank and Zeckhauser, 2009). For example, Stanford University has implemented a managed approach to competition, requiring employees to select among a variety of insurance options for which the University contributes a fixed dollar amount to the premium. This serves to encourage employees to select the lowest-cost option and encourages price competition among the different insurers (Enthoven and Talbott, 2004).

This section presents the comments of David R. Reimer on the effect health insurance exchanges have had on premiums and inflation in Wisconsin. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

The potential for health insurance exchanges Reimer (2009) discussed the potential for health insurance exchanges as an option for lowering costs and improving quality of care in the context of the experience of Dade County, Wisconsin. Health insurance exchanges, which provide health insurance consumers with access to information related to a number of competing health plans, can address the issues of cost and quality if three conditions are present. First, the exchange must overcome the problem of adverse selection; second, the number of consumers using the exchange must be large enough to encourage insurers to participate; and third, consumers must have an incentive to purchase the lowest-cost plans.

State government employees in Wisconsin have access to county-based health insurance exchanges, presenting a number of health insurance options ranging from the least-expensive HMO to the fee-for-service standard plan. For example, premiums for the tier 1 HMO option in 2009 are limited to $31 per month for an individual, while a tier 2 HMO has premiums that are more than twice as expensive—$69 per month. The tier 3 plan is $164 per month, again more than twice as much as the next cheapest option. There is an incentive present to choose the lowest-cost HMO because employees must pay much of the extra cost of any of the higher-cost options. Of the 72 counties in Wisconsin, Dane County has the largest population of state employees (i.e., potential enrollees). Likely as a result of the large potential enrollee population, the premiums for plans in Dane County are much lower than in other counties. For example, the tier 1 HMO option costs $528 per month for an individual (2009), while the premium in other counties is as high as $628. In addition, the inflation rate in Dane County has been much lower than the rate in other counties in Wisconsin.

Additional estimates A review of the literature related to the potential savings associated with health insurance exchanges yielded few papers. One paper, however, estimated the possible savings associated with having the government operate a public plan option alongside other options in an exchange (Berenson et al., 2009b). Based on assumptions including the fact that the public plan would pay providers based on locally available prices, and would offer a set of package options, the authors estimated that significant savings could be possible due to lower government administrative costs and lower payment rates. More specifically, Berenson and colleagues (2009) estimated that the public plan could save the government $17 billion in 2010, and about $224 billion from 2010–2019. When including savings to the private sector, the 10-year savings estimate jumps to $412 billion.

Estimates comparison Given that only one estimate is at the national level and is for a public health insurance plan, and the other estimate focuses on premium differences among exchanges within a single state, a comparison of the cost estimates is not feasible. However, evidence from Wisconsin and Stanford suggest that, if certain conditions are met, managed competition has the potential to lower health spending.

Value-Based Insurance Design

Cost-sharing arrangements with health insurance enrollees have generally been constant for each service even though the effectiveness and value of each service differs in general and may even differ from patient to patient (Chernew et al., 2007). Value-based insurance design (VBID) is grounded on the concept that, by pricing different services according to their effectiveness, consumers can be encouraged to use those services that have higher value (Choudhry, 2009). Of note, the Government Accountability Office (GAO) (2007) has found that Medicare beneficiaries in 12 specific areas who saw general practitioners that treated a disproportionately large share of high-cost patients “were more likely to have been hospitalized, more likely to have been hospitalized multiple times, and more likely to have used home health services.”

This section presents discussions by Niteesh K. Choudhry and Lisa Carrara of the potential savings achievable by implementing VBID nationally. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

The potential for value-based insurance design (VBID) Choudhry (2009) and Carrara (2009) presented two methods by which health insurance can be designed to encourage value-based healthcare utilization. Choudhry discussed the potential for designing patient cost sharing for medications so as to encourage the patient to consume services that have higher value than other services, and he estimated a savings of more than $2 billion if VBID were applied to five common conditions. Carrara described the potential of designating high-performing specialists based on measures of clinical quality and efficiency as a method of directing consumers to make healthcare decisions based on the overall value of care, rather than just price alone, estimating a 3 to 4 percent savings in a customer’s annual claims the first year.

Choudhry presented results from various studies studying the impact of providing lower cost sharing for prescription drugs. Two studies looked at the effects of lowering cost sharing for postmyocardial infarction drugs, and found that Medicare could save $2,453 per patient over a lifetime (Choudhry et al., 2008), and commercial insurers could save $1,181 over 3 years (Choudhry et al., 2007). In addition, Rosen and colleagues (2005) found that Medicare could save $922 per patient over a lifetime by lowering cost sharing for diabetes drugs.

Using data from the literature and from the Medical Panel Expenditure Survey, Choudhry estimated that a 1 percent cost reduction brought about by applying VBID to just five common medical conditions would amount to more than $2 billion in savings. However, he does note several limitations to this estimate. First, because the true effect of VBID on healthcare expenditure is unknown, this analysis relies on estimates derived from economic models and published literature. Second, the use of relative rates as a basis for calculating national savings estimates may be inappropriate if the cost savings from copayment reductions do not accrue at a constant rate. Third, payers who already set copayments at a very low level are unlikely to use VBID, and thus this estimate may overestimate the impact of VBID. Finally, the national expenditure estimates used for this analysis, by necessity, aggregate groups of conditions into single disease categories, such as “heart disease” and do not distinguish between patients of different disease severities.

Carrara described how tiered networks could be designed to influence consumer choice of physician, and how they may lead to lower use of health services across 12 specialty categories of care2 designated by Aetna. With such a model, she estimated a customer may save between 3 percent and 4 percent on claims in the first year, offset by a service fee charge.

Additional estimates A review of the literature found few papers that provided estimates of the potential savings resulting from VBID. RAND estimated that VBID targeting medications for six chronic conditions could reduce spending up to $1.2 billion over 10 years in Massachusetts alone (Eibner et al., 2009). Chernew and colleagues (2007) discussed some programs that have experimented with differential cost sharing based on value. One such program, by Pitney Bowes, lowered copayments for diabetes, asthma, and hypertension drugs, and reported 1-year savings of $1 million for their plan (Hensley, 2004). Other employers have programs that provided lower copayments for certain diabetes medications in an attempt to encourage diabetic patients to use those particular medications (Chernew et al., 2007).

UnitedHealth Group (2009b) estimated potential savings of $37 billion (2010–2019) from implementation of a program designed to provide Medicare beneficiaries with information on quality and efficiency variations among providers. Savings estimates were based on UnitedHealth Group’s experience with their quality measurement system combined with incentives to choose high-quality, lower-cost providers. In extrapolating to the Medicare program, UnitedHealth Group assumed such a program would be voluntary and therefore assumed conservative levels of participation (UnitedHealth Group, 2009b). The GAO (2007) reported that an insurer that placed more efficient physicians in a special network saw premium decreases of 3 to 7 percent compared to those less efficient. The GAO also reported that the “sentinel” effect, or the effect of being monitored and examined, reduced spending by as much as 1 percent. Finally, the State Employee Group Insurance Program in Minnesota adjusted patients’ out-of-pocket costs at the point of service based on the cost of the clinic used, and saved 13 percent (Moracco, 2009).

Estimates comparison A comparison of cost estimates cannot be conducted for value-based insurance as there currently is very little information out there on the potential national cost savings associated with such a benefit design. However, the savings estimates reported by Carrara and the GAO from physician profiling are similar; it is worth noting that Aetna was one of the insurers included in the GAO study.

Administrative Simplification

An estimated 31 percent of national health expenditures is consumed by administrative costs (Woolhandler et al., 2003). Nonstandardized formularies, forms, and reporting requirements from multiple payers demand significant time and attention from providers (Casalino et al., 2009a). Given the rising expenditures on health and the need to lower costs, administrative simplification has been viewed as a potential area of significant savings opportunity.

This section presents the discussions of Robin J. Thomashauer and David S. Wichmann. Thomashauer discussed the potential savings possible from simplifying credentialing and standardizing administrative exchange rules, and Wichmann described the savings possible from using technology to simplify administrative burdens. Results from other studies are then presented and discussed. Finally, the cost estimates from the various studies are compared.

The potential savings from payer harmonization and coordination Thomashauer (2009) discussed efforts being made to facilitate payer collaboration and process consolidation. One such effort is the Universal Provider Datasource (UPD), which is a single uniform system designed to collect self-reported provider information, which is then used for such purposes as credentialing. Another effort underway is the Committee on Operating Rules for Information Exchange (CORE), which is working to define rules designed to facilitate administrative data exchange and increase interoperability. She estimated that the industry could save approximately $3 billion over 3 years if the first phase of the CORE project is implemented across the country (IBM Global Business Services, 2009). The first phase of CORE rules include requirements for eligibility and benefits data as well as requirements for exchanging that data, enabling providers to more easily receive information verifying an individual’s eligibility for a particular insurance plan.

Wichmann (2009) presented an estimate of $332 billion in administrative savings over the next decade based on the application of technology to administrative activities (UnitedHealth Group, 2009b). More detail on these 12 options and the potential savings associated with each are presented in Table A-5. Of the $332 billion in savings, about 50 percent is estimated to accrue to providers, 20 percent to the government, and 30 percent to other payers. Specific contributors to costs that were targeted in each option include excessive manual processing, duplicate data entry, and paper distribution of information among others. Where the options potentially overlapped each other, an attempt was made to account for potentially duplicative costs. Some caveats associated with estimating savings related to administrative costs include the fact that administrative costs represent a small portion of total healthcare spending; in addition, savings from associated reductions in wasteful medical spending might be possible.

TABLE A-5. Options for Achieving Administrative Cost Savings.

TABLE A-5

Options for Achieving Administrative Cost Savings.

Additional estimates The Commonwealth Fund (2009) estimated that a national health insurance exchange in conjunction with a public plan could offer a reduction in administrative costs of $337 billion over the next 10 years. Based on results from a number of studies examining streamlined billing, Russo (2009) estimated that such an initiative on a national basis would save $35 billion per year. As many other estimates of savings in this category were discussed previously, please refer to Section II, Session 3, for additional discussion.

Estimates comparison As it has been estimated that in small physician practices more than $247,500 per year was spent on unnecessarily complex or redundant administrative tasks; $19,444 per year was spent on phone calls with pharmacies resolving drug formulary issues; $38,761 was spent per year verifying patient coverage, copayments, and deductibles for thousands of varying health plans; and $9,248 was spent per year resubmitting denied claims (MGMA, 2004), there is significant opportunity for savings through administrative simplification. Although estimates provided by Thomashauer (2009), Wichmann (2009), and the Commonwealth Fund (2009) are not directly comparable owing to the targeting of different means of simplification, they likely have some degree of overlap.

Session 5: Community-Based and Transitional Care Strategies

Chronic illness impacts not only patients, families, and providers, it also heavily impacts healthcare expenditures. Given fragmentation of the healthcare system, care management, palliative care, and community programs have been identified as clear options to facilitating improvements in outcomes and spending.

The presenters in this session focused on community-based and transitional care strategies, including:

  • Care management for medically complex patients;
  • Palliative care; and
  • Wellness and community programs.

Care Management for Medically Complex Patients

Fragmentation is a central characteristic of our healthcare delivery system and contributes to poor quality care, patient dissatisfaction, medical errors, redundant care, and rising health spending (Cebul et al., 2008; The Commonwealth Fund, 2008; IOM, 2001). The problems resulting from the systemic lack of care coordination are compounded for patients with multiple chronic conditions (Vogeli et al., 2007). Central to ongoing discussions on transforming the payment system to promote accountability through such mechanisms as bundled payments and accountable care organizations (Fisher et al., 2009; de Brantes et al., 2009) is the goal of promoting care coordination, especially for patients with chronic illness, as those with five or more chronic conditions account for two-thirds of the recent rise in Medicare spending (Thorpe and Howard, 2006).

This section discusses the analyses of Kenneth E. Thorpe on the potential savings that could be achieved by improving care management. Other estimates are also provided and discussed.

Potential savings from improved care management Thorpe (2009) examined the possible benefits that could arise from improving care coordination and identified policies that could help achieve this goal. Although they did not provide a comprehensive, national estimate of potential cost savings, they did provide a number of suggestive examples. Improved care coordination could help reduce the $12 billion MedPAC estimated is spent on potentially avoidable hospital readmissions every year (Miller, 2008). Findings from a recent study on frail elders in transitional care suggest a 10-year investment of $25 billion could lead to $100 billion in savings over the same period (Naylor et al., 2004). If the use of palliative care services could be increased to 7.5 percent of hospital discharges nationally, perhaps more than $37 billion could be saved over the next decade (Meier, 2009). Finally, an investment of $10 per person per year could yield more than $16 billion in medical cost savings within 5 years (Levi, 2009).

Thorpe (2009) described how the provision of coordinated care could be rewarded with three payment reforms: primary care reimbursement, bundled payments, and bonus pools. If Thorpe’s ideas for community health teams were implemented, primary care practices could be encouraged to establish formal relationships with these community health teams via a per-person per-month payment for each dually eligible patient. Participation in the reimbursement program would be contingent on the practice meeting specific National Committee for Quality Assurance medical home standards, and further financial incentives could be designed to foster quality improvements. To reduce the costs associated with hospital readmissions, Thorpe proposed that payments be bundled to cover all acute services for admission as well as Medicare-covered post-acute care for 30 days postdischarge. They also suggested that hospitals with above-average readmission rates receive reduced payments. Finally, he suggested that the formation of bonus pools could encourage primary care practices and community health teams staff to improve health outcomes and reduce unnecessary care.

Additional estimates A more extensive review of the literature is available in prior discussions on Owens (2009) and Ferris (2009). It is worth noting here that Berenson and colleagues (2009) estimate that care coordination for dually eligible Medicare and Medicaid beneficiaries could result in a 10-year (2010–2019) savings of $201 billion. For more details, please consult the summaries described in Section II, Session 2 (Owens) and Section III, Session 2 (Ferris).

Estimates comparison Thorpe (2009) did not present a national estimate for cost savings, and thus no direct comparison was performed.

Palliative Care

There is a substantial literature finding evidence supporting the notion that increased palliative care can have positive benefits across a number of areas, such as physical and psychological symptoms; family caregiver well-being; patient, family, and consulting physician satisfaction; support for families in crisis; planning for safe transitions out of hospitals; and family satisfaction and bereavement outcomes (Morrison and Meier, 2004). Given that the costs for care in the last year of life represent more than 25 percent of spending in Medicare (Berenson et al., 2009b) and that additional spending at the end of life does not buy higher-quality care (Yasaitis et al., 2009), there is significant potential for palliative care to improve outcomes and reduce healthcare costs.

This section summarizes the presentation of Diane E. Meier. Other estimates and a comparison of these estimates are also provided.

Potential savings from increased palliative care Meier (2009) described the role for palliative care programs in addressing the cost and quality problems in the U.S. healthcare system. By her calculations, potential savings from increased use of palliative care is approximately $5 billion per year. She also described the quality improvements that these programs could bring about, and the factors that might limit patients’ access to them.

Meier reasoned that, of the approximately 30 million annual hospitalizations in the United States, palliative care could be provided for 5 to 8 percent of these hospitalizations, as 2 percent of all hospitalizations end in death (AHRQ, 2002) and 3 to 6 percent of hospitalizations are for very sick patients who are discharged alive (Siu et al., 2009). Based on recent studies, the per-patient costs saved by palliative care consultation are $2,659 (Morrison et al., 2008). Currently, 53 percent of U.S. hospitals have palliative care programs (Goldsmith et al., 2008), and this type of care reaches only 1.5 percent of their hospitalizations. Hence, current savings attributable to palliative care is $1.2 billion. If the proportion of hospitals with palliative care programs increases to 90 percent and these programs reach at least 7.5 percent of hospitalizations, savings would increase to $6 billion (Goldsmith et al., 2008; Morrison et al., 2008; Siu et al., 2009). The marginal savings attributable to increased palliative care is the difference between potential ($6 billion) and current ($1.2 billion) savings.

However, to achieve these savings, a number of barriers to accessing this type of care would have to be surmounted, such as lack of physician and nursing education, financial incentives discouraging workforce development and organizational commitment, lack of an evidence base guiding quality care, and need for adequate compensation and loan forgiveness opportunities to attract young professionals into the field (Meier, 2009).

Additional estimates A subsequent literature review found that there is a small but growing literature on the cost-saving potential of extended palliative care on which other national estimates appear to be based. A core group of papers including Morrison and colleagues (2008) and Smith and colleagues (2003) appear influential. For example, a recent report by Berenson and colleagues (2009) cited an overlapping set of papers and concluded that, if a modest change to clinical decision making for patients in end-of-life care could be made, savings to Medicare could amount to $6 billion in 2010 and $91 billion over 10 years.

UnitedHealth Group (2009a) provided a notable estimate outside the peer-reviewed literature. They found that a program providing information to guide patients and their families in making medical decisions that included palliative care at the end of life could produce about $18 billion in savings between 2010 and 2019.

Estimates comparison The estimates of Meier (2009) of approximately $5 billion in annual savings and Berenson and colleagues (2009) of $6 billion in 2010 are similar. This is perhaps unsurprising as they draw on similar resources. It is not unreasonable to expect that other existing estimates based on the scientific literature would be of similar magnitude. The estimate by UnitedHealth Group (2009a) is not directly comparable because of the estimates encompass different time frames.

Wellness and Community Programs

As previously discussed, prevention has long held value in health care. As nearly 40 percent of all deaths in the United States are due to behavioral causes, attention on prevention has encompassed obesity, vaccinations, and cancer screening (Mokdad et al., 2004). Although some have argued that prevention can save costs from the prevention of illness, others have cited evidence to the contrary (Cohen et al., 2008; Elmendorf, 2009).

Although many preventive services are clinical in nature, as described in Flottemesch (2009) during the May workshop, this section focuses on the estimates presented by Jeffrey Levi, who examined the potential for community-based programs to deliver cost savings. Competing estimates are presented and discussed as well.

Potential savings from wellness and community programs Levi presented results from a collaborative study finding that a small investment in preventive services could significantly reduce U.S. net expenditures on health (TFAH, 2008). Focusing on programs that target communities or at-risk segments of communities, the Trust for America’s Health (TFAH) found that an investment of $10 per person per year in proven community-based programs to increase physical activity, improve nutrition, and prevent smoking and other tobacco use could reduce net expenditures on health care by more than $16 billion annually within 5 years. Out of the total possible savings, at least $5 billion represents savings to Medicare, $2 billion represents savings to Medicaid, and $9 billion represents savings to private payers.

The two main components to the study were a comprehensive literature review of community-based prevention studies and a model that would calculate potential returns to these preventive services investments. For the literature review, the TFAH consulted with the New York Academy of Medicine and identified 84 studies of community-based programs and policy changes that could be identified as public health interventions. Many important modeling assumptions were derived from this literature review. For the calculations, researchers at the Urban Institute developed a model focused on three elements: individual-level spending on Medicare for selected preventable diseases, potential savings from reducing the prevalence of these diseases, and distribution of these potential savings across payers. The potential national savings calculated using this model are presented in Table A-6 and Table A-7. State-level results and other findings are available in Prevention for a Healthy America (TFAH, 2008).

TABLE A-6. National Return on Investment of $10 per Person.

TABLE A-6

National Return on Investment of $10 per Person.

TABLE A-7. Distribution of Payer Savings from an Investment of $10 per Person.

TABLE A-7

Distribution of Payer Savings from an Investment of $10 per Person.

The researchers described a number of limitations to these estimates. They noted that savings were calculated in a way that excluded future changes in medical technology. The modeling also excluded spending on infrastructure that would be required to implement these preventive programs. Limitations like these suggest that the estimated benefits to increased preventive services may be overstated. However, as the researchers generally took a conservative approach to their model inputs by assuming higher costs and lower benefits whenever possible, they concluded that their estimates likely understate the potential benefits to increased preventive services.

Additional estimates Although there are many evaluations of the effect of wellness programs on health and costs, it appears that the TFAH (2008) result is unique as a comprehensive national estimate of potential savings from these community-based wellness programs. A substantial majority of papers found in our search for comparable national estimates were papers based on data collected from small-scale interventions. Furthermore, most of these papers investigated the effect of work-based wellness programs. Parks and Steelman (2008) conducted a meta-analysis of these studies, but the authors did not provide an estimate of the potential savings these wellness programs could have on costs. A review by Pelletier (2005) of 12 studies published between 2000 and 2004 also did not contain an estimate that could be used to compare the TFAH (2008) result.

Although not focused on community wellness programs, the Commonwealth Fund’s estimates of $255 billion from reducing tobacco use and $406 billion from reducing obesity are of related interest. For additional discussion on the potential savings and costs related to preventive services, please refer to the section on missed prevention opportunities in Section II.

Estimates comparison Comparable estimates were not identified. However, that the TFAH (2008) found cost savings from these programs is consistent with the findings in most papers that wellness programs do reduce expenditures. Furthermore, while there is some overlap between the conditions targeted by community programs and primary clinical preventive services, the estimated annual savings of $16 billion from the former would likely complement the estimated savings of $7 billion from increased primary preventive services presented by Flottemesch (2009). Finally, as Woolf (2009) noted, even though a particular intervention may not be cost saving it may indeed be cost-effective and improve quality and quantity of life at an acceptable price.

Session 6: Entrepreneurial Strategies and Potential Changes in the State of Play

The value placed on innovation is seen throughout the healthcare industry. Emerging strategies, such as techniques to minimize artificial variability and technology, may have the ability to lower costs and improve outcomes.

The presenters in this session explored entrepreneurial strategies and potential changes in the state of play, including:

  • Managing variability in healthcare delivery;
  • Retail clinics; and
  • Technological innovation.

Managing Variability in Healthcare Delivery

The potential savings from reducing hospital inefficiency were discussed in previous sections of this working paper. Achieving some of these potential savings might be possible with the implementation of variability methodology.

In general, the variability in patient flow through the care delivery process can be characterized in one of two ways: natural or artificial. An example of natural variability is the flow of patients admitted to a hospital unit through the ED. Many emergencies are random events and considered uncontrollable. However, artificial variability is controllable and introduced by extrinsic factors such as scheduling. Artificial variability can be found in the flow of elective admissions (e.g., elective surgical, catheterization lab, and oncology admissions) to a hospital if the scheduling of these patients does not take into account the impact on the rest of a hospital’s resources such as inpatient units, ED beds, and diagnostic services. For example, improperly scheduling these patients can lead to ED overcrowding, boarding, and diversion, and can lead to unnecessary competition for inpatient beds between elective and ED admissions. Although the science of operations management does not have any solution for artificial variability, variability methodology has been developed as a tool to specifically address artificial variability. At its core, variability methodology involves identification, quantification, and elimination of artificial variability so the remaining variability can be managed using the standard operations management tools mentioned above (Joint Commission Resources, 2009).

Below, the potential cost savings from the widespread implementation of variability methodology estimated by Sandeep Green Vaswani are summarized. Results from a literature review seeking other estimates are provided and a comparison of these estimates is made.

Potential savings from clinical service engineering applications In describing the potential for variability methodology to address artificial variability in patient flow and thereby reduce healthcare expenditures, Vaswani estimated nationwide annual savings from the implementation of variability methodology as being between $35 billion and $112 billion.

The preliminary estimate was based on two factors: (1) the number of beds in the U.S. healthcare system that could be closed through better management by implementing variability methodology, and (2) the cost of operating those beds. Vaswani (2009) first considered a scenario in which all hospital admissions would come through the ED. Results from operations management suggest that inpatient bed occupancy could be increased from the current 65 percent to 80 percent without causing excessive waiting times (Litvak, 2005). Allowing for hospital admissions to include elective admissions, bed occupancy could be increased to over 90 percent by implementing variability methodology (Litvak, 2005). Taking a conservative route, Vaswani and colleagues (2009) assumed an increase in occupation rate to 80 percent and the closing of unneeded beds. Multiplying the number of closed beds by a low estimate ($250,000) and high estimate ($800,000) of their operating costs (Butterfield, 2007) yielded the national cost savings estimate. This estimate assumed no growth in demand. Additional assumptions to account for growth, which is likely, led to annual savings estimates in the range of $39 billion to $121 billion. Notably, current capacity can absorb up to 25 percent growth of inpatient demand without cost increases or bed closures. Their estimate also assumed a full staffing rate of 100 percent. For sake of argument, if 5 percent of staffed beds are actually unstaffed, then the annual savings estimates are in the range of $26 billion to $82 billion.

Additional estimates Variability methodology is a relatively recent innovation (Litvak and Long, 2000). It appears that the authors and their colleagues at the Management of Variability Program at Boston University have conducted the majority of research on, and evaluations of, its implementation. Hence, the estimated national savings estimate from implementing variability methodology in Litvak and colleagues (2005) is likely unique in the literature.

The operations management literature on hospital efficiency was also searched for comparable estimates. Broad studies and reviews such as those performed by Ling and colleagues (2002) and Jack and Powers (2004) provided neither comparable estimates nor reference to other papers that were close enough to the investigation described above. However, the discussion of hospital efficiencies in the context of estimates provided by Milstein (2009) and Pittenger (2009) are relevant and likely complement, and potentially overlap, the estimates provided by Vaswani.

Estimates comparison Given the above discussion, there are no directly comparable estimates of potential national savings. As mentioned, it may be informative to put the estimates presented by Vaswani’s findings in context with the previous discussion on clinical inefficiency. During this workshop, Pittenger (2009) categorized savings opportunities from application of the Virginia Mason Production System into operational, clinical, and patient safety opportunities. Yet clearly variability methodology also addresses issues of efficiency of resource use. If both these strategies are employed, the savings realized will likely result from overlapping interactions. These are, of course, imperfect comparisons, and the degree of overlap cannot be estimated based on the information provided here.

Retail Clinics

Retail clinics appeared in the medical market in 2000. They are typically staffed by mid-level providers with remote physician oversight and have the aim of providing a limited set of simple clinical services. Given their lower operating costs, it may be possible that the increased use of these clinics could reduce healthcare costs in the United States.

In this section, we discuss the savings calculation presented by N. Marcus Thygeson (2009). A comparison of other existing estimates is provided at the end of this section.

Potential savings from expansion of retail clinic use Thygeson (2009) investigated the impact of retail clinics on healthcare costs and quality. He concluded that the maximum savings that could be achieved by expanding the use of these clinics was $7.5 billion per year. However, he identified a number of factors that could drive these potential savings to as low as $2 billion per year.

To obtain the estimate of national savings, the author first determined how many medical episodes could be properly treated in retail clinics across the United States. Using data from retail clinics in Minnesota, he calculated expected per-episode savings. Supposing that all retail clinic-eligible visits would be made at retail clinics instead of physician offices and emergency departments, he found that the upper bound for potential savings was $7.5 billion per year (Table A-8).

TABLE A-8. Estimated National Savings Conversion of all U.S. Retail Clinic-Eligible Visits.

TABLE A-8

Estimated National Savings Conversion of all U.S. Retail Clinic-Eligible Visits.

The author stated three caveats to these estimates. First, previous work has shown that, at least among the insured population, a reduction in retail clinic-eligible episodes treated in EDs was not observed. This suggests that the contribution to savings from patients’ decreased ED use may well be lower than calculated. Second, over 85 percent of retail clinics are located in the 50 largest metropolitan statistical areas in the United States. It is not clear if savings could be achieved in more rural locations. Third, the estimate does not account for individuals who would not have sought treatment from a healthcare provider in the absence of a retail clinic. There is some evidence that the presence of a retail clinic may induce individuals to increase their demand for health care (Wang et al., in press). Hence, this behavioral response suggests that the $7.5 billion savings may be an overestimate. Finally, the analysis above ignores the potential competitive response on the part of established healthcare providers.

Additional estimates A search of the peer-reviewed literature found no articles that could provide a competing national estimate of the impact of increased use of retail clinics on healthcare costs. However, a recent study found that retail clinics provide less costly treatment compared to physician offices and urgent care centers for three common illnesses, with no apparent adverse effect on quality of care or delivery of preventive care (Mehrotra et al., 2009). It was also estimated that encouraging growth of retail clinics could yield savings up to $6 billion over a decade in Massachusetts (Eibner et al., 2009). Scott (2006), one of the more highly-referenced reports outside the peer-reviewed literature, also noted some evidence of the potential for savings but did not provide a national savings estimate. However, as retail clinics are a relatively recent phenomenon, it has been noted that “credible data on the clinics’ impact on the quality and cost of care” are sparse (Alexander, 2008).

Estimates comparison Although no other national estimates of the potential savings available from use of retail clinics are available for comparison, emerging evidence indicates that use of convenience clinics is rapidly rising and may be a unique source of cost savings.

Technological Innovation

While technology has been oft cited as a driver of spending growth (Kaiser Family Foundation, 2007), technology and innovation have also empowered cost-lowering applications in health care such as telehealth and telemedicine. The Care Coordination/Home Telehealth (CCHT) national pilot program implemented by the Department of Veterans Affairs (VA) provides an informative case study on the use of telemedicine as a powerful tool to improve outcomes while lowering costs. CCHT was guided by the institutional experience of the VA and findings from a randomized-controlled study of chronic care patients using video home telehealth systems (Johnson et al., 2000). The experiment found that the technology was effective, well-received by patients, helped maintain quality care, and had cost-saving potential. The pilot program combined these telehealth methods with care coordination efforts in a chronic care model that combined patient self-management and an algorithm used to choose appropriate home telehealth technologies (Lorig et al., 2001; Ryan et al., 2003).

Adam Darkins’ presentation on the use and cost improvements achieved by the VA through implementation of CCHT is described here. Other savings estimates are presented and compared as well.

Potential savings from technological innovation Darkins (2009) presented findings from a VA case report on their CCHT national pilot program. The aim of the program was to provide more appropriate care to patients with chronic conditions who might benefit from care provided outside the usual outpatient clinic appointment paradigm. Based on results from this pilot program, Darkins estimated that national implementation of CCHT could result in annual cost savings between $1.7 billion and $2.2 billion (22 percent and 48 percent of total healthcare costs) for the target population.

Darkins and colleagues (2008) reported a number of impressive outcomes from their pilot program: a 19 percent reduction in hospital admissions, a 25 percent reduction in lengths of stay, an 86 percent mean patient satisfaction score, and no measured diminution of quality. The annual cost of providing CCHT was $1,600 per patient, which represented large savings relative to in-home care via nursing teams ($13,121) or purchasing nursing home care on the commercial market ($77,745). Basing his calculation on the above findings, the author estimated that cost savings could fall between $1.7 billion and $2.2 billion per year for the target population. Table A-9, below, provides details on the author’s cost-savings calculations; these estimates cannot be summed as the target groups are not discrete.

TABLE A-9. Examples of Crude Estimates of Cost Reductions That May Be Realizable Through Implementation of Care Coordination/Home Telehealth (CCHT) Outside the Department of Veterans Affairs.

TABLE A-9

Examples of Crude Estimates of Cost Reductions That May Be Realizable Through Implementation of Care Coordination/Home Telehealth (CCHT) Outside the Department of Veterans Affairs.

Additional estimates Generally speaking, this intervention falls into the category of care coordination initiatives, which is discussed at length in the review of Owens (2009) and Ferris (2009). However, this case report also addresses the potential of technological innovation to reduce costs, which a few recent papers have examined. Vo (2008) estimated that, after a 6-year rollout period, $3.6 billion in savings could be achieved through implementing telehealth technology on a national scale. In particular, he claimed that physician-to-physician consultations mediated by telehealth technology would reduce unnecessary or redundant tests by 45 percent. Pan and colleagues (2008) estimated that the implementation of telehealth systems in emergency rooms, prisons, nursing home facilities, and physicians offices across the United States could achieve $4.3 billion in annual savings. Nearly all other papers surveyed noted the impact of limited implementation and did not attempt to quantify potential national savings.

Estimates comparison Although the other savings estimates from Vo (2008) and Pan and colleagues (2008) are not directly comparable, it is striking that they are of the same order of magnitude. Furthermore, the VA’s experience with the use of home telehealth technology for patients with chronic conditions adds to a growing body of evidence that home telehealth has the potential to reduce costs (Polisena et al., 2009). Also as mentioned above, please refer to the Owens (2009) and Ferris (2009) literature reviews for a broader discussion of potential savings from increased care coordination.

Summary Table of Estimates

Topic/PresenterPresenter EstimateRelevant Comparisons EstimatesRemarks
UNDERSTANDING THE TARGETS
Session 1: Unnecessary Services
Overuse of Services beyond Evidence-Established Benchmarks
 Amitabh Chandra8% annual reduction in both Medicare costs and mortality10.8% to 25.5% reduction in costs, holding quality constant (Rosko and Mutter, 2008)

$65.1 billion (3.4% of U.S. health spending) spent on eight selected wasteful services (Bentley et al., 2008)

Overuse of urinalyses, electrocardiograms, and x-rays totaled an estimated annual direct medical cost of up to $194.0 million (Merenstein et al., 2006)

$600 billion due to excess medical and surgical services; avoidable emergency department use has been estimated to cost $21.4 billion nationally and overuse of antibiotics to cost $1.1 billion annually; $300.0 million in annually spending on unnecessary MRI scans for back pain (Delaune and Everett, 2008)

$5.1 billion annually could be saved from a 50 percent decline in unnecessary visits for common conditions—headaches, back pain, and benign breast conditions; $6.5 billion in annual savings from reducing unnecessary MRI testing for back pain and headaches (Mecklenburg and Kaplan, 2009)
Cost estimates of overuse of clinical services difficult to compare directly given the inclusion of different services in each estimate
Use of Services beyond Benchmarks where Evidence Is Not Established
 Elliott S. Fisher$47.8 to $53.9 billion (18% to 20%) annual cost savings to Medicare (2005/6 dollars)28.9% cost savings to Medicare (Wennberg et al., 2002)Authors’ study analyzes disaggregated data from a more recent period
Choice of Higher Cost Services over Evidence-Established Benchmarks
 David Wennberg$125 billion (up to 5%) net cost savingsN/AFew studies on shared decision-making provide assessment of costs (Leatherman and Warrick, 2008)
Session 2: Inefficiently Delivered Services
Medical Errors and Redundant Tests
 Ashish Jha$16.6 billion (2004 dollars) in direct medical costs for medical errors

$8.2 billion for redundant tests
$17 to $29 billion in total costs (Institute of Medicine (IOM), 2000) for medical errors

$4.6 billion (2000 dollars) in national health care costs for 18 injuries (Zhan and Miller, 2003)
As Zhan and colleagues (Zhan and Miller, 2003) based their estimates on coding for medical errors in administrative data, which likely underestimated the incidence of errors, while Jha utilized incidence rates and costs from studies generally relying on comprehensive chart review-based data, the latter’s estimates may be more accurate
Care Fragmentation
 Mary Kay Owens$240 billion average annual savings by 2013 from coordinated care program$200.5 billion over 10 years (2010–2019) from care coordination for Medicare and Medicaid dually-eligibles (Berenson et al., 2009)Owens (2009) is only cost estimate that extrapolated results to entire population; other studies found no savings from care coordination (Fireman et al., 2004; Peikes et al., 2009)
Inefficient Use of Higher Cost Providers
 Gary S. Kaplan$5.1 billion in annual savings from reduction in unnecessary visits; $6.5 billion annually form unnecessary MRIs; $8.3 billion for use of lower-cost providers; $2.3 billion from substituting low-cost telephone or computer-based visits for conventional visits for chronic conditionN/AMultiple studies support findings of improved quality and lower costs from use of mid-level practitioners, though none offer national savings estimates (Eibner et al., 2009; Hatem et al., 2008; Hooker, 2002; Roblin et al., 2004)
Inefficiencies in Physician Offices and Hospitals
 William F. Jessee$6.4 billion (2007 dollars) by reducing inefficiencies in physician officesN/AN/A
 Arnold Milstein2% decrease in overall U.S. healthcare spending if all hospitals achieve same performance and cost levels as top 12% (MedPAC, 2009)$400.0 billion over 10 years for Medicare and $1.3 trillion for private payers over the same decade if U.S. hospitals reduced their inpatient costs to the level of Thedacare (Toussaint, 2009)

$19.4 billion annually from elimination of non-value added activities (Hafer, 2009)
Differences likely arise from different methodologies, sources of data, and time horizons
Session 3: Excess Administrative Costs
Estimates of Excess Administrative Costs
 James G. Kahn, Lawrence P. Casalino and James L. Heffernan$168 to $183 billion in total excess BIR spending$209 billion in total excess administrative spending (Woolhandler et al., 2003)Joint workshop estimate uses micro-level approach while Woolhandler estimate uses macro-level approach
Regulatory and compliance-imposed costs beyond benchmarks
 Peter K. Smith$87.9 billion in potential savings from reducing documentation requirements of nursesN/AN/A
Potential Reduction in Administrative Expenses
 Andrew L. Naugle$20 to $23 billion in annual potential savings to commercial payers$63 to $75 billion in excess BIR spending by private insurers (Kahn, 2009)*Differences likely arise due to different definitions of administrative costs and different methods for calculating administrative costs
Session 4: Prices That Are Too High
Hospital Service Prices
 Cory S. CappsIncrease in annual national healthcare expenditures of $10 to $12 billion due to hospital consolidationsN/AN/A
Prices of Medications
 Jack Hoadley$9 billion in total annual savings from a 5% across-the-board price reduction (excluding government purchasers that already receive significant discounts)$10 billion annual savings for Medicare Part D spending from increased used of generics (CBO, 2008a)

$21.9 billion in annual savings for Medicare Part D if prices reduced to Federal Supply Schedule (Gellad et al., 2008)

$156 billion savings over the next decade from reducing the prices for medications paid by Medicare Part D for dually eligibles by 30 percent to Medicaid price levels both target pharmaceutical manufacturers (U.S. House of Representatives Committee on Oversight and Government Reform, 2008)
Differences in estimates due to the magnitude of the price reduction utilized in each study
Prices of Durable Medical Equipment
 Thomas J. Hoerger$2.8 billion annual savings from reduction in Medicare reimbursements, fraud, and waste for DMESpending on DME $19 billion less than expected, relative to wealth (Farrell et al., 2008)Differences in estimates likely due to differences in the study population—one estimate is for Medicare while the other is nationwide spending
 Mark E. WynnInitial bids for national competitive bidding program 26% lower than fee schedules
Prices of Medical Devices
 Jeffrey C. Lerner$4.7 billion (2008 dollars) in savings for medical device price negotiationsN/AN/A
Session 5: Missed Prevention Opportunities
Savings from Increased Primary and Secondary Prevention
 Thomas J. Flottemesch$7 billion annual spending reduction from increased primary prevention

$3.3 billion annual spending increase from increased secondary prevention
$1.6 trillion between now and 2023 from prevention and early intervention for seven common chronic diseases (DeVol et al., 2007)

$255 billion in savings over 11 years from reduction in tobacco use; $406 billion over same time period from reduction in obesity (The Commonwealth Fund, 2009)

$191 billion from interventions aimed at preventing diabetes among those at highest risk (Berenson et al., 2009)

Annual excess costs attributable to smoking and conditions related to obesity and being overweight at $567 to $161 billion and $200 billion, respectively; the costs of poorly controlled diabetes were $22 billion while non-adherence cost another $100 billion (PriceWaterhouseCoopers, 2009)

No evidence of cost savings (CBO, 2004b; Delaune and Everett, 2008; Elmendorf, 2009; Goetzel et al., 2005; Mattke et al., 2007; Russell, 2009)
For most preventive services, many studies suggest expanded utilization leads to higher, not lower, medical spending overall; however, some targeted prevention interventions have been found to be cost-saving
Savings from Increased Tertiary Prevention
 Michael P. Pignone$45 billion annual spending reduction from increased tertiary preventionPlease see Flottemesch (2009) for more detailsPlease see Flottemesch (2009) for more details
BIR = billing and insurance-related; DME = durable medical equipment; SDM = shared-decision making.
* Estimate presented during May workshop.
STRATEGIES THAT WORK
Session 1: Knowledge Enhancement-Based Strategies
Comparative Effectiveness Research
 Carolyn M. ClancyN/A$480 billion over 10 years (2010–2019) (Collins et al., 2009)Given uncertainty in predicting adoption, others (Berenson et al., 2009; CBO, 2007) have noted potential for savings but declined to provide specific estimates
Evidence-Based Clinical Protocols
 Lucy A. Savitz$2 billion annual savings for evidence-based protocol for treating febrile infants$175 billion over 10 years (2010–2019) from implementation of an integrated medical management program in Medicare and application of evidence-based standards to reimbursement policies (UnitedHealth Group, 2009a)Estimates are not directly comparable as one is based on savings from a single clinical protocol; the other estimates saving for federal spending
Electronic Health Records with Decision Support
 Rainu Kaushal$1 to $2.7 million annually per hospital after an initial investment from adoption of computerized physician order entry (Massachusetts Technology Collaborative & New England Healthcare Institute, 2009)$77 billion in annual savings due to efficiency gains; $371 billion for hospital systems ($142 billion physician offices) over 15 years when including gains from safety (Hillestad et al., 2005)Significant variation exists in the estimates of savings associated with adoption of EHRs and HIT depending on the time horizon analyzed, the type of technology being examined, and the extent to which the authors assume the technology will be adopted
$86,400 per provider over five years from adoption of EHRs in the ambulatory setting (Wang et al., 2003)$180 billion over 10 years from investment in HIT (Collins et al., 2009)

$800 billion spillover effects from adoption of EHR (Russo, 2009)

$97 billion in 10-year savings from adoption of EHRs (Berenson et al., 2009)

Likely no cost savings (CBO, 2008)
Session 2: Care Culture and System Redesign-Based Strategies
Improved Provider Profile and Use
 Michelle J. LynN/A$8.3 billion in savings if half of outpatient visits for uncomplicated patients could be handled capably by qualified non-physicians (Mecklenburg and Kaplan, 2009)*

$16 billion in savings from community-based wellness programs (Trust for America’s Health, 2008)

$2 to $7.5 billion in savings from retail clinics (Thygeson, 2009)*
Though the interventions are related, broadly speaking, the savings estimates are not directly comparable
 Jason HwangN/A
Care Site Efficiency and Productivity Initiatives and Incentives
 Kim R. Pittenger$57.8 billion in savings from widespread implementation of Virginia Mason Production SystemPlease see Milstein (Milstein, 2009) for more detailsPlease see Milstein (2009) for more details
 Sandeep Green Vaswani$35 to $112 billion annual savings from national implementation of Variability Methodology in hospitals
Care Site Integration Initiatives
 Timothy G. Ferris$0.6 and $1.5 billion for Medicare over a two year period from implementation of care delivery model targeting the highest risk patients$367.4 billion in total savings to federal government over ten years from bundle of interventions (UnitedHealth Group, 2009a)

$175 billion in savings from patient-centered medical homes over ten years (Collins et al., 2009)

$14.8 billion over the next decade for Medicare and Medicaid from lowering payment for potentially preventable readmissions within 15 days of discharge to 60 percent of the usual payment (Berenson et al., 2009)
Estimates are not directly comparable due to specificity of the intervention described in Ferris; regardless of the approach taken, all the reviewed papers endorse the concept of care coordination as a potential method of improving health and care coordination; please Owens (2009) for more details
Antitrust Interventions
 Roger FeldmanN/AN/APlease see Capps (2009) for more details
Promoting Information Technology Interoperability/Connectivity
 Ashish Jha$81 billion through improvements in HIT safety and efficiency (Hillestad et al., 2005)

$337 billion during a 10-year implementation period and annual savings of nearly $78 billion in each subsequent year (amounts measured in 2003 dollars) (Walker et al., 2005)
Please see Kaushal (2009) for more detailsBoth of these studies have been subject to significant methodological critiques; please see Kaushal (2009) for more details
Service Capacity Restrictions
 Frank A. SloanN/AN/AAuthor suggested that effectiveness of capacity restrictions depends on other policy decisions on cost containment; recent work (Grabowski et al., 2003; Ho, 2007) support the notion that CON programs have not succeeded in cost containment
Medical Liability Reform
 Randall R. Bovbjerg$20 billion (0.9%) of annual health spending could be saved with conventional tort reform$210.0 billion in savings from reduction in defensive medicine (PriceWaterhouseCoopers, 2009)PriceWaterhouseCoopers’ Health Research Institute estimate far exceeds bounds established in majority of econometric research publications on this topic
Session 3: Transparency of Cost and Performance
Transparency in Prices
 John SantaN/AN/AN/A
Transparency in Comparative Value of Treatment Options
 G. Scott GazelleN/AN/AN/A
Transparency in Comparative Value of Providers
 Paul B. GinsburgN/A$14.5 billion (2010–2019) from sharing quality data in Medicare (UnitedHealth Group, 2009b)N/A
Transparency in Comparative Value of Hospitals and Integrated Systems
 Peter K. Lindenauer$2.5 to $5 billion in annual savings from public reporting requirements related to hospitalsN/AN/A
Transparency in Comparative Value of Health Plans
 Margaret E. O’KaneN/AN/AN/A
Session 4: Payment and Payer-Based Strategies
Bundled and Fee-for-Episode Payments
 Amita Rastogi$165 billion from utilization of bundled payment for 13 specific conditions in commercial population$96.4 billion over 5 years for Medicare if shift to episode-of-care based payments (Schoen et al., 2007)Difficult to compare savings estimates due to focus on different conditions and populations
Managed Competition
 David R. ReimerN/A$17.4 billion in 2010 (federal savings) due to operation of a public plan option in a health insurance exchange (Berenson et al., 2009)N/A
Value-Based Insurance Design
 Niteesh K. Choudhry$2 billion if VBID applied to five common conditionsN/AN/A
 Lisa Carrara3% to 4% savings in patient’s claims in the first year by designating specialists based on high quality and efficiency3% to 7% decrease in premiums from use of more efficient providers (GAO, 2007)

$37 billion (2010–2019) from implementation of a program designed to provide Medicare beneficiaries with information on quality and efficiency variations among providers (UnitedHealth Group, 2009a)
Aetna (from which Carrara’s estimate were drawn) was one of the insurers included in the GAO study; as neither the Carrara or GAO estimate translated savings into dollar amounts, direct comparison are not possible
Administrative Simplification
 David S. Wichmann$322 billion based on application of technology to administrative activities (UnitedHealth Group, 2009b)$337 billion in administrative savings over 10 years due to a national health insurance exchange with a public plan option (The Commonwealth Fund, 2009)Estimates are not directly comparable due to targeting of different means of simplification, though there is some degree of overlap
 Robin Thomashauer$3 billion if CORE is implemented nationwide (IBM Global Business Services, 2009)
Session 5: Community-Based and Transitional Care Strategies
Care Management for Medically Complex Patients
 Kenneth E. Thorpe$75 billion savings over ten years from investment in transitional care from frail elders (Naylor et al., 2004)Please see Owens (2009) and Ferris (2009) for more detailsPlease see Owens (2009) and Ferris (2009) for more details
Palliative Care
 Diane E. Meier$4.8 billion in annual savings from increased palliative care$6.4 billion in savings in 2010 (Berenson et al., 2009)

$18 billion in savings between 2010 and 2019 (UnitedHealth Group, 2009a)
Meier and Berenson et al. estimates draw on overlapping literature; UnitedHealth Group estimate difficult to compare given decade long estimate
Wellness and Community Programs
 Jeffrey Levi$16 billion in annual savings within five years$7 billion in annual savings from increased primary preventive services (Flottemesch, 2009)*Estimates not directly comparable since Flottemesch analyzed clinical preventive services while Levi analyzed community wellness and prevention programs
Session 6: Entrepreneurial Strategies and Potential Changes in the State of Play
Retail Clinics
 N. Marcus Thygeson$2 to $7.5 billion in annual savings from increased utilization of retail clinicsN/AMultiple studies support findings of improved quality and lower costs from use of retail clinics, though none offer national savings estimates (Eibner et al., 2009; Mehrotra et al., 2009)
Technological Innovation
 Adam Darkins$1.7 billion in annual cost savings from increased usage of Care Coordination/Home Telehealth$3.6 billion in savings from national implementation of telehealth technology (Vo, 2008)

$4.3 billion in annual savings from widespread implementation of telehealth systems (Pan et al., 2008)
Estimates not directly comparable given different interventions in different settings

NOTE: CORE = Committee on Operating Rules for Information Exchange; EHR = electronic health record; HIT = health information technology; MedPAC = Medicare Payment Advisory Commission; VBID = value-based insurance design.

*

Estimate presented during May workshop.

SUMMING THE LOWER BOUND ESTIMATES

To provide an informal contextual perspective on the magnitude and distribution of the excess healthcare costs estimated from the workshop presentations and supplemental literature review, the staff of the Institute of Medicine’s Roundtable on Value & Science-Driven Health Care considered the estimates cited in the background paper and identified the lowest estimate within each category of excess expenditure considered. After adjustment to 2009 expenditure levels, these estimates were summed and are indicated on the preceding table with a condensed summary in Box A-1. It should be emphasized that these are virtually all unvalidated extrapolations, based on assumptions from limited observations, and, in the face of obvious overlaps, duplications and uncertainties in the component estimates. They are therefore offered purely for illustrative purposes and to prompt the follow-on analyses necessary for a clearer understanding of the nature, magnitude, and interrelationships of excess health expenditures in the United States, as well as of the strategies necessary to address them.

Box Icon

BOX A-1

Excess Cost Domain Estimates: Lower bound totals from workshop discussions. Overuse: services beyond evidence-established levels Discretionary use beyond benchmarks

Examples of the follow-up analyses required include the following questions and issues:

  • Where are there large differences in estimates addressing similar issues, what are the methodologic differences, and how can they be accommodated or revised to improve the estimates?
  • Which areas and topics need the most additional work, and are there other areas and topics to be addressed?
  • To minimize double counting among categories, and account for intervention synergy, how might the crosswalk delineating areas and degrees of overlap be best approached?
  • Which benchmarks in the variety of topics covered within this summary reflect the most appropriate benchmark levels to guide further analyses?
  • To what degree can cost findings based on national Medicare data be applied to other populations such as those commercially insured?
  • How might additional analyses be further refined to ensure accuracy of the analytics and capture of the significant dimensions and nuances of the areas covered?
  • What additional research is needed to identify the specific, actionable interventions and the steps needed to achieve net savings?

REFERENCES

  1. Best AM. Best’s Market Share Reports (One-Year Premiums and Loss Study) 2009. [accessed August 26, 2009]. http://www​.ambest.com​/sales/marketshare1ypls/
  2. Aaron HJ. The costs of health care administration in the United States and Canada—Questionable answers to a questionable question. New England Journal of Medicine. 2003;349(8):801–803. [PubMed: 12930934]
  3. AHRQ (Agency for Healthcare Research and Quality) Healthcare Cost and Utilization Project: Hospitalization in the United States. 2002. [accessed August 9, 2009]. http://www​.ahrq.gov/data​/hcup/factbk6/factbk6c.htm.
  4. AHRQ (Agency for Healthcare Research and Quality) Outcomes of Community Health Worker Interventions. 2009. [accessed August 9, 2009]. http://www​.ahrq.gov/CLINIC​/tp/comhworktp.htm.
  5. Alexander KJ. Health plans embrace retail clinics. Managed Care. 2008;17(3):32–34. 43. [PubMed: 18447157]
  6. American Diabetes Association. Direct and Indirect Costs of Diabetes in the United States. 2009. [accessed September 2, 2009]. http://www​.diabetes.org​/diabetes-statistics​/cost-of-diabetes-in-us.jsp.
  7. American Medical Association. 5 years of tort reform: Lone star success story. 2008. [accessed August 26, 2009]. http://www​.ama-assn.org​/amednews/2008/09/15/edsa0915.htm.
  8. Andes S, Metzger LM, Kralewski J, Gans D. Measuring efficiency of physician practices using data envelopment analysis. Managed Care. 2002;11(11):48–54. [PubMed: 12491859]
  9. Arcury TA, Preisser JS, Gesler WM, Powers JM. Access to transportation and health care utilization in a rural region. Journal of Rural Health. 2005;21(1):31–38. [PubMed: 15667007]
  10. Austin A, Gravelle J. Does price transparency improve market efficiency? Implications of empirical evidence in other markets for the health sector. [accessed August 26, 2009]. http://stinet​.dtic.mil​/oai/oai?&verb​=getRecord&metadataPrefix​=html&identifier=ADA471238.
  11. Baicker K, Chandra A. Medicare spending, the physician workforce, and beneficiaries’ quality of care. Health Aff (Millwood) 2004;Suppl Web Exclusives:W4-184–197. [PubMed: 15451981]
  12. Baker LC, Fisher ES, Wennberg JE. Variations in hospital resource use for Medicare and privately insured populations in California. Health Affairs (Millwood) 2008;27(2):w123–w134. [PubMed: 18270221]
  13. Bates DW, Spell N, Cullen DJ, Burdick E, Laird N, Petersen LA, Small SD, Sweitzer BJ, Leape LL. The costs of adverse drug events in hospitalized patients. Adverse Drug Events Prevention Study Group. Journal of the American Medical Association. 1997;277(4):307–311. [PubMed: 9002493]
  14. Bentley TG, Effros RM, Palar K, Keeler EB. Waste in the U.S. health care system: A conceptual framework. Milbank Quarterly. 2008;86(4):629–659. [PMC free article: PMC2690367] [PubMed: 19120983]
  15. Berenson R, Holahan J, Blumberg L, Bovbjerg R, Waidmann T, Cook A. How we can pay for health care reform. 2009. [accessed August 26, 2009]. www​.urban.org/UploadedPDF​/411932_howwecan-pay.pdf.
  16. Blanchfield B, Heffernan J, Osgood B, Sheehan R, Meyer G. Simplifying the billing process: Still an opportunity for significant cost efficiency. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  17. Bovbjerg R. The role of liability reform. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop, July 16–18; Washington, DC. 2009.
  18. Bradley B. Tax reform’s lesson for health care reform. New York Times. 2009 Aug 30:WK9.
  19. Butterfield SM. A New Rx for Crowded Hospitals: Math. Stacey Butterfield Interviews Eugene Litvak, Ph.D, ACP Hospitalist. 2007.
  20. Capps C. The approximate effect of hospital consolidation on national healthcare expenditures. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  21. Capps C, Dranove D. Hospital consolidation and negotiated PPO prices. Health Affairs (Millwood) 2004;23(2):175–181. [PubMed: 15046141]
  22. Carrara L. Tiered provider works impact on value-based benefit design. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  23. Casale A, Paulus RA, Selna MJ, Doll MC, Bothe AE, McKinley KE. ProvenCare: A provider-driven pay-for-performance program for acute episodic cardiac surgical care. Annals of Surgery. 2007;246(4):613–621. [PubMed: 17893498]
  24. Casalino L, Nicholson S, Gans D, Hammons T, Morra D, Karrison T, Levinson W. What does it cost physician practices to interact with health insurance plans? Health Affairs (Millwood) 2009;28(4):w533–w543. [PubMed: 19443477]
  25. Casalino L, Nicholson S, Gans D, Hammons T, Morra D, Levinson W. What does it cost physician practices to interact with payers. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  26. CBO (Congressional Budget Office) How increased competition from generic drugs has affected price and returns in the pharmaceutical industry. 1998. [accessed August 26, 2009]. http://www​.cbo.gov/doc​.cfm?index=655&type​=0&sequence=1.
  27. CBO (Congressional Budget Office) Limiting Tort Liability for Medical Malpractice. Washington, DC: Congressional Budget Office; 2004.
  28. CBO (Congressional Budget Office) Would Prescription Drug Importation Reduce US Drug Spending? 2004. [accessed September 10, 2009]. http://www​.cbo.gov/doc​.cfm?index=5406&type=0.
  29. CBO (Congressional Budget Office) Prices for Brand-Name Drugs Under Selected Federal Programs. Washington, DC: Congressional Budget Office; 2005.
  30. CBO (Congressional Budget Office) Research on the Comparative Effectiveness of Medical Treatments. 2007. [accessed August 26, 2009]. http://www​.cbo.gov/ftpdocs​/88xx/doc8891/12-18-ComparativeEffectiveness.pdf.
  31. CBO (Congressional Budget Office) Evidence on the Costs and Benefits of Health Information Technology. 2008. [accessed September 10, 2009]. http://www​.cbo.gov/ftpdocs​/91xx/doc9168/05-20-HealthIT.pdf.
  32. CDC (Center for Disease Control and Prevention) Advance Data from Vital and Health Statistics No. 388. 2007. [accessed August 26, 2009]. www​.cdc.gov/nchs/data/ad/ad388.pdf.
  33. Cebul RD, Rebitzer JB, Taylor LJ, Votruba ME. Organizational fragmentation and quality care in the U.S. healthcare system. Journal of Economic Perspectives. 2008;22(4):93–113. [PubMed: 19791306]
  34. Center for the Evaluative Clinical Sciences. Project topic brief: Preference-sensitive care. Dartmouth Atlas: 2005.
  35. Chandra A. Saving money and lives. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  36. Chassin MR, Kosecoff J, Park RE, Winslow CM, Kahn KL, Merrick NJ, Keesey J, Fink A, Solomon DH, Brook RH. Does inappropriate use explain geographic variations in the use of health care services? A study of three procedures. Journal of the American Medical Association. 1987;258(18):2533–2537. [PubMed: 3312655]
  37. Chaudhry B, Wang J, Wu S, Maglione M, Mojica W, Roth E, Morton SC, Shekelle PG. Systematic review: Impact of health information technology on quality, efficiency, and costs of medical care. Annals of Internal Medicine. 2006;144(10):742–752. [PubMed: 16702590]
  38. Chernew ME, Rosen AB, Fendrick AM. Value-based insurance design. Health Affairs (Millwood) 2007;26(2):w195–w203. [PubMed: 17264100]
  39. Chirikos TN, Sear AM. Measuring hospital efficiency: A comparison of two approaches. Health Services Research. 2000;34(6):1389–1408. [PMC free article: PMC1089087] [PubMed: 10654837]
  40. Choudhry N. Value based insurance designs. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  41. Choudhry NK, Avorn J, Antman EM, Schneeweiss S, Shrank WH. Should patients receive secondary prevention medications for free after a myocardial infarction? An economic analysis. Health Affairs (Millwood) 2007;26(1):186–194. [PubMed: 17211028]
  42. Choudhry NK, Patrick AR, Antman EM, Avorn J, Shrank WH. Cost-effectiveness of providing full drug coverage to increase medication adherence in post-myocardial infarction Medicare beneficiaries. Circulation. 2008;117(10):1261–1268. [PMC free article: PMC2722033] [PubMed: 18285564]
  43. Clancy C. Comparative effectiveness research. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  44. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in hospitalized patients. Excess length of stay, extra costs, and attributable mortality. Journal of the American Medical Association. 1997;277(4):301–306. [PubMed: 9002492]
  45. CMS (Centers for Medicare & Medicaid Services) National health expenditure projections 2008–2018. 2009. [accessed June 11, 2009]. http://www​.cms.hhs.gov​/NationalHealthExpendData​/downloads/proj2008.pdf.
  46. Cohen JT, Neumann PJ, Weinstein MC. Does preventive care save money? Health economics and the presidential candidates. New England Journal of Medicine. 2008;358(7):661–663. [PubMed: 18272889]
  47. Coleman EA, Parry C, Chalmers S, Min SJ. The care transitions intervention: Results of a randomized controlled trial. Archives of Internal Medicine. 2006;166(17):1822–1828. [PubMed: 17000937]
  48. Collins SR, Nuzum R, Rustgi SD, Mika S, Schoen C, Davis K. How health care reform can lower the costs of insurance administration. The Commonwealth Fund; 2009. [accessed August 30, 2010]. http://www​.commonwealthfund​.org/~/media/Files​/Publications/Issue%20Brief​/2009/Jul​/Admin%20Costs/1299​_Collins_how_hlt_care​_reform_can_lower_costs_ins_admin_v2​.pdf. [PubMed: 19618542]
  49. Commonwealth Fund. The Path to a High Performance US Health Care System: A 2020 Vision and the Policies to Pave the Way. 2009. [accessed August 26, 2009]. http://www​.commonwealthfund​.org/Content/Publications​/Fund-Reports​/2009/Feb/The-Path-to-a-High-Performance-US-Health-System.aspx.
  50. Commonwealth Fund Commission on a High Performance Health System. Why Not the Best? Results from the National Scorecard on US Health System Performance. New York: The Commonwealth Fund; 2008.
  51. Congressional Research Service. Does price transparency improve market efficiency? Implication of empirical evidence in other markets for the health sector. 2007. [accessed August 26, 2009]. http://www​.fas.org/sgp/crs/misc/RL34101​.pdf.
  52. Conover C. Health care regulation: A $169 billion hidden tax, policy analysis no 527. 2004. [accessed June 2009]. http://www​.cato.org/pubs/pas/pa527.pdf.
  53. Cromwell J, McCall N. Medicare participating heart bypass demonstration: Final report to the Health Care and Financing Administration. Waltham: Health Economics Research; 1998.
  54. Cromwell J, Dayhoff DA, Thoumaian AH. Cost savings and physician responses to global bundled payments for Medicare heart bypass surgery. Health Care Finance Review. 1997;19(1):41–57. [PMC free article: PMC4194487] [PubMed: 10180001]
  55. Currie J, MacLeod W. First do no harm? Tort reform and birth outcomes. Quarterly Journal of Economics. 2008;123(2):795–830.
  56. Cutler DM, DeLong JB. Obama can cure health care’s ills. 2008. [accessed August 26, 2009]. http://www​.forbes.com​/2008/10/27/obama-healthcare-reform-oped-cx​_dmc_bd_1028cutlerdelong.html.
  57. Danzon PM. Liability for Malpractice, Handbook of Health Economics. Culyer AJ, Newhouse JP, editors. Amsterdam: Elsevier; 2000. pp. 1339–1404.
  58. Danzon PM, Furukawa MF. International prices and availability of pharmaceuticals in 2005. Health Affairs (Millwood) 2008;27(1):221–233. [PubMed: 18180499]
  59. Darkins A. Care coordination/Home Telehealth. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  60. Darkins A, Ryan P, Kobb R, Foster L, Edmonson E, Wakefield B, Lancaster AE. Care coordination/home telehealth: The systematic implementation of health informatics, home telehealth, and disease management to support the care of veteran patients with chronic conditions. Telemedicine Journal and e-Health. 2008;14(10):1118–1126. [PubMed: 19119835]
  61. de Brantes F, Rastogi A, Gosfield A, Emery D, Machado E. Bundled and fee-for-episode payments: Lessons learned from the Prometheus payment model. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  62. Delaune J, Everett W. Waste and inefficiency in the U.S. health care system clinical care: A comprehensive analysis in support of system-wide improvements. New England Health Care Institute. 2008
  63. Deloitte, Touche LLP. Comparative effectiveness: Perspectives for consideration. 2009. [accessed August 26, 2009]. http://www​.nejm.org.
  64. Department of Health and Human Services. A comparison of prices for power wheelchairs in the Medicare program. 2004. [accessed April 2009]. http://www​.oig.hhs.gov​/oei/reports/oei-03-03-00460.pdf.
  65. Department of Health and Human Services. Medicare home oxygen equipment: Cost and servicing. 2006. [accessed June 11, 2009]. http://www​.oig.hhs.gov​/oei/reports/oei-09-04-00420.pdf.
  66. DeVol R, Bedroussain A, Charuworn A, Chatterjee K, Kim IK, Kim S. An unhealthy America: The economic burden of chronic disease charting a new course to save lives and increase productivity and economic growth. Santa Monica, CA: Milken Institute; 2007.
  67. Dopson S, Fitzgerald L. Knowledge to action: Evidence based health care in context. New York: Oxford University Press; 2005.
  68. Dranove D, Kessler D, McClellan M, Satterthwaite M. Is more information better? The effects of “report cards” on health care providers. Journal of Political Economics. 2003;111(31):555–588.
  69. Dubay L, Kaestner R, Waidmann T. The impact of malpractice fears on cesarean section rates. Journal of Health Economics. 1999;18(4):491–522. [PubMed: 10539619]
  70. EBM (Evidence-Based Medicine) Working Group. Evidence-based medicine: A new approach to teaching the practice of medicine. Journal of the American Medical Association. 1992;268(17):2420–2425. [PubMed: 1404801]
  71. Edmonds C, Hallman GL. Cardiovascular care providers: A pioneer in bundled services, shared risk, and single payment. Texas Heart Institute Journal. 1995;22(1):72–76. [PMC free article: PMC325213] [PubMed: 7787473]
  72. Eibner C, Hussey P, Ridgely M, McGlynn E. Controlling Health Care Spending in Massachusetts: An Analysis of Options. 2009. [accessed August 2009]. http://www​.rand.org/pubs​/technical_reports/2009/RAND_TR733​.pdf.
  73. Elmendorf DW. Personal communication: Letter to the Honorable Nathan Deal. 2009.
  74. Enthoven AC, Talbott B. Stanford University’s experience with managed competition. Health Affairs (Millwood) 2004;23(6):136–140. [PubMed: 15537591]
  75. Farrell D, Jensen E, Kocher B, Lovegrove N, Melhem F, Mendonca L. Accounting for the Cost of U.S. Healthcare. Washington, DC: McKinsey Global Institute; 2008.
  76. Feldman R. Antitrust policy in healthcare: Searching for a strategy that works. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  77. Ferris T. Medicare demonstration project. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  78. Fireman B, Bartlett J, Selby J. Can disease management reduce health care costs by improving quality? Health Affairs (Millwood) 2004;23(6):63–75. [PubMed: 15584100]
  79. Fischer MA, Avorn J. Economic implications of evidence-based prescribing for hypertension: Can better care cost less? Journal of the American Medical Association. 2004;291(15):1850–1856. [PubMed: 15100203]
  80. Fischer MA, Vogeli C, Stedman M, Ferris T, Brookhart MA, Weissman JS. Effect of electronic prescribing with formulary decision support on medication use and cost. Archives of Internal Medicine. 2008;168(22):2433–2439. [PubMed: 19064827]
  81. Fisher E, Bronner K. Potential savings: Insights from regional benchmarks. The Healthcare Imperative Lowering Costs and Improving Outcomes Workshop; Washington, DC. 2009.
  82. Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL. The implications of regional variations in Medicare spending. Part 1: The content, quality, and accessibility of care. Annals of Internal Medicine. 2003;138(4):273–287. [PubMed: 12585825]
  83. Fisher ES, McClellan MB, Bertko J, Lieberman SM, Lee JJ, Lewis JL, Skinner JS. Fostering accountable health care: Moving forward in Medicare. Health Affairs (Millwood) 2009;28(2):w219–w231. [PMC free article: PMC2656392] [PubMed: 19174383]
  84. Flottemesch T. Costs from missed cost-saving primary and secondary prevention opportunities. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  85. Frank RG, Zeckhauser RJ. Health insurance exchanges—Making the markets work. New England Journal of Medicine. 2009;361(12):1135–1137. [PubMed: 19625711]
  86. FTC (Federal Trade Commission) Order Dismissing Complaint, Docket #9326. 2008. [accessed August 12, 2009]. http://www​.ftc.gov/os​/adjpro/d9326/index.shtm.
  87. Gabel JR. Congress’s health care numbers don’t add up. New York Times. Aug 26, 2009. p. A23.
  88. GAO (Government Accountability Office) Medicare: Focus on Physician Practice Patterns Can Lead to Greater Program Efficiency: Report to Congressional Committees. Washington: U.S. Government Accountability Office; 2007.
  89. Garber K. Democrats avoid tort reform in healthcare debate. 2009. [accessed August 26, 2009]. http://www​.usnews.com​/articles/news/politics​/2009/08/27/democrats-avoid-tort-reform-in-healthcare-debate.html.
  90. Gazelle G. Transparency in comparative value of treatment options. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  91. Gellad WF, Schneeweiss S, Brawarsky P, Lipsitz S, Haas JS. What if the federal government negotiated pharmaceutical prices for seniors? An estimate of national savings. Journal of General Internal Medicine. 2008;23(9):1435–1440. [PMC free article: PMC2517993] [PubMed: 18581187]
  92. Genetic Pharmaceutical Association. Economic analysis of generic pharmaceuticals 1999–2008: $734 billion in health care savings. Arlington: 2009.
  93. Ginsburg PB. Shopping for price in medical care. Health Affairs (Millwood) 2007;26(2):w208–w216. [PubMed: 17284467]
  94. Ginsburg P. Provider price and quality transparency. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  95. Girosi F, Meili R, Scoville R. Extrapolating evidence of health information technology savings and costs. Santa Monica, CA: RAND Corporation; 2005.
  96. Goetzel RZ, Ozminkowski RJ, Villagra VG, Duffy J. Return on investment in disease management: A review. Health Care Finance Review. 2005;26(4):1–19. [PMC free article: PMC4194913] [PubMed: 17288065]
  97. Goldsmith B, Dietrich J, Du Q, Morrison RS. Variability in access to hospital palliative care in the United States. Journal of Palliative Medicine. 2008;11(8):1094–1102. [PMC free article: PMC2941669] [PubMed: 18831653]
  98. Grabowski D, Ohsfeldt R, Morrisey M. The Effects of CON repeal on Medicaid nursing home and long term care expenditures. Inquiry. 2003;40(2):146–157. [PubMed: 13677562]
  99. Hafer M. Personal communication: Simpler healthcare estimates. Washington, DC: Sep 23, 2009.
  100. Hartman M, Martin A, McDonnell P, Catlin A. National health spending in 2007: Slower drug spending contributes to lowest rate of overall growth since 1998. Health Affairs (Millwood) 2009;28(1):246–261. [PubMed: 19124877]
  101. Hatem M, Sandall J, Devane D, Soltani H, Gates S. Midwife-led versus other models of care for childbearing women. Cochrane Database of Systematic Reviews. 2008;(4):CD004667. [PubMed: 18843666]
  102. Heffernan J, Blanchfield B, Osgood B, Sheehan R, Meyer G. Simplifying the billing process: Still an opportunity for significant cost efficiency. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  103. Hellinger FJ, Encinosa WE. The impact of state laws limiting malpractice damage awards on health care expenditures. American Journal of Public Health. 2006;96(8):1375–1381. [PMC free article: PMC1522105] [PubMed: 16809580]
  104. Hensley S. From “one size fits all” to tailored co-payments (Online) 2004. [accessed August 26, 2009]. http://online​.wsj.com​/article/SB108732170963037581.html.
  105. Hillestad R, Bigelow J, Bower A, Girosi F, Meili R, Scoville R, Taylor R. Can electronic medical record systems transform health care? Potential health benefits, savings, and costs. Health Affairs (Millwood) 2005;24(5):1103–1117. [PubMed: 16162551]
  106. Himmelstein DU, Woolhandler S. Hope and hype: Predicting the impact of electronic medical records. Health Affairs (Millwood) 2005;24(5):1121–1123. [PubMed: 16162553]
  107. Ho V. Does certificate of need affect cardiac outcomes and costs? The James A. Baker III Institute for Public Policy. 2007. [PubMed: 17340175]
  108. Hoadley J. How can we tell if prescription drug prices are too high?. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  109. Hoerger T. Prices that may be too high: Durable medical equipment. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  110. Hollingsworth B. Non-parametric and parametric applications measuring efficiency in health care. Health Care Management Science. 2003;6(4):203–218. [PubMed: 14686627]
  111. Hooker RS. A cost analysis of physician assistants in primary care. Journal of the American Academy of Physicians Assistants. 2002;15(11):39–42. 45, 48 passim. [PubMed: 12474431]
  112. Hwang J. Decentralizing health care delivery by using community workers. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  113. IBM Global Business Services. CAQH CORE Phase 1 Measures of Success Study. 2009. [accessed 2009]. http://www​.caqh.org/COREIBMstudy.php.
  114. IOM (Institute of Medicine) To Err Is Human: Building a Safer Health Care System. Washington, DC: National Academy Press; 2000.
  115. IOM (Institute of Medicine) Crossing the Quality chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press; 2001.
  116. Jack BW, Chetty VK, Anthony D, Greenwald JL, Sanchez GM, Johnson AE, Forsythe SR, O’Donnell JK, Paasche-Orlow MK, Manasseh C, Martin S, Culpepper L. A reengineered hospital discharge program to decrease rehospitalization: A randomized trial. Annals of Internal Medicine. 2009;150(3):178–187. [PMC free article: PMC2738592] [PubMed: 19189907]
  117. Jack EP, Powers TL. Volume flexible strategies in health services: A research framework. Production and Operations Management. 2004;13(3):230–244.
  118. Jensen E. How does U.S. spending on health care compare to international benchmarks? What are areas of specific excess spending and opportunities for cost savings?. The Healthcare Imperative Lowering Costs and Improving Outcomes Workshop; Washington, DC. 2009.
  119. Jessee W. Avoidable costs from delivery inefficiencies: Physician offices. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  120. Jha AK. Costs from mistakes. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  121. Jha AK, Chan DC, Ridgway AB, Franz C, Bates DW. Improving safety and eliminating redundant tests: Cutting costs in U.S. Hospitals. Health Affairs (Millwood) 2009;28(5):1475–1484. [PubMed: 19738266]
  122. Johnson B, Wheeler L, Deuser J, Sousa K. Outcomes of the Kaiser Permanante Tele-Home Health research project. Archives of Family Medicine. 2000;9(1):40–45. [PubMed: 10664641]
  123. Johnson LL, Becker RL. An alternative health-care reimbursement system—Application of arthroscopy and financial warranty: Results of a 2-year pilot study. Arthroscopy. 1994;10(4):462–470. discussion 471–462. [PubMed: 7945644]
  124. Joint Commission Resources. Managing patient flow: Strategies and solutions. 2nd edition. Oakbrook Terrace: Joint Commission International; 2009.
  125. Kahn J. Billing and insurance-related administrative costs: Burden to health care providers. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  126. Kaiser Family Foundation. How changes in medical technology affect health care costs. 2007. [accessed August 26, 2009]. http://www​.kff.org/insurance​/snapshot/chcm030807oth.cfm.
  127. Kaplan G. Better and faster health care is more affordable. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  128. Kaushal R. The value of EHRs. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  129. Kessler D, McClellan M. Do doctors practice defensive medicine? Quarterly Journal of Economics. 1996;111(2):353–390.
  130. Klein S, McCarthy D. Gundersen Lutheran Health System: Performance Improvement through Partnership Case Study: Organized Health Care Delivery System. 2009. [accessed 2009]. http://www​.commonwealthfund​.org/~/media/Files​/Publications/Case%20Study​/2009/Aug​/1307_McCarthy_Gundersen​_Lutheran_case_study_v2.pdf.
  131. Krishnan R. Market restructuring and pricing in the hospital industry. Journal of Health Economics. 2001;20(2):213–237. [PubMed: 11252371]
  132. Krishnan R, Krishnan H. Effects of hospital mergers and acquisitions on prices. Journal of Business Research. 2003;56:647–656.
  133. Krugman P. Administrative Costs (Online) 2009. [accessed August 26, 2009]. http://krugman​.blogs​.nytimes.com/2009/07​/06/administrative-costs/
  134. Kuehn BM. States explore shared decision making. Journal of the American Medical Association. 2009;301(24):2539–2541. [PubMed: 19549964]
  135. Laurant M, Reeves D, Hermens R, Braspenning J, Grol R, Sibbald B. Substitution of doctors by nurses in primary care. Cochrane Database of Systematic Reviews. 2005;(2) [PubMed: 15846614]
  136. Leatherman S, Warrick L. Effectiveness of decision aids: A review of the evidence. Medical Care Research Review. 2008;65(6 Suppl):79S–116S. [PubMed: 19015379]
  137. Lerner J. Medical device pricing: Getting paid beyond tsk, tsk!. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  138. Levi J. Community prevention and health care costs. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  139. Lindenauer P. Transparency as a strategy to improve the value of hospital care. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  140. Ling LX, Benton W, Leong GK. The impact of strategic management decisions on community hospital performance. Journal of Operations Management. 2002;20(4):389–408.
  141. Litvak E. Optimizing patient flow by managing its variability, Front office to front line: Essential issues for health care leaders. Vol. 5. Oakbrook Terrace: Joint Commision Resources; 2005. pp. 91–111.
  142. Litvak E, Long MC. Cost and quality under managed care: Irreconcilable differences? American Journal of Managed Care. 2000;6(3):305–312. [PubMed: 10977431]
  143. Lorig KR, Sobel DS, Ritter PL, Laurent D, Hobbs M. Effect of a self-management program on patients with chronic disease. Effective Clinical Practice. 2001;4(6):256–262. [PubMed: 11769298]
  144. Lyn M, Silberberg M, Michener JL. Caregiver profile, efficiency, and team care. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  145. Mankiw G. Medicare has lower administrative costs? 2009. [accessed August 29, 2009]. http://gregmankiw​.blogspot​.com/2009/07/does-medicare-have-lower-administrative​.html.
  146. Massachusetts Technology Collaborative & New England Healthcare Institute. Saving lives, saving money in practice: Strategies for computerized physician order entry in Massachusetts hospitals. 2009. [accessed August 26, 2009]. http://web3​.streamhoster​.com/mtc/cpoe2009.pdf.
  147. Mattke S, Seid M, Ma S. Evidence for the effect of disease management: Is $1 billion a year a good investment? American Journal of Managed Care. 2007;13(12):670–676. [PubMed: 18069910]
  148. Mechanic RE, Altman SH. Payment reform options: Episode payment is a good place to start. Health Affairs (Millwood) 2009;28(2):w262–w271. [PubMed: 19174388]
  149. Mecklenburg R, Kaplan G. Excess costs attributable to inefficient use of caregivers. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  150. MedPAC (Medicare Payment Advisory Commission) Report to Congress: Promoting Greater Efficiency in Medicare. 2007. [accessed June 2009]. http://www​.medpac.gov​/documents/jun07_EntireReport.pdf.
  151. MedPAC (Medicare Payment Advisory Commission) A Data book: Healthcare Spending and the Medicare Program. 2008. [accessed 2009]. www​.medpac.gov/documents​/Jun08DataBook_Entire_report.pdf.
  152. MedPAC (Medicare Payment Advisory Commission) Report to Congress. 2009. [accessed September 2009]. http://www​.medpac.gov​/documents/Mar09_EntireReport.pdf.
  153. Mehrotra A, Liu H, Adams JL, Wang MC, Lave JR, Thygeson NM, Solberg LI, McGlynn EA. Comparing costs and quality of care at retail clinics with that of other medical settings for 3 common illnesses. Annals of Internal Medicine. 2009;151(5):321–328. [PMC free article: PMC2805258] [PubMed: 19721020]
  154. Meier D. Palliative care improves quality, reduces costs. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  155. Menduno M. Apothecary.Now. Hospital Health Network. 1999;73(7):34–38. 40, 32. [PubMed: 10451637]
  156. Mercer. CCNC/ACCESS Cost Savings, Analysis for State Fiscal Years 2003, 2004, 2005 and 2006. 2007. [accessed August 26, 2009]. http:/www​.communitycarenc.com.
  157. Merenstein D, Daumit GL, Powe NR. Use and costs of nonrecommended tests during routine preventive health exams. American Journal of Preventive Medicine. 2006;30(6):521–527. [PubMed: 16704947]
  158. MGMA (Medical Group Management Association) MGMA. 2004. [accessed August 26, 2009]. http://www​.medpac.gov​/documents/Jun08_EntireReport.pdf.
  159. Miller H. Using value-based payment to improve outcomes and control costs. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  160. Miller ME. Report to Congress: Reforming the Delivery System. Washington, DC: Medicare Payment Advisory Commission; 2008.
  161. Milstein A. How much lower are risk-adjusted costs per case in low cost hospitals with high quality scores. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  162. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States, 2000. Journal of the American Medical Association. 2004;291(10):1238–1245. [PubMed: 15010446]
  163. Moracco N. Personal communication: Community dialogue on value based benefit design. Washington, DC: Sep 10, 2009.
  164. Morrison RS, Meier DE. Clinical practice. Palliative care. New England Journal of Medicine. 2004;350(25):2582–2590. [PubMed: 15201415]
  165. Morrison RS, Penrod JD, Cassel JB, Caust-Ellenbogen M, Litke A, Spragens L, Meier DE. Cost savings associated with U.S. hospital palliative care consultation programs. Archives of Internal Medicine. 2008;168(16):1783–1790. [PubMed: 18779466]
  166. National Center for Health Statistics. About Healthy People 2010. 2009. [accessed August 26, 2009]. http://www​.cdc.gov/nchs​/about/otheract/hpdata2010/abouthp​.htm.
  167. Naugle A. Reducing payer business development expense through healthcare reform. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  168. Naylor MD, Brooten DA, Campbell RL, Maislin G, McCauley KM, Schwartz JS. Transitional care of older adults hospitalized with heart failure: A randomized, controlled trial. Journal of the American Geriatric Society. 2004;52(5):675–684. [PubMed: 15086645]
  169. O’Connor AM, Llewellyn-Thomas HA, Flood AB. Modifying unwarranted variations in health care: Shared decision making using patient decision aids. Health Affairs (Millwood) 2004;Suppl Web Exclusives:VAR63–72. [PubMed: 15471770]
  170. O’Connor AM, Rostom A, Fiset V, Tetroe J, Entwistle V, Llewellyn-Thomas H, Holmes-Rovner M, Barry M, Jones J. Decision aids for patients facing health treatment or screening decisions: Systematic review. British Medical Journal. 1999;319(7212):731–734. [PMC free article: PMC28223] [PubMed: 10487995]
  171. O’Connor AM, Stacey D, Entwistle V, Llewellyn-Thomas H, Rovner D, Holmes-Rovner M, Tait V, Tetroe J, Fiset V, Barry M, Jones J. Decision aids for people facing health treatment or screening decisions. Cochrane Database of Systematic Reviews. 2003;(2):CD001431. [PubMed: 12804407]
  172. O’Kane M. Transparency: Necessary but not sufficient. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  173. Owens M. Identifying and quantifying the cost of uncoordinated care: Opportunities for savings and improved outcomes. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  174. Pan E, Cusack C, Hook J, Vincent A, Kaelber DC, Bates DW, Middleton B. The value of provider-to-provider telehealth. Telemedicine Journal and e-Health. 2008;14(5):446–453. [PubMed: 18578679]
  175. Paris V, Docteur E. Pharmaceutical pricing and reimbursement policies in Canada. Paris: Organisation for Economic Co-operation and Development; 2006.
  176. Parks KM, Steelman LA. Organizational wellness programs: A meta-analysis. Journal of Occupational Health Psychology. 2008;13(1):58–68. [PubMed: 18211169]
  177. Pauly MV, Burns LR. Price transparency for medical devices. Health Affairs (Millwood) 2008;27(6):1544–1553. [PubMed: 18997210]
  178. Peikes D, Chen A, Schore J, Brown R. Effects of care coordination on hospitalization, quality of care, and health care expenditures among Medicare beneficiaries: 15 randomized trials. Journal of the American Medical Association. 2009;301(6):603–618. [PubMed: 19211468]
  179. Pelletier KR. A review and analysis of the clinical and cost-effectiveness studies of comprehensive health promotion and disease management programs at the worksite: Update vi 2000–2004. Journal of Occupational Environmental Medicine. 2005;47(10):1051–1058. [PubMed: 16217246]
  180. Pignone M. Treatment costs from missed tertiary cost-saving prevention opportunities. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  181. Pittenger K. Using production system methods in medical practice: Improving medical costs and outcomes. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  182. Polisena J, Coyle D, Coyle K, McGill S. Home telehealth for chronic disease management: A systematic review and an analysis of economic evaluations. International Journal of Technology Assessment Health Care. 2009;25(3):339–349. [PubMed: 19619353]
  183. PriceWaterhouseCoopers. The Price of Excess: Identifying Waste in Healthcare Spending. 2009. [accessed September 20, 2009]. http://www​.pwc.com/us​/en/healthcare/publications​/the-price-of-excess.html.
  184. Purvis L. Rx watchdog report: Trends in manufacturer prices of prescription drugs used by medicare beneficiaries: 2008 Year-end update. Washington, DC: AARP Public Policy Institute; 2009.
  185. Rastogi A. Lessons learned from the Prometheus payment model. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  186. Riemer D. Effective health insurance exchanges: The Dane County, Wisconsin model. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  187. Roblin D, Howard D, Becker E, Adams E, Roberts M. Use of midlevel practitioners to achieve labor cost savings in the primary care practice of an MCO. Health Services Research. 2004;39(3):607–626. [PMC free article: PMC1361027] [PubMed: 15149481]
  188. Romer C. The Economic Case for Health Care Reform. Executive Office of the President Council of Economic Advisers; 2009.
  189. Rosen AB, Hamel MB, Weinstein MC, Cutler DM, Fendrick AM, Vijan S. Cost-effectiveness of full Medicare coverage of angiotensin-converting enzyme inhibitors for beneficiaries with diabetes. Annals of Internal Medicine. 2005;143(2):89–99. [PubMed: 16027450]
  190. Rosko MD, Mutter RL. Stochastic frontier analysis of hospital inefficiency: A review of empirical issues and an assessment of robustness. Medical Care Research and Review. 2008;65(2):131–166. [PubMed: 18045984]
  191. Russell LB. Preventing chronic disease: An important investment, but don’t count on cost savings. Health Affairs (Millwood) 2009;28(1):42–45. [PubMed: 19124852]
  192. Russo M. The $3 trillion question: What health care reform can save for families, businesses and taxpayers. United States Public Interest Research Groups; 2009.
  193. Ryan P, Kobb R, Hilsen P, Nedd N. Opening a window of opportunity through technology and coordination: A multisite case study. Telemedicine Journal and e-Health. 2003;9(3):265–271. [PubMed: 14611694]
  194. Sakowski JA, Kahn JG, Kronick RG, Newman JM, Luft HS. Peering into the black box: Billing and insurance activities in a medical group. Health Affairs (Millwood) 2009;28(4):w544–w554. [PubMed: 19443478]
  195. Salkever DS. Regulation of prices and investment in hospitals in the U.S. Handbook of Health Economics. 2000;1B:1489–1535.
  196. Santa J. Price transparency in cost care. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  197. Savitz L. The intermountain way to positively impact costs and quality: the EB-CPM for managing febrile infants. The Healthcare Imperative, Lowering Costs and Improving Outcomes Workshop; Washington, DC. 2009.
  198. Schoen C, Guterman S, Shih A, Lau J, Kasimow S, Gauthier A.2007Bending the Curve: Options for the Commonwealth Fund Commission on a High Performance Health System http://www​.commonwealthfund​.org/~/media/Files​/Publications/Fund%20Report​/2007/Dec​/Bending%20the%20Curve​%20%20Options%20for​%20Achieving%20Savings​%20and%20Improving​%20Value%20in%20U%20S​%20%20Health%20Spending​/Schoen_bendingthecurve_1080​%20pdf.pdfaccessed September 19, 2009.
  199. Scott M. Health care in the express lane: The emergence of retail clinics. Oakland: California HealthCare Foundation; 2006.
  200. Scott R. The direct medical costs of healthcare-associated infections in US Hospitals and the benefits of prevention. 2009. [accessed August 26, 2009]. http://www​.cdc.gov/ncidod​/dhqp/pdf/Scott_CostPaper.pdf.
  201. Sherlock D. Administrative expenses of health plans. 2009. [accessed August 26, 2009]. http://www​.bcbs.com/issues​/uninsured/Sherlock-Report-FINAL​.pdf.
  202. Sidorov J. It ain’t necessarily so: The electronic health record and the unlikely prospect of reducing health care costs. Health Affairs (Millwood) 2006;25(4):1079–1085. [PubMed: 16835189]
  203. Sirovich B, Gallagher PM, Wennberg DE, Fisher ES. Discretionary decision making by primary care physicians and the cost of U.S. health care. Health Affairs (Millwood) 2008;27(3):813–823. [PMC free article: PMC2438037] [PubMed: 18474975]
  204. Sisko A, Truffer C, Smith S, Keehan S, Cylus J, Poisal JA, Clemens MK, Lizonitz J. Health spending projections through 2018: Recession effects add uncertainty to the outlook. Health Affairs (Millwood) 2009;28(2):w346–w357. [PubMed: 19240055]
  205. Siu AL, Spragens LH, Inouye SK, Morrison RS, Leff B. The ironic business case for chronic care in the acute care setting. Health Affairs (Millwood) 2009;28(1):113–125. [PubMed: 19124861]
  206. Sloan F. Reducing healthcare spending by reducing service capacity: evidence and policy options. The Healthcare Imperative, Lowering Costs and Improving Outcomes Workshop; Washington, DC. 2009.
  207. Sloan FA, Shadle JH. Is there empirical evidence for “defensive medicine”? A reassessment. Journal of Health Economics. 2009;28(2):481–491. [PubMed: 19201500]
  208. Sloan F, Entman S, Reilly B, Glass C, Hickson G, Zhang H. Tort liability and obstetricians’ care levels. International Review of Law and Economics. 1997;17(2):245–260.
  209. Smedley B, Stith A. Unequal Treatment: Confronting Racial and Ethnic Disparities in Care. Washington, DC: The National Academies Press; 2003.
  210. Smith P. Enhancing clinical data as a knowledge utility. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  211. Smith TJ, Coyne P, Cassel B, Penberthy L, Hopson A, Hager MA. A high-volume specialist palliative care unit and team may reduce in-hospital end-of-life care costs. Journal of Palliative Medicine. 2003;6(5):699–705. [PubMed: 14622449]
  212. Sorrel AL. AMA analysis reaffirms: Tort reforms work. 2008. [accessed August 26, 2009]. http://www​.amaassn.org​/amednews/2008/03/03/prsa0303.htm.
  213. Stone PW, Chapman RH, Sandberg EA, Liljas B, Neumann PJ. Measuring costs in cost-utility analyses. Variations in the literature. International Journal of Technology Assessment Health Care. 2000;16(1):111–124. [PubMed: 10815358]
  214. Strunk BC, Devers KJ, Hurley RE. Health plan-provider showdowns on the rise. Issue Brief Center for Studying Health System Change. 2001;(40):1–4. [PubMed: 11603408]
  215. Stuart ME, Macuiba J, Heidrich F, Farrell RG, Braddick M, Etchison S. Successful implementation of an evidence-based clinical practice guideline: Acute dysuria/ urgency in adult women. HMO Practice. 1997;11(4):150–157. [PubMed: 10176516]
  216. TFAH (Trust for America’s Health) Prevention for a Healthy America. 2008. [accessed August 26, 2009]. http:​//healthyamericans​.org/reports/prevention08/
  217. Thomas EJ, Studdert DM, Newhouse JP, Zbar BI, Howard KM, Williams EJ, Brennan TA. Costs of medical injuries in Utah and Colorado. Inquiry. 1999;36(3):255–264. [PubMed: 10570659]
  218. Thomashauer R. Industry collaboration: Simplifying administrative complexity. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  219. Thorpe K. Community health teams and transitional care strategies. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  220. Thorpe KE, Howard DH. The rise in spending among Medicare beneficiaries: The role of chronic disease prevalence and changes in treatment intensity. Health Affairs (Millwood) 2006;25(5):w378–w388. [PubMed: 16926180]
  221. Thygeson M. Retail clinics-cost saving or just convenient. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  222. Toussaint J. Personal communication: ThedaCare estimates. Washington, DC: Sep, 2009. 2009.
  223. Town RJ, Wholey DR, Feldman RD, Burns LR. Hospital consolidation and racial/income disparities in health insurance coverage. Health Affairs (Millwood) 2007;26(4):1170–1180. [PubMed: 17630461]
  224. Tu HT, Ginsburg PB. Losing ground: Physician income, 1995–2003. Track Reports. 2006;(15):1–8. [PubMed: 16791996]
  225. UnitedHealth Group. Federal Health Care Cost Containment—How in Practice Can It Be Done? Working Paper #1. Minneapolis: UnitedHealth Group; 2009.
  226. UnitedHealth Group. Health Care Cost and Containment—How Technology Can Cut Red Tape and Simplify Health Care Administration-Working Paper #2. Minneapolis: United Health Care for Reform and Modernization; 2009.
  227. U.S. House of Representatives Committee on Oversight and Government Reform. Medicare Part D: Drug pricing and manufacturer windfalls. 2008. [accessed September 10, 2009]. http://oversight​.house​.gov/documents/20080724101850.pdf.
  228. Vaswani S. Managing variability in healthcare delivery. The Healthcare Imperative, Lowering Costs and Improving Outcomes Workshop; Washington, DC. 2009.
  229. Vo AH. The Telehealth Promise: Better Health Care and Cost Savings for the 21st Century. UTMB. 2008
  230. Vogeli C, Shields AE, Lee TA, Gibson TB, Marder WD, Weiss KB, Blumenthal D. Multiple chronic conditions: Prevalence, health consequences, and implications for quality, care management, and costs. Journal of General Internal Medicine. 2007;22(Suppl 3):391–395. [PMC free article: PMC2150598] [PubMed: 18026807]
  231. Vogt W, Town R. How Has Hospital Consolidation Affected the Price and Quality of Hospital Care? Princeton: Robert Wood Johnson Foundation; 2006. [PubMed: 22051574]
  232. Wachter RM, Pronovost PJ. The 100,000 lives campaign: A scientific and policy review. Joint Commission Journal of Quality and Paitent Safety. 2006;32(11):621–627. [PubMed: 17120921]
  233. Wagner EH, Sandhu N, Newton KM, McCulloch DK, Ramsey SD, Grothaus LC. Effect of improved glycemic control on health care costs and utilization. Journal of the American Medical Association. 2001;285(2):182–189. [PubMed: 11176811]
  234. Walker J, Pan E, Johnston D, Adler-Milstein J, Bates DW, Middleton B. The value of health care information exchange and interoperability. Health Affairs (Millwood) 2005;Suppl Web Exclusives:W5-10-W15–18. [PubMed: 15659453]
  235. Wang MC, Ryan G, McGlynn EAMA. Why do patients seek care at retail clinics and what alternatives did they consider. American Journal of Managed Care. In press. [PMC free article: PMC3303882] [PubMed: 20142442]
  236. Wang SJ, Middleton B, Prosser LA, Bardon CG, Spurr CD, Carchidi PJ, Kittler AF, Goldszer RC, Fairchild DG, Sussman AJ, Kuperman GJ, Bates DW. A cost-benefit analysis of electronic medical records in primary care. American Journal of Medicine. 2003;114(5):397–403. [PubMed: 12714130]
  237. Wennberg D. The deep dive. DMAA Annual Meeting; September 17; Las Vegas, Nevada. 2007.
  238. Wennberg JE, Fisher ES, Skinner JS. Geography and the debate over Medicare reform. Health Affairs (Millwood) 2002;Supp Web Exclusives:W96–114. [PubMed: 12703563]
  239. Wennberg JE, Bronner K, Skinner JS, Fisher ES, Goodman DC. In-patient care intensity and patients’ ratings of their hospital experiences. Health Affairs (Millwood) 2009;28(1):103–112. [PMC free article: PMC2729566] [PubMed: 19124860]
  240. Werner RM, Asch DA. The unintended consequences of publicly reporting quality information. Journal of the American Medical Association. 2005;293(10):1239–1244. [PubMed: 15755946]
  241. Wichmann D. Health care cost containment. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; July 16–18; Washington, DC. 2009.
  242. Wilson JF. Can disease prevention save health reform? Annals of Internal Medicine. 2009;151(2):145–148. [PubMed: 19620172]
  243. Winslow CM, Solomon DH, Chassin MR, Kosecoff J, Merrick NJ, Brook RH. The appropriateness of carotid endarterectomy. New England Journal of Medicine. 1988;318(12):721–727. [PubMed: 3279313]
  244. Woolf S. The price paid for not preventing diseases. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  245. Woolhandler S, Campbell T, Himmelstein DU. Costs of health care administration in the United States and Canada. New England Journal of Medicine. 2003;349(8):768–775. [PubMed: 12930930]
  246. Wynn M. Pricing of durable medical equipment: Lessons from bidding projects. The Healthcare Imperative: Lowering Costs and Improving Outcomes Workshop; May 21–22; Washington, DC. 2009.
  247. Yaggy SD, Michener JL, Yaggy D, Champagne MT, Silberberg M, Lyn M, Johnson F, Yarnall KS. Just for us: An academic medical center-community partnership to maintain the health of a frail low-income senior population. Gerontologist. 2006;46(2):271–276. [PubMed: 16581892]
  248. Yasaitis L, Fisher ES, Skinner JS, Chandra A. Hospital quality and intensity of spending: Is there an association? Health Affairs (Millwood) 2009;28(4):w566–w572. [PMC free article: PMC2768577] [PubMed: 19460774]
  249. Zhan C, Miller MR. Excess length of stay, charges, and mortality attributable to medical injuries during hospitalization. Journal of the American Medical Association. 2003;290(14):1868–1874. [PubMed: 14532315]

Footnotes

1

Monopsony is defined as the market condition that exists when there is one buyer.

2

The 12 specialty categories include cardiology, cardiothoracic surgery, gastroenterology, general surgery, obstetrics and gynecology, orthopedics, otolaryngology, neurology, neurosurgery, plastic surgery, urology, and vascular surgery.

Copyright © 2010, National Academy of Sciences.
Bookshelf ID: NBK53915

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (5.4M)

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...