• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Circ Cardiovasc Qual Outcomes. Author manuscript; available in PMC Jan 1, 2012.
Published in final edited form as:
PMCID: PMC3057895

Lifetime Costs of Medical Care after Heart Failure Diagnosis

Shannon M. Dunlay, MD, MSc,* Nilay D. Shah, PhD, Qian Shi, PhD, Bruce Morlan, MS, Holly VanHouten, BA, Kirsten Hall Long, PhD, and Véronique L. Roger, MD, MPH*



Heart failure (HF) care constitutes an increasing economic burden on the healthcare system, and has become a key focus in the healthcare debate. However, there are limited data on the lifetime healthcare costs for individuals with HF after initial diagnosis.

Methods and Results

Olmsted County residents with incident HF from 1987–2006 were identified. Direct medical costs incurred from the time of HF diagnosis until death or last follow-up were obtained using population-based administrative data through 2007. Costs were inflated to 2008 U.S. dollars using the general Consumer Price Index (CPI). Inpatient, outpatient, and total costs were estimated using a two-part model with adjustment for right censoring of data. Predictors of total costs were examined using a similar model. A total of 1054 incident HF patients were identified (mean age 76.8 years, 46.1% male). After a mean follow-up of 4.6 years, 765 (72.6%) patients had died. The estimated total lifetime costs were $109,541 (95% CI $100,335–118,946) per person, with the majority accumulated during hospitalizations (mean $83,980 per person). After adjustment for age, year of diagnosis, and comorbidity, diabetes mellitus and preserved ejection fraction (≥50%) were associated with 24.8% (p=0.003) and 23.6% (p=0.041) higher lifetime costs, respectively. Higher costs were observed at initial HF diagnosis and in the months immediately before death in those surviving >12 months after diagnosis.


HF imposes a significant economic burden, primarily related to hospitalizations. Variations in cost over a lifetime can help identify strategies for efficient management of patients, particularly at the end of life.

Keywords: community, cost, heart failure, epidemiology, health services research


Heart failure (HF) imposes a staggering economic burden on the healthcare system. In developed countries, the cost of HF care constitutes 1–2% of overall healthcare spending1. In the United States, HF consumes more Medicare dollars than any other diagnosis2, and total costs for HF care were estimated at $34.8 billion dollars in 20083. HF costs have risen over time, and may continue to rise due to an aging population and rise in the prevalence of HF due to improved survival after diagnosis4. However, little is known about lifetime medical costs among individual patients with HF and how these costs are accrued from the time of diagnosis until death.

Published cost studies in HF have primarily included hospitalized patients or trial participants, have used prevalent HF cases that were not validated, or followed patients for a very brief period of time. To the best of our knowledge, there have been no studies examining costs among incident HF patients, as these data are often difficult to obtain. As HF is a disease characterized by periodic exacerbations5, it is likely that accrual of costs varies over the course of a lifetime. Further, while studies have estimated that hospitalization costs are responsible for 65–70% of total HF costs1, 6, far less is known about the costs associated with outpatient care. Finally, little is known about clinical characteristics associated with higher or lower lifetime costs. These data would be particularly informative to guide interventions and public health approaches to lower healthcare related costs in HF.

HF imposes a major economic burden in the U.S., but there is insufficient data on the lifetime costs of HF care from the time of initial diagnosis, the temporal distribution of costs and resource use, and the patient characteristics predictive of high or low costs. Accordingly, we aimed to examine these questions by evaluating the cumulative direct medical costs of care from the time of HF diagnosis among a community HF cohort.


Study Design

This is a cohort study conducted in Olmsted County, Minnesota. The population in the county was estimated at 137,521 according to the 2005 U.S. Census, 50% of whom are female, and 90% are Caucasian. Population-based research is possible in Olmsted County as the county is relatively isolated from other urban centers, and there are few providers, the largest of which is the Mayo Clinic. Medical records from all sources of care for county residents are extensively indexed and linked via the Rochester Epidemiology Project, a centralized system. Billing data for outpatient visit, hospitalization, and surgery, have been collected and organized through the Olmsted County Healthcare Expenditure and Utilization Database (OCHEUD) for each person living in the county7. This extensive framework allows patients to be followed passively using their medical record information and billing data, provided they have provided Minnesota Research Authorization when moving into the county. Historically, more than 97% of persons have given Minnesota Research Authorization. Thus, the population-based data presented herein are representative of the Olmsted County population.

Patient Identification

Olmsted County residents with a potential HF diagnosis from 1987–2006 were identified by International Classification of Diseases, Ninth Revision (ICD9) code 428 (HF). Codes are assigned based on physician diagnoses during outpatient visits or at hospital discharge. From all patients with ICD9 code 428, a random subset were selected to undergo case validation and data abstraction. The index date for HF was defined as the first evidence of HF in the medical record. Patients with a first diagnosis of HF prior to the study period or before moving to Olmsted County were excluded. Cases were validated using methods previously described4. Experienced nurse abstractors reviewed records to ensure each met Framingham criteria and had a physician’s diagnosis of HF present. When this method was utilized previously, the inter-abstractor agreement was 100%4, indicating these methods of classification are highly reproducible. In total, 80% of ICD9 code 428’s met criteria for Framingham heart failure, and 90% had a physician’s diagnosis of heart failure present. During the design of the cohort, other ICD9 codes for HF were investigated but were found to be very low yield so were not included in the study.

Patient Baseline Characteristics

Baseline patient characteristics were abstracted from the medical record. Physician’s diagnosis was used to define hyperlipidemia, chronic obstructive pulmonary disease (COPD), and peripheral vascular disease. Cerebrovascular disease was defined by physician’s diagnosis. A random sample of 50 patients with diagnoses of cerebrovascular disease were investigated with manual abstraction to determine the nature of the cerebrovascular disease. Smoking status was classified as ‘current’, ‘prior’, or ‘never’. Hypertension was defined by a physician diagnosis of hypertension in the medical record or systolic blood pressure >140 mm Hg or diastolic blood pressure >90 mm Hg. Diabetes mellitus was defined by fasting blood glucose levels or use of insulin and/or oral hypoglycemic medications. Myocardial infarction (MI) was defined using standard epidemiological criteria8. Body mass index (BMI) was calculated using the weight and height at HF diagnosis. Ejection fraction (EF) closest to the time of HF diagnosis (and within one year) was collected from available imaging. Laboratory values closest to the time of HF diagnosis (and within 1 year in all cases) including hemoglobin and creatinine were obtained from the medical record. Anemia was defined as hemoglobin <12 mg/dL in women and <13 mg/dL in men. Creatinine clearance was estimated using the Modification of Diet in Renal Disease equation9.

Patient-Level Cost Data

Each time a patient presents for a medical encounter billing data is generated, reflecting the direct medical costs of care. These costs for county residents have been organized since 1987 through the Olmsted County Healthcare Expenditure and Utilization Database (OCHEUD)7. This unique database provides a standardized inflation-adjusted estimate of the costs of each service given at providers within the community. The value of each unit of service has been adjusted to national cost norms by use of accepted valuation techniques10. Physician services are valued using Medicare reimbursement rates. All other resource utilization adjusts billed charges by using hospital cost-to-charge ratios and wages indices. The medical costs from all healthcare encounters within the county on persons in the cohort were obtained from the time of HF diagnosis through 2007 (censored at the time of death or last follow-up). All costs, including physician costs, accrued during a hospital admission were categorized as inpatient costs; other costs were considered outpatient. Observed outpatient costs were categorized according to the American Medical Association’s Current Procedural Terminology (CPT) codes. Costs that were unable to be classified were categorized separately. The costs of outpatient prescription medications and nursing home care were not captured.

Statistical Analysis

Patient baseline characteristics are presented as means with standard deviations or frequencies and percentages. Cumulative costs for the cohort (n=1054) during the study period are presented as the mean cost per patient and interquartile range.

Lifetime costs were estimated on all patients using the methods of Tian and Huang11. This method employs a flexible two-part model to analyze lifetime medical costs which are often highly skewed with some zero costs. Inverse probability weighting was used to account for informative right-censoring for those alive at study end. The first part of the model is a logistic regression model that estimates the probability of having positive costs; the second part of the model uses generalized linear methods to model the lifetime costs given the occurrence of positive cost. Inpatient and outpatient costs were estimated separately based on fitted two-part models. Confidence intervals of estimated costs were generated by bootstrapping method12, 13.

Similarly, a two-part model was employed to evaluate the independent predictors of lifetime costs. Clinical variables of interest were entered into a model predicting inpatient, outpatient, and total costs. For each variable, an estimate of the percentage change in costs associated with each clinical characteristic is generated. Each of the models was estimated using a generalized linear model with a gamma distribution and a log link. The gamma distribution was determined using the modified Park test14.

Missing data were minimal (<2%), with the exception of EF (26% missing within one year of HF diagnosis). Multiple imputation was used to input the EF values for patients with missing values. Specifically, five datasets with imputed values of EF were created and analyzed, and the results were combined using Rubin’s rules15. Analysis was performed using SAS Version 9.3.1 (Cary, North Carolina), Splus version 8 (TIBCO Software Inc., Palo Alto, CA), and R (2008, R Foundation for Statistical Computing, Vienna, Austria). A p value <0.05 was used as the level of significance.


Study Population

A total of 1054 incident HF patients were randomly sampled from 1987–2006, distributed evenly throughout the study period. The baseline characteristics of the study population are shown in Table 1. They were elderly (mean age 76.8 years), and 46.1% were men. After a mean follow-up of 4.6 years (median 3.7 years, range 0–20.8 years), 765 (72.6%) patients had died.

Table 1
Baseline Patient Characteristics (n=1054)

Longitudinal Medical Costs After HF Diagnosis

The total costs for the cohort during the study period were $100,967,086, and the distribution of inpatient and outpatient costs are shown in Table 2. The majority of costs were due to hospitalizations (77.0%), with an average of $73,762 per person. The highest proportion of inpatient costs were due to room and board (mean $31,516 per person), procedures (mean $9097 per person), and evaluation and management (mean $7542 per person). The total outpatient costs were $23,221,517, an average of $22,032 per person. The highest proportion of outpatient costs were due to evaluation and management (mean $5541 per person) and procedures (mean $7406 per person). Among evaluation and management, the majority of costs were due to office visits. Dialysis accounted for the highest proportion of procedural costs, though these costs were accumulated by only 31 patients. Among imaging studies, computed tomography (CT) and magnetic resonance imaging/angiography (MRI/MRA) scanning accounted for 29.6% of imaging costs, followed by echocardiography (20.8%). Internal cardiac defibrillator (ICD) costs were minimal at only 2% of total costs in this population (97 of 1054 patients, 9.2% affected), likely reflecting that this is a newer technology unavailable for much of the study period. In addition, many in the population had preserved EF and would not qualify for use for primary prevention of sudden cardiac death.

Table 2
a Longitudinal Direct Medical Costs For 1054 HF Patients

Estimated Lifetime Costs From HF Diagnosis Until Death

Lifetime costs were estimated for all patients using available costs, and are shown in Table 3. The estimated total lifetime costs were $109,541 (95% CI $100,335–118,946) per person. The majority of costs were accumulated during hospitalizations (mean $83,980 per person), with fewer costs accrued in the outpatient setting (mean $25,560).

Table 3
Estimated Lifetime Costs from Time of Heart Failure Diagnosis Until Death Among 1054 Patients

In an ancillary analysis, we examined the lifetime medical costs after HF diagnosis among those who died during follow-up (n=765), and thus had their entire lifetime costs captured. Their mean total lifetime costs were $94,847 per person. As expected, mean total costs were lower than those estimated for the entire cohort, as the duration of follow-up was shorter if we included only those that died. The majority of costs were accumulated during hospitalizations (mean $75,817 per person), with fewer costs accrued in the outpatient setting (mean $19,030 per person).

Predictors of Lifetime Costs

The multivariable predictors of lifetime costs are shown in Table 4. Diabetes mellitus was associated with a 32% increase in inpatient and 25% increase in total costs, while preserved EF (≥50%) was associated with a 21% increase in inpatient and 24% increase in total costs compared with reduced EF, though survival was similar in patients with preserved and reduced EF (p=0.213). Cerebrovascular disease was associated with decreases in inpatient (17%), outpatient (37%), and total costs (22%). Among a random sample of 50 patients with a diagnosis of cerebrovascular disease, 25 (50%) had a prior stroke (80% ischemic, 20% hemorrhagic), 20 (40%) had prior transient ischemic attack (TIA), and 7 (14%) had prior carotid endarterectomy (5 occurred after stroke or TIA). In 3 of the 50 cases (6%) there were no prior stroke, TIA, or carotid endarterectomy documented in the medical record. Sixteen of the 25 (64%) patients with prior stroke had a residual neurologic deficit. Inpatient, outpatient, and total costs were lower when patients were older at HF diagnosis. A creatinine clearance <30 mL/min was associated with 87% higher outpatient costs, which likely reflect the use of dialysis in some patients in this population. Hypertension was associated with 23% lower outpatient costs, but similar inpatient and total costs. This was only significant when adjusting for creatinine clearance, reflecting the relationship between hypertension, renal dysfunction, and dialysis. Being diagnosed with HF from 2002-06 was associated with lower total costs compared with previous time periods. However, though the model used attempts to adjust for incomplete follow-up, it may simply be that lower adjusted costs estimated in these patients are due to inadequate adjustment for right censoring, as a larger proportion of patients diagnosed in this time period were alive at study end (and therefore adjusted by the model) compared with other periods.

Table 4
Multivariable Predictors of Lifetime Costs

Distribution of Costs Over the Remainder of Life After HF Diagnosis

The distribution of costs from the time of HF diagnosis until death by month for those surviving 36–48 months is shown in Figure 1A. This figure demonstrates that costs are high at the time of initial diagnosis, likely reflecting that 56% of community patients are diagnosed with HF while hospitalized16. Then, costs decrease and remain relatively low until the end of life, where costs again increase. A similar distribution in costs was observed for those surviving from 12–36 and 48–72 months after diagnosis. HF patients surviving <12 months also exhibited high initial costs, but maintained a higher level of costs for the remainder of their lifetime, without a significant rise at the end (Figure 1B).

Figure 1Figure 1
Distribution of Costs over Lifetime

To assess the impact of the HF diagnosis on costs, we compared the costs in the year prior to HF diagnosis (ending 3 months prior to HF diagnosis) to the year beginning with HF diagnosis. Markedly higher costs were accrued following heart failure diagnosis, with mean total costs per person of $8219 in the year prior vs. $34372 in the year following HF diagnosis (p<0.001).


The U.S. spent an estimated $2.6 trillion on healthcare in 2009, representing more than 17.6% of our gross domestic product17. Heart failure care constitutes a large portion of this growing economic burden, as it consumes more Medicare dollars than any other diagnosis2 and cost an estimated $37 billion in 200918. Reduction of costs is a major priority, and HF will thus continue to be a central target given its high associated costs. However, limited data exist to inform interventions and policy on the distribution and predictors of healthcare costs over the lifetime of patients diagnosed with HF. Herein, among a population-based community cohort with incident HF followed longitudinally, we have shown that the lifetime costs of HF care are high, and are primarily related to hospitalizations. Diabetes and preserved EF are independently associated with higher total costs. Costs are accrued unevenly over time, with high-cost periods at the time of initial HF diagnosis and in the final months of life.

While the annual costs of caring for the HF population in the U.S. are estimated to be high, little is known about the cumulative costs of caring for HF patients following diagnosis. Russo et al examined costs in the final two years of life among 47 advanced HF patients with low EF enrolled in the Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure (REMATCH) trial, and reported a mean cost of $156,168, of which >50% were accumulated in the last 6 months of life19. Notably, these data pertained to a limited number of trial participants with low EF, who do not represent the general HF population20. Among 343 patients with prevalent HF enrolled in the Cardiovascular Health Study, 10-year medical costs were estimated at $54,704, with an average cost per year alive of $10,83221. Participants with HF incurred greater costs than study participants without HF. However, the use of prevalent cases precluded the examination of costs from the time of diagnosis. Herein, we quantified the direct medical costs of care in a large, community-based cohort of HF patients followed longitudinally from the time of diagnosis. Among the >70% of the cohort who died during the study period, total lifetime costs were high, at a mean of nearly $100,000 per person accumulated over an average of 4 years from diagnosis until death. It is important to note that healthcare delivery in this setting is largely integrated so that information is readily available to providers of multiple different disciplines. This may allow improved coordination of care and result in minimization of redundancy in testing and evaluation. Thus, care in this type of environment, may reflect lower costs than would be incurred in some systems where care of patients may be more fragmented. However, while absolute numbers may differ, there is no reason to believe that the pattern of expenditures would differ in other settings.

Hospitalizations have been estimated to account for at least two-thirds of HF costs1, 6, 22. Our data reveal that, on average, a staggering 79% of lifetime costs of HF care are accumulated during hospitalization. Similarly, among all 1054 HF patients followed over the course of the study, representing nearly 5000 person-years of follow-up, $77.7 of $101.0 million (77%) were due to hospitalizations. While room and board was the greatest contributor to inpatient costs (43% of inpatient costs), consistent with prior studies23, procedures, imaging and laboratory testing accounted for an additional 32% of inpatient costs for the cohort, thereby constituting opportunities for cost-saving when clinically appropriate. The high economic burden that hospitalizations impose on the healthcare system in the U.S. has been an area of recent focus, in particular due to the high rate of readmissions observed among HF patients24,25. While more than 50% of admissions among HF patients may be preventable26, currently hospitalizations among HF patients remain frequent and costly.

Identifying patient factors associated with higher costs may be helpful to target those patients for cost-saving interventions. We recently reported on the predictors of lifetime hospitalizations among HF patient in Olmsted County16, and found that male sex, COPD, diabetes mellitus, anemia, and renal insufficiency were each associated with an increased risk of hospitalization. When examining clinical variables for their association with lifetime costs herein, diabetes mellitus and preserved EF were critical factors associated with higher lifetime costs, with each having an estimated increase of 25% and 24%, respectively. The clinical characteristics associated with risk of hospitalization and lifetime costs may differ, as comorbidities conveying adverse prognosis could have variable impact on lifetime costs even though they increase the risk of hospitalization. This certainly could explain, for example, why renal insufficiency was associated with higher hospitalization risk previously16, but not higher lifetime total costs herein, as it conveys worse prognosis, and thus a shorter lifespan by which to accumulate costs. With diabetes, the increase in total costs was due to an increase in hospitalization costs, which is consistent with our prior work demonstrating that diabetes was associated with a 53% increase in hospitalization risk16. However, preserved EF was also predictive of increased total costs, with a significant 21% increase in inpatient and insignificant 33% increase in outpatient costs. Liao et al reported that among 881 elderly HF patients, 5-year costs were similar for patients with reduced and preserved EF after adjusting for other factors 27. Patients with preserved EF have an increased frequency of many comorbidities, which could result in longer hospital length of stay and increased total costs, though this remains to be established. The presence of cerebrovascular disease at the time of HF diagnosis was associated with lower cumulative resource utilization including lower inpatient (17%), outpatient (37%), and total (22%) lifetime costs after adjusting for age and other comorbidities. Approximately half of these patients had prior stroke, a large proportion with residual neurological deficits. Perhaps, these patients underwent less extensive medical evaluation due to their neurological status and/or wishes. Further work is needed to understand this unexpected finding.

While it has previously been recognized that the costs of caring for selected HF patients at the end of life are high19, to the best of our knowledge, no studies have assessed costs from the time of initial HF diagnosis in order to identify a pattern of cost accrual over the lifetime after diagnosis. Herein, patients surviving >12 months after initial diagnosis accumulate higher costs at the time of initial diagnosis and in the months immediately prior to death, with lower costs accrued in the interim. However, patients surviving <12 months after diagnosis follow an alternate pattern, with high initial costs of care, followed by a steady level of cost accrual for the remainder of life that appears higher than those surviving longer periods of time. The costs of HF care appear to be particularly high in the immediate months prior to death, a time that has been associated with increased hospitalizations and resource use16, 19. Discussions between patients and clinicians regarding prognosis and the appropriateness of therapies and wishes for end of life care are of particular importance, and could result in cost-saving due to improved resource allocation.

Some limitations should be acknowledged in interpreting these data. This study included only direct medical costs, and did not include outpatient medication costs or indirect costs of care. Costs may have been underestimated for some patients if they received medical care outside of the community, either due to emigration or care during periods of travel. However, when residency status was examined during the study period using the resources of the Rochester Epidemiology Project28, <2% of the sample moved away from Olmsted County and the surrounding communities, reflecting a negligible degree of emigration. While we acknowledge that the results pertain to one single geographical area and that geographical variations in care exist, the ability to examine lifetime costs of care in a stable population provides insights of clinical and societal relevance. We did not compare lifetime costs of care with a control group of patients without HF to accurately determine whether HF alone accounted for the increase in costs. However, the costs of care post-HF diagnosis were markedly higher in the same patients compared with pre-diagnosis. An accurate control group for cost comparison is difficult, as the impact of other comorbidities and life expectancy would also need to be taken into account. After adjustment for comorbidities, those patients diagnosed with heart failure from 2002-06 had significantly lower costs compared with other years. This may be in part due to incomplete adjustment for right-censoring of data despite our use of rigorous methodology, as a much higher proportion of patients diagnosed with HF from 2002-06 were alive at study end compared with other time periods. This is further exacerbated by the fact that a large proportion of them have not experienced the high-cost end of life period. Finally, the study population is primarily Caucasian, and further studies are needed in communities where the racial and ethnic composition may differ. However, there are several strengths including the use of a validated incident population of HF patients followed longitudinally, and the inclusion of both inpatient and outpatient cost data in our analysis.

The implications of these data are three-fold. First, the longitudinal cost of caring for HF patients is high, and primarily driven by inpatient care. This suggests preventing hospitalizations and reducing the cost of hospitalizations are keys to reducing the total costs of care HF patients. Second, diabetes mellitus and preserved EF were each associated with a large increase in lifetime costs, even after adjusting for other factors. Diabetic persons, in particular, represent a segment of the HF population who are at greater risk for hospitalization and resource use, and may be a target population for further investigation and intervention. Third, the costs of caring for patients with HF are highest at the time of initial diagnosis and at the end of life. Judicious allocation of resources at the end of life based on patient wishes and needs are required to decrease costs.

In summary, the costs of caring for HF patients in the community is high, and primarily due to hospitalizations. Costs are higher at diagnosis and end of life. Cost-saving measures targeted at reducing hospitalizations, particularly in patients with diabetes and preserved EF, and reducing end of life costs, may have the greatest impact.


We would like to thank Kay Traverse R.N. for her assistance with data collection.

Funding Sources: Dr. Dunlay was funded by a National Institute of Health Ruth L. Kirschstein National Research Service Award (T32 HL07111-31A1) and Dr. Roger by an NIH RO1 (HL72435). This study was also made possible by the Rochester Epidemiology Project (Grant #R01-AR30582) from the National Institute of Arthritis and Musculoskeletal and Skin Diseases) and Grant Number 1 UL1 RR024150 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH), and the NIH Roadmap for Medical Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH. Information on NCRR is available at http://www.ncrr.nih.gov/. Information on Reengineering the Clinical Research Enterprise can be obtained from http://nihroadmap.nih.gov


Conflict of Interest Disclosures None.


1. Stewart S, Jenkins A, Buchan S, McGuire A, Capewell S, McMurray JJ. The current cost of heart failure to the National Health Service in the UK. Eur J Heart Fail. 2002;4:361–371. [PubMed]
2. O’Connell JB, Bristow MR. Economic impact of heart failure in the United States: time for a different approach. J Heart Lung Transplant. 1994;13:S107–112. [PubMed]
3. Heart Disease and Stroke Statistics 2008 Update. Available at: http://www.americanheart.org/downloadable/heart/1200078608862HS_Stats%202008.final.pdf.
4. Roger VL, Weston SA, Redfield MM, Hellermann-Homan JP, Killian J, Yawn BP, Jacobsen SJ. Trends in heart failure incidence and survival in a community-based population. Jama. 2004;292:344–350. [PubMed]
5. Gheorghiade M, Pang PS. Acute heart failure syndromes. J Am Coll Cardiol. 2009;53:557–573. [PubMed]
6. Berry C, Murdoch DR, McMurray JJ. Economics of chronic heart failure. Eur J Heart Fail. 2001;3:283–291. [PubMed]
7. Leibson CL, Katusic SK, Barbaresi WJ, Ransom J, O’Brien PC. Use and costs of medical care for children and adolescents with and without attention-deficit/hyperactivity disorder. Jama. 2001;285:60–66. [PubMed]
8. Roger VL, Killian J, Henkel M, Weston SA, Goraya TY, Yawn BP, Kottke TE, Frye RL, Jacobsen SJ. Coronary disease surveillance in Olmsted County objectives and methodology. J Clin Epidemiol. 2002;55:593–601. [PubMed]
9. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130:461–470. [PubMed]
10. Lave JR, Pashos CL, Anderson GF, Brailer D, Bubolz T, Conrad D, Freund DA, Fox SH, Keeler E, Lipscomb J, Luft HS, Provenzano G. Costing medical care: using Medicare administrative data. Med Care. 1994;32:JS77–89. [PubMed]
11. Tian L, Huang J. A two-part model for censored medical cost data. Stat Med. 2007;26:4273–4292. [PubMed]
12. Version R Package Version 1.2–36. 2009. Bootstrap-R (S-Plus) Functions [computer program]
13. Davison AC, Hinkley DV. Bootstrap Methods and their Applications. Cambridge: Cambridge University Press; 1997.
14. Manning WG, Mullahy J. Estimating log models: to transform or not to transform? Journal of health economics. 2001;20:461–494. [PubMed]
15. Rubin D. Multiple imputation for nonresponse in surveys. New York: J Wiley and Sons; 1987.
16. Dunlay SM, Redfield MM, Weston SA, Therneau TM, Hall Long K, Shah ND, Roger VL. Hospitalizations after Heart Failure Diagnosis in the Community. J Am Coll Cardiol. 2009;54:1695–702. [PMC free article] [PubMed]
17. Meyer H. Healthcare in Crisis. Medical Economics. 2009;86:18–24. [PubMed]
18. Heart Disease and Stroke Statistics 2009 Update. American Heart Association; 2009.
19. Russo MJ, Gelijns AC, Stevenson LW, Sampat B, Aaronson KD, Renlund DG, Ascheim DD, Hong KN, Oz MC, Moskowitz AJ, Rose EA, Miller LW. The cost of medical management in advanced heart failure during the final two years of life. J Card Fail. 2008;14:651–658. [PubMed]
20. Lloyd-Williams F, Mair F, Shiels C, Hanratty B, Goldstein P, Beaton S, Capewell S, Lye M, McDonald R, Roberts C, Connelly D. Why are patients in clinical trials of heart failure not like those we see in everyday practice? J Clin Epidemiol. 2003;56:1157–1162. [PubMed]
21. Liao L, Anstrom KJ, Gottdiener JS, Pappas PA, Whellan DJ, Kitzman DW, Aurigemma GP, Mark DB, Schulman KA, Jollis JG. Long-term costs and resource use in elderly participants with congestive heart failure in the Cardiovascular Health Study. Am Heart J. 2007;153:245–252. [PubMed]
22. Malek M. Health economics of heart failure. Heart. 1999;82(Suppl 4):IV11–13. [PMC free article] [PubMed]
23. Hauptman PJ, Swindle J, Burroughs TE, Schnitzler MA. Resource utilization in patients hospitalized with heart failure: insights from a contemporary national hospital database. Am Heart J. 2008;155:978–985. e971. [PubMed]
24. Epstein AM. Revisiting readmissions--changing the incentives for shared accountability. N Engl J Med. 2009;360:1457–1459. [PubMed]
25. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360:1418–1428. [PubMed]
26. Michalsen A, Konig G, Thimme W. Preventable causative factors leading to hospital admission with decompensated heart failure. Heart. 1998;80:437–441. [PMC free article] [PubMed]
27. Liao L, Jollis JG, Anstrom KJ, Whellan DJ, Kitzman DW, Aurigemma GP, Mark DB, Schulman KA, Gottdiener JS. Costs for heart failure with normal vs reduced ejection fraction. Arch Intern Med. 2006;166:112–118. [PubMed]
28. Melton LJ., 3rd History of the Rochester Epidemiology Project. Mayo Clin Proc. 1996;71:266–274. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...