Nonprofit Hospitals’ Approach to Community Health Needs Assessment
Kenneth R. McLeroy, PhD, James N. Burdine, DrPH, and David Matarrita-Cascante, PhDAbstract
Objectives. We sought a better understanding of how nonprofit hospitals are fulfilling the community health needs assessment (CHNA) provision of the 2010 Patient Protection and Affordable Care Act to conduct CHNAs and develop CHNA and implementation strategies reports.
Methods. Through an Internet search of an estimated 179 nonprofit hospitals in Texas conducted between December 1, 2013, and January 5, 2014, we identified and reviewed 95 CHNA and implementation strategies reports. We evaluated and scored reports with specific criteria. We analyzed hospital-related and other report characteristics to understand relationships with report quality.
Results. There was wide-ranging diversity in CHNA approaches and report quality. Consultant-led CHNA processes and collaboration with local health departments were associated with higher-quality reports.
Conclusions. At the time of this study, the Internal Revenue Service had not yet issued the final regulations for the CHNA requirement. This provides an opportunity to strengthen the CHNA guidance for the final regulations, clarify the purpose of the assessment and planning process and reports, and better align assessment and planning activities through a public health framework.
The Patient Protection and Affordable Care Act of 2010 includes a provision requiring all nonprofit hospitals to conduct a community health needs assessment (CHNA) and develop an implementation strategies plan. Nonprofit hospitals must conduct a CHNA at least every 3 years and implement strategies to address identified priority needs.1 The Internal Revenue Service (IRS), the bureau responsible for the regulation and enforcement of Section 9007 of the Affordable Care Act, provides general guidelines to nonprofit hospitals regarding the CHNA requirement.1 Included in this requirement are identifying and prioritizing community health needs, inventorying resources, developing an implementation strategies report to address health needs, and involving stakeholders with public health knowledge and expertise and leaders, representatives, or members of medically underserved, low-income, and minority populations in the community.1
Very little research has been conducted on nonprofit hospitals’ approach to the CHNA requirement, perhaps because of its relative newness. Using CHNA and implementation strategies reports developed by nonprofit hospitals in Texas, we evaluated and analyzed various CHNA methods, report components, and influential factors. In addition, we assessed CHNA and implementation strategies report quality by using a public health framework.
COMMUNITY HEALTH ASSESSMENT
Assessment is one of public health’s 3 core functions.2 Community health assessment (CHA) is an important aspect of the assessment function and is critical in the development, implementation, and evaluation of effective health improvement programs and policies. (With concern to the nonprofit hospital IRS requirement, we will refer to the assessment process and report as “community health needs assessment [CHNA],” since that is how it is identified in the regulation. However, there are issues and potentially negative consequences to perceiving these processes as “needs” assessment.3,4 When not referring to the IRS regulation, we will refer to the assessment process as “community health assessment [CHA].”)
There are varying definitions of and approaches to CHA. When one uses a public health framework, CHA processes generally include the following: gathering and analyzing quantitative and qualitative data; using data to identify health issues; using broad social determinants of health to identify influences on health issues, including environment, behavior, socioeconomics, and culture; identifying resources and resource gaps; identifying health disparities; engaging and mobilizing the community; organizing and sharing findings; setting health priorities; developing an action plan to address health priorities; implementing action plans; and providing opportunities for continual feedback with community members.4,5–17
Because of often-divergent viewpoints between medical and public health disciplines, it is not known how nonprofit hospitals will approach components of the CHNA requirement and to what extent processes will represent a public health framework. Professional training and the way in which community and health are perceived often differ significantly between medical and public health disciplines. Public health employs a broad conceptualization of health that includes biology and genetics, individual behavior, social environment, physical environment, and health services.17 The medical model emphasizes diagnosis, treatment, and care of the patient, and the public health model stresses prevention and health promotion for the population. The medical paradigm places emphasis on medical care, whereas the public health paradigm includes medical care, but also focuses on the broader social determinants of health, including environmental, behavioral, socioeconomic, and cultural factors.18,19
Research strongly suggests that clinical measures, such as quality of and access to health care, contribute little to overall health compared with other factors. One model, used in US county health rankings, attributes 20% of health to clinical care. The remaining factors—social and economic factors, health behavior, and physical environment—contribute 40%, 30%, and 10%, respectively.18
Previously, public health agencies, often in collaboration with other community partners and stakeholders, have largely led CHA efforts. The importance of community participation and mobilization in CHA processes includes the following: recognizing the community as a unit of identity; building on strengths and resources; facilitating collaborative partnerships; emphasizing locally relevant problems and an ecological perspective; promoting power sharing, colearning, and capacity building; improving cultural sensitivity, reliability, and validity through quality community participation; increasing community trust and ownership; developing community systems through a cyclical and interactive process; disseminating the findings and knowledge; and enhancing sustainability.20–22
Using a public health framework supports identifying broader determinants of health; adopting clinical and nonclinical strategies to address health issues; identifying clinical and nonclinical community resources; sharing expertise and tasks among multiple organizations, agencies, and disciplines; leveraging resources; and increasing the probability of reducing health disparities and improving population health status.
NONPROFIT HOSPITALS AND COMMUNITY BENEFIT
The legislation tied to the CHNA regulation largely came about because of concerns as to whether nonprofit hospitals are meeting minimum community benefit standards needed to maintain tax-exempt status. Tax-exempt status for nonprofit hospitals translates to billions of tax dollars saved as well as charitable contributions.23,24 However, studies show that there is little difference between community benefit provided by nonprofit and for-profit hospitals, raising the question of whether nonprofit hospitals merit this continued distinguished status.23,25,26 Another concern is the type of community benefit nonprofit hospitals provide, with the majority of community benefit expenditures going to direct patient care.26
The CHNA and other community benefit requirements may provide opportunities to create linkages between medicine and public health, address disparities, engage the broader community, improve population health, and provide greater benefit to society.18,19,27,28 However, there is little evidence that this is the perspective widely held by nonprofit hospitals. Hospitals often view health through a different lens than public health and this raises the question as to whether hospitals alone have the necessary training, perspective, and resources to properly assess community health needs and identify appropriate strategies to have an effective influence on population health.
COMMUNITY HEALTH ASSESSMENT EVALUATION CRITERIA
In public health planning, the supposition is that quality plans lead to quality programs, which increase the likelihood of improved health outcomes.29 Because of the vague IRS guidance, there is a large degree of subjectivity while one is evaluating the quality of CHNA and implementation strategies reports. However, criteria exist for the evaluation of quality public health plans, using the draft IRS requirements as well as criteria gleaned from the literature based on quality characteristics of CHAs and plans for adolescent pregnancy prevention, obesity prevention, and sustainability planning. These include
Partner and stakeholder involvement,1,29–34
Organizational structure and personnel considerations,1,30,32–34
Definition of community,1,30,32
Examination of data,1,7,29–34
Identification and prioritization of issues,1,7,30,32–34
Examination of causation of problem,7,32,33
Consideration of local context,29–33
Identification of assets or resources,1,7,29,31–33
Clear goals and measurable objectives,1,7,29,31–34
Action plan or strategies to address issues,1,32,34
Evidence-based strategies,31–33,39
Reflects social determinants of health,29,31,32
Description of the process,1,7,29–32
Feasibility and sustainability,1,29–31,33,34
Evaluation of plan,1,29–31,33,34 and
Accessibility of plan.1,7,29–33
These criteria provide a framework for evaluating the quality of CHNA and implementation strategies reports. See Table 1 for further description of these community health assessment and planning evaluation criteria.
TABLE 1—
Evaluation Criteria for Community Health Assessment and Community Health Planning: United States, 2010–2013
| Evaluation Criteria | IRS Draft Requirements1 | Dunet and Butterfoss31; Dunet et al.29 | Parra-Medina et al.33 | Sridharan et al.34 | Barnett30 | Myers and Stoto7 | Catholic Health Association32 |
| Partner and stakeholder involvement | Identify organizations or parties with whom the hospital collaborated or contracted assistance in conducting the CHNA; describe those with public health expertise included; describe those with medically underserved, low income, and minority populations included; describe planned collaboration to address health needs between the hospital and other facilities or organizations | Meaningful stakeholder involvement; balanced representation (that reflects community) | Develop in collaboration with agencies or organizations in the community | Communication mechanism between stakeholders; proof of interagency collaboration | Shared ownership of community health; community engagement; shared accountability and regional governance | When possible, conduct assessment in collaboration with other hospitals, local health departments and community partners; form assessment team or advisory committee that represents community; seek community input that reflects the racial, ethnic, and economic diversity of the community; validate priorities with community input | |
| Organizational structure and personnel considerations | Identify resources and programs the hospital plans to commit to address the health needs | Include advisory committee or similar structure | Organizational structure to oversee implementation; identifies staff needs | Institutional oversight | Form assessment team or advisory committee that includes key staff within the organization | ||
| Definition of community | Definition of the community served; description of how that was determined | Defining community (jurisdictional issues) | Define community to include primary and secondary service areas and the types of patients the hospital serves (age, gender, conditions treated) | ||||
| Examination of data (data source, methods) | A description of the data and other information used in the assessment; methods of collecting and analyzing data and information | Systematic examination of data; data are from reliable sources | Include prevalence data | Data-driven planning process | Data collection and analysis | Clearly identify data sources (e.g., citations to graphs or tables); present data in meaningful subgroups of population (e.g., to assess health disparities) | Base the assessment on review of public health data collected by government agencies and other authoritative sources; consider the following types of information: demographics, health indicators, health risk factors, access to health care, and social determinants of health; collect community input using 1 or more of the following methods: community forums, focus groups, interviews, or surveys; analyze data collected by using comparisons with other communities and with federal or state benchmarks and, when available, trends within the community |
| Identification and prioritization of issues | Identify significant community health needs; prioritized description of the significant health needs | Clearly state the problem | Priority setting | Include the most important aspects of the community’s health | Identify from 3 to 10 priorities; align priorities with organizational, state, and national priorities; give priority to persons who are low-income and disadvantaged | ||
| Examination of causation of problem | Delineate causation of the problem | Clearly indicate the relationships among related health indicators | Look for disparities and contributing causes of health problems; understand root causes of needs being addressed | ||||
| Consideration of local context | Match strategies to population; integration of strategies into existing programs and infrastructure | Identify local factors that contribute to the problem; consistent with other local programs | Alignment opportunities | Coordinate hospital and community strategies to ensure the most effective use of resources; build on existing programs and other community assets when possible | |||
| Identification of assets or resources | Description of potential resources to address the significant health needs | Assess existing resources | Assess assets and resources | Provide sufficient focus on positive characteristics (e.g., as well as negative) | Use knowledge of community assets in determining priorities | ||
| Clear goals and measurable objectives | Description of potential measures to address the significant health needs | Goals for changing health status; SMART objectives; objectives logically related to goals | Clearly define program objectives; objectives are written in measurable format; activities outlined support objectives | Revisit goals | Clearly state goals and purpose of CHA | For each prioritized need, identify the goal to be achieved, measurable objectives(s), indicators for determining whether objectives were met, and evaluation measures | |
| Action plan or strategies to address issues | Develop implementation strategy that corresponds to the health needs identified through the CHNA (how issue will be addressed) | Established processes or procedures to ensure agencies fulfill responsibilities | Update the implementation strategy upon major changes in community health status and at least every 3 years | ||||
| Evidence-based strategies | Strategies based on scientific evidence | Program based on current research | Investigate evidence-based approaches to ensure effective use of hospital and community resources | ||||
| Reflects social determinants of health | Objectives include multiple ecological levels | Identify a range of possible interventions | |||||
| Description of the process | Description of the assessment process and methods, process, and criteria used to identify and prioritize health needs, input provided by partners, and why health priorities not addressed | Documentation of rationale for strategies selected; describe how partners will be involved | Sufficiently documents the process and methods used to create the CHA | Document how priorities were identified and who was involved in setting priorities | |||
| Feasibility and sustainability | Locate, maintain, and sustain resources | Realistic; strategies for seeking funding | Funding and sustainability | Strategic investment and funding patterns | |||
| Evaluation of plan | Plan to evaluate impact | Evaluation plan | Describe how will be evaluated and how findings will be used | Continued data collection plan to assess progress toward goals | Monitoring and evaluation | ||
| Accessibility of plan | Make report publicly available; post report on the hospital’s Web site, or Web site of other collaborating organization | Understandable; useful; designed to elicit interest and support of reader; wide distribution of plan | Readability | Public reporting: federal, state, and local issues | Use consistent format; include a summary and detailed versions; well organized and easy to find content; easy to understand; available online; include appropriate links; easily photocopied; includes narrative and graphic representation of key findings; use similar structure or data elements as other community planning tools | Distribute report to all partners and contributors; make the implementation strategy publicly available |
Notes. CHA = community health assessment; CHNA = community health needs assessment; IRS = Internal Revenue Service; SMART = specific, measurable, attainable, realistic, and time-bound.
The IRS regulations have yet to be finalized; however, proposed regulations guided the CHNA process and CHNA and implementation strategies report components for the first 3-year period. Currently, the guidance is broad, open to interpretation, and allows a fair amount of latitude and flexibility. Some adaptability and tailoring is important, as each community and nonprofit hospital is different in terms of resources, demographics, health issues, partners, history, and other contextual factors that contribute to how community organizations and members work together, make decisions, identify issues and resources, and address issues. However, without more specific guidance or evaluation criteria, the requirement’s usefulness, applicability, and potential to improve community outcomes is unknown.
The purpose of this study was to gain a better understanding of how nonprofit hospitals are fulfilling the IRS requirement to conduct and develop CHNA and implementation strategies reports. Initial steps to improve understanding included (1) creating a mechanism for evaluating CHNA and implementation strategies reports using a public health framework and the IRS guidance, (2) evaluating and scoring a sample of Texas nonprofit hospital reports by using this mechanism, (3) providing an initial overview of Texas nonprofit hospitals’ reports, and (4) beginning to identify key characteristics and factors that resulted in CHNA and implementation strategies reports of greater quality.
METHODS
We conducted an Internet-based search for Texas nonprofit hospital CHNA and implementation strategies reports, between December 1, 2013, and January 5, 2014, which the IRS requires be made publicly available.35 Both the CHNA and implementation strategies report sections were required to meet study inclusion criteria. We only included reports with both sections in the evaluation and analysis. We located 135 CHNA reports developed by Texas nonprofit hospitals. Forty had not yet included the implementation strategies report section and we excluded these from the sample. Thus, we reviewed, evaluated, and scored 95 CHNA and implementation strategies reports. This study accounted for approximately 53% of the nonprofit hospital population in Texas.
We evaluated CHNA and implementation strategies reports with the 16 criteria items described in Table 1 and scored each item by using a 6-point scale, borrowing from a scale used to assess state plans for obesity prevention.29,31 Each criteria item that was not addressed was scored zero; items that were low quality or had no detail were scored 1; items that were low quality and included very limited detail were scored 2; items that were partially or variably addressed were scored 3; items that were sufficiently addressed (good, solid job) were scored 4; and items that were addressed with high quality and detail were scored 5. To increase validity of the framework used to evaluate reports, the framework was reviewed by 4 researchers and practitioners in community health, health care services, and health policy. We reviewed each CHNA and implementation strategies report as a whole and then evaluated and scored each factor independently to the extent possible. There is currently no evidence in the literature that certain factors are more important than others, so we weighted the 16 factors equally. The cumulative score of these 16 items produced a total report score, which we used as an indicator of CHNA and implementation strategies report quality. To reduce subjectivity, 2 trained reviewers evaluated and scored 10% (n = 10) of the reports, which were randomly selected from the 95 CHNA and implementation strategies reports. The intraclass correlation coefficient of 0.187 represents strong agreement among raters.
Other characteristics of the CHNA and implementation strategies reports collected included the year the CHNA was conducted or published (if these differed, we used year of publication), whether the CHNA process was staff- or consultant-led, and whether the hospital partnered with a local health department to conduct the assessment. We obtained these data from the CHNA and implementation strategies reports. We also gathered characteristics of the hospital and the community in which the hospital was located, including the rural–urban continuum code of the county,36 hospital size based on number of beds, whether the hospital was religious or faith-based, whether the hospital was part of a health care system or a stand-alone facility, the presence of a city or county health department,37 and median county income.38 We obtained other hospital characteristic data from CHNA and implementation strategies reports and hospital Web sites.
We generated descriptive statistics, including range, median, and mean for hospital characteristics, CHNA and implementation strategies report characteristics, and evaluation criteria. We also used Pearson correlation and Spearman rank correlation to determine relationships between variables. Several variables are ordinal, so the results of Spearman are reported. Finally, we used robust standard errors univariate regression and multiple linear regression forward and backward variable selection method to identify hospital, community, and report characteristics that made significant contributions to variability in the CHNA and implementation strategies report quality score as well as to build a best-fit model. The forward and backward variable selection method resulted in the same model. Using robust standard errors takes into account the presence of outliers and failure to meet normality and heteroscedasticity assumptions.39 We did not include evaluation criteria in regression analyses, as these scores were the basis for the total report score, so there are inherent associations. We used Stata version 12 to conduct all analyses (StataCorp LP, College Station, TX).
RESULTS
Eighty-two percent of the hospitals (n = 78) were located in metropolitan areas (rural–urban continuum code 1–3). Hospitals ranged in size from 6 hospital beds to 1109 beds (mean = 245; median = 134). Sixty-one percent (n = 58) were faith-based nonprofit hospitals and 84% (n = 80) were part of a larger health care system.
The majority of the reports (82%; n = 78) were conducted and published in 2013, with 13% (n = 12) published in 2012 and 5% (n = 5) in 2011. Forty-five percent of the CHNA processes were staff-led; 55% were consultant-led (n = 43; n = 52). Contrary to the IRS requirements, only 13% of the reports (n = 12) evaluated collaborated with a local health department in a meaningful way.
Each evaluation criterion ranged widely on the 6-point scale (0 = not addressed, to 5 = high quality). The criteria with the highest mean scores were examination of the data using reliable sources and multiple data collection sources and methods (3.35), the feasibility and sustainability of plans (3.14), and report readability and accessibility to the public (3.01). The criteria with the lowest mean scores were the identification of issues or strategies that considered social determinants of health (1.15), use of evidence-based strategies (1.34), consideration of local contextual factors (1.40), and examination of contributing causes to health issues (1.70). Table 2 shows the range and mean of each evaluation criteria item.
TABLE 2—
Evaluation Criteria Range and Mean From Community Health Needs Assessment and Implementation Strategy Report Evaluation Among Nonprofit Hospitals: Texas, 2010–2013
| Criteria Item | Range | Mean |
| Partner and stakeholder involvement | 1–5 | 2.50 |
| Organizational structure and personnel considerations | 1–5 | 2.38 |
| Definition of community | 1–5 | 2.92 |
| Examination of data (data source, methods) | 1–5 | 3.35 |
| Identification and prioritization of issues | 1–5 | 2.84 |
| Examination of causation of problem | 0–4 | 1.70 |
| Consideration of local context | 0–5 | 1.40 |
| Identification of assets or resources | 0–5 | 2.93 |
| Clear goals and measurable objectives | 0–5 | 2.23 |
| Action plan or strategies to address issue | 0–4 | 2.51 |
| Evidence-based strategies | 0–5 | 1.34 |
| Reflects social determinants of health | 0–4 | 1.15 |
| Description of the process | 0–5 | 2.56 |
| Feasibility and sustainability | 0–4 | 3.14 |
| Evaluation of plan | 0–4 | 2.22 |
| Accessibility of plan | 1–5 | 3.01 |
| Total score | 11–61 | 38.2 |
Notes. Ratings: 0 = not addressed; 1 = low quality or no detail; 2 = quality weak or very limited detail; 3 = quality or detail partial or variable; 4 = good, solid quality or good detail; 5 = high quality or highly detailed.
Using the summative values of 16 evaluation criteria items, total report scores ranged from 11 to 61 (possible high = 80). The mean total score was 38.2; the median total score was 40.0. The majority of reports fell in the midscoring range: 16 report scores (16.8%) ranged from 11 to 27; 49 report scores (51.6%) ranged from 28 to 46; and 30 report scores (31.6%) ranged from 47 to 61.
Associations Among Hospitals, Reports, Evaluation Criteria, and Total Scores
We expected some strong variable correlations; presence of local health departments, larger hospitals, and higher median county income had positive associations with urbanity (ρ = 0.6581; ρ = 0.4043; and ρ = 0.5996; respectively). Faith-based hospitals were more likely to be part of a health care system versus a stand-alone facility (ρ = 0.4829). The year the assessment was conducted was strongly associated with hospital location and median county income (ρ = 0.4985 and ρ = 0.5558; respectively); interestingly, nonmetropolitan-based hospitals and lower-income counties were likely to conduct an earlier assessment (in 2011 or 2012).
Consultant-led assessments were positively associated with how well the CHNA process was described (ρ = 0.4549) and report readability and accessibility to the public (ρ = 0.5449). Collaborating with a local health department had a strong, positive association with involvement of partners and stakeholders in the CHNA process (ρ = 0.5896) and examining contributing causes of problems (ρ = 0.4867). Faith-based hospitals had a strong, negative association with report readability and accessibility to the public (ρ = −0.4100) and the extent to which hospitals appeared to provide organizational support and personnel for the CHNA process (ρ = −0.4696).
Hospital characteristics, including hospital size and system-based hospitals, had very weak positive associations with the total CHNA and implementation strategies report score (ρ = 0.0467 and ρ = 0.0338; respectively). Metro-located hospitals had a moderately weak positive association with the overall report score (ρ = 0.1086), whereas faith-based hospitals had a moderately weak negative association with the overall report score (ρ = −0.1285). The presence of a county or city health department had a very weak positive association (ρ = 0.0163) and median county income had a moderately weak positive association (ρ = 0.1175) with the total report score. Characteristics of the report had somewhat stronger relationships to report quality: consultant-led assessments and collaborating with a local health department to conduct the assessment were positively associated with total scores (ρ = 0.3362 and ρ = 0.2542; respectively). The year the assessment was conducted had a very weak positive association with total CHNA score (ρ = 0.0262).
Most report evaluation criteria were strongly associated with total CHNA score. Those most strongly associated with total score were identification of existing assets or resources to contribute to health needs (ρ = 0.7872), identification of issues or strategies that considered social determinants of health (ρ = 0.7618), examination of contributing causes to problems (ρ = 0.7434), creation of action plans and strategies to address identified issues (ρ = 0.6974), identification of evidence-based strategies (ρ = 0.6965), examination of the data using reliable data sources and multiple data collection sources and methods (ρ = 0.6612), development of an evaluation plan (ρ = 0.6543), development of clear goals and measureable objectives (ρ = 0.6402), and feasibility and sustainability of plans (ρ = 0.6254). Although most of the evaluation criteria were strongly associated with the assessment report total score, some correlational variations may suggest that some factors are more important to overall quality and provide evidence for weighting these criteria in future studies.
Factors With Influence on Total Score
We removed independent variables with little influence on the total score from the model to reduce relationship complexity and avoid distortion by extraneous variables (P > .25). See Table 3 for the univariate analysis summary of all variables. We fit the full model with the remaining possible predictors. The final model, which included the independent variables staff-led CHNA processes and collaboration with a local health department, fit significantly better than the intercept-only model (F2,92 = 16.19; P < .001; R2 = 0.2918). The final model accounted for 29.2% of the variability in the total report score. See Table 4 for the final model with adjusted and unadjusted regression coefficients and 95% confidence intervals (CIs). With this model, staff-led CHNA processes were associated with a 10.3-point decrease in total CHNA and implementation strategies report score (95% CI = −14.225370, −6.390098). Partnering with a local health department was associated with a 12.1-point increase in total report score (95% CI = 5.617092, 18.622170).
TABLE 3—
Summary of Univariate Analysis for All Variables From Community Health Needs Assessment and Implementation Strategy Report Evaluation Among Nonprofit Hospitals: Texas, 2010–2013
| Variables | No. | F | LR | P | R2 | AIC | BIC |
| Staff-led | 95 | 16.40227 | 15.43105 | <.001 | 0.1499262 | 699.9147 | 705.0225 |
| Partnered with LHD | 95 | 6.084728 | 6.020703 | .015 | 0.0614093 | 709.3251 | 714.4328 |
| Faith-based hospital | 95 | 2.441325 | 2.461656 | .122 | 0.0255793 | 712.8841 | 717.9919 |
| Hospital size | 95 | 0.5023405 | 0.5117626 | .480 | 0.0053725 | 714.834 | 719.9418 |
| Metro | 95 | 0.3834641 | 0.3909053 | .537 | 0.0041063 | 714.9548 | 720.0626 |
| Year | 95 | 0.0527114 | 0.0538298 | .819 | 0.0005665 | 715.2919 | 720.3997 |
| Health care system member | 95 | 0.0419957 | 0.0428892 | .838 | 0.0004514 | 715.3029 | 720.4106 |
| Median county income | 95 | 0.0309696 | 0.0316303 | .861 | 0.0003329 | 715.3141 | 720.4219 |
| LHD in county or city | 95 | 0.0148948 | 0.0152139 | .903 | 0.0001601 | 715.3306 | 720.4383 |
Notes. AIC = Akaike information criterion; BIC = Bayesian information criterion; F = F statistic; LHD = local health department; LR = likelihood ratio.
TABLE 4—
Final Model With Unadjusted and Adjusted Regression Coefficients From Community Health Needs Assessment and Implementation Strategy Report Evaluation Among Nonprofit Hospitals: Texas, 2010–2013
| Variable | Unadjusted (95% CI) | Adjusted (95% CI) |
| Staff-led | −7.955277 (−12.032750, −3.877800) | −10.30774 (−14.225370, −6.390098) |
| Partnered with local health department | 7.628514 (0.711742, 14.545290) | 12.12004 (5.617092, 18.622170) |
Notes. CI = confidence interval. F = 16.19; P < .001; R2 = 0.2918.
DISCUSSION
Although initial evaluation was limited to Texas nonprofit hospitals, these results provide preliminary insight to understanding nonprofit hospitals’ approach to the CHNA requirements. Results from this study suggest that collaborating with a local health department and working with a consultant appear to improve CHNA and implementation strategies report quality. As we largely used a public health model to evaluate the reports, a strong association between quality and local health department partnerships might be expected. We might expect communities with more local resources to have higher-quality assessment reports. However, variables presumably indicative of high-resource areas, such as the presence of a local health department, higher median income, larger hospital size, health care system membership, and metropolitan-located hospitals, had relatively weak associations.
This study suggests that consultants, at least at this point in time, may be better suited to lead CHNAs than hospital staff. Implementation strategies were largely composed of activities in which hospitals were already involved. It is not known to what extent hospitals may build on or expand these activities, rather than continue business as usual. Many implementation strategies reports included Medicaid 1115 Waiver Delivery System Reform Incentive Program (DSRIP) or similar-type projects. As the name suggests, hospitals are incentivized for meeting DSRIP project milestones. Although some DSRIP project areas have potential to have an impact on population health, the vast majority are medical interventions focused on patient care.40
There is a newfound focus on population health in health care through various policies and health care initiatives, including the Institute for Healthcare Improvement triple aim, primary care and public health integration, accountable care organizations, the Prevention and Public Health Fund, and new community benefit requirements for nonprofit hospitals. The CHNA process provides an opportunity to begin making such improvements.19,41,42 Nonetheless, hospitals struggled to include issues and strategies that reflected and accounted for broader social determinants of health. Hospital reports also performed poorly in identifying evidence-based strategies to address health issues, considering local contextual factors, and examining contributing causes to problems. If we are to hope for population health improvements through these methods, hospitals cannot continue to do what they are doing.
This is not to suggest that hospitals select issues and implement strategies outside the scope of their mission and capabilities. Health and social issues are intertwined and incredibly complex. Single-shot, unilateral approaches to health issues are not generally successful.43 Policies and programs, largely because of lack of resources, frequently offer superficial solutions to these issues rather than recognize that health issues are the result of social and economic inequalities.
Rosenberg referred to hospitals as “a necessary community institution strangely insulated from the community.”44(p349) Findings from this study provide support for the federal government to strengthen and clarify the final CHNA guidance and the purpose of the CHNA and implementation strategies processes and reports. Furthermore, mandating the use of a public health framework through the CHNA regulations can better align processes with other assessment and planning activities (e.g., federally qualified health center needs assessment, health department accreditation), resulting in opportunities for hospitals to create strategic partnerships with public health systems, social service agencies, and other public and private agencies and organizations. These collaborations can provide opportunities for hospitals, agencies, and other organizations to escape their disciplinary silos, align assessment processes, share data and resources, enhance organizational and disciplinary strengths, and, ultimately, increase the likelihood of improving population health.
Limitations
Findings should be interpreted with caution because of the scope of the study and sample size. This study is a representation of CHNA and implementation strategies reports in Texas and accounted for approximately 53% of the nonprofit hospital population in Texas. Though limited to Texas, we think these results are applicable to nonprofit hospitals in other states.
The variables in the regression model have wide confidence intervals. A larger sample of hospitals, preferably in multiple states, should be used to replicate the study. Relationships in the regression model are limited to the independent variables included; additional hospital, community, and other factors important to report quality may not have been considered. Such contextual factors will be investigated in greater depth in upcoming case study research.
This was an initial step in evaluating reports with 1 primary evaluator and 2 secondary evaluators for 10% of the reports. In future studies, 1 or more independent evaluators should review reports to measure and enhance interrater reliability.
Future Implications
Future studies should validate the public health framework and the scale as a method for evaluating CHNA and implementation strategies reports. Upon validation and study replication, a tool could be developed, tested, and further validated to evaluate assessment and planning processes, which could be used for tracking and reporting by practitioners and policymakers.
Acknowledgments
A special thank you to J. Charles Huber Jr for his data analysis and interpretation assistance and guidance. We are grateful to Gary M. Young for providing his expertise on the Internal Revenue Service community health needs assessment regulations as well as feedback on the evaluation framework. Thank you to Patricia Calzada and Bernard Appiah for providing secondary reviews and scoring of assessment and planning reports.
Human Participant Protection
Approval for this study was granted by the Texas A&M human subjects protection program office of research compliance.
Facebook
Twitter
Google+