This study examined approaches that nonprofit hospitals use to evaluate community benefit activities in the Community Health Needs Assessment/Implementation Plan (CHNA/IP) process. Content analysis of CHNAs/IPs completed between 2018 and 2021 from a 20% stratified random sample (n = 503) of US nonprofit hospitals. A coding sheet was used to record details about the evaluation content reported by hospitals in their CHNAs/IPs. Evaluation was coded into 4 categories: (1) no mention of evaluation; (2) description of evaluation without reporting any measures; (3) reporting reach (number of people served) only; and (4) reporting social/health outcomes. For logistic regression analyses, categories 1 and 2 were grouped together into "no evaluation measures" and categories 3 and 4 were grouped into "evaluation measures" for binary comparison. Multinomial logistic regression was also used to individually examine categories 3 and 4 compared with no evaluation measures. While a majority of nonprofit hospitals (71.4%, n = 359) mentioned evaluation in their CHNAs, almost half (49.7%, n = 250) did not report any evaluation measures. Among the 50.3% (n = 253) of hospitals that reported evaluation measures, 67.2% (n = 170) only reported reach. Fewer than 1 in 5 hospitals (16.5%, n = 83) reported social/health outcomes. Hospitals that hired a consultant (adjusted odds ratio [AOR] = 1.61; 95% confidence interval [CI], 1.08-2.43) and system members (AOR = 1.76; 95% CI, 1.12-2.75) had higher odds of reporting evaluation measures. Using hospitals that reported no measures as the base category, system members (AOR = 7.71; 95% CI, 2.97-20.00) also had significantly higher odds of reporting social/health outcomes, while rural locations had lower odds (AOR = 0.43; 95% CI, 0.20-0.94). Although hospitals are required to evaluate the impact of actions taken to address the health needs identified in their CHNAs, few hospitals are reporting social/health outcomes of such activities. This represents a missed opportunity, as health/social outcomes could be used to inform the allocation of resources to maximize community benefits and the expansion of successful community initiatives.
Read full abstract