Association Between Hospital Consolidation and Loss of Pediatric Inpatient Services

Abstract

This study examines the association between hospital consolidation and loss of pediatric inpatient services.

Inpatient pediatric services in the US have become increasingly centralized in recent decades.^1,2 This trend is in part due to closure of pediatric inpatient units among small, rural hospitals.³ An important yet unexamined contributing factor might be the recent acceleration in hospital consolidation.⁴ As hospitals consolidate, they may close pediatric units to reduce costs and avoid duplication of services. Alternatively, hospital consolidation may allow small hospitals to keep inpatient units open through efficient sharing of pediatric-focused resources and expertise.⁵ We sought to characterize whether hospital consolidation, as measured by newly reported membership in a health system, is associated with the loss of inpatient pediatric services.

Methods

We performed a longitudinal study of US hospitals using the 2011 to 2020 American Hospital Association annual surveys. We included all general and pediatric hospitals located in the 50 states and Washington, DC, that responded to the American Hospital Association survey in any of these years. We first categorized hospitals by their self-reported provision of pediatric inpatient services and measured the prevalence of those services over time. We then performed a longitudinal analysis in the subset of hospitals that provided inpatient pediatric services during any year in which they were not already a member of a health system. Hospitals were followed up until they joined a system or until the end of the study. We used logistic regression with generalized estimating equations to analyze the association between joining a health system and loss of inpatient pediatric services within 5 years. We also performed 2 sensitivity analyses. First, we repeated the analysis while varying the follow-up length. Second, we fit a Cox proportional hazards model, limiting the analysis to hospitals eligible in the first study year (2011) and treating system membership as a time-varying covariate. All models controlled for hospital size and status as a critical access hospital. We fit models using R version 4.2.0 (The R Project). We reported this study in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.⁶ This study used only publicly available hospital level data and therefore was deemed not to be human participants research by the University of Pittsburgh institutional review board.

Results

The study team examined 5104 unique hospitals representing 46 841 hospital-years. Provision of inpatient pediatric services declined over time, from 2023 of 4876 in 2011 (41.5%) to 1485 of 4551 in 2020 (32.6%) (Figure 1). The primary analysis included 1088 unique hospitals and 3862 hospital-years with 235 hospitals joining a system during the study period (21.6%). Joining a hospital system was associated with a loss of inpatient pediatric services within 5 years (adjusted odds ratio, 1.57; 95% CI, 1.26-1.96; P < .001). In the sensitivity analyses varying the follow-up period, this association remained statistically significant at 4 years but not at shorter follow-up lengths, although all point estimates were similar (Figure 2). In the proportional hazards analysis, the association between joining a health system and loss of inpatient pediatric services was not significant (adjusted hazard ratio, 1.23; 95% CI, 0.95-1.61; P = .13), although the point estimate was qualitatively consistent with the primary analysis.

Figure 1. Provision of Inpatient Pediatric Services by the American Hospital Association Survey Respondents, 2011 to 2020.

Figure 2. Association Between Hospital Consolidation and Loss of Pediatric Inpatient Services, by Follow-up Length in Years.

Discussion

Fewer US hospitals are providing inpatient pediatric services over time. Hospital consolidation, as measured by newly reported membership in a health system, was associated with closure of existing inpatient pediatric services within 5 years, potentially exacerbating this trend. These findings are limited by the nature of a survey-based dataset, which is vulnerable to non-response bias, though the response rate for this survey is annually quite high. Future research should explore how inpatient unit closures are affecting observed access to pediatric acute care, particularly in the context of known geographic variation in access. Research should also address whether the effects of consolidation vary by the different ways hospitals may consolidate (eg, buyouts vs mergers), as well as the effects of consolidation on family experiences, child health outcomes, and costs. Policy makers should consider the potential loss of inpatient pediatric services when reviewing proposed hospital mergers and acquisitions.

Supplement 1.

Data sharing statement