Pediatric Hospital Services Within a One-Hour Drive: A National Study

This study’s data are provided through the novel Health Systems and Provider Database.⁶  Created through the Agency for Healthcare Research and Quality’s Comparative Health Systems Performance Initiative, the Health Systems and Provider Database combines data from >20 governmental and proprietary sources and contains all active US hospitals, physicians, practices, and health systems.⁶  Hospital information derives from the Centers for Medicare and Medicaid Services’ Provider of Services file, SK&A, and the American Hospital Association’s Annual Survey.⁶  We first tabulate whether each service type is present within US states and counties. We then use the 2-step floating catchment area method to calculate the proportion of counties within a one-hour drive of each service type and corresponding percentile beds per 100 000 pediatric-aged population.^7–9  This accessibility method accounts for multiple hospitals within an area, the size and density of the pediatric-aged population, and average daily drive times; it was developed to help the Department of Health and Human Services assess Health Professional Shortage Areas.^7–9  We performed calculations using Stata (Stata Corp, College Station, TX) and iGeolise software. The Institutional Review Boards of Boston Children’s Hospital and Harvard University approved this study.

In 2016, 4278 US hospitals were capable of providing at least one type of pediatric hospital service (Table 1; Fig 1). General acute care hospitals (4015) provided 75.0% of the 131 685 beds available for pediatric-aged patients in the United States, whereas children’s hospitals (263) provided 25.0%. Those living in most counties (87.0%, 89.8%, and 90.6%) could reach general newborn, emergency, and pediatric inpatient services within a one-hour drive, respectively (green areas in maps). However, those living in a minority of counties (49.5%, 39.6%, and 36.1%), were within a one-hour drive of NICU, pediatric emergency, or PICU services, respectively (blue areas in maps). Even among counties with pediatric hospital services within a one-hour drive, bed availability could vary 10-fold (lighter versus deeper shading in maps). For example, for general pediatric inpatient services, counties in the 10th percentile had a single bed per 100 000 pediatric-aged population, whereas those in the 90th percentile had 55; for PICU services, counties in the 10th percentile had a single “swing” bed (a bed that could be purposed for either general pediatric inpatient or PICU services but not both simultaneously) per 100 000 pediatric-aged population, whereas those in the 90th percentile had 15.

In this study, we illustrate the degree to which infants, children, and adolescents live within a one-hour drive of pediatric services. Aspects of our findings may be expected: general hospital-based pediatric services are more geographically distributed than their advanced counterparts. Even among those living within a one-hour drive of inpatient pediatric services, such services tend to be thinly present, which confirms the notion that our nation’s pediatric hospital service capacity can saturate quickly.¹⁰  Thus, it may be sobering to realize that the need to drive more than one hour to reach hospital-based pediatric services is more the rule than the exception. This reality should be taken into account as our nation’s hospitals continue to respond to routine and extraordinary circumstances. In future studies, researchers should examine how the geographic accessibility of pediatric hospital resources has been changing over time, especially in light of the fact that hospitals are increasingly owned by systems that can connect inpatient services to outpatient ones.