Background: In medically and socially complex patients, population health programs that address health-related needs can impact utilization of the emergency department (ED) and inpatient hospitalizations. Understanding the impact of and predictors for preventable ED visits in the spina bifida (SB) population may help guide population health strategies for these patients requiring complex medical care.
Methods: We reviewed patients (<18-years-old) in Pediatric Health Information System (PHIS) with ICD-10-CM diagnosis code (Q05x) for SB who presented for ED, observation, or inpatient visits from 2016-2023. Absence of clinical and imaging charges was used to define low-value ED visit (i.e. care that could be delivered in lower-cost settings). This was compared to control patients with identical criteria except diagnosis of N13 for obstructive and reflux uropathy. Demographic variables, prior year ED utilization, encounter diagnoses, and hospital characteristics were compared between low-value ED vs non-low-value ED visits. Mixed effects logistic regression was used to model odds of low-value ED visit and adjust for repeat encounters.
Results: 20.7% of SB ED visits were categorized as low-value versus 12.0% in ED visits of the non-SB genitourinary controls (p<0.001). Estimated costs for all ED-related visits by the SB cohort was $225,294,630, with low-value visits accounting for 3.8% ($8,491,144) of the total. Patients presenting for low-value SB ED visits were more likely younger, Hispanic/Latino, non-White, with public insurance, lower household income, urban/suburban zip code, genitourinary encounter diagnosis, from west census region hospitals, and from hospitals with higher volume of SB ED visits (all p<0.001), Table 1. On multivariable mixed effects logistic regression, low-value ED visits were associated with younger age (OR 1.05, 95% CI 1.04 - 1.06, p<0.001), Hispanic/Latino ethnicity (OR 1.21, 95% CI 1.06 - 1.39, p=0.006), black (OR 1.35, 95% CI 1.16 - 1.58, p<0.001) and other (OR 1.15, 95% CI 1.00 - 1.33, p=0.044) race, public insurance (OR 1.14, 95% CI 1.01 - 1.29, p=0.033), and genitourinary encounter diagnosis (OR 1.16, 95% CI 1.04 - 1.30, p=0.009), Table 2. Gender and prior ED utilization was not associated with low-value ED visit (p=0.44). Odds of low-value ED visit across hospitals for a standard patient ranged from 0.31 to 5.36 (Figure 1).
Conclusions: Across PHIS hospitals, 1 in 5 ED visits by pediatric SB patients are low-value, which is significantly higher than non-SB populations. Younger age, Hispanic/Latino ethnicity, black and other race, public insurance, and genitourinary encounter diagnosis were all associated with higher odds for low-value ED visit. Finally, there was wide variation in odds for low-value ED visit across hospitals that warrants further investigation to elucidate best practices for healthcare delivery systems.
