Primary Care Follow-up After Pediatric Intensive Care Unit Hospitalization

Abstract Rationale: Timely primary care follow-up after pediatric intensive care unit (PICU) hospitalization may be an important opportunity to support recovery after critical illness. However, there are no national data evaluating the rate of post-PICU follow-up. Thus, we sought to determine the rate of and factors associated with primary care provider (PCP) follow-up after PICU hospitalization. Methods: Using Medicaid and commercial claims from the T-MSIS Analytic Files and Merative MarketScan Commercial Database, we identified children (age 0-18 years) with a PICU hospitalization during January 2016-December 2019. We fit a multivariable logistic regression model predicting the probability of timely primary care follow-up (within 7 days of discharge from PICU hospitalization) with the following covariates: age (< 1; 1-4; 5-9; 10-14; 15-18 years); sex; complex chronic conditions (neuromuscular, cardiovascular, respiratory, renal, gastrointestinal, hematologic/immunologic, metabolic, genetic, malignancy, technology dependent, transplant); length of hospitalization (< 3; 3-7, 8-14, 15-30, > 30 days); acute organ dysfunction during PICU hospitalization; Social Vulnerability Index (SVI) quartile; Gini Index quartile; and quartile of PCP availability. We also report the rate of subspecialty follow-up within 7 days. Analyses were conducted separately by payer type. Results: We identified 333,931 and 41,223 PICU hospitalizations for publicly and privately-insured children, of which 149,033 (44.6%) and 19,037 (46.2%) were females; and 178,306 (53.4%) and 27,295 (66.2%) had a complex chronic condition. Overall, 78,259 (23.4%) and 15,248 (37.4%) of publicly and privately-insured children, respectively, had primary care follow-up within 7 days of PICU hospitalization discharge, while 29,774 (8.9%) and 4,099 (9.9%) had subspecialty follow-up, respectively. Among all hospitalizations, younger age, lack of complex chronic conditions, and longer hospitalizations were associated with a higher likelihood of timely PCP follow-up. Among hospitalizations for Medicaid-insured children, residence in neighborhoods with the highest quartile of the SVI (indicating greatest socioeconomic disadvantage) was associated with a higher predicted probability of timely PCP follow-up (lowest vs highest quartile of SVI: 23.1% vs 24.7%; p<0.01). Among hospitalizations for privately-insured children, residence in areas with the highest quartile of PCP availability was associated with higher predicated probability of timely PCP follow-up (lowest vs highest quartile of PCP availability; 35.7% vs 41.3%; p<0.01). Conclusions: In this US cohort of nearly 400,000 critically ill children with Medicaid and private insurance, less than 40% had PCP follow-up within 1 week of hospital discharge. Efforts to increase follow-up should focus on older children, those with chronic conditions, and those with public insurance to ensure continuity in care.