Drivers of Pediatric Readmissions: Clinical Complexity and Social Factors in a Multicenter Analysis

Hospital surgical readmission within 30 days is a key institutional performance metric and a major contributor to health care costs.1,2 Existing studies focus on specific procedures or on clinical risk factors, with limited attention to the interaction between clinical complexity and social determinants of health.3–5This study examined 30-day pediatric surgical readmissions across 50 US hospitals to identify clinical and social predictors. Socioeconomic context was assessed using the Child Opportunity Index (COI), a neighborhood-level proxy for child-relevant opportunity structures across access to health, education, environmental resources, and socioeconomic domains.6 This study seeks to address gaps in understanding the drivers of pediatric surgical readmissions, thereby informing clinical workflows and quality improvement interventions.A retrospective cohort study was conducted using Pediatric Health Information System (PHIS) data for pediatric surgical patients discharged from 50 children’s hospitals from January 1, 2021, through September 30, 2024 (N = 303 876). Exclusions included emergency visits, ambulatory surgery, observation stays, and nonoperative encounters. Cardiac surgical and psychiatric cases were excluded because of their complexity and high readmission rates.7The primary outcome was 30-day all-cause readmission.Descriptive statistics characterized the cohort. Missing data were excluded from analysis, and denominators are presented for variables with missingness. Because of large sample sizes, statistical significance for differences in readmission rates was not assessed by univariable analysis. Multivariable logistic regression identified independent predictors of readmission adjusting for confounding, including demographics (age, sex, race, and ethnicity), insurance primary payor, household income, COI, hospital region, complex chronic conditions (CCCs), and surgical category. Variables in the multivariable analysis were selected a priori based on clinical interest and included as categorical covariates. Multicollinearity was assessed using variance inflation factors. Mixed-effects logistic regression with a random effect for PHIS hospital was attempted; however, the model did not converge. Results are presented as adjusted odds ratios (AORs) and 95% CIs, with P less than .05 considered significant. Analyses were performed using Stata softwareOf 303 876 patients, 28 718 (9.5%) were readmitted within 30 days. Children (11.8%) and neonates (11.5%) had higher readmission rates than adolescents (8.3%). Pacific Islander patients had the highest rate (11.7%), and self-pay patients had the lowest (6.4%). Children from the highest-income households had lower readmission rates (8.8%) (Table 1).Hospitals in the Northeast had the highest readmission rate (10.5%), and those in the West and Southwest had the lowest (8.9%).Self-pay patients were significantly less likely to be readmitted (AOR, 0.87; 95% CI, 0.78–0.96; P = .006). Procedures involving the lymphatic and hemic system (AOR, 2.12; 95% CI, 1.91–2.35; P < .001) and vascular system (AOR, 2.01; 95% CI, 1.85–2.18; P < .001) showed the highest readmission odds. CCCs, particularly malignancy (AOR, 4.47; 95% CI, 4.31–4.64; P < .001) and gastrointestinal conditions (AOR, 1.67; 95% CI, 1.62–1.73; P < .001), were the strongest predictors (Table 2).Nearly 1 in 10 pediatric surgical patients were readmitted within 30 days, underscoring ongoing challenges in postoperative continuity and discharge planning. Clinical and procedural complexity were the key drivers. Consistent with the published literature, younger children, children with CCCs, and those undergoing high-intensity procedures are particularly vulnerable8–10 for readmission. Malignancy was the strongest predictor of readmission (AOR, 4.47). However, this finding should be interpreted with caution because our all-cause outcome measure could not distinguish unplanned readmissions from planned admissions, such as those for chemotherapy.Self-pay and privately insured patients had lower odds of readmission. Although we observed differences in readmission patterns across racial and ethnic groups, these findings should be interpreted with caution. Race and ethnicity are complex social constructs that may reflect broader structural, socioeconomic, and health care system factors; however, these elements were not directly measured in our analysis.Explanations for lower readmissions among children residing in lower COI neighborhoods, such as reduced access to follow-up care, remain speculative. Without direct measures of health care use, access barriers, or patient-level social stressors, it is difficult to determine whether observed differences reflect true variation in outcomes or differences in care-seeking patterns.The limitations of this study include analyzing all causes rather than unplanned or planned readmissions and the lack of hospital-level variables. Cardiac surgical and psychiatric cases were excluded because of their complexity and inherently higher readmission rates.The COI is not specifically designed to assess health care system factors or access to specialty surgical follow-up. Structural barriers including insurance coverage nuances, care coordination gaps, and hospital-specific discharge practices are not incorporated into the index.These findings highlight opportunities for improved perioperative risk stratification and targeted quality improvement. Lower readmission rates among self-pay patients may reflect barriers to return care rather than fewer complications, underscoring the need for strong discharge education and proactive follow-up.Higher readmission risk after lymphatic/hemic and vascular procedures points to priority areas for enhanced postoperative surveillance, whereas children with CCCs, especially malignancy and gastrointestinal disease, require coordinated, multidisciplinary perioperative care and individualized discharge planning.This study identified the independent association among key demographics, socioeconomic and clinical factors, and 30-day readmission. These findings support a risk-stratified, equity-informed perioperative care model aimed at reducing avoidable readmissions.