Inpatient observation after phototherapy ≥35-week newborns: a retrospective study of bilirubin rebound

Introduction: phototherapy (PT) is an effective treatment for neonatal hyperbilirubinemia; however, controversy persists regarding the need for in-hospital observation after discontinuation to detect possible “rebound.” The aim was to evaluate the frequency with which hospitalization is prolonged solely for bilirubin monitoring after discontinuing PT and its usefulness in identifying clinically significant rebound. Material and methods: a retrospective analytical cohort study was conducted in a tertiary hospital (2019–2022), including newborns ≥35 weeks admitted for hyperbilirubinemia and treated with PT. Two strategies after discontinuation of PT were compared: inpatient observation versus discharge with outpatient follow-up. The primary outcome was the detection of rebound, defined as the return of bilirubin to the phototherapy threshold within 72-96 hours. Results: 155 infants were included; 60 (38.7%) remained hospitalized for monitoring and 95 (61.2%) were discharged with outpatient follow-up. Baseline characteristics, bilirubin levels, and duration of PT were similar between groups. The total hospital stay was longer in the observation group (mean 2.57 days) than in the immediate discharge group (1.75 days), but this difference was not statistically significant (p=0.053). The overall incidence of rebound was 3.2% (5/155): one case was detected during in-hospital observation and four during outpatient follow-up (48–72 hours). No adverse events were reported. Conclusions: clinically significant rebound after phototherapy discontinuation was uncommon and rarely detected during immediate inpatient observation. In selected low-risk term newborns with reliable outpatient follow-up, routine prolongation of hospitalization for rebound monitoring may be unnecessary, whereas higher-risk infants may still require closer surveillance. These findings should be interpreted cautiously given the retrospective single-center design and the limited number of rebound events.