Prehospital delay, the referral pathway, and bowel resection in childhood intussusception: a 10-year retrospective cohort study

Delayed presentation is thought to raise the risk of bowel resection in childhood intussusception, but most evidence is unadjusted. We estimated the association between prehospital delay (symptom onset to hospital presentation) and bowel resection and examined the role of prior outside-hospital care. This retrospective single-centre cohort study included episodes of confirmed childhood intussusception at a tertiary children’s hospital (2016–2026). Delay was analysed per doubling and in categories (< 24, 24 to < 48, 48 to < 72, ≥72 h); the outcome was bowel resection. Relative risks (RR) were estimated by modified Poisson regression, adjusted for age and sex (primary analysis) and additionally for recorded anatomical type, which is ascertained largely through operation (secondary analysis). The E-value assessed unmeasured confounding. Among 7,675 episodes (224 resections, 2.9%), resection rates across delay categories were 1.5%, 3.1%, 3.7%, and 8.3%. The RR per doubling of delay was 1.50 (95% CI 1.35–1.68) adjusted for age and sex, 1.18 (1.09–1.27) with anatomical type added, and 1.63 unadjusted; recorded type was unspecified in 79.5% of episodes. Prior outside-hospital care (31.2% of 7,403 episodes with an ascertainable pathway) was associated with longer delay (adjusted ratio of geometric mean delay 1.58, 1.51–1.65) and with resection (5.3% versus 1.8%; age- and sex-adjusted RR 2.70, 2.07–3.54, falling to 2.15, 1.62–2.86 after adjustment for delay). Adjusting the delay estimate for referral pathway gave 1.38 (1.23–1.55). Longer prehospital delay was associated with bowel resection, but the size of the association depended on how anatomical type was handled (RR per doubling 1.18 to 1.50). Prior outside-hospital care was associated with longer delay and with resection; these observational analyses cannot establish mediation or show where delay accumulates. The findings support prospective evaluation of diagnostic and referral pathways rather than showing that any pathway change would reduce resections. Not applicable.