Predictors of Early Postoperative Complications Following Pediatric Circumcision: A Multicenter Retrospective Cohort Study

Abstract

Background: Circumcision is one of the most common procedures performed in male children. Although it is generally considered safe, some children develop complications soon after the procedure. The factors that contribute to these complications are still not fully understood, particularly across the different techniques and medical or non-medical reasons for circumcision.

Objective: To determine independent predictors of early postoperative complications following pediatric circumcision, and to derive an internally validated four‑item complication risk score.

Methods: We reviewed the records of 850 male children aged 0–12 years who underwent circumcision at five centres over a period of 36 months. The study focused on complications that occurred within 30 days after the procedure, including bleeding, infection, wound dehiscence, injury to the glans or meatus, poor cosmetic results, and urinary retention. The factors that remained significant in the multivariable logistic regression were each given one point. These points were then combined to create a simple risk score, which was validated using bootstrap resampling.

Results: Of 850 procedures, 48 (5.6%, 95% CI 4.3–7.4%) had an early postoperative complication, rising sharply with age from 0.0% in neonates to 6.8% in older children (p=0.018). Age>1 year (aOR 4.2), trainee operator (aOR 3.8), surgical technique (aOR 3.0), and medical indication (aOR 2.5) independently predicted complications (all p≤0.014); general anaesthesia did not (p=0.312) and was excluded. The four‑item score achieved an apparent AUC of 0.73 (0.65–0.81), optimism‑corrected to 0.68. At the Youden‑optimal cut‑off (≥3), sensitivity was 79.2% and specificity 56.1%; a higher cut‑off (≥4) raised specificity to 91.5% and positive predictive value to 20.0%, with sensitivity 35.4%.

Conclusion: In our study, early complications after circumcision were more common among older children, those undergoing the procedure for medical reasons, those treated with a surgical rather than device-assisted technique, and those operated on by trainees. We developed a four-item risk score based on these factors, but it needs to be tested in other groups of children before it can be used in routine clinical practice.

Journal of Paediatric Surgeons of Bangladesh (2026) Vol. 17 (2): 18-24