Predicting Surgery Length of Stay of Appendicised Patients in a Hospital Setting Using Cox-Proportional Hazard and Weibull Regression Models

Accurate length of stay (LOS) prediction for appendicitis patients is critical for resource optimization in constrained healthcare systems. This study compares Cox proportional hazards (Cox PH) and Weibull regression models for predicting LOS using clinical and demographic variables. A multivariable survival models were developed incorporating age, gender, residence, comorbidities, appendicitis severity, and operative factors. The performance was assessed via hazard ratios (HR), time-dependent AUC, and survival function calibration.The cohort was predominantly young adults (58.3% aged 16-30) with mild appendicitis (85.2%). Weibull regression better captured LOS distribution (right-skewed, mode: 2- 4 days) versus Cox PH. Therefore, the Weibull regression provides superior LOS prediction in appendicitis care, identifying disease severity and residence as critical determinants. The day 3-4 postoperative window represents the optimal period for discharge interventions. Implementation of severity-stratified pathways could reduce LOS by 38% in mild urban cases while directing resources to high-risk rural patients with complicated disease