Resection Rates and Predictors of Resectability of Pancreatic Tumors at Mulago Hospital: A Retrospective Cross-Sectional Study

Background: Pancreatic tumors are among the most lethal malignancies globally, with surgical resection being the only curative option. However, in low-income countries, most patients present at advanced stages, limiting surgical eligibility. Accurate preoperative prediction of resectability is essential for effective surgical planning. This study evaluated clinical, biochemical, and radiological predictors of pancreatic tumor resectability at Mulago National Referral Hospital (MNRH). Methods: We conducted a retrospective cross-sectional study of 100 patients diagnosed with pancreatic tumors between January 2021 and December 2024. Demographic, clinical, biochemical, and radiological data were collected. Resection rates were calculated. Associations between independent variables and resectability were first explored using independent-samples t-tests and crude odds ratios. Predictors were then assessed with univariate and multivariate logistic regression models to obtain adjusted odds ratios. P < 0.05 at 95% confidence interval were considered to be statistically significant. Results: The overall resection rate was 21%. Although 55% of patients were deemed resectable on imaging, 61.8% were found unresectable intra-operatively due to vascular invasion or metastases. Most tumors (92%) were located in the pancreatic head, and 51% were stage III or IV. In multivariate analysis, only tumor size >4 cm was significantly associated with irresectability (AOR = 0.054, 95% CI: 0.015-0.193, p < 0.001). Conclusion: Despite favorable imaging, many patients are unresectable at surgery. Tumor size is an independent predictor of resectability. Enhanced diagnostic imaging is needed to improve preoperative assessment in low-resource settings.