Development and validation of multivariable predictive models for recurrence and mortality following nonoperative management of sigmoid volvulus

Sigmoid volvulus is a challenging condition, and deciding between elective surgery or expectant management can be complex. The aim of this study was to develop a tool for predicting the risk of recurrent sigmoid volvulus and all-cause mortality within 1 year following initial nonoperative management. This is a retrospective cohort study using Medicare claims data from 2016 to 2018 of beneficiaries admitted urgently/emergently for volvulus, undergoing colonic decompression and discharged alive without surgery (excluding those discharged to hospice). The primary outcomes were recurrent sigmoid volvulus and all-cause mortality within 1 year. Proportional hazards models and logistic regression were employed to identify risk factors and develop prediction equations, which were subsequently validated. Among the 2078 patients managed nonoperatively, 36.1% experienced recurrent sigmoid volvulus and 28.6% died within 1 year. The prediction model for recurrence integrated age, sex, race, palliative care consultations and four comorbidities, achieving area under the curve values of 0.63 in both the training and testing samples. The model for mortality incorporated age, palliative care consultations and nine comorbidities, with area under the curve values of 0.76 in the training and 0.70 in the testing sample. This study provides a straightforward predictive tool that utilizes easily accessible data to estimate individualized risks of recurrent sigmoid volvulus and all-cause mortality for older adults initially managed nonoperatively. The tool can assist clinicians and patients in making informed decisions about such risks. While the accuracy of the calculator was validated, further confirmation through external validation and prospective studies would enhance its clinical utility.