Readmissions after major gynecologic oncology surgery.

Objectives: To examine the underlying indications, timing, and risk factors associated with unplanned hospital readmissions after major surgery for a gynecologic malignancy.

Methods: This is a retrospective database cohort study utilizing the National Surgical Quality Improvement Program database (NSQIP). The association between risk factors with respect to 30-day unplanned readmission was modeled using logistic regression. Timing of readmission and the primary reason of readmission was abstracted from the database.

Results: Overall, the unplanned readmission rate was 6.5% (832/12,804). On multivariate analysis, operative time≥3h (OR 1.39, p<0.001), open abdominal surgery (OR 2.2, p<0.001), any complication prior to discharge (OR 1.6, p<0.001), two or more additional surgical procedures (OR 1.34, p=0.003), or cervical cancer as the site of primary disease (OR 1.30, p=0.05) were noted to be independent predictors of readmission. To provide a convenient calculation of overall probability of readmission, we developed a nomogram of factors significantly predicting readmission. Overall, infections were a cause of 45% of the readmissions. Surgical Site Infections were the most common reason, accounting for 29.2% of all readmissions. A majority of the readmissions (approximately 75%) were within two weeks of discharge from the hospital.

Conclusions: Efforts to reduce readmission rates should focus on identifying patients at a high risk of readmission and reducing surgical site infections. Additionally, prospective evaluation of interventions targeted at reducing readmissions should focus on the first two weeks after discharge from the hospital.