Use of routine CT-SCANS to detect severe postoperative complications after pancreato-duodenectomy.

BACKGROUND: To evaluate the performance of CT-scans performed one week after pancreato-duodenectomy (PD) to detect severe postoperative complications requiring an invasive treatment. PATIENTS AND METHODS: This monocentric retrospective study was conducted on data collected between 2005 and 2013. Patients undergoing PD underwent CT-scan with IV contrast at the end of the first postoperative week. The results of the CT-scans were analyzed to evaluate the usefulness of this procedure. The main assessment criterion was the occurrence of type-III complication (or greater) according to the Dindo-Clavien classification. RESULTS: In total, 138 patients were included. The mortality rate was 2.2%. The postoperative complication rate was 57.2%. The pancreatic fistula rate was 19.6%; 46 patients (33.3%) presented with a severe complication. A total of 138 CT-scans were analyzed: 44 (31.8%) were abnormal, 94 (68.2%) were normal. Among patients with abnormal CT-scans, 17 (39%) presented with a severe complication requiring an invasive treatment. Among the 94 patients with normal CT-scans, 14 patients (15%) presented a severe postoperative complication. Evaluation of the performance of the CT-scans at the end of the first postoperative week found a sensitivity of 55%, a specificity of 75%, a positive predictive value of 39%, and a negative predictive value of 85%. CONCLUSION: Systematic CT-scans performed at the end of the first postoperative week do not effectively detect severe complications after PD and do not help to prevent them.