Readmissions after pancreatic and liver surgery

Abstract

Background: Hepatopancreaticobiliary surgery accounts for the most complex gastrointesitinal surgery performed. With the advent of advancing technical skills, application of laparoscopic techniques and increased acceptance of complex surgical technique, the complexity of these cases continue to increase and evolve. When pursuing such operations, it is critical to evaluate the morbidity associated with these procedures. A surrogate marker of this can be seen with the need for readmission after comple HPB surgery. Herein, we report our single instiution data on the readmission rates and measures to address this. Methods: We conducted a review of readmissions under the division of HPB surgery for pancreatic and liver surgery between 2014–2017. Our focus was on patients who were readmitted within 30 days of their index operation. Detailed retrospective chart review was performed with respect to patient clinical demographics, operative details, morbidity, postoperative complications, readmission date as well as readmission. The cohort was then divided into patients who were admitted within 7 days of dismissal to those who were readmitted after 7 days. Further, amongst patients with readmissions, we identified patients who had short term (≤ 3days) for reason for readmission. Results: There were a total of 124 pancreatic surgery readmissions of 847 operations performed (14.6%). Of these, 63 patients had a length of stay for their index hospitalization of ≤7 days. Of these patients, 27 (42.8%) were readmitted within one week of dismissal and 23 (85%) of these patients had a readmission length of stay ≤3 days. Reasons for short term readmission (≤3 days) were pancreatic fistula and/or intra-abdominal fluid collection requiring percutaneous drainage in 11 patients (48%), dehydration in 4 patients (17%), wound infection and delayed gastric emptying in 3 patients respectively (13%) and wound infections in 2 patients (8%). 61 of 124 patients (49%) had an original length of stay at index hospitalization of > 7 days and 24 (39.3%) had a readmission within 7 days. There was a trend toward significance in the length of initial hospitalization influencing readmission comparing 7 days, although this did not reach statistical significance (p=0.07). There were a total of 65 liver readmissions of 765 operations performed (8.5%). Of these, 38 patients had a length of stay for their index hospitalization of ≤7 days. Of these 38 patients, 19 (38.0%) were readmitted within 7 days and 18 patients (47%) had a short term length of stay and 14(73.7%) had a readmission LOS of under 7 days. Twenty seven patients had a LOS for their index hospitalization of > 7 days and 15 (44.1%) had a readmission within 7 days. Again, no statistically significant difference was found between patients who were admitted for 7 days at their index hospitalization with respect to readmission (p = 0.68). Conclusion: Much emphasis has been placed on readmissions following HPB surgery including assessment of risk factors, and development of validated risk scores. We have shown that for patients admitted within 7 days of dismissal majority of patients required short term readmissions of ≤3 days. Identifying these patients can help target a population of who will benefit from preplanned tests and imaging procedures after their dismissal with the eventual goal to reduce short term readmissions, patient dissatisfaction, and hospital cost.