Presenter: David Brauer MD, MPHS | Washington University, St. Louis Background: The surgical management of hepatopancreatobiliary (HPB) and gastric malignancies regularly occurs at major referral centers and is associated with a high rate of postoperative readmissions. Prior studies have shown that care fragmentation following surgery, including readmission to a hospital other than the index surgical hospital (an ‘outside hospital’ (OSH)), is associated with worse outcomes. No prior study has examined hospital-level variables associated with outcomes following readmission in this patient population. Methods: Patients undergoing HPB or gastric oncologic surgeries were identified from select State Inpatient Databases from the Healthcare Cost and Utilization Project from 2006-2014. Follow-up occurred up to the earliest of 90 days after discharge, receipt of chemotherapy, or discharge to hospice. The primary outcome was readmission mortality, with secondary outcomes of total readmission inpatient cost and total readmission inpatient length of stay. Annual hospital surgical volume was calculated using HCUP data for oncologic and non-oncologic HPB and gastric surgeries. Additional hospital-level variables were linked using the American Hospital Association Annual Survey Database. Due to data use agreements, hospitals with less than 10 visits during the inclusion period were excluded. Kruskal-Wallis test, logistic regression, and Youden’s index were used. Results: 31,526 patients were discharged following HPB or gastric oncologic surgery. 7,536 (24%) were readmitted within 90 days to a total of 636 hospitals. 28% of readmissions (n=2,124) were to OSH. For patients readmitted to OSH, 90-day postoperative mortality was 50% higher than mortality for patients readmitted to the index surgical hospital (8.0% vs 5.4%; OR 1.5, 95% CI 1.2 – 1.9). Looking solely at the hospital the patient was readmitted to, 188 hospitals had 10 or more readmissions during the inclusion period, accounting for 75% of all readmissions. Hospital size, measured by bed count, was not associated with readmission mortality, cost, or length of stay. However, a hospital's annual volume of HPB and gastric surgeries was associated with readmission mortality: patients were 37% more likely to die during their readmission if they were readmitted to a hospital with less than 100 combined annual HPB and gastric procedures (6.4% vs. 4.7%; OR 1.37, 95% CI 1.1 – 1.7). When comparing large volume (100 or more surgeries annually) and small volume hospitals, total 90-day readmission inpatient cost was 23% higher at large volume hospitals (median $15,000 vs $12,000; p<0.01). This could be a reflection of the costs of achieving rescue through surgical intervention, as significantly more surgeries were performed at larger hospitals (8.3% vs 5.3%, p<0.01). Total 90-day readmission length of stay was no different between the groups (median 6 days; p=0.82). Conclusion: Alongside sound clinical judgement, this data should be used to inform the decision-making process for where a patient should be readmitted following HPB and gastric surgery. To promote successful rescue after complications and readmission following surgery, efforts should be made to reduce care fragmentation and to direct readmissions to hospitals performing 100 or more annual HPB and gastric surgeries.
Read full abstract