Purpose: MTL is a composite outcome measure based on routine administrative data defined as (a) postoperative mortality and/or (b) postoperative transfer to another hospital and/or (c) length of hospital stay ≥ the prespecified time period. Aim of the present study was to investigate MTL for profiling hospitals on surgical performance in colorectal cancer surgery, using data from the national registers of the German Society of General and Visceral Surgery (DGAV) and to determine the time interval for length of stay with the highest accuracy regarding major complications (Clavien-Dindo grade ≥ 3).
Methods: All patients undergoing colorectal cancer resection between January 2010 and February 2017 were included. MTL rates were calculated and compared to well-established single outcome measures using multivariate regression analysis. For each outcome measure, postoperative complications were tested regarding their predictability.
Results: Data from 14,978 patients were analyzed. Length of stay was significantly prolonged if postoperative complications occurred (p < 0.0001). Thirty-day mortality and the indication for a transfer to another hospital mainly resulted from cardiopulmonary complications. MTL occurs significantly more often than any of the single-outcome parameters. The time interval of 22 days demonstrated the highest accuracy regarding severe complications (Clavien-Dindo grade ≥ 3).
Conclusions: MTL reflects the complete spectrum of postoperative complications. Compared to individual surgical outcome parameters, MTL may have a better discriminatory power and is therefore suitable to mirror surgical quality. Because of its high accuracy regarding surgical major morbidity, 22 days is the best cut-off for length of stay within the German healthcare system.
Keywords: Colorectal cancer; Colorectal surgery; Composite outcome measure; Morbidity; Quality.