Inversed relationship between completeness of follow-up and coverage of postoperative complications in gallstone surgery and ERCP: a potential source of bias in patient registers.

Abstract

To analyse the completeness in GallRiks of the follow-up frequency in relation to the intraoperative and postoperative outcome. Population-based register study. Data from the national Swedish Registry for Gallstone Surgery and Endoscopic Retrograde Cholangiopancreatography (ERCP), GallRiks. All cholecystectomies and ERCPs recorded in GallRiks between 1 January 2006 and 31 December 2014. Outcomes for intraprocedural as well as postprocedural adverse events between units with either a 30-day follow-up of ≥90% compared with those with a less frequent follow-up (<90%). Between 2006 and 2014, 162 212 cholecystectomies and ERCP procedures were registered in GallRiks. After the exclusion of non-index procedures and those with incomplete data 152 827 procedures remained for final analyses. In patients having a cholecystectomy, there were no differences regarding the adverse event rates, irrespective of the follow-up frequency. However, in the more complicated endoscopic ERCP procedures, the postoperative adverse event rates were significantly higher in those with a more frequent and complete 30-day follow-up (OR 1.92; 95% CI 1.76 to 2.11). Differences in the follow-up frequency in registries affect the reported outcomes as exemplified by the complicated endoscopic ERCP procedures. A high and complete follow-up rate shall serve as an additional quality indicator for surgical registries.