The transition to data-driven quality metrics: determining the optimal surveillance period for complications after surgery

Muhammad Ali Chaudhary, Wei Jiang, Stuart Lipsitz, Zain G. Hashmi, Tracey P. Koehlmoos, Peter Learn, Adil H. Haider, Andrew J. Schoenfeld

Research output: Contribution to journalArticlepeer-review

7 Citations (Scopus)

Abstract

Background: Thirty-day complications frequently serve in the surgical literature as a quality indicator. This metric is not meant to capture the full array of complication resulting from surgical intervention. However, this period is largely based on convention, with little evidence to support it. This study sought to determine the optimal surveillance period for postsurgical complications, defined as the shortest period that also encompassed the highest proportion of postsurgical adverse events. Methods: TRICARE data (2006-2014) were queried for adult (18-64 y) patients who underwent one of 11 surgical procedures. Patients were assessed for complications up to 90 d after surgery. Kaplan–Meier curves, linear spline regression models at each incremental postsurgical day, and adjusted R-squared values were used to identify critical time point cutoffs for the surveillance of complications. Optimal length of surveillance was defined as the postsurgical day on which the model demonstrated the highest R-squared value. A supplemental analysis considered these measures for orthopedic and general surgical procedures. Results: One lakh ninety-eight patients met the inclusion criteria. A total of 21.8% patients experienced at least one complication during the follow-up period, with 59% occurring within the first 15 d. Kaplan–Meier curves for complications showed a demonstrable inflection before 20 d and 14-15 d possessed the highest R-squared values. Conclusions: In this analysis, the optimal surveillance period for postsurgical complications was 15 d. While the conventional 30-d period may still be appropriate for a variety of reasons, the shorter interval identified here may represent a superior quality measure specific to surgical practice.

Original languageEnglish
Pages (from-to)332-337
Number of pages6
JournalJournal of Surgical Research
Volume232
DOIs
Publication statusPublished - Dec 2018
Externally publishedYes

Keywords

  • Complications
  • Quality metrics
  • Surgical quality

Fingerprint

Dive into the research topics of 'The transition to data-driven quality metrics: determining the optimal surveillance period for complications after surgery'. Together they form a unique fingerprint.

Cite this