Comparison of two postoperative complication grading systems after treatment of stifle and shoulder instability in 68 dogs

OBJECTIVE: (1) To adapt and apply the Clavien–Dindo (aCD) postoperative complication grading system to dogs experiencing complications following a single orthopedic procedure. (2) To compare the reliability of the Clavien–Dindo system to the Cook complication grading system. STUDY DESIGN: Retrospective study. SAMPLE POPULATION: Sixty‐eight client‐owned dogs. METHODS: Scenarios derived from complications following TightRope stabilization of the stifle and shoulder were graded by four ACVS‐boarded surgeons using two systems; the Cook 3‐point scale and the aCD 5‐point scale. Because the aCD system distinguishes complications from outcomes (“sequelae” or “failure to cure”), two data sets were created: one with (n = 76) and without (n = 67) inclusion of “sequelae” and “failure to cure” cases. Interobserver reliability was evaluated using intraclass correlation coefficient (ICC) calculations. RESULTS: Seventy‐six scenarios from 68 records were evaluated. The ICC of the aCD system was 0.620 consistent with moderate reliability. The reliability of the Cook system was good, with an ICC of 0.848. Exclusion of cases with “sequelae” or “failure to cure” resulted in excellent reliability of the aCD system (ICC = 0.975) and good reliability of the Cook systems (ICC = 0.857). CONCLUSION: The aCD grading system was less reliable than the Cook system when evaluating all cases but more reliable when evaluating cases of complications excluding “sequelae” and “failures to cure”. CLINICAL SIGNIFICANCE: The Cook grading system is reliably good in grading postoperative complications in dogs. The aCD system can also be used to assess postoperative complications with excellent reliability but is less reliable when distinguishing complications from other postoperative outcomes.