BACKGROUND: Cardiac surgery lacks a uniform method for reporting postoperative morbidities, limiting the comparison of studies and meta-analyses. In other surgical specialties, the Clavien-Dindo classification (CDC) was successfully implemented as an outcome reporting method with an equally promising derivative, the Comprehensive Complications Index (CCI). This study aims to verify whether these classifications accurately represent the complexity of postoperative outcomes in cardiac surgery.
METHODS AND RESULTS: Data on 38,257 adult patients who underwent cardiac surgery at six university hospitals between October 2010 and December 2018 was prospectively collected in a provincial clinical database. Adaptations to the CDC were made for medically treated complications in cardiac surgery requiring antiarrhythmics, vasopressors, and transfusion of 1 or 2 units of blood, which were recoded as grade I complications. Postoperative complications until hospital discharge were recoded using the associated CDC grade as determined by an expert consensus of attending cardiac surgeons, and the CCI was calculated. Spearman’s rho was performed to evaluate correlations between the CDC and CCI with hospital stay, ICU stay, and surgery times. In this cohort (73% men, mean age 67±11 years), 72% of patients underwent coronary artery bypass graft surgery and/or valve surgery in 41% of patients. Other operations constituted 12% of cases. Median hospital stay was 6 days [Q1, Q3: 5, 9], while ICU stay was 2 days [1, 3]. The CDC grading was: none in 22.5%, grade I in 26.5%, grade II in 24.1%, grade IIIa in 7.1%, grade IIIb in 1.5%, grade IVa in 13.6%, grade IVb in 1.6%, and grade V in 3.1%. Median CCI was 21 [9, 34]. Only 6 patients (0.016%) had a score superior to 100. The CDC and CCI had a correlation of 0.976. When removing cases of patient deaths, the CDC and CCI correlated with ICU stay by 0.474 and 0.517 (p < 0.001), respectively, whereas correlations with hospital LOS were 0.577 and 0.620 (p < 0.001), respectively. The CDC and CCI also positively correlated with surgery duration (0.243 and 0.254, p < 0.001), cardiopulmonary bypass time (0.247 and 0.263, p < 0.001), and cross-clamp time (0.203 and 0.218, p < 0.001).
CONCLUSION: The CDC and CCI strongly correlated with hospital and ICU stay. They also correlated with surgery times, suggesting that longer, more complex operations are associated with increased incidence of postoperative complications. This adapted version of the CDC accurately reflects the complex postoperative course of adult cardiac surgery patients with one simple scale and has direct, useful applications to outcome reporting and quality improvement initiatives.