Unwarranted variation in training in isolated coronary artery bypass grafting in the United Kingdom

Several studies have reported comparable outcomes between trainees and consultants performing isolated coronary artery bypass grafting (CABG). However, the extent to which trainee operating opportunities vary between consultants and hospitals after accounting for patient risk profile remains unclear.

Adult patients undergoing first-time, elective or urgent isolated CABG in the United Kingdom between 2010 and 2019 were identified from the National Adult Cardiac Surgery Audit database. Procedures in which a trainee was recorded as the primary operator were compared with those with consultant. Three-level multilevel logistic regression models quantified hospital- and consultant-level variation after adjustment for perioperative factors.

A total of 127,795 patients were included; a trainee was recorded as the primary operator in 32,997 cases (25.8%). After propensity score matching, trainee primary operator cases had longer bypass and cross-clamp times (85 vs 81 minutes and 50 vs 48 minutes, respectively; both p < 0.001) but were not associated with higher in-hospital mortality (0.9% vs 1.2%), postoperative dialysis (1.4% vs 1.8%), or return to theatre (3.7% vs 3.8%).In the final multilevel model, residual variation in trainee primary operator status was observed at both hospital and consultant levels. Hospitals accounted for a larger proportion of variation (variance partition coefficient [VPC] 43.7%) than consultants (VPC 25.1%), with the rest of the variation attributed to patient preoperative characteristics. The median odds ratio (MOR) was 7.76 at the hospital level and 4.72 at the consultant level, indicating that, for two otherwise similar patients, moving from a lower- to a higher-training hospital would multiply the odds of trainee primary operator involvement by almost eight.

Substantial unwarranted variation exists in trainee operating opportunities for isolated CABG at the hospital and consultant levels in the United Kingdom. This variation persists after adjustment for patient case mix and operative factors, supporting the need for more standardised, transparent, and equitable monitoring of operative exposure across institutions.