“…While we found the ACRI and BHIS models demonstrated good predictive power of SSI development, we concurred with other studies identifying that NNIS is a poor predictor of SSI after cardiac surgery. 13 21 Both the ACRI and BHIS models were validated in procedure-specific populations (valve 21 and CABG, 22 respectively) which may be why they performed less well than the B-SIR tool which was developed specifically to include patients undergoing various forms of cardiac surgery. This can also suggests that risk profile assessment varies based on the institutions' patient population as previously identified.…”