A multicenter case-control pilot study was conducted (71/213) to know the predictive value of two risk scales for nosocomial wound infections, NNIS and Toronto score, applied to coronary artery bypass graft surgery in six hospitals in Santiago- Chile, from 1998 to 2002. A straight correlation of wound infection risk and Intensive Care Unit stay beyond three days, sub-group of the Toronto scale, was identified. Toronto scale, according to its four groups of clinical risk, is therefore an effective tool to adjust rates and make comparisons between hospitals that perform this type of surgery