Application of the New York State PTCA mortality model in patients undergoing stent implantation

This study applied the New York State conventional coronary angioplasty (PTCA) model of clinical outcomes to evaluate whether it has relevance in the current era of stent implantation. The model was developed in 62 670 patients treated with conventional PTCA from 1991 to 1994 to risk adjust mortality and bypass surgery after PTCA. Since then, stents have become the dominant form of intervention. Whether that model remains relevant is uncertain.

All patients undergoing stenting at the Mayo Clinic from 1995 to 1998 were analyzed for in-hospital mortality, bypass surgery performed after attempted stenting, and longer-term mortality. No patients were excluded. The New York model was used to risk adjust and predict in-hospital and follow-up mortality. There were 3761 patients with 4063 procedural admissions for stenting; 6,472 target vessel segments were attempted, and 96.1% of procedures were successful. With the New York multivariable risk factor equation, 79 in-hospital deaths were expected (1.95%); 66 deaths (1.62%) were observed. The New York model risk score in a logistic regression model was the most significant factor associated with in-hospital mortality (OR, 1.86; P<0.001). During a mean follow-up of 1.2+/-1.0 years, there were 154 deaths. Multivariable analysis documented 6 factors associated with subsequent mortality; New York risk score was the most significant (chi(2)=16.64, P=0.0001).

Although the New York mortality model was developed in an era of conventional angioplasty, it remains relevant in patients undergoing stenting. The risk score derived from that model is the variable most significantly associated with not only in-hospital but also longer-term outcome.