Logistic regression yields an adjusted odds ratio that approximates the adjusted relative risk when disease incidence is rare (<10%), while adjusting for potential confounders. For more common outcomes, the odds ratio always overstates the relative risk, sometimes dramatically. The purpose of this paper is to discuss the incorrect application of a proposed method to estimate an adjusted relative risk from an adjusted odds ratio, which has quickly gained popularity in medical and public health research, and to describe alternative statistical methods for estimating an adjusted relative risk when the outcome is common. Hypothetical data are used to illustrate statistical methods with readily accessible computer software.

For patients with multivessel disease, CABG continues to be associated with lower mortality rates than does treatment with drug-eluting stents and is also associated with lower rates of death or myocardial infarction and repeat revascularization.

Higher-volume operators and hospitals continue to experience lower risk-adjusted PCI outcome rates.

Patients undergoing coronary stenting who receive IR experience more adverse outcomes even in the era of drug-eluting stents. This has implications for choice of procedure and post-procedural monitoring.

The risk score accurately predicted in-hospital death for PCI procedures using future New York data. Its performance in other patient populations needs to be further studied.

We would like to thank Bollati et al for their interest in our study. 1 Their concerns about the validity of our study are (1) that the logistic regression analyses used to risk-adjust the outcomes are based on stepwise models, and (2) that there is no c statistic reported for the propensity model used to test for selection bias.We do not have space for a discussion of the pros and cons of stepwise analyses, but when we developed models based on all available independent variables, the results were almost identical to the findings with stepwise analyses. For example, when all variables were used in the model for subsequent revascularization, the hazard ratio changed from 1.50 (PϽ0.001) to 1.51 (PϽ0.001).With respect to point 2, as indicated in one of the references 2 cited by Bollati et al, the warning sign from the c statistic in a propensity model is a very high value (close to 1.0), because this means that there are few pairs of patients undergoing the interventions who have similar sets of risk factors. Our c statistic was 0.60, which indicates that the populations of on-pump and off-pump patients are not so dissimilar that they cannot be compared after adjustment. It should also be noted that in addition to the type of propensity analysis mentioned by Bollati et al, we matched patients exactly based on characteristics related to outcomes, and this analysis yielded very similar findings to the risk-adjusted analysis.Furthermore, Bollati et al suggest that our finding that off-pump surgery has superior short-term outcomes (lower mortality and complication rates) and worse long-term outcomes (higher subsequent revascularization rates) is counterintuitive and indicative of invalid analyses. First, because off-pump surgery is associated with more incomplete revascularization and fewer grafts per diseased vessel, it is not surprising that there were fewer short-term problems (as fewer vessels were attempted) and more longer-term problems associated with both incomplete revascularization and lower graft patency. Second, our findings are very consistent with those of randomized clinical trials referenced in our study.In conclusion, we are confident that the validity of our study 1 is not compromised by the methods used or the statistics resulting from the application of those methods. DisclosuresNone. Edward

Background No simplified bedside risk scores have been created to predict long-term mortality after coronary artery bypass graft (CABG) surgery. Methods and Results The New York State’s Cardiac Surgery Reporting System was used to identify 8,597 patients who underwent isolated CABG surgery in July-December 2000. The National Death Index was used to ascertain patients’ vital status through December 31, 2007. A Cox proportional hazards model was fit to predict death following CABG surgery using pre-procedural risk factors. Then points were assigned to significant predictors of death based on the values of their regression coefficients. For each possible point total, the predicted risks of death at years 1, 3, 5, and 7 were calculated. It was found that the 7-year mortality rate was 24.2% in the study population. Significant predictors of death included age, body mass index, ejection fraction, unstable hemodynamic state or shock, left main coronary artery disease, cerebrovascular disease, peripheral arterial disease, congestive heart failure, malignant ventricular arrhythmia, chronic obstructive pulmonary disease, diabetes, renal failure, and history of open heart surgery. The points assigned to these risk factors ranged from 1 to 7; and possible point totals for each patient ranged from 0 to 28. The observed and predicted risks of death at years 1, 3, 5, and 7 across patient groups stratified by point totals were highly correlated. Conclusions The simplified risk score accurately predicted the risk of mortality following CABG surgery, and can be used for informed consent and as an aid in determining treatment choice.

The risk index appears to be a valuable tool for predicting patient risk when applied to another year of New York data. It should now be tested against other risk indexes in a variety of geographical regions.