Risk of Cardiac Events After Surgery
Abstract & Commentary
By Michael H. Crawford, MD, Professor of Medicine, Chief of Clinical Cardiology, University of California, San Francisco. Dr. Crawford reports no financial relationships relevant to this field of study.
This article originally appeared in the October 2011 issue of Clinical Cardiology Alert. It was peer reviewed by Ethan Weiss, MD. Dr. Weiss is Assistant Professor of Medicine, Division of Cardiology and CVRI, University of California, San Francisco. Dr. Weiss is a scientific advisory board member for Bionovo.
Sources: Gupta PK, et al. Development and validation of a risk calculator for prediction of cardiac risk after surgery. Circulation 2011;124:381-387. Grover FL, Edwards FH. Objective assessment of cardiac risk for noncardiac surgical patients: An up-to-date simplified approach. Circulation 2011;124:376-377.
Perioperative cardiac events are the leading cause of surgical mortality. Thus, there has been considerable interest in predicting which patients are at highest risk. However, current risk prediction schemes have significant limitations. Thus, Gupta and colleagues analyzed the American College of Surgery National Surgery Quality Improvement Program database to determine factors associated with perioperative myocardial infarction (MI) or cardiac arrest, and to develop a risk calculator. Data from 2007 through 2008 were collected from about 200 hospitals in the United States. The 2007 data on more than 200,000 surgeries were used as the derivation set and the 2008 data (also > 200,000 surgeries) were used for validation. Only trauma patients, transplant patients, and those patients younger than 16 years old were excluded. The database included patients with aortic (2.1% of population), cardiac (0.3%), and peripheral arterial surgery (8.3%). Perioperative MI or cardiac arrest was seen in 0.65% of the validation group. Multivariate logistic regression analysis identified five significant predictors of MI or cardiac arrest: type of surgery, functional status, elevated creatinine, American Society of Anesthesiology class, and older age. In the validation set, a risk calculator based on these five factors had an area under the receiver operating curve (AUC) of 0.87, compared to the Revised Cardiac Risk Index (RCRI) of 0.75. In those patients undergoing aortic or noncardiac vascular surgery, the AUC was 0.75 vs the RCRI value of 0.59. The authors concluded that their cardiac risk calculator surpasses the performance of the RCRI and should simplify the informed consent process.
This paper has elevated the bar on cardiac event risk prediction in patients undergoing surgery of all types, but especially non-cardiovascular surgery. Notably, it uses a computer-based direct logarithmic regression model to determine the risk of MI or cardiac arrest like the STS or Euro-score do for cardiac surgery, rather than a point score system like the RCRI and most older schema (i.e., Goldman, Detzky). You can access the calculator at www.surgicalriskcalculator.com/miorcardiacarrest. The data entry is very simple and involves only five variables. 1. the American Society of Anesthesiology class (1-5): normal healthy patient (1), mild systemic disease, severe systemic disease, severe plus life threatening and moribund (5). 2. functional class as independent or partially or fully dependent (0 or 1); 3. creatinine > 1.5 (0, 1); 4. type of surgery classified as high risk (aortic, brain, hepatobiliary) or moderate (all others); and 5. age as a continuous variable.
One strength of this new schema is that it was derived from > 200,000 patients as compared to > 4000 for the RCRI and it performed better. Also, it includes newer laparoscopic surgeries and it is organ based. In addition, it has better discriminatory power, especially for vascular surgery, as compared to the RCRI.
There are limitations to the surgical risk calculator. The major ones for cardiology consultations are the lack of inclusion of information on stress test results, echocardiogram results, history or evidence of arrhythmias, beta-blocker use, and prior revascularization. These data may significantly alter risk, especially in the patient with known or suspected coronary artery disease. The definition of MI, especially in cardiac surgery patients, is problematic. In this study, they chose three times the upper limit of a normal troponin to exclude demand induced events. This is reasonable, but may be a strength or a weakness depending on your point of view.
I believe this is a significant step forward in this area and believe this new schema should replace the RCRI, especially for most patients undergoing non-cardiovascular surgery. It could be a starting point for evaluating patients undergoing cardiovascular surgery or those with known vascular disease, but other information will need to be considered in these patients before making a final decision on the risk of surgery. Unfortunately this study did not shed any light on the use of preoperative beta-blockers or stress tests, which remain controversial. The editorial by Grover and Edwards emphasizes that risk calculators are only one part of clinical decision making and are definitely not the whole enchilada. They believe this type of objective data will help with patient and family discussions, and consultations with other providers. In conclusion, we now have a better risk predictor tool, but the cardiology consultation has not been replaced by a computer yet.