The impact of preoperative stable angina pectoris on postoperative cardiovascular outcomes in patients with previous myocardial infarction (MI) who underwent major noncardiac surgery is not well studied. We studied patients with previous MI who underwent elective major noncardiac surgeries within the American College of Surgeons–National Surgical Quality Improvement Program (2005 to 2011). Primary outcome was occurrence of an adverse cardiac event (MI and/or cardiac arrest). Multivariable logistic regression models evaluated the impact of stable angina on outcomes. Of 1,568 patients (median age 70 years; 35% women) with previous MI who underwent major noncardiac surgery, 5.5% had postoperative MI and/or cardiac arrest. Patients with history of preoperative angina had significantly greater incidence of primary outcome compared to those without anginal symptoms (8.4% vs 5%, p = 0.035). In secondary outcomes, reintervention rates (22.5% vs 11%, p <0.001) and length of stay (median 6-days vs 5-days; p <0.001) were also higher in patients with preoperative angina. In multivariable analyses, preoperative angina was a significant predictor for postoperative MI (odds ratio 2.49 [1.20 to 5.58]) and reintervention (odds ratio 2.40 [1.44 to 3.82]). In conclusion, our study indicates that preoperative angina is an independent predictor for adverse outcomes in patients with previous MI who underwent major noncardiac surgery, and cautions against overreliance on predictive tools, for example, the Revised Cardiac Risk Index, in these patients, which does not treat stable angina and previous MI as independent risk factors during risk prognostication.
A seminal study by Goldman et al in the late 1970s, for the first time, quantified the effect of preexisting co-morbidities on postoperative cardiac outcomes in patients who underwent major noncardiac surgery. These investigators later updated their cardiac risk score in the late 1990s (Lee et al, Revised Cardiac Risk Index [RCRI]) to reflect the general decrease in risk of adverse cardiac events, a tributary of advances in surgical and anesthesia practices. The RCRI is an extremely useful and adept tool for point-of-care cardiac risk stratification of patients undergoing noncardiac surgery. RCRI was developed for “all-comers” undergoing major noncardiac surgery and thus, not unlike other predictive tools, may be limited in its discriminatory ability in extreme situations such as in multimorbid patients with preexisting cardiac pathology. For example, although the presence of stable angina in a patient with history of myocardial infarction (MI) is suggestive of greater ischemic burden, it does “not” add toward the risk of postoperative cardiac complications as determined by RCRI. As a result, stable anginal symptoms in such patients are often not addressed appropriately at the time of preoperative risk stratification. Against this background, we aimed to evaluate the association between the presence of stable anginal symptoms and risk for adverse postoperative cardiovascular outcomes in patients with a history of MI scheduled to undergo a major noncardiac surgery using the American College of Surgeons–National Surgical Quality Improvement Program (ACS-NSQIP) database.
Methods
The ACS-NSQIP is an ACS initiative supporting the collection of 30-day risk-adjusted data to facilitate the assessment of outcome measures after surgery. Details of the ACS-NSQIP ( www.acsnsqip.org ) have been described before. Briefly, it is a validated outcomes registry designed to provide feedback to participating hospitals on 30-day risk-adjusted surgical morbidity and mortality. The database consists of deidentified data on demographics, perioperative risk factors, intraoperative variables, and 30-day postoperative outcome for adult patients who underwent major surgical procedures in both the inpatient and outpatient setting at participating centers. Trained surgical clinical reviewers capture these data using a variety of methods including medical chart abstraction and telephone interviews with the attending surgeon and the patient. Comprehensive training of the reviewers, audit of participating sites, regular conference calls, and an annual meeting ensure data quality.
For the present study, we queried the ACS-NSQIP (2005 to 2011) Participant User Files focusing on 15 major cancer and noncancer surgical procedures ( Appendix ). The study population comprised patients who underwent 1 of the 15 elective noncardiac procedures who had a concomitant documented history of a recent MI (<6 months, n = 1,568). Primary exposure variable of interest was history of preoperative stable angina defined as the presence of dull, diffuse, substernal chest discomfort precipitated by exertion or emotion and relieved by rest or nitroglycerine within 4 weeks before surgery. Primary outcome measure was occurrence of an adverse cardiac event (postoperative MI or cardiac arrest). Secondary end points included length of stay (LOS), reintervention, readmission, and mortality. We performed multivariable-adjusted analyses to evaluate the association between the presence of preoperative stable angina pectoris and the postoperative outcomes described previously. Covariates included age at surgery, gender, race, body mass index, smoking status, hypertension, diabetes mellitus, congestive heart failure, cerebrovascular disease, serum creatinine, and procedure type.
Finally, we quantified the predictive accuracy of the RCRI and the modified RCRI (treating preoperative angina and MI as separate inputs in the score) in predicting an adverse cardiac outcome in the present patient population using the area under the curve (AUC) analysis. All input variables were given a score of 0 or 1 (0 = no, 1 = yes), in accordance with the original RCRI described. The Mantel–Haenszel test was used to assess the statistical significance of the difference in the discriminatory ability between the 2 models. An institutional review board waiver was obtained before conducting this study, in accordance with institutional regulation when dealing with deidentified administrative data. All statistical analyses were performed using SPSS 21.0 (Chicago, Illinois), with a 2-sided significance level set at p <0.05.
Results
We identified 1,568 patients (35.8% women; median age 70 years [interquartile range 62 to 77]) with history of previous MI who underwent 1 of the 15 major noncardiac surgical procedures. Baseline characteristics are listed in Table 1 . Overall, 35% study participants had coexisting diabetes, whereas hypertension was present in 88% of the study participants.
| Variable | All Patients (n=1,568) | Preoperative Angina Pectoris | P-value | |
|---|---|---|---|---|
| No (n = 1,319) | Yes (n = 249) | |||
| Age; Median (IQR) ∗ | 70 (62-77) | 70 (62-77) | 70 (62-77) | 0.894 |
| Women | 561 (35.8%) | 467 (35.4%) | 94 (37.8%) | 0.484 |
| Body mass index (Kg/m 2 ) | ||||
| <18.5 | 67 (4.3%) | 57 (4.4%) | 10 (4.0%) | 0.177 |
| 18.5-25 | 474 (30.5%) | 410 (31.3%) | 64 (25.9%) | |
| 25.1-30 | 550 (35.4%) | 464 (35.5%) | 86 (34.8%) | |
| >30 | 464 (29.8%) | 377 (28.8%) | 87 (35.2%) | |
| Race/Ethnicity | ||||
| White | 1370 (87.4%) | 1151 (87.3%) | 219 (88.0%) | 0.071 |
| Black | 109 (7.0%) | 94 (7.1%) | 15 (6.0%) | |
| Hispanic | 58 (3.7%) | 46 (3.5%) | 12 (4.8%) | |
| Asian/others | 31 (2.0%) | 28 (2.1%) | 3 (1.2%) | |
| Smoker | 569 (36.3%) | 481 (36.5%) | 88 (35.3%) | 0.774 |
| Hypertension | 1376 (87.8%) | 1146 (86.9%) | 230 (92.4%) | 0.015 |
| Diabetes Mellitus | 549 (35.0%) | 464 (35.2%) | 85 (34.1%) | 0.772 |
| Congestive Heart Failure | 182 (11.6%) | 116 (8.8%) | 66 (26.5%) | <0.001 |
| Cerebrovascular Disease † | 454 (29.0%) | 374 (28.4%) | 80 (32.1%) | 0.253 |
| Creatinine >2mg/dl | 146 (9.3%) | 125 (9.5%) | 21 (8.5%) | 0.721 |
| Vascular, Intraperitoneal or Intrathoracic Surgery | 1540 (98.2%) | 1293 (98.0%) | 247 (99.2%) | 0.296 |
In univariable analyses, overall incidence of any cardiac complication was 5.5% (n = 87), whereas that of postoperative MI was 3.6% (n = 57). Patients with history of preoperative angina pectoris had significantly greater incidence of postoperative cardiac complications and MI (cardiac complications, 8.4% [n = 21]; MI, 6% [n = 15], Figure 1 ) compared to those without anginal symptoms (cardiac complications, 5% [n = 66], p = 0.035; MI, 3.2% [n = 42], p = 0.036). In secondary outcomes, reintervention rates (22.5% [n = 56] vs 11% [n = 145], p <0.001) and LOS (median: 6 vs 5 days; p <0.001) were also higher in patients with preoperative angina compared to those without angina. There was no difference in 30-day mortality rate (p = 0.109; Figure 1 ). Similar findings were observed on subgroup analyses of patients who underwent vascular surgery only (n = 1,149) with a higher incidence of postoperative MI in patients with versus without preoperative angina pectoris (6% vs 2.8%).
