Hospital Patterns of Medical Management Strategy Use for Patients With Non–ST-Elevation Myocardial Infarction and 3-Vessel or Left Main Coronary Artery Disease

Ralf E. Harskamp, MD; Tracy Y. Wang, MD, MHS, MSc; Deepak L. Bhatt, MD, MPH; Stephen D. Wiviott, MD; Ezra A. Amsterdam, MD; Shuang Li, MS; Laine Thomas, PhD; Robbert J. de Winter, MD, PhD; Matthew T. Roe, MD, MHS


Am Heart J. 2014;167(3):355-362. 

In This Article


In this hospital-level analysis among U.S. hospitals with full revascularization capabilities, we found that in contemporary practice, close to 20% of patients with NSTEMI did not receive in-hospital revascularization—despite angiographic demonstration of significant 3VD/LMD—and there was no significant change in these patterns over a 5-year period. Additionally, we observed wide variability in the rates of MMS use among hospitals without substantial differences in patient case-mix and predicted mortality risk. Reasons for preferring MMS use over revascularization in this selected patient population appear to be influenced more by co-morbidities and high-risk features than by hospital characteristics. Overall, in-hospital outcomes and discharge treatments were similar among hospital tertiles of MMS use, except for referral to cardiac rehabilitation programs.

Prior studies in UA/NSTEMI patients with significant coronary disease identified during angiography from clinical trial databases have shown that the use of a MMS is associated with an increased risk for long-term mortality when compared with in-hospital revascularization.[3,11] U.S. registry data, as well as a pooled analysis from various UA/NSTEMI trials confirm these observations.[14,15] In a secondary analysis from the Superior Yield of the New Strategy of Enoxaparin, Revascularization, and Glycoprotein IIb/IIIa Inhibitors (SYNERGY) trial, the post-discharge mortality curves for patients with significant coronary disease demonstrated during angiography diverged after 90 days, with the highest long-term mortality rates observed in patients who were medically managed without revascularization.[3] Similar findings were observed in a recent analysis from the Early Glycoprotein IIb/IIIa Inhibition in Non-ST-Segment Elevation Acute Coronary Syndrome (EARLY ACS) trial with unadjusted mortality rates through 1 year that were consistently higher among medically managed patients compared with those who underwent revascularization.[11] Data from these two analyses confirm the high-risk of mortality associated with the use of MMS for UA/NSTEMI patients with significant coronary disease identified during angiography. Nevertheless, these analyses did not focus specifically on patients with multi-vessel coronary disease. Furthermore, current practice patterns regarding the management of such patients have not been profiled in routine practice.

Therefore, while we know from prior studies that many patients who present with UA/NSTEMI undergo angiographic stratification and/or revascularization, our study shows that even in hospitals with full revascularization capacities, MMS is still frequently utilized for patients who are known to benefit the most from revascularization, namely those with 3VD including those with significant left main involvement. While the reasons for the observed hospital-level variation in the use of MMS are uncertain, we noted a number of important findings that provide insight into our findings. First, we found that hospitals in the lowest tertile of MMS use had higher annual CABG volumes for NSTEMI patients despite having the lowest overall annual volume of NSTEMI patients among the tertiles. While self-evident, the local expertise of cardiothoracic surgeons certainly would be expected to influence the decision-making regarding surgical revascularization for individual patients. In contrast, the local expertise of interventional cardiologists would also be expected to influence revascularization decision-making for these patients—a finding magnified by the observation of a temporal increase in the use of PCI for patients in this analysis with a concomitant decrease in the use of CABG. Second, we found a higher percentage of uninsured and African-American patients among hospitals with a higher percentage of MMS use. Hospitals with high MMS use were more likely to be academic medical centers that typically are known to treat a higher proportion of uninsured and minority patients, so this observation may relate to natural differences in patient populations among centers, but these findings require further study as patient socioeconomic status may have a complex interplay regarding the decision-making process for revascularization. Third, although discharge medications did not differ among hospital tertiles, there is an inverse relationship between cardiac rehabilitation referral and the use of MMS, which may reflect the impact of health care policy making, as rehabilitation is typically part of a routine care map in post-CABG and post-PCI patients, but not in medically managed patients. Alternatively, these differences may be due to patient selection, as patients who are eligible for coronary revascularization would also be expected to be more favorable candidates to participate in cardiac rehabilitation programs due to enhanced mobility and functional status (typical criteria for deciding upon revascularization in the first place). Finally, the predicted risk of patient-level mortality did not differ among hospital tertiles, so patient risk status did not appear to relate to the hospital-level patterns of use of revascularization for multi-vessel coronary disease. Therefore, we observed wide variability and no unambiguous threshold or standard for revascularization in patients with 3VD/LMD at the hospital-level, despite excluding prior CABG patients who typically present challenging revascularization decisions.

Study Limitations

Our study had several limitations. First, important comorbidities that influence revascularization decisions, such as active malignancies, severe pulmonary disease, and poor mental and functional status, were not collected. Second, revascularization following hospital discharge was not captured, and confounding due to competing risk (patients that die soon after hospital arrival, while being considered for revascularization) was not accounted for in this analysis. Third, due to the limitations of the case report form, multi-vessel disease was defined as a stenosis of ≥50%, which may have included non-flow limiting intermediate lesions (50–70%) for which revascularization may not indicated. Fourth, detailed angiographic features that may have influenced the decision-making process, such as vessel size, target vessel quality, and the extent of diffuse distal coronary artery disease (poor targets for revascularization), were not collected. Fifth, post-discharge long-term clinical outcomes are not assessed in the ACTION Registry–GWTG. Finally, participation in the registry is voluntary for hospitals. As a result, certain geographical areas are under-represented and, consequently, our hospital-level analysis may not be reflective of the broader group of hospitals with full revascularization capabilities in the U.S.