NCDR® Research Network: Focus on Post-MI Aspirin, Diabetes, and PCI Outcomes Reporting
Patterns of Discharge Aspirin Dosing in 213,344 U.S. Patients after Acute Myocardial Infarction: Results from the NCDR
In this arena, U.S. and non-U.S. practice patterns often differ. For example, low-dose aspirin is more commonly used outside the U.S., although debate continues regarding proper ASA dosing for preventing CV outcomes. However, U.S. guidelines have evolved, and we see a trend toward lower dosing recommended at post-myocardial infarction (MI) hospital discharge. After percutaneous coronary intervention (PCI), the recent ACCF/AHA update to the unstable angina/non-ST-segment elevation myocardial infarction (NSTEMI) guidelines state that it is reasonable to use 81 mg per day of aspirin in preference to higher maintenance doses (Class IIa; Level of Evidence B; J Am Coll Cardiol 2012;60:645-81).
According to first author Hurst M. Hall, MD, The University of Texas Southwestern Medical Center, Dallas, TX, experts based this change on studies suggesting similar efficacy to higher doses, but reduced bleeding risk. So far, the recommended use of aspirin for general cardiopreventive medical management (75-162 mg) and post-CABG (100-325 mg) remain unchanged.
- Dr. Hurst's review, designed to describe current practice patterns, encompassed 213,344 acute MI patients (including 127,586 with NSTEMI) from 525 centers in the ACTION Registry®-GWTG™ between January 2, 2007 and March 31, 2011.
- Dosing at discharge was 325 mg (high dose) in 61 percent, 162 mg in 3 percent, and 81 mg in 36 percent.
- Fifty-one percent of patients were discharged on high-dose ASA alone. Among patients discharged on clopidogrel, 69 percent also received high-dose ASA, likely reflecting guidelines at the time of the study. Surprisingly, 44 percent of patients on triple therapy (ASA, clopidogrel, warfarin) were discharged on high-dose ASA.
- Slightly less than half of coronary artery bypass graft (CABG) patients were discharged on high-dose ASA, and 45 percent of patients managed medically were discharged on high-dose ASA, both contrary to recommendations.
- Among patients with an in-hospital major bleeding event (4 mg/dl drop in hemoglobin, ICH, or RBC transfusion), surprisingly 57 percent were prescribed high-dose ASA at discharge, similar to the percentage (64 percent) of patients without a major bleeding event.
- A 25-fold variance in high-dose ASA use observed across hospitals was largely disconnected from other quality metrics.
Dr. Hall concluded that high-dose ASA is over-utilized in the U.S., notably in high-risk subgroups without guideline-based indications.
Impact of Diabetes on Clinical Findings, Management, and In-Hospital Outcomes in NSTEMI and STEMI: An Analysis from the NCDR
Although patients with diabetes mellitus (DM) experience worse outcomes following MI compared to non-diabetics, a decade has passed since the last large trial comparing outcomes. "We thought that this important subject should be revisited to see if there has been any improvement in mortality for diabetic patients presenting with MI," said Jorge F. Saucedo, MD, professor of medicine at University of Oklahoma, Oklahoma City, OK.
The analysis encompassed 150,292 NSTEMI patients and 93,569 STEMI patients from more than 462 US sites with PCI capability in the ACTION Registry®-GWTG™, and stratified them into those with and without DM. It evaluated the relationship between in-hospital mortality and DM status.
- NSTEMI and STEMI patients with DM were slightly older than those without DM, more likely to be female, and had a higher BMI (kg/m2): 27.5 (no DM) versus 30.3 (DM) among NSTEMI patients and 27.8 (no DM) versus 30.1 (DM) in STEMI. Additionally, patients with DM were much more likely to have triple-vessel disease (NSTEMI no DM 31.0 percent vs. DM 43.5 percent; STEMI no DM 25.2 percent vs. DM 34.9 percent). (The p value was <0.0001 for all comparisons between No DM vs. DM groups in STEMI and NSTEMI.)
- DM patients were more likely to have heart failure (EF <25 percent) and less likely to receive PCI (p <0.0001 for both).
- Mortality was significantly (p <0.0001) higher for patients with DM in both NSTEMI (no DM 2.9 percent, DM 3.9 percent) and STEMI patients (no DM 3.2 percent, DM 5.0 percent). Combined death or MI was similarly (p <0.0001) higher for DM patients (3.7 percent/4.6 percent NSTEMI; 3.9 percent/5.9 percent STEMI).
- Major bleeding occurred more often (p <0.0001) in the presence of DM (NSTEMI 7.3 percent/11.7 percent; STEMI 8.9 percent/11.9 percent).
- The adjusted risk for in-hospital mortality was increased about 14 percent in the NSTEMI DM group and 17 percent in the STEMI DM group.
Dr. Saucedo, in an interview with CardioSource WorldNews, commented that NSTEMI DM patients' lower likelihood of receiving PCI (41.1 percent vs. 48.8 percent for no DM) may be attributed to reluctance of interventionalists to send this group, which had much worse glomerular filtration rate, to the cath lab.
"DM patients with NSTEMI and STEMI have a higher risk of mortality and bleeding complications. Novel treatment strategies that target diabetic patients with MI are warranted," he said.
State-Mandated Public Reporting of Percutaneous Coronary Intervention Outcomes is Associated with Better Outcomes: A Report from the NCDR
Four U.S. states (New York, Pennsylvania, Massachusetts, New Jersey) mandate the collecting and public reporting of PCI outcomes. Matthew A. Cavender, MD, research fellow at Brigham and Women’s Hospital, Boston, MA, noted that prior data suggest that such public reporting may result in lower revascularization rates in high-risk individuals, but the available studies are limited in scope.
To evaluate the effect of mandatory public reporting on patient selection and in-hospital mortality with PCI, Dr. Cavender reviewed CathPCI Registry® data on PCI patients from non-public reporting states (NRS) (n = 1,184,544) and public reporting states (RS) (n = 164,775).
- Predicted procedural mortality, based on the NCDR Risk Model, was similar for NRS (1.40 percent) and RS (1.37 percent) (p = 0.86). Actual mortality, however, was 1.41 percent for NRS but significantly lower – 1.20 percent – for RS (p <0.001).
- For elective PCI, predicted procedural mortality was again similar (0.36 percent NRS vs. 0.37 percent RS; p = 0.25), but once again divergent for actual mortality (0.37 percent NRS vs. 0.25 RS; p = 0.003).
- Predicted mortality with PCI for acute coronary syndromes (ACS) was higher for the RS (0.98 percent vs. 1.09 percent; p <0.001), but turned out actually to be lower (0.90 percent vs. 0.83 percent) (p = 0.002).
- In PCI for cardiogenic shock, the predicted mortality was higher for RS (24.83 percent vs. 25.98 percent) (p <0.001), but actual mortality was similar RS (27.92 percent vs. 26.76 percent) (p = 0.14).
- Adjusted models also showed significantly lower in-hospital mortality with public reporting for all patients and for those receiving elective PCI, PCI for ACS, or cardiogenic shock.
- Mandatory reporting was also associated with lower red blood cell transfusion rates (NRS 2.7 percent vs. RS 2.4 percent), reduced contrast use (creatinine clearance] >3: NRS 30.4 percent vs. RS 29.7 percent), and higher rates of discharge prescription for most medications (aspirin, beta-blockers, statins, thienopyridines). (All p values < 0.01.)
Dr. Cavender concluded that patients undergoing PCI in states with mandatory public reporting have similar predicted mortality, but lower actual mortality than non-mandatory reporting states. Further studies are needed to elucidate the mechanisms for the observed associations.
< Back to Listings