Abstract 1352: In-Hospital Assessment of Hemoglobin A1c in Patients With Diabetes Treated for Myocardial Infarction: Variability Between Hospitals and Relationship With Established Quality Indicators
BACKGROUND It is unknown whether hospitals that provide higher quality care for acute MI are also more likely to focus on management of coexisting conditions, such as diabetes (DM). While the American Diabetes Association recommends hemoglobin A1c (A1C) assessment for hospitalized patients (pts) with DM, the frequency of in-hospital A1C assessment and its relationship with established quality metrics in the setting of MI have not been described.
METHODS AND RESULTS TRIUMPH is a prospective multicenter registry enrolling 2755 consecutive MI pts between 2005–07, of whom 724 (26%) had DM. A1C assessment was defined as in-hospital measurement or chart documentation from the 3 months prior to admission. Quality indicators (QIs) included use of aspirin, beta blockers, ACEI/ARB (if LVEF ≤40%), and statins at discharge; composite QIs for MI care were calculated for each pt. A1C assessment was performed in 504 DM pts (70%). The variability in A1C assessment was substantially greater than the variability in QI use (range between hospitals 33–97% vs 78–97%, p=0.007; Figure⇓), and higher hospital QI use was not associated with greater likelihood of A1C assessment either before (OR 1.12, 95% CI 0.98–1.28 for A1C assessment per 1% increase in QI performance) or after adjustment for patient factors (OR 1.12, 95% CI 0.94–1.34).
CONCLUSIONS Assessment of glycemic control in MI pts with DM is highly variable across hospitals, and is not related to quality of MI care. This may reflect uncertainty regarding the value of A1C-guided DM therapy for MI pts, lack of consistent focus on DM management by the cardiology community, or relegation of DM management to the outpatient setting at some hospitals.