2011 National Meeting

3100 — Adherence to Quality Measures and Mortality among Veterans with Congestive Heart Failure

Trivedi AN (Providence VA Medical Center) , Jiang L (Providence VA Medical Center), Friedmann PD (Providence VA Medical Center), Wu WC (Providence VA Medical Center)

All VA medical centers (VAMCs) and most private-sector hospitals report on the quality of care provided to patients with congestive heart failure (CHF). Limited evidence supports that adherence to CHF quality indicators predicts improved health outcomes. We examined the association between adherence to 10 inpatient CHF quality indicators and subsequent 30-day and 1-year mortality.

For a cohort of hospitalized veterans assessed for one or more CHF quality indicators from 2001 to 2007, we linked individual-level External Peer Review Program data to VA utilization, vital status , laboratory and vital signs data. We constructed logistic regression models predicting 30-day and 1-year mortality following the performance assessment, adjusting for age, sex, race, comorbidities, physiologic and laboratory values, and clustering within VAMC. To determine whether associations between performance measures and mortality varied by clinical risk, we performed stratified analyses by tertiles of blood pressure, creatinine-clearance, potassium, and predicted probability of death.

Characteristics of the cohort (n = 64,321) were: Mean age: 70.7 years, 20.9% black, mean SBP 124, Cr-clearance 58.3, and number of comorbidities: 6.3. We observed no significant relationship between receipt of the performance indicator and 30 and 365 day mortality for 8 of 10 quality indicators. The two exceptions were ACE-inhibitor/angiotensin-receptor blocker (ACE-I/ARB) use prior to CHF admission (OR for 30-day mortality 0.84, 95%CI 0.71-1.00; and OR for 1-year mortality 0.88, 95%CI 0.79-0.99, for persons receiving the indicator compared to those not receiving) and ACE-I/ARB prescription on discharge (OR for post 30 day mortality 0.68, 95%CI 0.53-0.86; OR for 1-year mortality 0.76, 95%CI 0.68-0.86, for persons receiving the indicator compared to those not receiving). The association of ACE-I/ARB use and mortality was stronger in groups with higher risk of mortality and lower creatinine-clearance.

With the important exception of performance indicators assessing ACE-I/ARB use, the inpatient CHF quality indicators reported by VAMCs do not predict lower mortality at 30 days or one year.

Given the proliferation of performance measures for CHF and other medical conditions, the VA should prioritize those measures that are strongly linked to patient outcomes. Construction of new performance measures that predict improved outcomes in congestive heart failure are needed.