Performance Measures and Outcomes for Heart Patients
Fonarow GC, Abraham WT, Albert NM, et al. Association between performance measures and clinical outcomes for patients hospitalized with heart failure. JAMA. 2007 Jan 3;297(1):61-70
As our population ages, more emphasis will be placed on issues surrounding efficient and evidence-based care. Heart failure, which accounted for 3.6 million hospitalizations in 2003 and has an overall prevalence of 5 million, will be at the forefront of public policy. As pay for performance (P4P) and standards of care become increasingly prevalent, the medical community will need to scrutinize the standards by which we are measured.
The American College of Cardiology and the American Heart Association (ACC/AHA) developed guidelines for the treatment and care of patients with heart failure. These measures include heart failure discharge instructions, evaluation of left ventricle (LV) function, angiotensin converting enzyme (ACE) inhibitors or angiotensin II receptor antagonist (ARB) for LV dysfunction, adult smoking cessation counseling, and anticoagulation at discharge for patients with atrial fibrillation. Adherence to these performance measures should be based on evidence.
The authors’ goal was to determine the validity of these guidelines. The Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF) registry allowed for the documentation and follow-up of patients adhering to the heart failure guidelines as set forth by the ACC/AHA. The study assessed the relationship between these guidelines and clinical outcomes, including 60- to 90-day mortality and a composite end point of mortality or rehospitalization.
In this study the OPTIMIZE-HF registry was used as the source of prospective data collection. Ten percent of eligible patients were randomly selected from the registry between March 2003 and December 2004 from 91 hospitals. Eligibility for the OPTIMIZE-HF registry included patients 18 and older admitted for worsening heart failure or significant heart failure during their hospital stay. The performance measure of discharge instruction, smoking cessation, and anticoagulation were measured for all eligible patients. Patients with an ejection fraction of 40% or less, or moderate to severe systolic function, were included for the ACE inhibitor/ARB performance measure. One measure not included was treatment with beta-blockers at discharge. The authors included beta-blockers at discharge with metrics similar to those described for ACE/ARB criteria.
The conformity rates and process-outcome links were then determined for the performance measures and beta-blocker treatment as it related to 60- to 90-day mortality/rehospitalization.
The study focused on a random follow-up cohort of 5,791 patients from 91 hospitals. This was similar to the OPTIMIZE-HF cohort of 48,612 patients in 259 hospitals. Demographically, the average cohort’s age was 72, 51% male and 78% white, with 42% of patients diagnosed with ischemic heart disease and 43% with diabetes mellitus. These results were similar to the demographics of the overall OPTIMIZE-HF registry.
Of the eligible patients in the follow-up cohort, 66% (4,010) received complete discharge instructions. Eighty-nine percent of eligible patients (4,664) had their left ventricular function evaluated. For those patients with documented left ventricular systolic dysfunction (2,181), 83% were given an ACE inhibitor or ARB at discharge. Patients who had a diagnosis of atrial fibrillation were discharged with anticoagulation at a rate of 53%, and 72% of patients were counseled on smoking cessation. As compared with ACE inhibitors/ARB, similar results (84%) were seen for beta-blockers at discharge.
Only two of the five ACC/AHA performance measures were predictive of decreasing morbidity and mortality/rehospitalization in unadjusted analysis: patients discharged on ACE inhibitors/ARBs (odds ratio, 0.51; 95% CI 0.34–0.78; P- .002) and smoking cessation counseling. Beta-blockers, not a formal part of the ACC/AHA guidelines, were also a predictor of lower risk of both mortality and rehospitalization (odds ratio, 0.73; 95% CI, 0.55-0.96; P-0.02)