Comprehensive coronary artery disease care in a safety-net hospital: results of Get With The Guidelines quality improvement initiative

Adherence to published coronary artery disease (CAD) guidelines is suboptimal, particularly among minorities and the poor. While hospital-based quality-improvement programs may increase the use of evidence-based therapies, little data exist regarding the impact of such programs in sociodemographically disadvantaged (vulnerable) populations. Vulnerable patients in the United States are cared for primarily within the safety-net health system, which comprises urban public hospitals and outpatient community health centers. Denver Health is an example of an integrated system that encompasses both types of facilities. To assess evidence-based medication use in CAD patients after initiation of an inpatient quality-improvement program at Denver Health. We reviewed the medical records of 499 patients with angiographically proven CAD who were hospitalized between July 1998 and December 2002. Patients were prospectively identified through a multidisciplinary intervention led by a nurse manager, and their records were input retrospectively into the American Heart Association's Get With The Guidelines patient management tool. The association's program, which recommends initiating 4 cardioprotective drug classes while patients are hospitalized, was started 2 years into the observation period (August 2000). Treatment rates were compared over the ensuing years. We evaluated temporal trends in discharge use of 4 drugs: (1) betablockers, (2) angiotensin-converting enzyme inhibitors (ACEIs), (3) hydroxymethylglutaryl coenzyme A reductase inhibitors (statins), and (4) aspirin. We calculated the proportion of eligible patients (no documented contraindication) who were prescribed each drug category as well as the proportion who received all 4 drug categories, our principal composite outcome. If any one drug was absent, the composite criterion was considered unmet. We observed progressive improvement in discharge use of the 4- drug composite: 18% in 1998-1999 (95% confidence interval [CI], 12%-25%), 50% in 2000 (95% CI, 37%-63%), 62% (95% CI, 54%-70%) in 2001, and 72% (65%-79%) in 2002 (P