Rosenthal GE, Quinn LM, Harper DL, Pine M.
AHSR FHSR Annu Meet Abstr Book. 1995; 12: 135-6.
Section of General Medicine, VA Medical Center, Cleveland, OH 44106, USA.
PROBLEMS AND OBJECTIVES. Initiatives to compare hospital outcomes have been implemented in recent years. However, the impact of current initiatives on hospital quality and patient outcomes is poorly studied. The current study was conducted to examine changes in hospital mortality that have occurred since the initiation of Cleveland Health Quality Choice (CHQC), a regional initiative to compare performance in 30 hospitals. DATA AND METHODS. Since 1991, CHQC has provided hospitals with comparative mortality data for eight diagnoses: acute myocardial infarction (AMI), congestive heart failure (CHF), pneumonia, stroke, obstructive airway disease, gastrointestinal hemorrhage, coronary bypass surgery and bowel resection. Data are adjusted for severity of illness using validated diagnosis-specific models (ROC curve areas, 0.81-0.91). Risk-adjusted mortality rates were determined for all hospitals, in aggregate, over four successive periods of data collection and analysis that included 101,060 patients. RESULTS AND CONCLUSIONS. Risk-adjusted mortality declined for seven of eight diagnoses; in weighted regression analyses, declines over time were significant (p<.05) for CHF and pneumonia. Among all patients, risk-adjusted mortality rates were 7.4% in Period 1 and 6.8%, 6.8% and 6.5% in Periods 2, 3 and 4, respectively; a weighted regression analysis estimated that mortality declined 0.30% per period (p=.07). Among diagnoses, changes ranged from -0.50% (CHF) to 0.00% (AMI) per period and generally reflected declines in observed mortality rates with stable predicted rates. IMPLICATIONS FOR AUDIENCE. The findings demonstrate a general trend of declining hospital mortality over time associated with an initiative to provide comparative outcomes data to hospitals and purchasers. Although additional analyses are needed to determine if the declines are due to changes in discharge triage practices, shifts in admission to lower mortality hospitals, or improvements in the process of care, the study provides preliminary evidence of the potential benefit of disseminating valid risk-adjusted outcomes data to health care providers.
Publication Types:
Keywords:
- Behavior
- Data Collection
- Heart Failure
- Hospital Mortality
- Hospitals
- Humans
- Information Dissemination
- Myocardial Infarction
- Patient Discharge
- Pneumonia
- ROC Curve
- Regression Analysis
- Weights and Measures
- mortality
- hsrmtgs
Other ID:
UI: 102215684
From Meeting Abstracts