Petersen NJ, Ashton CM, Souchek J, Menke TJ, Collins TC, Foster W, Yu HJ; Association for Health Services Research. Meeting.
Abstr Book Assoc Health Serv Res Meet. 1999; 16: 43-4.
Veterans Affairs Medical Center, Houston, TX 77030, USA.
RESEARCH OBJECTIVE: Monitoring the performance of healthcare systems is essential as they undergo major changes in the way they provide medical care. The Department of Veterans Affairs (VA) is one of the largest providers of health care in the U.S., with over 150 acute care facilities in 22 service networks. The objective of this study was to compare facilities in the VA system on the short-term risk - adjusted mortality rate in 3 conditions for which poor care might be associated with poor outcomes. STUDY DESIGN: Patients hospitalized in fiscal years 1995 to 1997 for acute myocardial infarction (MI), acute stroke, and surgical repair of hip fracture were identified from the VA Patient Treatment File of discharges. VA's Beneficiary Identification and Records Locator Subsystem provided the date of death for patients who died. Admission date was used as the starting point for counting days until death for MI and stroke patients, while date of surgery was used for hip fracture patients. Kaplan-Meier estimates were used to compute unadjusted 30- and 90-day mortality rates for MI and stroke and 180-day rates for hip fracture repair. Cox regression models provided proportional hazards estimates of age-adjusted rates. Outliers were identified by comparing each network with the medial mortality rate. For each VA hospital with 10 or more patients in the disease-specific cohort, we computed unadjusted and age-adjusted rates. PRINCIPAL FINDINGS: Overall, adjusted 30-day and 90-day mortality rates in 1997 for acute MI were 11.5% and 14.5%. There were significant differences by age, with 30-day and 90-day rates of 6.2% and 7.8% in patients under 65, 14.5% and 18.5% in patients 65-74, and 24.9% and 31.0% in patients 75 and older. Age-adjusted 30-day rates ranged from 7.8% to 14.6% across the 22 service networks. Variation also existed among the VA medical centers (VAMCs), with 30-day mortality rates ranging from 0% to 24.9%. For acute stroke, the 1997 30-day and 90-day age-adjusted rates were 10.8% and 15.2%. As with MI, there were significant differences by age (30-day mortality of 7.6% for under 65, 10.6% for 65-74, and 16.3% for 75 and older; 90-day mortality of 9.8% for under 65, 15.5% for 65-74, and 24.4% for 75 and older). In the service networks, the lowest 30-day rate was 5.3% while the highest was 14.5%. Three VAMCs had 0% 30-day mortality while the highest rate among facilities hospitalizing at least 10 stroke patients was 28.2%. System-wide, the 180-day adjusted mortality rate in 1997 for hip fracture repair was 14.3%. Mortality for patients 75 and older (21.9%) differed significantly from that of patients under 65 (7.1%). The range across the 22 networks was 5.3% to 21.3%. Whereas seven VAMCs had 0% mortality, the highest adjusted rate among the VA hospitals was 33.2%. CONCLUSIONS: Wide variations exist across the VA service networks and across VA hospitals in the mortality rates for patients hospitalized for the 3 conditions. These differences are not due to differences in the age-distribution of patients, but may be due to differences in process-of-care or to additional differences in case mix. IMPLICATIONS FOR POLICY, DELIVERY OR PRACTICE: These data are used as a screening tool to help identify hospitals with effective practices as well as hospitals with possible problems. Additional work needs to be done to determine if differences in the quality of care are responsible for the large differences in mortality among facilities.
Publication Types:
Keywords:
- Hospitals
- Hospitals, Veterans
- Humans
- Myocardial Infarction
- Stroke
- United States
- United States Department of Veterans Affairs
- Veterans
- instrumentation
- mortality
- therapy
- utilization
- hsrmtgs
Other ID:
UI: 102194534
From Meeting Abstracts