Frank F. Song, PhD, Department of Quality Improvment Projects, Virginia Health Quality Center, 4510 COx Road, Suite 400, Glen Allen, VA 23060, 804-289-5320, fsong@vaqio.sdps.org

Quality Improvement Organizations (QIO) have been working collaboratively with hospital staff to improve the quality of care for Medicare patients by increasing the use of evidence-based care processes for hospitalized Acute Myocardial Infarction (AMI) patients.

All AMI quality indicators (QI) improved substantially between 1998 and 2002 in Virginia and at national level. However the variation of QI rates among providers is also substantial. In addition, it is not known if improvements in QI rates have transformed into the intended outcome of reduced mortality rates among AMI patients. It is also very important for QIOs and providers to know how much of the variation in AMI mortality rates among providers can be explained by the variance in healthcare QI measures.

This paper examines the impact of process of care measures on hospital mortality by using Virginia Medicare claims data from 1999 to 2002. All Medicare patients admitted with AMI during this period in Virginia’s 80 acute care hospitals form the study population. Time series and logistic regression models are used to assess the impact of QI changes and QI rates on in-hospital mortality respectively. The Charlson comorbidity index and related patient demographic factors are used to control for variation in patient risk factors across hospitals. Results of the analyses are reported and policy implications discussed.

Learning Objectives: