Date of Award

Summer 1998

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Program/Concentration

Urban Services--Health Services

Committee Director

Clare Houseman

Committee Member

Lindsay Rettie

Committee Member

Stacey Plichta

Committee Member

Janice Belcher

Committee Member

Paul Champagne

Abstract

As U.S. health care expenditures top the $1 trillion mark, there is increased interest in measuring the performance of health care providers. For bottom line oriented payors such as government and business, the focus is on measuring cost. As hospitals account for over one-third of health care expenses, hospital cost per admission is a common measure of performance.

Many environmental and organizational factors come into play in determining hospital cost per admission. This research examines several of these factors, using Raymond Zammuto's model of organizational effectiveness assessment. Using Zammuto's framework, this research looks at the relationship of social, physical, and biological factors to cost per admission. Social factors include: hospital teaching status; ownership; patient socioeconomic status; and community poverty level. Physical factors include: hospital location; bed size; staff size; number of services offered; presence of specialty and tertiary services; and presence of obstetrical services. Biological factors are patient age and community elderly.

Although it is generally accepted that hospital cost per admission should be adjusted to account for differences among hospitals in patient complexity (i.e., case mix) and outpatient volume, not all adjustment methodologies take cost of living differences into account. To test the impact of adjusting for cost of living differences in addition to case mix and outpatient volume differences, this research uses three versions of the dependent variable: (1) cost per admission adjusted for case mix, outpatient volume, and cost of living; (2) cost per admission adjusted for case mix and outpatient volume only; and (3) unadjusted cost.

The study population consists of 85 general acute care hospitals in the Commonwealth of Virginia. 1994 Annual Historical Filing data submitted to the former Virginia Health Services Cost Review Council (now Virginia Health Information) were used.

Multivariate linear regression analysis of the cost per admission adjusted for case mix, outpatient volume, and cost of living indicates that patient age (percent of hospital patients age 65+), the presence of obstetrical services, and hospital bed size are significant variables. Larger hospital bed size is related to higher cost per admission. Larger percent of hospital patients age 65+ and the presence of obstetrics are related to lower cost per admission.

Bivariate and multivariate analyses demonstrate that variables such as hospital location and community poverty level have a great impact on cost per admission when adjusted for case mix and outpatient volume only. If a cost of living adjustment is also made to the cost per admission calculation, the location related variables are not signifiant due to the relationship between those variables and cost of living. It is concluded that a cost of living adjustment should be made in addition to case mix and outpatient volume adjustments when studying hospital cost per admission.

The multivariate linear regression model for cost per admission adjusted for case mix, outpatient volume, and cost of living accounts for 30.9% of the cost per admission variance. Other factors such as physician practice patterns and hospital management policies play an important role in hospital performance. These factors have been beyond the scope of this study but merit additional research.

DOI

10.25777/2nr2-vf07

ISBN

9780599059474

Download

Share

COinS