MAUSKOPF J, ORZOL S, TOOLEY J, POLSKY D, GLICK HA; International Society of Technology Assessment in Health Care. Meeting.

Annu Meet Int Soc Technol Assess Health Care Int Soc Technol Assess Health Care Meet. 2001; 17: abstract no. 160.

Research Triangle Institute, Research Triangle Park, N.C., University of Pennsylvania, Philadelphia, PA, Pharmacia, Skokie, IL.

Introduction - Obtaining diagnosis-specific estimates of cost of hospitalization or cost per day of hospitalization in countries outside the U.S. can be time consuming and expensive. We report results of an imputation procedure for these costs that uses cost data for a limited number of diagnoses and published U.S. Medicare DRG relative values.Methods - We sought to obtain hospital cost and length of stay (LOS) estimates for the 49 (out of 186) most frequently observed inpatient diagnoses in a cardiovascular trial in 5 countries (223 out of 245 [49x5] obtained, 91%). We used OLS regression to predict these outcomes; explanatory variables were 4 country dummies and 3 relative value measures from the U.S., the Medicare DRG relative weight, DRG arithmetic mean LOS, and DRG geometric mean LOS. We used the results of these regressions to impute cost and LOS for the country- and diagnosis-specific hospitalizations for which data were not obtained.Results - The combination of the country dummies and the DRG data had adjusted r-square values ranging between 0.47 to 0.56 for costs and between 0.25 to 0.30 for length of stay. The bulk of the explanatory power was derived from the DRG variables (adjusted r-square values ranging between 0.45 to 0.54 for costs and between 0.17 to 0.29 for length of stay). The DRG relative weights failed to add significant predictive information to either the arithmetic (p = 0.62) or the geometric (p = 0.28) mean LOS, but both LOS measures added or tended to add significant information to each other (p ranging between 0.004 and 0.07).Conclusions - For the 5 countries studied, US relative measures of LOS were strongly predictive of the relative costs of hospitalization. Their ready availability may allow one to collect costs for a subset of hospitalizations and impute costs for the remaining hospitalizations.

Publication Types:

• Meeting Abstracts

Keywords:

• Costs and Cost Analysis
• Diagnosis-Related Groups
• Ethnic Groups
• Family
• Hospital Costs
• Hospitalization
• Medicare
• United States
• economics
• ethics
• hsrmtgs

Other ID:

• GWHSR0002813

UI: 102274492

From Meeting Abstracts