Regression analysis of health care charges with heteroscedasticity

We examine the effects of a prospective drug utilization review and patients' characteristics on total in-patient and out-patient health care charges. Our analysis of charges is complicated by the fact that the total health care charges are skewed. A log-transformation of these charges can normalize their distribution but may not stabilize their variance. To handle these problems, we propose a linear regression model with a non-constant variance (heteroscedasticity). Using results from a fitted linear regression model for log-transformed charges, we also discuss interpreting the regression coefficients in the original scale and estimating the total health care charges to individual patients. Employing these methods, we analyse total health care charges for drug utilization review patients with hypertension and identify patients' factors that are related to their total health care charges.

[1]  H. White.  A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity , 1980 .

[2]  J. Burridge,et al.  A note on nonregular likelihood functions in heteroscedastic regression models , 1994 .

[3]  D. Ruppert,et al.  Transformation and Weighting in Regression , 1988 .

[4]  J. F. Fitzgerald,et al.  Predicting Inpatient Costs With Admitting Clinical Data , 1995, Medical care.

[5]  A. Stewart,et al.  The MOS short-form general health survey. Reliability and validity in a patient population. , 1988, Medical care.

[6]  T. Breurch,et al.  A simple test for heteroscedasticity and random coefficient variation (econometrica vol 47 , 1979 .

[7]  J. Eisenberg,et al.  Views from funding agencies. Agency for Health Care Policy and Research. , 1998, Medical care.

[8]  N. Duan.  Smearing Estimate: A Nonparametric Retransformation Method , 1983 .

[9]  Differences Between Elderly Public Housing Tenants and Community Residents: A Case-Control Study , 1988 .

[10]  W. Manning,et al.  The logged dependent variable, heteroscedasticity, and the retransformation problem. , 1998, Journal of health economics.

[11]  W. Tu,et al.  A Wald test comparing medical costs based on log-normal distributions with zero valued costs. , 1999, Statistics in medicine.

[12]  R. J. Johnson,et al.  Hospital resource consumption among older adults: a prospective analysis of episodes, length of stay, and charges over a seven-year period. , 1994, Journal of gerontology.

[13]  C. McDonald,et al.  The Regenstrief Medical Record System: 20 years of experience in hospitals, clinics, and neighborhood health centers. , 1992, M.D. computing : computers in medical practice.

[14]  R. Halvorsen,et al.  The Interpretation of Dummy Variables in Semilogarithmic Equations , 1980 .