Estimating Lifetime or Episode-of-Illness Costs Under Censoring

Posted: 7 Jul 2011

See all articles by Willard G. Manning

Willard G. Manning

University of Chicago - Harris School of Public Policy

Anirban Basu

University of Chicago - Department of Medicine

Date Written: June 2010


Many analyses of health care costs involve use of data with varying periods of observation and right censoring of cases before death or the end of the episode of illness. The prominence of observations with no expenditure for some short periods of observation and the extreme skewness typical of these data raise concerns about the robustness of estimators based on inverse probability weighting (IPW) with the survival from censoring probabilities. They also cannot distinguish between the effects of covariates on survival and intensity of utilization, which jointly determine costs. In this paper, we propose a new estimator that extends the class of two-part models to deal with random right censoring, and more fully incorporates the information from the censored periods. Our model also addresses issues about the time to death in these analyses and separates the survival effects from the intensity effects. Using simulations we highlight our proposed estimator compared to the inverse probability estimator, which shows bias when censoring is large and covariates affect survival. We find our estimator to be unbiased and also more efficient for these designs. Using data from the Medicare-SEER files, we apply our method and compare it to the IPW method in estimating the differential effect of diagnosis grade on 10-year costs of prostate cancer patients. Prostate cancer is already the most commonly detected non-cutaneous malignancy among American men. It is estimated that more than 234,000 new cases of PC were diagnosed in 2006 and more than 27,000 men died of the disease. Patients with prostate cancer spend over $4B annually for prostate cancer treatments. Therefore, understanding trajectory of costs for this disease and how covariates affect these trajectories are essential for comparative effectiveness research in this field. Our results indicate that the IPW approach and ours generate similar results when looking at 2-year cost where there is no censoring and the differential effects of grade on survival are small. However, when looking at the 10-year costs, the estimated effects of grade of the cancer differ substantially across estimators. This appears to be the result of the differential effect of grade on survival over the longer period, which conforms to our results of bias in the IPW approach when there is an effect of treatment or other covariates on survival but those effects are partly masked due to heavy censoring. Anirban Basu is an Assistant Professor of Medicine at the University of Chicago and a faculty research fellow at the National Bureau of Economic Research. In his research, Dr. Basu strives to apply micro-econometric theory and models to health economic evaluations. He has extensive experience in modeling health expenditure data and has also worked on the theoretical and empirical foundations in cost-effectiveness analyses and value of information analyses in the context of prostate cancer, schizophrenia and diabetes. Dr. Basu is an Associate Editor for both Health Economics and the Journal of Health Economics and co-teaches courses on decision analysis and cost-effectiveness analysis.

Keywords: Censored costs, episode of illness, inverse probability weighting, survival versus intensity effects

Suggested Citation

Manning, Willard G. and Basu, Anirban, Estimating Lifetime or Episode-of-Illness Costs Under Censoring (June 2010). American Society of Health Economists (ASHEcon) Paper . Available at SSRN:

Willard G. Manning (Contact Author)

University of Chicago - Harris School of Public Policy ( email )

1155 East 60th Street
Chicago, IL 60637
United States
(773) 834-1971 (Phone)
(773) 702-1979 (Fax)

Anirban Basu

University of Chicago - Department of Medicine ( email )

5841 S. Maryland Ave
Chicago, IL 60637
United States
773 834 1796 (Phone)
773 834 2238 (Fax)


Here is the Coronavirus
related research on SSRN

Paper statistics

Abstract Views
PlumX Metrics