TY - JOUR
T1 - Estimating the costs attributable to a disease with application to ovarian cancer
AU - Etzioni, Ruth
AU - Urban, Nicole
AU - Baker, Mary
N1 - Funding Information:
4.1.2. SURVIVAL DATA. We obtained survival data from the SEER database. This program is sponsoredb y the National Cancer Institute, and collects data from nine population-based tumor registries, covering about 10% of the U.S. population. Data recorded include month and year of diagnosis, age at diagnosis, type of cancer, clinical stage at diagnosis, histology, and month and year of death. For the analysis, data consistedo f 9700 epithelial ovarian cancer cases,w ho were diag-
Funding Information:
The authors thankAd dy Tseng, Fran M&y, and Ann Fowler for assistancew ith programminga nd data managementA; mie Potoskyf in helpfulc ommunicationan d, with Gem/d Riley, fur makingt he combinedS EER-Medicare databasea vailabk to us; Bill Barlow and StephenT aplin for helpfuld iscussionsa, nd Eric Feuer, for sharing his work with Yoch~nanW ax with us which helpedt o shapet hisarticle. The authors are especiallyg ratefult o Lany Kesslerf m his in-depthr eadin of, and commento.sn , an earlier draft of rhis manuscript.T his work was suppurtde by Contract NOICN-05230 from the Division of Cancer Preventiona nd Control of the National Cancer
PY - 1996/1
Y1 - 1996/1
N2 - This article is concerned with the methodological issues that arise when estimating the expected costs attributable to a disease. In particular, the article considers methods appropriate for handling incomplete or censored cost and survival data, incorporating discounting, and computing attributable costs. After motivating the need for an estimate of the average, present value of the attributable costs, we present the Kaplan-Meier sample average (KMSA) estimator, which takes into account the censored nature of the data that are typically available. We investigate the statistical properties of the estimator and compare it to others employed in the literature, showing how certain methods for incorporating discounting can introduce bias. We demonstrate the utility of the estimator by applying it to estimation of the costs attributable to ovarian cancer, using data from a database linking Medicare claims with the Surveillance, Epidemiology, and End Results cancer registry. Our analysis suggests that the average, present value of the 15 year costs attributable to ovarian cancer is $21,285 for local stage cases and $32,126 for distant stage cases in 1990 dollars.
AB - This article is concerned with the methodological issues that arise when estimating the expected costs attributable to a disease. In particular, the article considers methods appropriate for handling incomplete or censored cost and survival data, incorporating discounting, and computing attributable costs. After motivating the need for an estimate of the average, present value of the attributable costs, we present the Kaplan-Meier sample average (KMSA) estimator, which takes into account the censored nature of the data that are typically available. We investigate the statistical properties of the estimator and compare it to others employed in the literature, showing how certain methods for incorporating discounting can introduce bias. We demonstrate the utility of the estimator by applying it to estimation of the costs attributable to ovarian cancer, using data from a database linking Medicare claims with the Surveillance, Epidemiology, and End Results cancer registry. Our analysis suggests that the average, present value of the 15 year costs attributable to ovarian cancer is $21,285 for local stage cases and $32,126 for distant stage cases in 1990 dollars.
UR - http://www.scopus.com/inward/record.url?scp=0029970392&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0029970392&partnerID=8YFLogxK
U2 - 10.1016/0895-4356(96)89259-6
DO - 10.1016/0895-4356(96)89259-6
M3 - Article
C2 - 8598518
AN - SCOPUS:0029970392
SN - 0895-4356
VL - 49
SP - 95
EP - 103
JO - Journal of Chronic Diseases
JF - Journal of Chronic Diseases
IS - 1
ER -