The daily costs of chemotherapy-associated adverse events: A cluster-based analysis

Abstract

6001 Background: Chemotherapy-associated adverse events (C-AEs) including neutropenia (N), anemia (A), thrombocytopenia (T), mucositis (M) and dehydration (D) are common and associated with increased hospital costs. However, C-AEs often occur in clusters, and as a consequence, it makes little clinical sense to discuss the costs associated with a single C-AE in isolation from other C-AEs. Methods: This study estimates the costs of inpatient and outpatient C-AEs based on a new approach to classifying C-AEs - the clinically relevant C-AE cluster. The study population included all cancer patients receiving chemotherapy (N=2,067) during August 1, 2000 and July 31, 2001 at The University of Pennsylvania Health System. For these patients, C-AE clusters were determined using clinical algorithms based on the start and end dates of defined adverse events and dates of chemotherapy administration. Regression models were developed using a loglinear variance adjusted estimation procedure, to better estimate the daily costs associated with the presence of these AE clusters, adjusting for age, sex, cancer diagnosis, metastatic disease, chemotherapy regimen, and comorbidities including AIDS,congestive heart failure, angina, arrhythmia, hypertension, asthma, renal failure, diabetes, and other conditions. Results: The presence of increasingly complex clusters increased the daily cost of hospitalization. The cost of an adjusted hospital day with N alone was $1704 (95% CI 1189 to 2220), and the adjusted daily hospital cost for A was $2035 (1863 to 2207), while a hospital day with A+N+T was $2381 (2190 to 2586), and A+N+T+D was $2852 (2314 to 3389). A hospital day with M alone was $910 (-691 to 2511), while M + (N and/or A,T) increased the costs to $2001 (1733–2272). Conclusions: Chemotherapy associated AEs often occur in clusters. Observing these clusters provides an excellent method for improving the estimate of expected costs from hospitalization. Models incorporating clinically meaningful C-AE clusters can be utilized to both estimate the expected total costs of chemotherapy-associated AEs and to explore more rational utilization of supportive care interventions aimed at reducing C-AEs. (Funded by a grant from Amgen, Inc.) Author Disclosure Employment or Leadership Consultant or Advisory Role Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Amgen