ID: 29696
Meeting / Value in Health Info: ISPOR 17th Annual International Meeting
Washington, D.C., USA
June, 2012
Code: PRM33
Disease: Multiple Diseases/No Specific Disease
Topic: Research on Methods (RM)
Topic Subcategory: MS
Title: VALIDATING A WEB-BASED, INCREMENTAL COST-EFFECTIVENESS SOFTWARE PROGRAM THAT IMPLEMENTS A MARKOV CHAIN MONTE CARLO (MCMC) ANALYSIS MODEL
Author(s):

McGhan WF, Khole T, Vichaichanakul K, Willey VJUniversity of the Sciences, Philadelphia, Philadelphia, PA, USA

Content:

OBJECTIVES: To evaluate a web-based software program which incorporates Markov Chain Monte Carlo (MCMC) analysis to compare the cost-effectiveness of any two treatments, allowing modifiable inputs of key variables. METHODS: A web-based software program was developed, which incorporates Markov Chain Monte Carlo (MCMC) analysis to compare the cost-effectiveness of any two treatments. The online software program was based on calculation methods described in “Decision Making in Health and Medicine” textbook from Hunink et al. The MCMC web-based program computes and graphically displays the results, using JavaScript algorithms and is available as freeware at www.healthstrategy.com. We compared the online results with analyses using Decision Maker software available from UMDNJ.edu. The variable inputs that can be modified in the web-based application include: state transition probabilities, number of patients, number of cycles, cost per state, and utility per state. RESULTS: The web-based tool creates plots of incremental costs versus incremental utilities, in cost-effectiveness quadrants; and if death is the absorbing state, also graphs life expectancy curves for two treatment comparisons. As an example of the similarity of findings, when considering three transition states per treatment, the online software versus the Decision Maker model results were as follows: treatment cost (means: $1417 vs. $1300 and standard deviations: 1706 vs. 1604); treatment effectiveness (means: 7.6 vs. 7.8 and standard deviations: 7.2 vs. 7.0). CONCLUSIONS: With this online MCMC program, the user can input their own therapy parameters, and then generate key means and standard deviations, incremental costs, incremental utilities, life expectancy curves, and incremental cost effectiveness ratios. MCMC has advantages over Markov cohort analyses because means and standard deviations can be generated from the MCMC calculations. This web-based application has potential benefit as a basic educational tool for students and health professionals interested in exploring these analytical approaches.

Back to HDS Software List