Cost-effectiveness evaluation of predictive molecular diagnostics using the example of Hereditary Non-Polyposis Colorectal Cancer (HNPCC)

Study Objective: Four different diagnostic strategies, with and without various molecular diagnostic tests, are compared and contrasted not only by years gained and the cost of therapy and diagnosis, but also by the cost-effectiveness of the diagnostic strategies. Methodology: A fictitious cohort of 100,000 people, whose genetic pre-disposition leading to the development of colorectal cancer corresponds to a representative average amongst the current population, will be studied from their 1st to their 85th year. This data will be then put through Markov models specifically developed for the study. At the end of the Markov process, it will then be possible to compile a cost-effectiveness report in regard to the various diagnostic and treatment strategies. Results: A tiered diagnosis (with family case history, micro-satellite instability, molecular diagnostic diagnosis of an index person and subsequent genetic analysis of all people at risk) represents the most cost-effective method at a rate of C 3,867 per year gained. The cost-effectiveness of a purely clinical diagnosis has a rate of C 4,397 per year gained and is followed by the cost of direct gene testing of people at risk from families at risk at a rate of (sic) 6,208. The worst level of cost-effectiveness, with a rate of C 15,705, was shown by nationwide gene screening. The incremental cost-effectiveness of Strategy IV and Strategy 11 is C 124,168 per gained year. Conclusions: With the scenarios put forward we can show that a 65% reduction in gene test costs is necessary in order for a cost-effective nationwide gene screening for HNPCC to take place. The break-even level, however, depends only on a few cost-effectiveness drivers such as screening and therapy costs, proportion of HNPCC of all colorectal cancer and discounting rate. Should these changes (e.g., through a restructured medical environment), then we would see such a change in the break-even cost of a gene test and that a cost-effective nationwide gene screening could be made plausible. In a final evaluation of the use of predictive molecular diagnostics, other dimensions (such as possible psychological problems and discriminatory risks) apart from cost-effectiveness should also be included.