Use of clinical history (CH) for Lynch syndrome (LS) risk assessment: An economic decision analysis.
1514 Background: Previous studies comparing screening strategies (STs) for LS concluded that immunohistochemistry (IHC)-based STs are most cost effective. However, these analyses generally exclude clinical history (CH)-based criteria due to concerns for reliability, e.g. Amsterdam (Ams), Bethesda (Beth), MMRpro (Mpro), MMRpredict (Mpre) and PREMM (PRE). We conducted a comprehensive comparison of all STs. Methods: Using assumptions from published reports we performed a cost effectiveness analysis (CEA) with a societal viewpoint using TreeAge software. 20 STs for proband (Pd) and general population (GP) screening in a cohort of 100,000 assumed a 3% LS prevalence (Prev) in Pd and 0.23% in GP, 5 first degree relatives (FDRs) per LS diagnosis (Dx), 50% LS Prev in FDRs, and 90% genetic testing (GT) compliance in Pds and GP and 52% in FDRs. We used the number of LS Dxs as effectiveness measure (EM). Sensitivity, accuracy, and incremental cost effectiveness ratios (ICERs) were calculated. Results: The ST of Mpro followed by GT is highly sensitive and has an accuracy of .997. Assuming 1-1.4 life years saved per Dx based on prior studies, this strategy is also cost effective. Strategies with greater sensitivity were dominated or not cost effective. Table shows results in order of increasing cost per Dx. Conclusions: This comprehensive comparison suggests that Mpro is an appropriate first step in screening for LS in Pds, and its routine use may be considered as a quality measure in the management of colorectal cancer. IHC + BRAF, GT should be reserved for Pds where CH is incomplete. [Table: see text]