| The objectives of this research were: to conduct quality-weighted random effects meta-analysis of the effect of mammography (MM) screening on breast cancer (BC) mortality, to create a framework to analyze the cost-effectiveness of MM screening, designed specifically to isolate the effect of MM on life expectancy (LE) with BC, and to run full sensitivity analyses of this model.; Methods. The relative risk (RR) of BC mortality in screened vs. control groups was estimated by incorporating a study-specific quality weight into inversevariance weighted random effects (RE) analysis. Cost-effectiveness of MM screening vs. no screening was analyzed on a base case of a white female population being screened annually from the age of 40, on a lifetime horizon from the societal perspective. The RR of BC mortality was used to estimate life expectancy with BC in screened women. Direct and indirect costs included MM, its workup and BC management. Extensive sensitivity analyses were conducted.; Results. Significant 20% and 22% reductions in BC mortality were observed in screened women of any age and above age fifty [RR: 0.78(95%CI=0.70--0.88)] respectively. CIs were narrower than those from RE analysis. Under fifty, the result reduced to a fixed effects analysis: 0.84(95%CI=0.73--0.97). No trend of effect with study quality was observed, although under fifty, seven studies were pooled before a significant effect was observed. MM is most cost-effective at age 60, (ICER: {dollar}38,876/QALY), being slightly less so at 40 (ICER: {dollar}47,991/QALY). The ICER (incremental cost-effectiveness ratio) increases steadily after age 65, reaching its highest estimation at age 80. The model was very sensitive to the RR of BC mortality with MM. SA also showed sensitivity to discount rate, but relative robustness to other parameters.; Conclusions. Quality scores may be used in a meta-analysis to account for between-study variance due to heterogeneity. Evidence of MM benefit under 50 may depend on the number and quality of studies included. The most cost-effective program tested was to start annual MM at age 40, switching to 2-year intervals at 50, and stopping at age 70. This framework can be used for the cost-effectiveness analysis of screening recommendations, and future policy analyses. |
