Finding the optimal mammography screening strategy: A cost-effectiveness analysis of 920 modelled strategies

Abstract

Breast cancer screening policies have been designed decades ago, but current screening strategies may not be optimal anymore. Next to that, screening capacity issues may restrict feasibility. This cost-effectiveness study evaluates an extensive set of breast cancer screening strategies in the Netherlands. Using the Microsimulation Screening Analysis-Breast (MISCAN-Breast) model, the cost-effectiveness of 920 breast cancer screening strategies with varying starting ages (40-60), stopping ages (64-84) and intervals (1-4 years) were simulated. The number of quality adjusted life years (QALYs) gained and additional net costs (in €) per 1000 women were predicted (3.5% discounted) and incremental cost-effectiveness ratios (ICERs) were calculated to compare screening scenarios. Sensitivity analyses were performed using different assumptions. In total, 26 strategies covering all four intervals were on the efficiency frontier. Using a willingness-to-pay threshold of €20 000/QALY gained, the biennial 40 to 76 screening strategy was optimal. However, this strategy resulted in more overdiagnoses and false positives, and required a high screening capacity. The current strategy in the Netherlands, biennial 50 to 74 years, was dominated. Triennial screening in the age range 44 to 71 (ICER 9364) or 44 to 74 (ICER 11144) resulted in slightly more QALYs gained and lower costs than the current Dutch strategy. Furthermore, these strategies were estimated to require a lower screening capacity. Findings were robust when varying attendance and effectiveness of treatment. In conclusion, switching from biennial to triennial screening while simultaneously lowering the starting age to 44 can increase benefits at lower costs and with a minor increase in harms compared to the current strategy.

Keywords: breast cancer; cost-effectiveness; mass screening; microsimulation modelling; screening strategies.

FIGURE 1

Cost‐effectiveness curve for scenarios with starting ages between 40 and 60 and stopping ages between 64 and 84, including efficiency frontier [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 2

Effects of internationally implemented strategies assuming 100% attendance. The United Kingdom (UK) and Malta (MA) triennial 50 to 69; Estonia (EST) and biennial 50 to 64; Belgium (BE), Germany (DE), Poland, Cyprus, Denmark, Finland, Latvia, Lithuania, Luxembourg, Norway, Poland, Slovenia and Switzerland biennial 50 to 69; the Netherlands (NL) and France (FR) biennial 50 to 74; Hungary (HU) biennial 45 to 65; Austria (AU) and Czech republic (CZ) biennial 45 to 69; Sweden (SW) biennial 40 to 69 [Color figure can be viewed at wileyonlinelibrary.com]