Using Decision-Analytic Modeling to Isolate Interventions That Are Feasible, Efficient and Optimal: An Application from the Norwegian Cervical Cancer Screening Program.

<bold>Background: </bold>Decision makers often need to simultaneously consider multiple criteria or outcomes when deciding whether to adopt new health interventions.<bold>Objectives: </bold>Using decision analysis within the context of cervical cancer screening in Norway, we aimed to aid decision makers in identifying a subset of relevant strategies that are simultaneously efficient, feasible, and optimal.<bold>Methods: </bold>We developed an age-stratified probabilistic decision tree model following a cohort of women attending primary screening through one screening round. We enumerated detected precancers (i.e., cervical intraepithelial neoplasia of grade 2 or more severe (CIN2+)), colposcopies performed, and monetary costs associated with 10 alternative triage algorithms for women with abnormal cytology results. As efficiency metrics, we calculated incremental cost-effectiveness, and harm-benefit, ratios, defined as the additional costs, or the additional number of colposcopies, per additional CIN2+ detected. We estimated capacity requirements and uncertainty surrounding which strategy is optimal according to the decision rule, involving willingness to pay (monetary or resources consumed per added benefit).<bold>Results: </bold>For ages 25 to 33 years, we eliminated four strategies that did not fall on either efficiency frontier, while one strategy was efficient with respect to both efficiency metrics. Compared with current practice in Norway, two strategies detected more precancers at lower monetary costs, but some required more colposcopies. Similar results were found for women aged 34 to 69 years.<bold>Conclusions: </bold>Improving the effectiveness and efficiency of cervical cancer screening may necessitate additional resources. Although efficient and feasible, both society and individuals must specify their willingness to accept the additional resources and perceived harms required to increase effectiveness before a strategy can be considered optimal. [ABSTRACT FROM AUTHOR]