Optimal healthcare decision making under multiple mathematical models: application in prostate cancer screening

Abstract

Important decisions related to human health, such as screening strategies for cancer, need to be made without a satisfactory understanding of the underlying biological and other processes. Rather, they are often informed by mathematical models that approximate reality. Often multiple models have been made to study the same phenomenon, which may lead to conflicting decisions. It is natural to seek a decision making process that identifies decisions that all models find to be effective, and we propose such a framework in this work. We apply the framework in prostate cancer screening to identify prostate-specific antigen (PSA)-based strategies that perform well under all considered models. We use heuristic search to identify strategies that trade off between optimizing the average across all models' assessments and being "conservative" by optimizing the most pessimistic model assessment. We identified three recently published mathematical models that can estimate quality-adjusted life expectancy (QALE) of PSA-based screening strategies and identified 64 strategies that trade off between maximizing the average and the most pessimistic model assessments. All prescribe PSA thresholds that increase with age, and 57 involve biennial screening. Strategies with higher assessments with the pessimistic model start screening later, stop screening earlier, and use higher PSA thresholds at earlier ages. The 64 strategies outperform 22 previously published expert-generated strategies. The 41 most "conservative" ones remained better than no screening with all models in extensive sensitivity analyses. We augment current comparative modeling approaches by identifying strategies that perform well under all models, for various degrees of decision makers' conservativeness.

Operational research techniques applied throughout cancer care services: a review

“…Hall et al (1992) investigate how best to choose between proposed projects to reduce the number of smokers in the USA. Stakeholders participated first in a modified Delphi process to determine appropriate Bertsimas et al (2016) Prostate Finding optimal fixed-interval screening strategies according to multiple models Multi-objective optimisation, local search heuristic decision criteria and then answered a questionnaire to decide on their relative importance. This led to development of a binary integer program which allocates funds between proposals based on budgetary constraints and how well the proposals meet the criteria.…”

“…One paper is about cervical cancer screening through pap smears (McLay, Foufoulides, & Merrick, 2010), one is for general cancer screening with an application to prostate cancer screening (Bertsimas, Silberholz, & Trikalinos, 2016) and the other is a model for general chronic diseases, with a case study for breast cancer screening (Rauner, Gutjahr, Heidenberger, Wagner, & Pasia, 2010). McLay et al (2010) and Rauner et al (2010) both develop new models to optimise screening policies, while Bertsimas et al (2016) combine recommendations from multiple existing models to find screening strategies that balance being optimal on average with those being optimal according to the most pessimistic model. Both Bertsimas et al (2016) and Rauner et al (2010) Rauner et al's (2010) model is that it can be used to optimise screening strategies across several diseases simultaneously.…”

“…McLay et al (2010) and Rauner et al (2010) both develop new models to optimise screening policies, while Bertsimas et al (2016) combine recommendations from multiple existing models to find screening strategies that balance being optimal on average with those being optimal according to the most pessimistic model. Both Bertsimas et al (2016) and Rauner et al (2010) Rauner et al's (2010) model is that it can be used to optimise screening strategies across several diseases simultaneously.…”

“…McLay et al (2010) develop a simulation-optimisation model, while both Rauner et al (2010) and Bertsimas et al (2016) formulate multi-objective optimisation problems that are solved with metaheuristics and a local search heuristic, respectively. McLay et al (2010) consider three different performance measures (number of cervical cancer deaths, number of life years lost due to cervical cancer and cervical cancer incidence), and optimise separately for each of them.…”

“…McLay et al (2010) consider three different performance measures (number of cervical cancer deaths, number of life years lost due to cervical cancer and cervical cancer incidence), and optimise separately for each of them. Rauner et al (2010) and Bertsimas et al (2016) assess models in terms of quality-adjusted life years; Rauner et al (2010) also set a budget constraint. Gu, Wang, and McGregor (2010) and Haase and Müller (2015) aim to determine the best locations for screening facilities.…”