Decision-analytic approaches to evaluating prevention policy alternatives

To achieve population health given uncertainty and budgetary constraints, policymakers must choose between interventions. We apply decision science to inform such decisions. We construct a microsimulation model of human papillomavirus (HPV) and cervical cancer, developing empirical calibration procedures to translate disease natural history uncertainty into uncertainty about policy-relevant outcomes. The model evaluates U.S. cervical cancer prevention given newer screening technologies and HPV vaccines. We find: age-based screening differing from current guidelines appears cost-effective; vaccinating pre-adolescents is most beneficial; and all women should continue screening. Using simulation models calibrated to epidemiological data from low resource countries, we identify influential parameters in country-specific decisions about cervical cancer screening. Through fieldwork and internationally available data, we quantify the costs of achieving patient adherence, laboratory processing and specimen transport, and patient time seeking care. Including these costs influences the cost-effectiveness of screening alternatives. Focus on programmatic aspects of service delivery is relevant to chronic disease prevention worldwide. To inform vaccine policy for diseases such as measles, we consider methods that complement simulation models. While dynamic transmission models often require unavailable data, longitudinal regressions can identify determinants of vaccine coverage and reductions in mortality due to vaccination. We find that vaccine coverage depends on health worker availability and sustained program investment and that coverage is associated with substantial mortality reductions. This approach is relevant for many vaccine-preventable diseases.