Using expert elicitation to estimate the potential impact of improved diagnostic performance of laboratory tests: a case study on rapid discharge of suspected non–ST elevation myocardial infarction patients

摘要

Abstract Early health technology assessment can provide insight in the potential cost‐effectiveness of new tests to guide further development decisions. This can increase their potential benefit but often requires evidence which is lacking in early test development stages. Then, expert elicitation may be used to generate evidence on the impact of tests on patient management. This is illustrated in a case study on a new triple biomarker test (copeptin, heart‐type fatty acid binding protein, and high‐sensitivity troponin [HsTn]) at hospital admission. The elicited evidence enables estimation of the impact of using the triple biomarker on time to exclusion of non–ST elevation myocardial infarction compared with current serial HsTn measurement (performed 0, 2, and 6?h after admission). Cardiologists were asked to estimate the effect of the triple biomarker on patient's discharge rates and interventions performed, depending on its diagnostic performance. This elicited evidence was combined with Dutch reimbursement data and published evidence into a decision analytic model. Direct hospital costs and patients' discharge rates were assessed for 3 testing strategies including this triple biomarker (ie, only at admission or combined with HsTn measurements after 2 and 6?h). Direct hospital costs of suspected non–ST elevation myocardial infarction patients using serial HsTn measurements are estimated at €1825 per patient. Combining this triple biomarker with HsTn measurements after 2 and 6?hours is expected to be the most cost‐effective strategy. Depending on the diagnostic performance of the triple biomarker, this strategy is estimated to reduce costs with €66 to €205 per patient (ie, 3.6%‐11.3% reduction). Expert elicitation can be a valuable tool for early health technology assessment to provide an initial estimate of the cost‐effectiveness of new tests prior to their implementation in clinical practice. As demonstrated in our case study, improved diagnostic performance of the triple biomarker may have benefits that should be further explored.