P-731 Costs and effects accounted for in economic evaluations of Medically Assisted Reproduction: a systematic review

Abstract Study question Which costs and effects must be considered in an economic evaluation of Medically Assisted Reproduction (MAR) and to which extent do published studies consider these? Summary answer Current economic evaluations of MAR focus on technical efficiency using narrow short-term perspectives and excluding relevant costs/effects, and hence, fail to provide adequate ‘value-for-money’ information. What is known already Due to limited budgets of health systems and patients, decision-makers need to understand which MAR interventions are cost-effective. However, it is complex to evaluate MAR from an economic perspective, as MAR treatment involves women, men, their relationship, and potential children born. Hence, it remains unclear which costs and effects related to these different stakeholders should be accounted for, and over which time horizon. In this study, we performed a systematic review of health economic studies in MAR and compared the results with a predefined value-framework providing 12 cost categories and 9 effect categories. Study design, size, duration A systematic review was performed of articles (published 2010-2021) in PubMed Central, Embase, Web of Science CC, CINAHL, Cochrane (CENTRAL), HTA and NHS EED. Using the PICO framework, full economic evaluations (cost-utility analysis (CUA), cost-benefit analysis (CBA), cost-minimization analysis (CMA) and cost-effectiveness analysis (CEA)) were selected and assessed. We identified 66 economic evaluations of MAR, including the analysis of 81 Incremental Cost-Effectiveness Ratios (ICER) as 15 studies presented more than one ICER. Participants/materials, setting, methods Our value-framework lists 12 cost-categories (infertility diagnosis, MAR, pregnancy, delivery, post-natal -care, basic/complicated childcare, care for involuntary childlessness, productivity losses (MAR/pregnancy/childcare), patient out-of-pocket expenses, broader long-term social costs) and 9 effect-categories (health child/parents short- and long-term, well-being child/parents short- and long-term, broader long-term social benefits) of MAR, indicating which costs/effects are relevant for different analytical methods (CEA/CUA/CBA), time-horizons (<1 year/1-5 years/>5 years) and perspectives (health system/patient/society). All ICERs were analysed based on their included costs/effects. Main results and the role of chance There is a large scope of costs and effects potentially relevant to understand the economic value of MAR. Of the 66 included studies, 88% performed a CEA whereas 2% performed a CUA. Most ICERs expressed cost-effectiveness as cost per live birth (52%) or cost per pregnancy (24%) whereas one study expressed the result as cost per QALY. 24% of all ICERs adopted the broader societal perspective, whereas 61% adopted the narrower health system perspective. Time horizons considered were very heterogeneous, with 68% not looking further than one year after treatment and 1% considering a time horizon of a lifetime. Only the direct costs of MAR and the short-term medical outcomes of born children were consistently accounted for in all ICERs. The following significant costs and effects were largely neglected: costs related to diagnostics, pregnancy and delivery (including complications), postnatal follow-up care and childcare; indirect costs (patient out-of-pocket expenses, productivity losses); indirect effects of infertility on the health of wish parents; longer term costs and effects of a live birth. These significant cost and effects were excluded due to the limited scope set by most studies, adopting short time horizons and narrow perspectives. Limitations, reasons for caution The systematic review relied on available databases of published scientific articles. Therefore, economic evaluations published in the grey literature (government reports, unpublished manuscripts, etc.) may have been missed. Only articles published in English were included. Wider implications of the findings As MAR has broad implications for many stakeholders, future studies should adopt more accurate costing-methods (using our value-framework) and focus on allocative efficiency. Hence, the debate about the value-for-money of MAR and about how large MAR budgets should be relative to those of other disease areas, will be better informed. Trial registration number Not applicable