RESUMEN
OBJECTIVE: Inclusion of relevant effectiveness and safety outcomes in economic evaluation of health technologies is required to aid efficient healthcare decisions. Our objective was to identify the key issues related to the inclusion of adverse events (AEs) in economic evaluation and explore perspectives for good practice recommendations to handle these issues. METHODS: We focused on the frequently encountered methodological issues related to the integration of AEs in economic evaluations of health technologies. We distinguished the following elements: the incorporation of AEs in decision models, the terminology of AEs, the estimation of AEs consequences in terms of quality of life (QoL) and costs, and the exploration of the uncertainty related to the impact of AEs on the economic results. RESULTS: We illustrated and discussed each of the identified issues by giving health technology assessment examples. We focused on the extent to which the integration of AEs in decision models can be improved by dealing with the lack of relevant real-world safety data, estimating the consequences of AEs (eg, for costs and QoL loss), exploring the impacts of AEs that are not adequately captured in current measurement of health-related QoL, and identifying the need for development of a good terminology of relevant types of AEs to be incorporated in economic evaluation. CONCLUSION: Based on a reflection the key methodological issues related to the incorporation of adverse drug events in economic evaluations, we suggested several recommendations to serve a starting point for health technology assessment agencies and researchers to develop good research practices in this field.
Asunto(s)
Análisis Costo-Beneficio , Calidad de Vida , Evaluación de la Tecnología Biomédica , Humanos , Análisis Costo-Beneficio/métodos , Efectos Colaterales y Reacciones Adversas Relacionados con Medicamentos/economía , Técnicas de Apoyo para la Decisión , Incertidumbre , Terminología como Asunto , Modelos EconómicosRESUMEN
BACKGROUND: Major depressive disorder (MDD) is associated with decreased patient well-being and symptoms that can cause substantial impairments in patient functioning and even lead to suicide. Worldwide, MDD currently causes the second-most years lived with disability and is predicted to become the leading cause of disability by 2030. Utility values, capturing patient quality of life, are required in economic evaluations for new treatments undergoing reimbursement submissions. We aimed to identify health state utility values (HSUVs) and disutilities in MDD for use in future economic evaluations of pharmacological treatments. METHODS: Embase, PubMed, Econlit, and Cochrane databases, plus gray literature, were searched from January 1998 to December 21, 2018, with no language or geographical restrictions, for relevant studies that reported HSUVs and disutilities for patients with MDD receiving pharmacological interventions. RESULTS: 443 studies were identified; 79 met the inclusion criteria. We focused on a subgroup of 28 articles that reported primary utility data from 16 unique studies of MDD treated with pharmacological interventions. HSUVs were elicited using EQ-5D (13/16, 81%; EQ-5D-3L: 11/16, 69%; EQ-5D-3L or EQ-5D-5L not specified: 2/16), EQ-VAS (5/16, 31%), and standard gamble (1/16, 6%). Most studies reported baseline HSUVs defined by study entry criteria. HSUVs for a first or recurrent major depressive episode (MDE) ranged from 0.33 to 0.544 and expanded from 0.2 to 0.61 for patients with and without painful physical symptoms, respectively. HSUVs for an MDE with inadequate treatment response ranged from 0.337 to 0.449. Three studies reported HSUVs defined by MADRS or HAMD-17 clinical thresholds. There was a large amount of heterogeneity in patient characteristics between the studies. One study reported disutility estimates associated with treatment side effects. CONCLUSIONS: Published HSUVs in MDD, elicited using methods accepted by health technology assessment bodies, are available for future economic evaluations. However, the evidence base is limited, and it is important to select appropriate HSUVs for the intervention being evaluated and that align with clinical health state definitions used within an economic model. Future studies are recommended to elicit HSUVs for new treatments and their side effects and add to the existing evidence where data are lacking.