A qualitative examination of the content validity and acceptability of the EQ-5D-Y-3L and EQ-5D-Y-5L in adolescents with type 1 diabetes mellitus in Ethiopia

Abstract

Abstract Background The EQ-5D-Y-3L and EQ-5D-Y-5L are widely used generic measures of health-related quality of life (HRQoL), including among children and adolescents with Type 1 Diabetes Mellitus (T1DM). However, evidence on the measures’ content validity and acceptability in low-income settings, particularly sub-Saharan Africa, remains limited. This study aimed to assess the content validity and acceptability of the EQ-5D-Y-3L and EQ-5D-Y-5L among adolescents with T1DM in Ethiopia. Methods Semi-structured qualitative interviews were conducted with adolescents aged 12–17 years with T1DM in Addis Ababa, Ethiopia. Purposive sampling was used to achieve breadth in age, sex, and duration of illness. Combined concept elicitation and cognitive interviews followed an interview guide exploring participants’ understanding of health, experiences completing the EQ-5D-Y-3L and EQ-5D-Y-5L, and comprehensibility, relevance, and comprehensiveness of the measures. Data were analysed using descriptive qualitative content analysis. Health concepts reported as important by the adolescents were mapped against the EQ-5D-Y-3L/5L descriptive system. Results Twenty-five adolescents participated; 13 were male. Adolescents independently described health in terms of physical health, emotional well-being, functioning, disease and dietary management, and social functioning. Most participants identified concepts aligned with EQ-5D-Y-3L/5L dimensions, and both measures were generally reported to be easy to understand and complete. However, some comprehension issues were noted, including broad interpretations on two of the five dimensions and difficulty selecting precise values on the EQ VAS. Participants endorsed the relevance of the measures to their ill health experience. Several important disease-specific aspects, such as diabetes self-management, dietary management, treatment burden, and supportive social relationships, were not captured. The recall period ‘today’ was considered short (inadequate) mainly due to the fluctuating nature of T1DM. When asked about preference for measures, most preferred the EQ-5D-Y-5L over EQ-5D-Y-3L due to its greater descriptive capability. Conclusion The EQ-5D-Y-3L and EQ-5D-Y-5L demonstrate good acceptability, and their content validity in terms of understanding and relevance was generally supported. However, issues related to comprehension and comprehensiveness were noted. These findings support the use of both generic measures of HRQoL in Ethiopian adolescents with T1DM if a broad measure of health is required. When these are supplemented by disease-specific measures, they may provide a more comprehensive assessment for adolescents with T1DM.