Not so NICE: Comparison of EQ-5D-3L and AQol-8D utility measures in first episode psychosis

Poster A59, Thursday, October 20, 11:30 am - 1:00 pm, Le Baron

Caragh Behan1, Sarah Masterson1, Eric Roche1, Laoise Renwick2, Catherine McDonough3, Brendan Kennelly4, Paul McCrone5, Mary Clarke1,6; 1DETECT EIS / St John of God Hospital, 2School of Nursing, University of Manchester, 3Cavan Monaghan Community Mental Health Service, Ireland, 4School of Economics, NUIG Galway, Ireland, 5King's Health Economics, King's College London, 6School of Medicine, University College Dublin, Ireland

Multiattribute utility measures such as the EQ-5D-3L, are used to generate Quality Adjusted Life Years (QALYs). The EQ-5D-3L is the preferred measure of many decision makers internationally for generating a QALY in cost-effectiveness analysis. QALYs are used to compare across interventions and conditions. While the EQ-5D-3L is useful in many conditions there are concerns about its sensitivity in conditions such as psychosis. We examined a sample of people aged 17-65 presenting with affective and non-affective first episode psychosis between 2012-2014 to five rural and urban catchment areas in Ireland and followed-up at one year . The sample was part of a cohort participating in an economic evaluation of an early intervention service. Of these, 227 people completed at least one utility questionnaire. 94 completed the EQ-5D-3L at baseline and one year, and 159 people completed both an EQ-5D and an AQol-8D at one year. The EQ-5D-3L was valued using the UK TTO set, and the AQoL-8D was valued using the Australian TTO. Data were examined using SPSS. Preliminary results showed that the EQ-5D has a higher mean than the AQOL-8D, and a marked ceiling effect with a clustering of values at one. The measures were poorly correlated both with each other. Further analysis will evaluate correlation with disease specific measures. Multiattribute utility instruments facilitate the comparison of outcomes across conditions and interventions, yet these instruments appear insensitive in psychosis. This has implications for the distribution of resources based on the results of economic evaluations of interventions for people with FEP.

Topic Area: First Episode Psychosis

Back to Poster Schedule