Feasibility of Testing Client Preferences for Accessing Injectable Opioid Agonist Treatment (iOAT): A Pilot Study

Sophia Dobischok,1 Rebecca K Metcalfe,1 Elizabeth Angela Matzinger,1 Kurt Lock,1,2 Scott Harrison,3 Scott MacDonald,3 Sherif Amara,4 Martin T Schechter,1,5 Nick Bansback,1,5 Eugenia Oviedo-Joekes1,5 1Centre for Health Evaluation & Outcome Sciences, Providence Health Care, Vancouver, BC, Canada; 2BC Centre for Disease Control, Provincial Health Services Authority, Vancouver, BC, Canada; 3Providence Health Care, Providence Crosstown Clinic, Vancouver, BC, Canada; 4SafePoint Supervised Consumption Site, Fraser Health Authority, Surrey, BC, Canada; 5School of Population and Public Health, University of British Columbia, Vancouver, BC, CanadaCorrespondence: Eugenia Oviedo-Joekes, Centre for Health Evaluation and Outcome Sciences, St. Paul’s Hospital, 575-1081 Burrard St, Vancouver, BC, V6Z 1Y6, Canada, Tel +1 604-682-2344 Ext. 62973, Fax +1-604-806-8210, Email eugenia.joekes@ubc.caPurpose: Injectable opioid agonist treatment (iOAT) is an effective treatment for opioid use disorder (OUD). To our knowledge, no research has systematically studied client preferences for accessing iOAT. Incorporating preferences could help meet the heterogenous needs of clients and make addiction care more person-centred. This paper presents a pilot study of a best-worst scaling (BWS) preference elicitation survey that aimed to assess if the survey was feasible and accessible for our population and to test that the survey could gather sound data that would suit our planned analyses.Patients and Methods: Current and former iOAT clients (n = 18) completed a BWS survey supported by an interviewer using a think-aloud approach. The survey was administered on PowerPoint, and responses and contextual field notes were recorded manually. Think-aloud audio was recorded on Audacity.Results: Clients’ feedback fell into five categories: framing of the task, accessibility, conceptualization of attributes and levels, formatting, and behaviour predicting questions. Survey repetitiveness was the most