A multi-methods and longitudinal study of patients' perceptions in injectable opioid agonist treatment: Implications for advancing patient-centered methodologies in substance use research

Background: Patients' perceptions are vital to the delivery and evaluation of substance use treatment. They are most frequently collected at one time-point and measured using patient satisfaction questionnaires or qualitative methodologies. Interestingly, the findings of these studies often diverge, as satisfaction scores tend to be highly positive, while qualitative findings suggest dissatisfaction and areas for improvement. This divergence limits current understandings of patients' perceptions and their potential change over time in treatment.

Objective: This study explores the relationship between open-ended positive and negative perceptions of treatment and patient satisfaction scores over time.

Methods: The RUTH (Research on the Utilization of Therapeutic Hydromorphone) prospective cohort study included 131 participants receiving injectable diacetylmorphine or hydromorphone in Canada's first injectable opioid agonist treatment (iOAT) program. The study collected the Client Satisfaction Questionnaire (CSQ-8) at eight time-points over an 18-month period. Following a multi-methods approach, the study complemented the CSQ-8 with open-ended positive and negative comments of iOAT. The research team analyzed these comments thematically at each time-point to develop positive and negative perception themes. We then used growth curve modeling to explore the relationship between positive and negative perception themes and patient satisfaction over time.

Findings: Over the eight time-points, six positive and eight negative perception themes emerged, broadly reflecting structural (e.g., expansion of iOAT), process (e.g., schedules), relational (e.g., interactions with providers), and outcome-related (e.g., met/unmet needs) perceptions of iOAT. On average, participants reported high satisfaction (grand mean = 29.2 out of 32), and scores did not significantly change over time. However, we did find significant unexplained variation within participants in their satisfaction trajectories and between participants in their initial satisfaction scores. In conditional growth curve models, the theme "unfavorable interactions with providers" had the strongest independent effect on overall satisfaction trajectories.

Conclusions: This study provides an example of how open-ended comments can be integrated with patient satisfaction questionnaire data to gather a comprehensive and patient-centered evaluation of substance use treatment. Considering the iOAT context specifically, relational dynamics and daily treatment access were significant predictors of patient satisfaction over time and may be attributes of iOAT that require further investigation.