Psychometric performance of tools to measure internalized drug use stigma among people who use drugs and people on medication-assisted treatment in Dar es Salaam, Tanzania

Background

People who use drugs (PWUD) and people who use medication-assisted treatment (MAT) to treat opioid use disorder face severe stigma. Stigma may manifest in delayed health-seeking behaviors, which adversely affects health and increases disease risk. Few validated measures assess internalized drug use stigma or MAT disclosure concerns in sub-Saharan Africa. This study examines the performance of internalized drug use stigma scales and characterizes the prevalence of these stigmas among PWUD and people who use MAT in Dar es Salaam, Tanzania.

Methods

We conducted a cross-sectional survey among 250 PWUD and 251 MAT clinic clients. Internalized drug use and MAT disclosure concerns were measured using scales adapted to the Tanzanian context. We conducted exploratory factor analysis to identify latent constructs and evaluate reliability of the scales. We then used confirmatory factor analysis to assess construct validity and reliability.

Results

The six-item internalized drug use stigma scale (IDUSS) was unidimensional, reflected acceptable goodness-of-fit, and acceptable reliability (Cronbach’s α = 0.79). The eight-item internalized drug use stigma scale for people using MAT (IDUSS-MAT) was bidimensional. Allowing error terms of related items to covary improved goodness-of-fit. The IDUSS-MAT demonstrated acceptable reliability (Cronbach’s α = 0.79). Among PWUD and MAT clinic clients, higher levels of stigma were associated with increased depression and reduced social support and quality of life.

Conclusions

Both the IDUSS and IDUSS-MAT demonstrated good construct validity and acceptable reliability. These validated scales provide the foundation for collecting data to inform the development of stigma reduction interventions in Tanzania and among similar Sub-Saharan populations of PWUD and the tools needed to assess those interventions.