A pilot study of implementing an adapted model for integration of interventions for people with alcohol use disorders in Tanzanian primary healthcare facilities.

Background: Ensuring that evidence-based interventions for people with alcohol use disorders (AUD) are acceptable, effective, and feasible in different socio-cultural and health system contexts is essential. We previously adapted a model of integration of AUD interventions for the Tanzanian primary healthcare system. This pilot study aimed to assess the impact on AUD detection and the acceptability and feasibility of the facility-based components of this model from the perspective of healthcare providers (HCPs).

Methods: This mixed-methods study comprised a pre-post quasi-experimental study and a qualitative study. The integrated model included training HCPs in managing AUD, introducing systematic screening for AUD, documentation of AUD service utilization, and supportive supervision. We collected information on the number of people identified for AUD three months before and after piloting the service model. A non-parametric trend test, a distribution-free cumulative sum test, was used to identify a change in the identification rate of AUD beyond that observed due to secular trends or, by chance, three months before and after implementing the integrated AUD facility-based interventions. The Mann-Kendal test was used to assess the statistical significance of the trend. We conducted three focus group discussions exploring the experience of HCPs and their perspectives on facilitators, barriers, and strategies to overcome them. The focus group discussions were analyzed using thematic analysis.

Results: During the pre-implementation phase of the facility-based interventions of the adapted AUD model, HCPs assessed 322 people for AUD over three months, ranging from a minimum of 99 to a maximum of 122 per month. Of these, 77 were identified as having AUD. Moreover, HCPs screened 2058 people for AUD during implementation; a minimum of 528 to a maximum of 843 people were screened for AUD per month for the three months. Of these, 514 screened positive for AUD (AUDIT ≥ 8). However, this change in screening for AUD was not statistically significant (p-value = 0.06). HCPs reported that knowledge and skills from the training helped them identify and support people they would not usually consider having problematic alcohol use. Perceived barriers to implementation included insufficient health personnel compared to needs and inconvenient health management information systems. HCPs proposed strategies to overcome these factors and recommended multisectoral engagement beyond the health system.

Conclusions: Although the change in the trend in the number of people screened for AUD by HCPs post-implementation was not statistically significant, it is still feasible to implement the facility-based components of the adapted integrated AUD model while addressing the identified bottlenecks and strategies for implementation. Therefore, a large-scale, adequately powered implementation feasibility study is needed. Findings from this study will be used to finalize the adapted model for integrating AUD interventions for future implementation and larger-scale evaluation.