Effectiveness of the brief negotiation intervention across latent profiles of risky alcohol behaviours among trauma patients at the Northern Zonal referral hospital in Tanzania
Background: Alcohol use contributes to injury, making trauma care an opportunity to address harmful drinking. In Tanzania, the Punguza Pombe Kwa Afya Yako (PPKAY) brief intervention reduced binge-drinking days among trauma patients at three months. However, its overall effect does not establish whether patients with different patterns of alcohol-related harms respond differently.Aims: To identify baseline profiles of risky alcohol-related behaviours and assess whether PPKAY effectiveness differed across profiles.Methods: We analysed data from a pragmatic randomised adaptive trial at Kilimanjaro Christian Medical Centre (2020-2025). Latent class analysis of ten Drinker Inventory of Consequences indicators included 674 patients. Classification-error-adjusted regression examined profile characteristics. Among 543 patients with recorded three-month follow-up, baseline-adjusted models compared pooled PPKAY arms with usual care and tested intervention-by-profile interactions. Binge-drinking days was the primary outcome; bootstrap-then-impute procedures estimated confidence intervals.Results: Two profiles were identified: lower-risk with driving-related harms (531; 78.8%) and pervasive high-risk behaviours (143; 21.2%). Younger age, higher alcohol-use severity, depressive symptoms and perceived injury severity characterised the high-risk profile. For binge-drinking days, PPKAY incidence rate ratios were 0.84 (95% CI 0.22-3.37) and 0.023 (95% CI 0.001-0.223), respectively. The interaction ratio was 0.028 (95% CI 0.001-0.456; p = 0.024). Interactions for drinking days and drinking amount were not statistically significant.Conclusions: PPKAY had a stronger estimated effect on binge-drinking days in the pervasive high-risk profile. PPKAY should remain available to all eligible trauma patients, while broader assessment of alcohol-related harms and depressive symptoms should be evaluated to identify patients needing additional support.Keywords: Alcohol drinking; Brief intervention; Latent class analysis; Treatment effect modification; Trauma; TanzaniaParent-trial registration: ClinicalTrials.gov NCT04535011
Authors
- Paige O’Leary (ORCID: https://orcid.org/0000-0003-3697-5953)
- João Ricardo Nickenig Vissoci (ORCID: https://orcid.org/0000-0001-7276-0402)
- Ashley J. Phillips (ORCID: https://orcid.org/0000-0002-8547-0719)
- Edwin Joseph Shewiyo (ORCID: https://orcid.org/0000-0002-2707-2720)
- Blandina Theophil Mmbaga (ORCID: https://orcid.org/0000-0002-5550-1916)
- Catherine A. Staton
- Victoria Masue
- Beatrice Tingo (ORCID: https://orcid.org/0009-0002-6423-6090)
- Francis Sakita
- Timothy Antipas
- Gen-theresia Masiku
- Kisse Kamwela
- Doreen O Msemakweli
- Linda Minja
- William Nkenguye
- Dativa Macha
- Joao Vitor Perez de Souza
- Alice Andongolile
- Anaelderrick Frederick Kimaro (ORCID: https://orcid.org/0000-0002-1284-3799)
- Erimina Marandu
- Octavia Lucian Materu
Institutions
- Duke University (US)
Publication Details
- Journal
- PsyArXiv (OSF Preprints)
- Published
- 2026-11-16
- DOI
- https://doi.org/10.31234/osf.io/8ve9d
- Primary Topic
- Substance Abuse Treatment and Outcomes
- Type
- preprint