Emergency department brief intervention reduced binge drinking days after injury in Tanzania
Emergency care in Moshi became the setting for a pragmatic test of whether a short, culturally adapted alcohol intervention can change behavior after an acute injury. In a randomized adaptive clinical trial at Kilimanjaro Christian Medical Centre, adults who disclosed alcohol use before the injury, screened positive on the Alcohol Use Disorder Identification Test, or had a positive breathalyzer were enrolled into the PPKAY program, short for Punguza Pombe Kwa Afya Yako, or Reduce Alcohol for your Health. The comparison was usual care. At 3 months, the pooled intervention strategy produced a measurable decline in binge drinking days, a result that matters because injury visits often expose a high-risk moment when drinking patterns are still fluid and behavior change may be more achievable than in routine outpatient care.
The trial screened 1,484 patients over the enrollment period from October 12, 2020, through April 14, 2023. Of those, 448 met criteria and consented. Randomization assigned 148 to usual care and 300 to the pooled intervention arm. Most participants were men, 94%, with a mean age of 36.4 years, and 59% identified as Chagga. The population was therefore largely adult working-age men presenting after injury, a group in which alcohol use can amplify trauma risk, complicate recovery, and drive repeat emergency visits.
A brief bedside conversation paired with text reminders
The intervention itself was deliberately simple. Nurses delivered a 15-minute brief intervention using motivational interviewing techniques, then reinforced the message with weekly text-based reminders after discharge. Some participants received standardized reminders, while others received personalized reminders, and both were grouped together for the primary stage comparison against usual care. The design reflects a practical idea in addiction medicine: a single counseling session may open the door, but repeated low-intensity prompts can keep the conversation alive after a patient leaves the hospital and re-enters everyday routines shaped by peers, stress, and social drinking norms.
The logic is biologically and behaviorally plausible. Acute injury can create a narrow window of receptivity, especially when patients can link alcohol use to a tangible consequence such as trauma. Motivational interviewing works by reducing defensiveness and strengthening personal reasons to change, rather than confronting a patient with directive advice. Text messages extend that effect by providing repeated cues for self-monitoring and recall, which can matter when alcohol use is driven by habit, opportunity, or immediate social pressure rather than a single isolated decision.
Fewer binge drinking days at 3 months
The primary endpoint was the number of binge drinking days in the previous 4 weeks at 3 months after discharge. In difference-in-differences analysis, the intervention arm showed a reduction in mean predicted binge drinking days of 1.2 days versus usual care, with a 95% confidence interval of -2.3 to -0.3 and a p value of 0.002. The signal is not subtle. It points to a real between-group separation on a behavior that is often difficult to shift, especially in settings where alcohol treatment resources are limited and stigma may suppress routine disclosure.
The trial used intention-to-treat analysis and blinded outcome assessors for follow-up, which strengthens the internal credibility of the findings. Even so, the outcome remained self-reported, and that creates room for social desirability bias, particularly because participants were not blinded. In alcohol research, self-report can drift toward what patients believe clinicians want to hear, especially after counseling. That concern does not erase the finding, but it does shape how firmly the result should be interpreted and why future phases of the adaptive trial remain important.
Why the intervention may matter in trauma care
Emergency departments are often the only clinical contact point for people who drink harmfully but do not seek treatment for alcohol use itself. An injury visit can be the first time a patient is confronted with the consequences of drinking in a medical setting rather than at home or in the community. For a setting like Tanzania, where the investigators framed the work around the absence of evidence-based culturally adapted interventions, a 15-minute nurse-led approach has obvious appeal. It is brief enough to fit emergency workflow, inexpensive relative to long-term specialty addiction care, and adaptable to hospitals with limited behavioral health staffing.
The text-message component may be especially valuable in this context. Weekly reminders can function as a low-cost booster after discharge, when a patient is no longer surrounded by clinicians and is back in the environments that reinforce drinking. In practical terms, that design tries to convert a one-time counseling session into a continuing nudge. The trial also suggests that culturally adapted messaging may help close the gap between generic alcohol counseling and communication that feels locally relevant and linguistically accessible.
What the adaptive design leaves for the next phase
This study reports Stage 1 of a larger adaptive trial. The next phase will compare PPKAY alone with personalized and standard boosters, which should help isolate how much of the observed effect comes from the counseling itself and how much comes from the ongoing text messages. That question is central for implementation. If the intervention remains effective without the booster, programs could simplify delivery. If the reminders account for a meaningful share of the benefit, then the messaging system becomes the active ingredient worth preserving.
Attrition did not differ between groups, and reasons for missing follow-up included loss to follow-up, withdrawal, and deaths. Even with those losses, the study preserved enough follow-up to detect a difference at 3 months. The trial registration number is ClinicalTrials.gov NCT04535011.
The result is clinically useful because it comes from a real emergency department, in a real injury population, with a model that can be delivered by nurses and reinforced outside the hospital. The remaining challenge is more operational than conceptual: identifying whether the intervention can hold its effect over longer follow-up and whether the text-based component is essential enough to justify routine scale-up in resource-constrained trauma care.