A Non-Inferiority Trial of an Electronic vs Face-to-face Brief Intervention After Alcohol Screening in Two Countries

Background and Study Purpose

As is well-documented by the World Health Organization (WHO), alcohol abuse remains a major health hazard worldwide, responsible for 5.3 percent of all deaths annually. It also contributes to over 200 conditions related to disease and injury, which include an array of behavioral and mental health disorders. There is now a substantial body of literature suggesting the effectiveness of Screening and Brief Intervention (SBI) as a strategy to identify people at high risk for alcohol use disorder, and to invite them, within the context of a short motivational interview, to reflect on their drinking patterns and develop a realistic and achievable plan to reduce their consumption. The screening component of SBI typically relies on the Alcohol Use Disorders Identification Test (AUDIT), which comprises two parts. The first is a 3-item assessment (the AUDIT-C); individuals who score 8 or higher are then invited to: (1) respond to 7 additional questions about their alcohol consumption (the AUDIT-7) and (2) participate in a brief intervention that is typically conducted by their primary health care provider or a staff member.

There are a variety of reasons, however, why brief interventions may fail to fulfill expectations of their positive effects. Eligible participants may be reluctant to engage fully in brief interventions because of the stigma attached to doing so; and primary health care providers may lack the time, attention, and confidence in their ability to conduct the brief intervention, or delegate responsibility for its administration to staff members who are insufficiently trained to deliver the intervention as intended, which includes establishing a sufficient rapport with the participant.

Consideration has thus been given to strategies to automate the delivery of brief interventions so that they can be self-administered electronically. Prototypes of this delivery mechanism are under development, and are sufficiently far advanced that they warrant investigations of their effectiveness. The key research question to be addressed by this study is:

• Are brief interventions, as electronically self-administered via an app on handheld devices, at least as effective in reducing the 30-day quantity and frequency of alcohol consumption as brief interventions administered in person by health care providers?

Note that the study's purpose is not to determine whether the administration of brief interventions via a handheld device is superior to that of an in-person administration, but to investigate whether the two modes of delivery are at least of equal effectiveness.

Study Setting

This study will be conducted in communities in two low- and middle-income countries: Zacatecas, Mexico and Alexandra Township, South Africa. In each site approximately seven to nine hospitals and health care clinics will participate in the study.

Institutional Recruitment Procedures

A country-specific study contractor will recruit the hospitals and clinics where the SBI will be delivered. All participating institutions will express their willingness to permit the contractor's trained facilitators (health educators) to approach patients in waiting areas and administer the AUDIT screening tool to them via an electronic device. For those patients who score 8 or higher on the AUDIT-C, the institutions will agree to:

• support the administration of the AUDIT-7;
• agree to and support the assignment of all participating clinics on alternating weeks to either the traditional in-person brief intervention or the electronic brief intervention. Thus all patients qualifying for the study in any given week will be assigned to one group, and all those in the subsequent week will be assigned to the other group.

Study Participant Requirements and Procedures

All patients who are at least 18 years old and score 8 or above on the AUDIT will be eligible and recruited for the study as long as they have not been screened in the past six months. They will be informed of their eligibility via the consent language presented in the app after the screening and before the study survey (those who do not qualify for the study do not see the consent language or the survey but go directly from the screening to their brief intervention). Specifically, the consent form on the device will inform them that:

• they are eligible to participate in a study on alcohol use patterns over time;
• they will be invited to answer a few more questions about their alcohol use;
• they will be asked to complete similar surveys in three months and again at six months, with various options, i.e., at the time of a follow-up clinic visit (for those with chronic health conditions), at a meeting with a clinic staff person at a convenient time and location, or via a phone interview in South Africa and via a link to the survey sent to them so they can complete it on their phone or other device at home or a phone interview in Mexico;
• they will receive an incentive at the time of each survey with each successive survey incentive of increased value compared to the previous one (small cash payments in Mexico; a mobile airtime credit at baseline and grocery vouchers at the two follow-up surveys in South Africa); and
• they will be asked to provide personal contact information to be used to recontact them at each of the two follow-up intervals.

Study Sample

All screened patients who are are aged 18 or older, who score 8 or higher on the AUDIT, and who have not been screened in the past 6 months will be eligible to participate in the study. Initial targets were to enroll 460 participants (230 in each brief intervention condition or study arm) in each country. At this point, Mexico has enrolled 563 participants (289 traditional BI; 274 electronic BI). Baseline data collection is still ongoing in South Africa.

Analysis Plan

In this non-inferiority trial, the null hypothesis is that the intervention (i.e., a BI delivered electronically via an app) is inferior to the comparison (i.e., an in-person brief intervention, or business as usual). This would be reflected in a statistically significant greater reduction in drinking alcohol post-intervention among those receiving an in-person BI compared to those receiving an electronic BI.