Community Implementation of SBIRT Using Technology for Alcohol Use Reduction in Mozambique: Scale-up Study Protocol

Abstract

Introduction:

Hazardous drinking imposes a major public health burden worldwide, especially in low-income countries like Mozambique. Implementing Screening, Brief Intervention, Referral to Treatment (SBIRT) to address problem drinking is recommended; evidence regarding the best strategies to implement SBIRT at scale are needed.

Methods:

Guided by the Reach Effectiveness Adoption Implementation Maintenance model, we will conduct a two-year, cluster-randomized, hybrid type 2 implementation-effectiveness trial in 12 districts in Mozambique evaluating implementation (reach, adoption, fidelity, and maintenance), clinical effectiveness (symptoms, functioining, co-morbidities) outcomes, and cost (ClinicalTrials.Gov: NCT03610815). Eight districts will be randomized to an application-based, mobile SBIRT condition and four to SBIRT Conventional Training and Supervision. Study arms will be delivered by clinic-based community health workers. The Consolidated Framework for Implementation Research will guide our mixed-methods assessments throughout the study.

Results:

The arm showing the better cost-effectiveness, will be scaled up in the other arms’ districts. During this 12-month scale-up phase, Ministry of Health personnel will be charged with trainings, clinical activities, and supervision in all 12 districts without research team support. The scale-up phase is critical to identify real-world facilitators and barriers tracking internal and external factors in clinics that continue using the superior arm and those that switch to that condition.

Next Steps:

Working in a multi-step process with stakeholders from multiple sectors, outcomes and lessons learned from our study will inform development of an implementation tool-kit to guide scaling-up SBIRT for community hazardous drinking services to other low-and middle-income countries and low resource settings in high income countries.

Introduction

Hazardous drinking, alcohol consumption that places individuals at risk for adverse health events, is a major public health challenge, particularly in low- and middle-income countries (LMICs)^1–3. Alcohol misuse is avoidable, yet results in high morbidity and mortality as the third leading risk factor for poor health globally^1–4. Harmful alcohol use contributes to more than 200 types of diseases and injuries. The impact is worst in poor populations and in LMICs where disease burden per liter of alcohol consumed is greater than in wealthy populations^2,5,6.

Mozambique is a low-income country where half of the population lives below the poverty line and disease burden is high. While research on alcohol use in Mozambique is scarce, recent studies reveal that binge drinking (episodes of excessive drinking; four or more drinks per episode for women and five or more drinks per episode for men) is frequent among drinkers. Two recent studies with Mozambicans ages 25–64,^7,8 found that 24.2–28.9% of women and 48.7–57.7% of men were current drinkers (having consumed at least one drink in the past year).^7,8 Approximately 60% of current drinkers consumed alcohol at least once or twice a week and more than 40% reported binge-drinking in the previous week. Three-quarters of drinkers reported consuming home-distilled traditional alcohol beverages with high alcohol content, which was strongly associated with binge drinking. Depression and death of a child were associated with hazardous drinking (over four on Alcohol Use Disorders Identification Test [AUDIT])⁹.

Mozambique’s HIV prevalence has increased by 14.8% since 2009¹⁰. Alcohol is implicated in HIV risk behaviors and poor HIV care adherence¹¹. HIV is the second largest category of alcohol-attributable Disability Life Years (DALYs); alcohol-attributable disease burden worsens if the impact of alcohol consumption on HIV incidence and course is considered, with alcohol being responsible for 6.4% of all deaths and 4.7% of all DALYs lost in the African region¹². Thus, decreasing hazardous drinking could improve HIV care outcomes.

The World Health Organization’s Mental Health Gap Action Programme Guidelines (mhGAP)^13,14 recommend using Screening, Brief Intervention, and Referral to Treatment (SBIRT)¹⁵ to reduce hazardous drinking. SBIRT is typically integrated into medical care to address substance use disorders¹⁶. However, there are challenges to implementing SBIRT with fidelity, obtaining the effectiveness observed in trials, and sustaining SBIRT¹⁷, particularly in LMICs¹⁸. Hazardous drinking services (HDS) in Mozambique are delivered in specialty clinics by psychiatric technicians trained in SBIRT and other mhGAP evidence-based HDS. With one specialty clinic per district, each with 50,000–150,000 inhabitants, most individuals with hazardous drinking are not served. Our study will address this challenge, working with the Mozambican Ministry of Health to task-shift SBIRT to community health workers (CHWs). We will build on the Ministry of Health’s mhGAP-Epilepsy¹⁹ Program and our recently funded community mental health trial (U19MH113203) linking study data to Mozambique’s electronic medical record (EMR) to track impact on co-morbidities, including HIV and Tuberculosis (TB).

Mobile health technology (mHealth)²⁰, such as the mSBIRT application²¹, is a promising tool for widespread, cost-effective, and sustainable delivery of services in LMICs^20,22–24. The mSBIRT application was designed for use by healthcare providers with patients. It assists providers in quickly assessing the patient’s risk due to substance use and guides providers through a brief intervention that is tailored to the patient’s responses. For LMICs like Mozambique, it is crucial to identify the most appropriate strategy to implement SBIRT to address community HDS. Capitalizing on Mozambique’s task-shifting strategies and commitment to inform scale-up of HDS with local treatment guidelines (SBIRT, mhGAP), this study will scale-up SBIRT in the community. We will compare mSBIRT with SBIRT Conventional Training and Supervision strategy (SBIRT-CTS) to determine the most cost-effective strategy for scale-up and problem-solve implementation barriers.

Community Implementation of SBIRT using Technology for Alcohol use Reduction in Mozambique (Community I-STAR Mozambique) comprises a two-year, cluster-randomized, hybrid implementation-effectiveness type-2 trial in 12 districts evaluated by mixed-methods (See Table 1 for study timeline). Our first two aims will be accomplished during the trial. Aim 1 is to conduct an implementation impact evaluation using the Reach Effectiveness Adoption Implementation Maintenance (RE-AIM) model and compare the adapted mSBIRT to SBIRT-CTS in terms of Reach (primary outcome), Adoption, Implementation Fidelity, and Maintenance over time. Aim 2 is to compare clinical effectiveness and cost-effectiveness of mSBIRT and SBIRT-CTS, overall, by gender, age and urbanicity²⁵. Aim 3 will be accomplished throughout the trial and the subsequent scale-up phase (scaling-up the most effective strategy to other districts) and will identify organizational and clinician factors that affect SBIRT implementation and effectiveness. We have described design solutions that are responsive to key challenges and advantages of the study in Box 1.

Table 1.

I-STAR Study Timeline and Milestones

MILESTONES	Year 1				Year 2				Year 3				Year 4				Year 5
	1	2	3	4	1	2	3	4	1	2	3	4	1	2	3	4	1	2	3	4
Start-up	x
Training personnel/IRB approvals/finalize measures	x
Formative Phase		x	x	x	x
Finalize SBIRT-CTS and mSBIRT training and supervision manuals and piloting		x	x	x
Finalize data collection/management and Manual of Procedures				x
Trainer of trainers/harmonize trainings					x
Effectiveness-lmplementation type 2 RCT (Aim 1 & 2)						x	x	x	x	x	x	x	x
RCT Implementation Phase						x	x	x	x
Training of CHWs and PCPs						x	x
Service, consultation, monitoring and feedback								x	x
RCT Sustainability Phase										x	x	x	x
Competency										x	x	x	x
Cost-Effectiveness Data Analysis										x	x	x	x	x	x
Scale-up of Most Effective Arm - cross over																x	x	x	x
Mixed-Methods Process Evaluation (Aim 3)							x	x	x	x	x	x	x	x	x	x	x	x	x
Conduct Focus Groups and Interviews							x	x	x	x	x	x	x	x	x	x	x	x	x
Ongoing Quan/Qual Data							x	x	x	x	x	x	x	x	x	x	x	x	x
Analysis
Finalize Tool-kit Development																		x	x	x
Policy Workshop																				x

Note: IRB – Institutional Review Board; SBIRT-CTS – Screening, Brief Intervention and Referral to Treatment Conventional Training and Supervision; mSBIRT – mobile health Screening, Brief Intervention and Referral to Treatment; CHW – Community Health Worker; PCP – Primary Care Provider

Box 1. Key Challenges, Advantages and Design Solutions of Community I-STAR Mozambique.

Key Challenges

• Limited access to hazardous drinking services – currently provided by psychiatric technicians at specialty clinics at the district level with only one specialty clinic per district

• Effectiveness, fidelity, and sustainability challenges in using SBIRT for hazardous drinking services in LMICs

• Mobile health technology versus conventional training and supervision untested

Key Advantages

• A committed Ministry of Health involved in implementation science research capacity building

• Task-shifting strategic policy already in place – Psychiatric Technicians, supervised by mental health specialists, provide mental health services in urban areas

• mSBIRT previously developed by our partners in the US

Design Solutions

• Implement SBIRT for prevention and treatment of hazardous drinking at the community level delivered by non-specialized providers in severely resource-limited settings

• Use mSBIRT as a cost-effective, sustainable delivery mechanism for evidence-based services in LMICs

• Use surrogate respondents to identify household members who may be at risk for hazardous drinking as part of a public health intervention

• Explore causal model/predictors of implementation success combining psychological theories of behavior change with organizational theories

• Use state-of-the-art implementation measures to examine implementation, effectiveness, and cost-effectiveness of two SBIRT delivery strategies during the phases of implementation and sustainability, followed by evaluation of scale-up of the more cost-effective strategy

We will determine which study arms is superior according to implementation impact, effectiveness (clinical and cost), and implementation process. We chose the RE-AIM framework for Aims 1 and 2 because it is widely-studied and identifies five constructs that affect the quality, speed, and public health impact of efforts to translate research into practice: Reach the intended target population; Efficacy/Effectiveness; Adoption by target staff, settings, institutions; Implementation Fidelity; and Maintenance of intervention effects²⁶. For a broader understanding of the implementation process (Aim 3), we chose the Consolidated Framework for Implementation Research (CFIR)²⁷ to guide qualitative exploration of factors affecting implementation. CFIR synthesizes implementation science frameworks into five domains: intervention characteristics; outer setting (organization’s economic, political, social context); inner setting; characteristics of individuals involved in implementation; and implementation process.

However, neither RE-AIM nor CFIR posit causal links among variables. Specifically, they do not describe how organization- and provider-level factors interact to affect implementation. To address this limitation, we borrow from and expand upon Williams and Glisson (Figure 1)²⁸, who combine organizational factors with psychological variables from the Theory of Planned Behavior²⁵. We propose that the most proximal variable to CHWs’ and supervisors’ use of SBIRT is their intentions to use it. Intentions are influenced by three determinants: attitudes (perceptions of SBIRT); descriptive norms (belief that people like me use SBIRT); injunctive norms (belief that important others expect me to use SBIRT); and self-efficacy (perception that providers can effectively perform SBIRT). We hypothesize that organizational factors influence the determinants of intentions and moderate the association between intentions and use of SBIRT. That is, even the strongest intentions may be thwarted if organizational resources do not support acting on those intentions.

Figure 1.

Applying Organizational and Psychological Theories to Implementation

Note. Adapted from Glasgow et al, 1999 & Williams and Glisson, 2014

Our hypotheses are:

Hypothesis 1: RE-AIM Implementation Impact Evaluation

• mSBIRT will yield significantly better RE-AIM implementation outcomes than SBIRT-CTS. The primary outcome is Reach (% access care/those in need). Secondary outcomes are Adoption (% clinic/staff using SBIRT), Implementation Fidelity (adherence to model/strategy), and Maintenance (use over two years) at months 12,18 and 24.

Hypothesis 2: RE-AIM Clinical Effectiveness and Cost-Effectiveness Evaluation

• 2a: Clients exposed to mSBIRT will have greater reduction in hazardous drinking and improved health-related quality of life than those exposed to SBIRT-CTS at months 12, 18 and 24.

• 2b: Hazardous drinking improvement due to SBIRT will lead to improvement in health-related outcomes, including HIV-related outcomes (e.g., viral load) for people living with HIV at months 12, 18 and 24.

• 2c: mSBIRT will be more cost-effective than SBIRT-CTS at months 12, 18 and 24.

Hypothesis 3: Implementation Process Evaluation

• 3a: Organizational urbanicity (urban, peri-urban, rural), culture and climate will affect clinicians’ attitudes, perceived norms, and self-efficacy regarding the use of SBIRT, which will in turn affect intentions to use SBIRT.

• 3b: Clinicians’ demographics and intentions to use SBIRT will affect adoption and use of SBIRT.

• 3c: Resources will moderate the association between clinicians’ intentions and implementation.

Methods

Overview

We obtained ethical approval from institutional review boards at a US academic institution and the Mozambican Institute for Health Education and Research. We will build on the Ministry of Health’s mhGAP-Epilepsy¹⁹ Program and our community mental health trial (U19MH113203) linking study data to the Mozambique’s EMR (e-saude), to track impact on HIV and TB outcomes. We will conduct a 2-year, cluster-randomized, hybrid implementation-effectiveness type-2 trial²⁹ in 12 districts. Eight districts will be randomized to mSBIRT and four to SBIRT-CTS. CHWs will deliver the intervention in both arms. The more effective strategy will be scaled up to the other arm in year three following the trial. Guided by the CFIR²⁷, mixed-methods assessments throughout the study will be used to understand factors affecting implementation. This study is embedded in a larger study scaling-up comprehensive mental health care ³⁰.

Study Sites

The study will be conducted in two provinces–Nampula and Sofala. Nampula province in the north is the most populous province in Mozambique (5,758,920 inhabitants) and contains both vast rural areas and the country’s third largest city. Sofala province (2,221,803 inhabitants) is in the central region. Research will be conducted in 12 districts (4 SBIRT-CTS; 8 mSBIRT) which constitute approximately 60% of the provinces’ populations (i.e. 4,788,430). Every district comprises clinics with diverse degrees of urbanicity, including clinics that are rural, peri-urban and urban. In both provinces, mental health services are presently only provided at district psychiatric clinics by 1–2 psychiatric technicians per district and 1–5 psychologists per province.

Each district has approximately seven primary care clinics staffed by 2–5 primary care providers. Approximately five CHWs per clinic attend to 250–400 families in their community annually, depending on each family’s health needs. The CHWs are the first level of contact in the community and they are part of the National Mozambique Health System. They are involved in infectious disease (Malaria, HIV, and TB) testing and treatment and maternal-child health services.

Study Procedures

Overview

Year one will be devoted to implementation³⁰, which will involve training and supervising staff, monitoring services and fidelity, providing implementation feedback to providers, and interviewing providers about implementation. In year two, we will assess sustainability of mSBIRT and SBIRT-CTS delivery³⁰. These implementation and sustainability phases occur within the context of the clustered randomized trial. In year three – the scale-up phase, – the more cost-effective strategy will be implemented in districts that received the less effective strategy.

Randomization of districts will be done 2:1 with eight districts in the mSBIRT and four in the SBIRT-CTS. Randomization will be stratified based on size (i.e., number of CHWs) of the district to ensure balance.

Regardless of trial arm, Ministry of Health procedures require that individuals who are reached, in need and provided with care, be seen weekly using SBIRT. Those reached and engaged in follow-up constitute retention. Each weekly contact with a provider is documented by the first two items from the AUDIT, the Short Form Health Survey 12 (SF-12),³¹ and the Mental Wellness Tool (mwTool; described below).

All patients and providers will provide written informed consent. While risks to provider participants are minimal, it is possible that patient participants could experience increased distress and suicidal ideation or report abuse. In accordance with the Ministry of Health safety procedures, local co-investigators (all clinicians) will be notified of an acute crisis within 24 hours. Additionally, we will convene a five-member Data Safety and Monitoring Board (DSMB) that will include an implementation scientist, statistician, mental health researcher with expertise in low-resource settings, Mozambique context expert, and ethicist. They will assist in monitoring adverse events and meet annually.

Implementation Phase

In the initial six months of the implementation phase, SBIRT and Motivational Interviewing experts will provide a one-week training for Ministry Mental Health specialists. Trainees will present five cases in group supervision to become certified national trainers and supervisors. These newly-minted trainers will train and supervise the CHWs. We will work with the trainers to adapt and pilot mSBIRT. We will also conduct a mixed-methods evaluation of the acceptability and feasibility of tablet use, and capture CHWs’ intentions of using mSBIRT.

As part of the larger mental health study ³⁰, all CHWs will be trained to use a tablet-based, comprehensive mental disorders assessment tool – the mwTool – that identifies presence of substance use disorders, common mental disorders, severe mental disorders, and acute risk of suicide. We will measure trainees’ knowledge (Information), attitudes towards people with hazardous drinking (Motivation), and self-efficacy (Skills) in providing SBIRT before and after training.^25,28

During the following six months of the implementation phase, the research team will provide active supervision through observation of role play and case presentation sessions. CHWs will complete three cases and submit transcripts of the case sessions to their supervisor. Supervisors will evaluate transcripts using a treatment fidelity checklist comprised of core intervention components and provide feedback to CHWs. If supervisors determine insufficient fidelity and competency, continued supervision and clinical support will ensue until competency is achieved.

The mSBIRT application guides providers to follow standardized treatment steps using interactive methods, including knowledge tabs with tips and sample scripts, and automated processes to enable providers to recognize any clinical “red flags.” Meta-data from the application will track individual provider implementation outcomes (adoption) and fidelity. Provider-level user data will indicate patients seen, key activity completion, and sessions per patient. If the user-data of any provider appears to deviate from the standard sequence of activities, this will be flagged for the supervisor intervention and corrective feedback to maximize fidelity. We will capture similar data in the SBIRT-CTS via provider and supervision checklists.

Sustainability Phase

The 12-month sustainability phase will begin once all CHWs and supervisors have achieved competency in either mSBIRT or SBIRT-CTS. The research team will monitor the CHWs, supervisors, and involved Ministry of Health personnel, but will not provide supervision unless requested or determined necessary based on performance.

We will sample six clinics in each arm from rural (n=2), peri-urban (n=2), and urban (n=2) areas. We will conduct 12 audio-recorded focus groups (1–1.5 hours each) with CHWs (one group per catchment area) and 24 semi-structured household interviews with patients and/or their family members (two interviews per catchment area).

Scale-Up Phase

In the scale-up phase, interviews will continue only in the six clinics from the superior arm: one focus group with CHWs at each clinic (n=6), and two household interviews per catchment area (n=12).

Treatment

The goal is for CHWs to screen every household in their catchment area annually using the mwTool. For those who score positively, CHWs will administer the AUDIT to determine a score that will guide intervention. Patients with an abstinence or low risk score (<5) will receive feedback and follow-up in 4 weeks. Those with scores 5–15 will be considered at-risk or have hazardous use and will receive a brief intervention weekly for 4 weeks. Patients who score equal to or above 16 in the harmful use or dependent (≥20) range will also receive the four-session intervention plus medical monitoring by a primary care provider or psychiatric technician. Before the fourth session, CHWs will administer the AUDIT again. The fourth session reassessment will determine whether patients will be discharged (abstinence or low risk), referred to an ongoing group intervention at the community level (same or improved, but not low risk), or treated by primary care provider or psychiatric technician (worsening symptoms).

SBIRT sessions last 45–60 minutes and involve screening and a brief intervention consisting of providing feedback and health information, engaging patient in motivational conversations, and negotiating a plan for change³². SBIRT incorporates principles of motivational interviewing – using an empathic style, supporting motivation and self-efficacy for change³³, building discrepancy, and creating an action plan. We will establish a Mozambican SBIRT/mSBIRT workgroup with specialists who have expertise treating substance use disorders to adapt the treatment. Our adaptation workgroup will provide input into the mSBIRT application itself and suggestions on the general content and iconography.

Measures

The study measures are organized by RE-AIM outcomes (Table 2). To determine the effectiveness and implementation outcomes, we will use the AUDIT and the mwTool. In both arms, the CHWs will conduct yearly household mental health and substance abuse screening with the mwTool which incorporates some AUDIT-C items as well as screeners for other mental health problems. Each CHW has a catchment area of 250–400 households³⁴. To decrease CHW burden, the CHWs will administer the mwTool to only two household members per home (aged 16 and older), but ask about other household members. Local research has shown that surrogate respondents provide valid information on alcohol intake⁸. CHWs will individually admister the full AUDIT^35,36 and SF-12^31,37–40 to anyone aged 16 and older screening positive on the mwTool for substance use disorders. (Table 2) Based on AUDIT scores, CHWs will deliver either mSBIRT or SBIRT-CTS, according to trial arm. The full AUDIT will be administered at sessions one and four to measure effectiveness.

Table 2.

RE-AIM Measures for Mixed-Methods Data Collection

DOMAIN	DEFINITION	PROVIDER*
Implementation Impact Outcomes
Reach *	Those who access care, or initiate SBIRT (numerator) out of the number of people in need (denominator) who are all individuals that screen positive in the mwTool (men ≥4; women ≥335,54-56) Characteristics of those reached vs. not* (gender, age, health outcomes and urbanicity)	Community Health Worker (CHW)/Patient (Pt)
Adoption- Clinic and Staff Level	Number of clinics/CHW using SBIRT (numerator) over those expected (denominator, all in each district)**.	Research Assistant (RA)/ CHW
Clinics	Characteristics of clinics adopting vs. not in both conditions** (Urbanicity: rural/urban; number of CHW; patients’ characteristics (i.e., comorbidities; access barriers); Evidence-Based Practice Attitudes Scale - 15 items57; Ready Set Change (organizational readiness for change)51; The Reach, Effectiveness, Adoption, Implementation, Maintenance (RE-AIM) Planning Tool, 27- item scale26,58,59; The Program Sustainability Assessment Tool, 40-items60,61)
Staff (Causal Model33,62)	Characteristics of CHWs adopting vs. not in both conditions** (age; gender; Work; Self-efficacy, 10 items63– 65; Intentions (2-items; 7-point scale) asking how willing and how likely one is to use each SBIRT component66,67; Attitudes towards each component of SBIRT (6 standard semantic differential scales)68; Norms or perception that other people like them will use the SBIRT (6 standard item stems); and Self-Efficacy (two statements) measuring CHWs’/supervisors’ perceptions of their skills and abilities to perform SBIRT
Implementation Fidelity **	Consistency of implementation across time in both conditions. Characteristics of Clinics/CHW as described above in Adoption	RA/CHW
Maintenance	Use over two years. Number of clinics using SBIRT** Adaptations to protocols** Characteristics of Clinics/CHW as described above in Adoption**	RA/CHW
Clinical Effectiveness Outcomes
Symptoms *	AUDIT54,56,69 – 10-item instrument developed by the WHO	CHW/Pt
Functioning *	Short Form Health Survey 12 (SF-12)37–40	CHW/Pt
Other Health Related	Through electronic medical records	CHW/Pt
Outcomes *	(e.g., HIV viral load, adherence to care for HIV/TB)
Cost Effectiveness Outcomes
Cost (Effort and Time) **	Cost as milestones are accomplished (Stages of Implementation Completion Scale (SIC; 8-stages of implementation: Readiness or Pre-Implementation, Adoption	RA/CHW
	or Implementation, and Sustainability/Competency)30 & Cost of Implementing New Strategies (COINS; Mapping Implementation Resources Using the SIC)70

^*recorded by CHWs or Supervisors as clinical record; de-identified aggregate collected every 6 months – using mSBIRT or REDCap

^**administered by Research assistant - REDCap (secure data capture/management web tool) will be used for data management

Note: Qualitative methods will assess all domains in satisfaction, barriers/facilitators, community engagement/priority given to alcohol services

In Aim 3, we will use mixed-methods to identify organizational and clinician factors that affect SBIRT implementation. Interview guides and analyses will be guided by our conceptual model combining the Theory of Planned Behavior with organizational constructs from the CFIR²⁷. We will triangulate quantitative and qualitative data to understand the factors influencing adoption and sustainability and facilitate translation of research findings into effective policy. In the “cross-over” phase, adoption and causal model measures will be administered to all participants. In each focus group, we will present our quantitative results in easy-to-interpret graphics and descriptions. We will ask participants to consider the results, adoption rate and associations between clinician (intentions) and organizational (culture, climate) factors and adoption.

Analytic Strategy

The cluster-randomized trial will address Hypotheses 1–2 regarding the implementation and effectiveness of the interventions across the 8 mSBIRT and 4 SBIRT-CTS districts. The hypotheses compare the interventions’ effect on outcomes measured for individuals and community health worker aggregated at the clinic level. We expect to include 62 clinics across these 12 districts, employing 417 CHWs who serve between 104,250–166,800 households. Clinic-level outcomes will be measured at baseline and then at 12, 18, and 24 months. Because clinics are nested within districts, we will use Hierarchical Linear Models (HLM) which will include random effects on each level and fixed effects for the intervention indicator and for the time variable.

Hypothesis 1 will test if mSBIRT is superior to SBIRT-CTS on each of the RE-AIM implementation outcomes using HLM including an indicator for intervention, time (baseline, 12, 18, 24 months), and intervention by time. The magnitude and statistical significance of the interaction term coefficient will assess how change in implementation outcomes over time differs by intervention arm. If the arms differ on baseline measures, we will include those measures as control variables in the HLM.

Hypothesis 2 investigates the interventions’ effect on clinical outcomes (2a-b) and on costs (2c). For Hypothesis 2a, we will use similar HLM (controlling for intervention and time) to estimate changes in hazardous drinking and health-related quality of life at months 12, 18 and 24 by SBIRT exposure. Overall intervention effects will be tested over time, and interactions of intervention by gender, age, and urbanicity of the patient will test effect modification. For Hypothesis 2b, we will use HLM (controlling for intervention and time) to predict change scores from baseline in health-related outcomes (viral load, HIV/TB care adherence) at months 12, 18 and 24 using change in hazardous drinking from baseline. We will access effect modification by gender, age, and urbanicity by including interaction terms. Finally, Hypothesis 2c will test if mSBIRT is more cost-effective than SBIRT-CTS in improving both implementation (Hypothesis 1) and clinical outcomes (Hypotheses 2a-b). Comparison of costs will be aggregated across the sustainability phase in year two. We will assess health-related quality of life using the SF-12 Health Survey, which can be converted to construct quality-adjusted life year (QALYs)⁴¹. QALYs is a measure of health-relate quality of life that incorporates mortality and morbidity estimates in a single index is extremely useful for decision makers due to comparability across conditions and interventions. Incremental cost-effectiveness ratios will identify the superior arm for the scale-up phase (numerator = difference in mean costs between arms; denominator = difference in reach and reduction in hazardous drinking). Uncertainty in costs and outcomes will be assessed with cost-effectiveness acceptability curves^42–44.

Power Analysis for Hypotheses 1 and 2.

Power was calculated based on participation of approximately 480 CHWs located across clinics in 12 districts anticipating possible drop-out of ~20%. Each CHW visits between 250–400 families in their community at least once a year resulting in a coverage of approximately 150,000 families (~75,000 families within each randomized condition). For hypothesis 1, to test differences in implementation outcomes between randomized study condition at the family level (e.g. % households screened), we will have 80% power to detect small to moderate effect sizes between the two conditions with Cohen’s d = 0.23–0.36 following standard power calculations for cluster randomized trials.⁴⁵ This power calculation was derived assuming an intraclass correlation (ICC) of families within the 12 randomized districts in the typical range for healthcare outcomes across randomized regions, ICC =0.02–0.05.^46,47 For hypothesis 2, we conservatively assume 3–5% of the population will be identified with hazardous drinking and will enroll in either intervention. Thus we expect 4,500–7,500 clients will enroll across the two randomized study arms nested within 480 CHWs who will provide treatment. Given this sample of treated individuals, we will have 80% to detect intervention differences of Cohen’s d=0.16 or larger. This power is calculated assuming ICC=0.40 for client outcomes within CHWs which is conservatively at the high end of ICCs found within health service providers in low-income settings.⁴⁸ All power calculations assume an alpha of 0.05 and that all tests are 2-tailed.

Hypothesis 3 examines organizational and clinician factors affecting implementation using data from the implementation, sustainability, and scale-up phases. Hypothesis 3a is that organizational factors including urbanicity, culture and climate will affect clinicians’ attitudes, perceived norms, and self-efficacy regarding the use of SBIRT, which will in turn affect their intentions to use SBIRT. Hypothesis 3b is that clinicians’ demographics and intentions to use SBIRT will affect adoption and use of SBIRT. To test both Hypotheses 3a-b, we will use structural equation modeling^49,50 to estimate the effect of organizational factors and clinicians’ demographics on determinants of intentions (attitudes, norms, and self-efficacy) and then on intentions (Figure 1). Organizational factors will include urbanicity: rural/urban; number of staff; patients’ characteristics and the Ready Set Change measure.⁵¹ Consistent with best practice, we will aggregate these orginaztional measures at the clinic level. Intentions and determinants of intentions to adopt SBIRT will be measured using validated self-report measures. (Table 2) Using structural equation modeling allows direct incorporation of clinic-level random effects and provides straightforward estimation of indirect effects. We will also use structural equation modeling to estimate the effect of intentions on adoption while controlling for the clinic and clinican characeristics described above. We will further test effect modification of the organizational factors on the association between intentions and adoption using interaction terms.

For qualitative analysis, interviews and focus groups will be digitally recorded, transcribed, and loaded into Nvivo software for analysis. To avoid bias, analysts will be blinded to implementation outcomes. Initial coding will draw on the CFIR. An inductive process of iterative coding will identify themes and categories. Analysts will then be unblinded to implementation outcomes to determine distinguishing constructs.

Finally, for mixed-methods data analysis, quantitative data is gathered prior to qualitative data and weighted equally (sequential: QUAN->QUAL)^52,53; the function is of complementarity and the process is connecting (having the qualitative data set build upon the quantitative data set)⁵².

Results

The more cost-effective strategy will be scaled up in the other districts using procedures outlined in the implementation phase. However, during the scale-up phase, trained Ministry of Health personnel will be charged with delivery, training and supervision of the superior strategy in all clinics without research team support. The scale-up phase is critical to identify real-world facilitators and barriers through process evaluation and continued tracking of internal and external factors in clinics that started in the superior arm and those that switch strategies.

Next Steps

We will work with stakeholders from multiple sectors to use the lessons learned from our study to develop an implementation tool-kit that can guide SBIRT for community HDS scale-up and dissemination in LMIC countries (U19MH113203, T32 MH096724, and D43 TW009675 country partners), New York State and City Departments of Health and Mental Health. We will review the data, craft a draft tool-kit, and present it to LMIC policy makers and WHO representatives. Policy maker feedback will help finalize the tool-kit. The tool-kit will be freely available to all research collaborators.

Highlights.

• This study evaluates implementation, clinical outcomes, and cost-effectiveness of SBIRT versus mobile health SBIRT (mSBIRT) for hazardous drinking at the community level through 100% task-shifted care.

• The use of mSBIRT is a cost-effective, sustainable delivery mechanism for evidence-based services in LMICs.

• This study explores a causal model and predictors of implementation success by combining psychological theories of behavior change with organizational theories.