On-site bundled rapid HIV/HCV testing in substance use disorder treatment programs: study protocol for a hybrid design randomized controlled trial

This efficacy-effectiveness-implementation hybrid study consists of three distinct phases. Hybrid study designs combine features of clinical trials, which examine the efficacy, effectiveness, and the implementation research, which examines the adoption and sustainability of service delivery processes [46, 47]. Hybrid designs are postulated to advance the adoption and use of evidence-based practices and enhance the potential benefits of effective interventions [48‐50]. In the “developmental phase” of this study, we will produce a protocol for the provision of on-site bundled rapid HIV/HCV testing in SUD treatment programs. This will be followed by a “trial” phase, during which we will test the efficacy of the on-site bundled rapid HIV/HCV testing approach in increasing receipt of HIV and HCV test results among SUD patients. Finally, during a “translation” phase, we will conduct formative research on the barriers and facilitators of the diffusion, adoption, implementation, and sustainability of on-site bundled rapid HIV/HCV testing in SUD treatment programs. Overall, we will develop and evaluate the efficacy of the bundled rapid testing intervention and explore the preliminary effectiveness and implementation considerations related to intervention scale up. This study protocol follows the SPIRIT statement (http://​www.​spirit-statement.​org/​spirit-statement/​).

The study will be conducted at Promesa Inc. of the Acacia Network, one of the largest community based SUD treatment programs in the Bronx in New York City [61]. It has the capacity to serve more than 1,700 patients at any point in time across all of its facilities. We will include (1) one facility, which provides methadone to abstinence treatment to patients with opioid dependence (methadone-to-abstinence ambulatory - MTAA). Methadone to abstinence is a medical treatment protocol where methadone is provided in gradually decreasing doses until the point of abstinence. We will also include (2) one facility, which provides behavioral treatments (for example, cognitive behavior approaches) to patients with addictions to opioids and other types of drugs (chemical dependence outpatient program – CDOP. Both MTAA and CDOP are co-located with a federally qualified health center that provides primary healthcare services. The Primary Care Health Center (PCHC) offers on-site laboratory-based enzyme immunoassay (EIA) HIV and HCV testing with delayed results. SUD treatment patients who test positive for either HIV or HCV are referred to relevant medical services. Typically, patients choose to stay in the FQHC co-located with MTAA and CDOP, but if they prefer, they can be referred to another FQHC or primary care health center of their choice.

Our starting point for the development of the on-site bundled rapid HIV/HCV testing intervention is grounded on existing HIV EBIs, which address testing procedures and linkages to care. The rapid HIV testing EBI we will adapt was shown to be highly effective in increasing receipt of HIV test results compared to referral for off-site HIV testing (80 % versus 18 %, respectively) [25]. Patients in this EBI first receive verbal information about the rapid HIV test, including a description of the rapid testing procedure, timing for and meaning of test results, and an explanation of the window period during which an antibody test might be negative despite the presence of HIV infection. Patients are then offered a rapid HIV test [62]. Our approach to linkage to care services will be based on a modified Anti-retroviral Treatment and Access to Services (ARTAS) and Hepatitis C Care Coordination Model (Hep-C CCM) [63‐66]. ARTAS is an individual-level, multisession, time-limited intervention with the goal of linking recently diagnosed persons with HIV to medical care soon after receiving a positive test result. ARTAS led to significantly higher linkage to HIV care compared to a “passive referral” standard of care [63, 64]. Hep-C CCM is a case manager-coordinated service that links patients to primary care and hepatology clinics. The intervention was efficacious in increasing both attendance at HCV clinical evaluations and knowledge of HCV [65, 66]. The protocol for on-site bundled rapid HIV/HCV testing will incorporate elements from these interventions. It will be developed through a combination of in-depth interviews, focus group discussions, interactive testing, and expert consultations (see Table 1).

Patients randomized to the intervention group will thus receive an offer to participate in on-site bundled rapid HIV/HCV testing. Those who accept this offer will receive pre-test information about testing procedures and education, be tested and receive results during the same visit as testing, and be provided post-test counseling and support services according to the protocol developed through our ADAPT-IT³ process. Patients with reactive HIV and/or HCV antibody test results will be actively linked immediately to a healthcare provider for further evaluation and confirmatory testing (see Fig. 1). Patients randomized to the control group will receive the standard of care (SOC). The SOC is venipuncture whole blood laboratory-based HIV and HCV testing with delayed results (enzyme immunoassay). Patients may not receive pre-test information, results, and post-test counseling, especially if non-reactive or negative test result(s). Linkage to care in the SOC is accomplished by referral to a healthcare provider within the participating SUD treatment organization or by passive referral to other health facilities. After the end of the follow-up period (see below), all patients in the control group will be offered access to the on-site bundled rapid HIV/HCV testing intervention. This will include all procedures (that is, rapid HIV/HCV testing, post-test counseling, and linkages to care), as appropriate. Trained study staff will obtain written informed consent from all patients prior to enrollment. Additionally, trained supervisors will monitor the safety and efficacy of the trial intervention on a regular basis. They will ensure that all study procedures are conducted and that data are generated, documented, and reported in compliance with the protocol and applicable regulations.

Participants will be assigned randomly to the intervention or to the control group at the time of the baseline visit. Randomization will be stratified by recruitment site (MTAA versus CDOP programs), location of primary care provider (study site versus other), and type of drug use (injection as a method of drug use versus non-injection drug use). The randomization will be done in blocks of randomized size (varying from four to 10) with computer-generated random numbers. This procedure ensures that the distribution of treatment site and type of drug use will be balanced across intervention and control groups. It also prevents study staff from learning the randomization pattern, which becomes obvious in a fixed block-size scheme. The study statistician will generate the randomization lists, which will be concealed in sequentially numbered, opaque, sealed envelopes with a signature across the sealing point. Trained research assistants, who are blind to the randomization schedule, will provide the random assignment to participants at the time of randomization.

Participants must be (1) able and willing to provide informed consent, (2) seeking or currently receiving SUD treatment services (excluding alcohol only) at the participating treatment programs, (3) at least 18 years old, (4) able to communicate in English, (5) willing to sign a release form that will allow medical record review (to corroborate self-reports of receipt of test results), (6) willing to provide locator information, (7) self report being HIV and HCV negative, or report not knowing their HIV and HCV status, and (8) not have received results of an HIV or HCV test initiated within the last 12 months. We will employ several recruitment strategies to facilitate enrollment to reach our target sample size, including trained research staff, posters, and collaborating with staff of the study site to identify interested patients. We will promote participant retention by collecting locator information, which will be used to contact participants to remind them of follow-up visits and to locate participants who cannot be reached. Participation in the study is voluntary, and participants may choose to end their involvement at any time.

There will be three assessments during this trial of on-site bundled rapid HIV/HCV testing. These will include a baseline assessment administered prior to randomization, a follow-up assessment conducted 1 month after randomization, and a final assessment conducted 3 months after randomization (see Fig. 1).

The primary outcome will be self-reported receipt of HIV and HCV test results. The primary outcome will be assessed one-month post-randomization. Our use of self-reported receipt of test results is consistent with the use of self-reports as the primary outcome for evaluating rapid HIV testing interventions [25].

We will also evaluate several secondary outcomes. These will include an additional measure of receipt of test results based on medical records. Study participants will complete release forms (as applicable) in order to grant permission to study staff to review their medical records, including HIV and HCV testing records. Other secondary outcomes will include self-reported sexual risk behaviors and drug use practices such as (i) counts of unprotected vaginal and anal sex acts in the past 3 months with any sex partner, (ii) use of any drugs in the past 3 months, and (iii) unsafe injection practices. Additional measures will evaluate linkage to care by measuring the proportion of patients who initiated HIV care or HCV care before the 3-month follow-up among those whose test results were positive for HIV and/or HCV. Linkage to care for HIV or HCV will be measured separately, to account for variations in barriers to initiating care, which may be greater for HCV than barriers to initiating HIV medical care [67].

In order to measure all outcomes, indicators, and covariates, each study assessment will collect detailed data from study participants. We will use audio computer-assisted self-interview (ACASI) to limit social desirability biases in reporting sensitive behaviors (for example, sexual behaviors) [68]. Interviewers will be trained in the use of the ACASI program and survey to assist respondents as needed. To protect against the risk of loss of confidentiality during and after the trial, all data will be maintained on a password-protected local server. Study datasets will be de-identified prior to analysis.

We will collect the following information:

Demographic characteristics of participants (at baseline), which will include age, sex, race and ethnicity, years of formal education, income, employment status, health insurance, living arrangement including homelessness, number of children (under 18), and history with the criminal justice system (that is, arrests and incarcerations). This information will be used to describe the study sample, to assess for any differences between intervention groups, and to test for selective attrition.

Sexual behaviors: We will collect data on the total number of sex partners in the past 3 months, total number of vaginal sex partners and anal sex partners, total number of unprotected vaginal and total number of anal sex partners, and total acts of unprotected vaginal/anal sex.

Drug use: We will ask about the frequency, duration, and amount of substance use, including alcohol, cannabis, methamphetamine, cocaine, heroin, club-drugs, and mis-used prescription drugs. We will assess injection drug use, including the types of drugs injected and the sharing of needles and other drug paraphernalia (for example, spoons and cotton balls). We will also ask about overdoses and treatment for overdose (that is, the use of naloxone). At baseline, we will measure these behaviors over the 3 months before the interview. These measures will be repeated at 1- and 3-month follow-ups, with adjustments to the recall period.

Utilization of drug treatment services: We will ask the following three questions at baseline and at both follow-up visits: (1) Are you in treatment right now? (2) What treatment are you undergoing? (for example, detox) and (3) How many days have you been in this treatment?

Linkage to medical care: We will also collect data on initiation of treatment among HIV and/or HCV positive patients, using the following questions: (1) Did you have an appointment with a medical provider for HIV services? (2) Did you have an appointment with a medical provider for HCV services? (3) Number of medical appointments made and appointments attended since the last interview, as well as initiation of ART and HCV treatment?

We designed the trial to estimate the effects of on-site bundled rapid HIV/HCV testing on receipt of test results separately in each of the two types of SUD treatment facilities (MTAA and CDOP). Based on preliminary data from the study sites for the standard of care, we assumed that 50 % of control group participants enrolled at the CDOP facility will receive their HIV and HCV results after 1 month. We assumed that 55 % of control group participants receive their HIV and HCV test results after 1 month in the MTAA program, which treats a higher proportion of PWIDs. Based on findings from the evidence-based rapid HIV intervention that informs this study [25], we assumed that 80 % of patients in the intervention group would receive their HIV and HCV test results within 1 month of randomization. To detect such intervention effects on the basis of a two-sample, chi-squared test for independent proportions with 80 % power and α = 0.05, we will require 78 patients in CDOP and 108 patients in MTAA facilities. We assumed an 80 % follow-up rate among participants enrolled at the MTAA program. We assumed a slightly lower follow-up rate among participants enrolled at the CDOP facility (75 %) because adherence to behavioral SUD treatment is lower among CDOP patients at Promesa Inc. These are conservative assumptions. For example, during the trial of the rapid HIV testing EBI, which informs this study, researchers achieved a 98 % 1-month follow-up rate [25]. In total, we obtained a total sample size of 239 participants (104 in CDOP and 135 in MTAA).

The primary outcome is a categorical variable denoting self-reported receipt of rapid HIV and rapid HCV test results at 1-month post-randomization. We will use a generalized estimating equation model, which includes covariates for type of drug use (injection/non-injection) and location of primary care provider (study site versus elsewhere). The model will also include indicator variables for HIV/HCV testing intervention status and the type of SUD treatment facility (MTAA versus CDOP), where the participants were recruited. The primary test of our study hypothesis will be a two-degree of freedom test of significance of the intervention across the two types of SUD treatment facilities. In post-hoc tests, the efficacy and effect size in each type of SUD treatment facility will be examined. To do so, we will include an interaction between variables identifying study groups (intervention versus control) and type of SUD treatment facility (MTAA versus CDOP). The primary analysis will be intent-to-treat, where everyone is included in the study group to which they were randomized, regardless of receipt of intervention. In sensitivity analyses, participants who have died or who have been lost to follow-up will be counted as not having tested. We will adjust for baseline covariates that may be associated with attrition from the study sample using inverse probability reweighting procedures as recommended for the GEE model [69].

The secondary outcomes related to linkage to HIV and HCV care, as well as to drug use, are categorical variables. They will be analyzed using techniques similar to those used for the primary outcome. This is also the case for indicators related to HIV and HCV prevalence among patients tested by bundled rapid testing versus standard of care. The number of unprotected sex acts over a period of time is a count variable and will be modeled using Poisson or negative binomial regression (if the variable is over-dispersed) or the zero-inflated version of these if necessary. Type of drug use (injection/non-injection) and the recruitment site (MTAA or CDOP) will be included as covariates. If other covariates predict missing data, they will also be included as control variables. All analyses of secondary outcomes will focus on the pooled data, that is, CDOP + MTAA programs. Because of the limited sample size, we will not conduct formal mediation analyses.

We will conduct formative research on the barriers and facilitators of the adoption, implementation, and sustainability of on-site bundled rapid HIV/HCV testing in SUDs treatment programs. To do so, we will collect qualitative data from three groups: patients, providers, and program managers who participated in the randomized trial. This will consist of two focus group discussions (FGDs) with each group, followed by a short questionnaire on individual perceptions. FGDs will be informed by an integrated conceptual framework, derived from the “Diffusion of Innovation” theory [56]. We will identify facilitators and barriers to scaling-up the bundled rapid HIV/HCV strategy, linking infected patients to medical care, and we will develop strategies to address these barriers.

The Columbia University Medical Center (CUMC) Institutional Review Board (IRB) will provide ethical oversight for the study. Ethical approval for this study has been granted from the CUMC IRB (Protocol Number: IRB-AAAN5869).

The final results from the study will be submitted for publication in a peer-reviewed journal. We will also disseminate our findings to the study site and study population.

The trial is at the planning stage, with the first six phases of the ADAPT-IT³ process already completed. Recruitment of study participants is tentatively scheduled to commence in March 2016.