The intervention was developed and piloted through a collaboration between the research team and an implementation team at an opioid treatment program (OTP) where patients were receiving DAA treatment. The APT Foundation in New Haven is among the largest OTPs in Connecticut, with over 5,000 patients. It provides primary medical care and, since 2015, has provided onsite HCV treatment using DAAs. Of the patients initiating treatment in this onsite model, 98% completed DAA treatment despite ongoing drug use reported by one quarter of patients in the treatment cohort (23). Retention in treatment for opioid use disorder (OUD) was greater for those whose HCV infection was treated in this model compared to those who chose not to pursue DAA treatment onsite (24). This high rate of treatment completion and retention in OUD treatment made the APT Foundation the ideal location to pilot test the proposed behavioral intervention.
Implementation team
In addition to the research team, a four-member implementation team from the APT Foundation was trained to conduct all aspects of the intervention. The implementation team member tasked with conducting the group session was a Licensed Drug and Alcohol Counselor.
Intervention development process
To guide the process of developing the pilot intervention, we used the Stage Model of Behavioral Therapies Research (25). According to this model, the process of developing and evaluating a behavioral intervention falls into two stages. Stage 1 focuses on the development process and is broken into two substages: the intervention is developed and manualized and intervention staff are trained in stage 1a and the intervention is pilot tested in stage 1b. Stage 2 involves testing the efficacy of the intervention. In this article, we describe the Stage 1 process.
Intervention structure
The intervention itself was conceived as a two-session process to be delivered during the course of DAA treatment at the APT Foundation. The brief, initial session used treatment-related phlebotomy to provide instruction on the essentials of safe injection followed by a longer, second session informed by the principles of the IMB Model to reinforce the visual learning experience. The first session uses principles of vicarious learning to improve knowledge of safer injection practices (26) while the second session was designed according to elements of social cognitive theory more broadly (27).
First Session: The 10-Minute Intervention (10-MI)
Studies designed to provide hepatitis B vaccination to active drug users provided the impetus for this brief intervention (28, 29). Ethnographic observations conducted in a non-medical settings confirmed earlier quantitative studies that revealed that study participants were often unaware of the basics of injection hygiene (19); however, they were amenable to learning more when safer injection education was offered in the context of blood draws for hepatitis testing. Treating the first session as a “teachable moment,” content for the first session was limited to what could be delivered during a 10-minute or shorter blood draw, with the intervention content to be delivered by a trained, non-clinical staff person narrating the protective actions that the phlebotomist took during a routine blood draw. The session covered five key steps: (1) preparing a clean space for injection, (2) finding a vein, (3) getting the skin ready for injection, (4) registering the needle in the vein, and (5) cleaning up after injection.
Second Session: The IMB Intervention
In order to build on the existing evidence base surrounding behavioral interventions targeting components of the IMB Model (i.e., knowledge, motivation, behavioral skills, self-efficacy), the research team conducted a review of the literature since 2000 that specifically focused on peer-reviewed articles about interventions to reduce HCV infection and reinfection risk behaviors. We extracted information about intervention components, relevant measures, and observed outcomes. Consistent with the IMB Model, the research team agreed that the second session would be a group session and would focus on reinforcing information presented in the first session, correcting misinformation, and building motivation and self-efficacy around safer injection practices.
The second session was designed to last an hour, based on the structure of other group sessions offered by the APT Foundation. The IMB Intervention had four sequential components. The first provided an interactive review of the information presented during the 10-MI. The second brainstormed reasons and social supports that motivate safer injection moving forward. The third built self-efficacy around safer injection by discussing and demonstrating key elements of syringe cleaning, drug splitting, and reducing exposure to blood on surfaces and injection equipment. The fourth provided harm reduction resources and referrals. The IMB Intervention incorporated interactive elements to increase engagement and ensure more effective transfer of information and skills.
Development of intervention materials
Based on this intervention structure and content, a plain-language script was developed for the first intervention session, highlighting key points to be covered and including supplemental information for facilitators to use when answering any patient questions. Some of the intervention handouts were adapted from existing harm reduction resources developed by the Chicago Recovery Alliance (30), CATIE (Canadian AIDS Treatment Information Exchange) (31), and the authors of the SHIELD (Self-Help In Eliminating Life-threatening Diseases) intervention, a peer-based intervention focused on adoption of safer sex and drug use behaviors among persons who use drugs (32). Materials were reviewed by the research team and then presented to the implementation team for their review and commentary. After all feedback was incorporated, the materials were finalized and approved by the Yale University Institutional Review Board prior to use in the pilot project.
Study protocol to test intervention
In keeping with the Stage Model of Behavioral Therapies Research, we conducted a Stage 1b pilot in order to test the feasibility of delivering the intervention and assess its effect on immediate and intermediate outcomes related to safer injection.
The research team decided on a pre-/post-test design, where measures of outcome variables were collected after each intervention session and again approximately 12 weeks and 24 weeks after HCV treatment completion, as depicted in Fig. 1. No control group was selected for this pilot study. Our initial goal was to recruit 75 participants.
Participant recruitment and data collection schedule
All participants were patients at the OTP who were planning to start treatment for HCV infection. Receipt of medications for OUD was not a criterion for study inclusion. In addition to the above, participants were eligible if they: (1) were over 18 years of age, (2) spoke English, and (3) provided informed consent. Participants were initially offered no monetary reimbursement for participation in the study, though participation in the second session satisfied group counseling requirements for those participants receiving OUD treatment. Later in the study, participants were all offered a monetary reimbursement totaling $60 provided on a stepped scale: $10 for completion of the baseline survey, $10 for completion of the survey after the 10-MI, $10 for completion of the first follow-up survey, and $30 for completion of the second and final follow-up survey. All participants who had consented to participate prior to the additional monetary reimbursement were contacted and reconsented to provide them with the same reimbursement retroactively. Rationale for the change in reimbursement is provided in the Results section.
Patients starting treatment for their HCV infection were approached by members of the implementation team, who assessed the patient’s interest in the study and conducted the informed consent process. Those who consented then completed the baseline assessment, which collected information on participant demographic characteristics, history of injection practices, and knowledge of syringe access and HCV infection.
After enrollment into the study, the 10-MI was delivered at approximately the fourth week of HCV treatment during a blood draw that is part of regular care to assess the impact of DAA treatment on HCV viral load. Afterwards, participants completed a brief assessment of safer injection knowledge, motivation, intention, self-efficacy, and behaviors (hereafter referred to as the safer injection assessment). Approximately two weeks later, participants were invited to attend the IMB Intervention, after which they repeated the safer injection assessment and completed a satisfaction survey. Upon completion of HCV treatment, participants again completed the safer injection assessment two more times. The timing of these follow-up assessments coincided with post-treatment viral load testing and therefore varied based on treatment regimen. Participants completed the first assessment around 12 weeks after treatment completion and the second assessment around 24 weeks after treatment completion. This schematic is depicted in Fig. 1.
Process Measures
We sought to identify the elements of the intervention that could be feasibly, acceptably, and reliably implemented. Feasibility data were collected pertaining to recruiting, engaging, retaining, and collecting longitudinal data from participants at each study visit. We assessed recruitment feasibility by comparing the number of individuals starting HCV treatment to the number who agreed to participate in the study. We also took note of common reasons given for not agreeing to participate. We assessed the feasibility of engagement in the intervention by charting the extent to which participants took part in intervention activities within the allotted time frame. We assessed retention feasibility by comparing the total number of participants who enrolled in the study to the number who completed all relevant benchmarks. Finally, we assessed the feasibility of data collection, as it was conducted, by examining the percentage of complete responses for all intervention data collection instruments.
Acceptability of the intervention was assessed by providing participants with a survey to rate their satisfaction with the intervention components and to comment on the elements of the intervention that they appreciated and those they thought should be changed. Participants completed the satisfaction survey after completing the IMB session.
For this study, we did not formally assess fidelity of implementation staff to deliver the intervention according to the written protocol. We did convene regular meetings with the implementation team to assess their experiences and challenges they felt prevented delivering the intervention according to the protocol.
Outcome measures
In keeping with the pilot setting and the IMB Model framework, a safer injection assessment was developed to collect data on a number of immediate, intermediate, and long-term outcomes and sought to capture changes over time in safer injection knowledge, motivation, self-efficacy, intention, and behavior.
In order to assess safer injection knowledge, we adapted questions from the results of previous studies with “True”, “False”, and “Don’t Know” as response options (20, 33). Participants were directed to select the “Don’t Know” option only in those cases where they could not hazard a guess as to the correct response. This battery of 10 questions covered the content of both intervention sessions.
To assess motivation to engage in safer injection in the future, we used an adapted version of the Treatment Self-Regulation Questionnaire from the larger Health-care, Self-Determination Theory Questionnaire (34, 35). We used items developed by Khatmi and colleagues (36) to inform our modification of the questionnaire items. The finalized scale included eight items, with a five-point Likert scale of response options: “Strongly Disagree”, “Disagree”, “Neither Agree Nor Disagree”, “Agree”, and “Strongly Agree”. In developing the scale, we were careful to ensure a balance of items related to intrinsic and extrinsic motivations to engage in safer injection.
The assessment of self-efficacy to practice safer injection steps adapted a version of the Health-care, Self-Determination Theory Questionnaire, modifying items from the Perceived Competence Scale (35). As before, items developed by Khatmi et al. (36) inspired our development of items for this scale. Items assessing self-efficacy included performing steps of safer injection (e.g., setting up injection equipment, finding a vein) and refusing to share syringes and other injection equipment. Based on work by Marotta and colleagues (37) and Latka and colleagues (38), the final instrument had 10 items and the same five-point Likert scale as the safer injection motivation questions.
We assessed engagement in injection drug use in the past three months. If participants indicated that they injected in the past three months, they were asked about the frequency of engaging in specific injection behaviors such as drug splitting, syringe cleaning, and syringe sharing. We asked a total of six questions about safer injection behaviors, with a five-point Likert scale of response options: “Never”, “Almost Never”, Half the Time”, “Almost Always”, and “Always”. We also aimed to assess the participants’ intentions to practice safer injection if they were to inject in the future. We developed a behavioral intention scale that asked participants to rate their likelihood of future engagement in the same six injection behaviors assessed in the behavioral scale, using the same Likert scale and response options.