Randomized Controlled Trial Protocol for Project BRIDGE: A Telephone-Administered Motivational Interviewing Intervention Targeting Risky Sexual Behavior in Older People Living with HIV

Abstract

Purpose:

By 2020, 70% of people living with HIV in the United States will be greater than 50 years of age. As many as 37% of sexually active older people living with HIV (OPLWH) engage in HIV transmission sexual behaviors. In spite of repeated calls for secondary prevention interventions to reduce condomless sex in OPLWH, no age-appropriate, evidence-based secondary prevention interventions exist for this group. Furthermore, many OPLWH face barriers to engaging in face-to-face secondary prevention services because of HIV- and age-related stigma, comorbid mental and physical health conditions that complicate travel, or geographic isolation. High rates of depression in OPLWH may further complicate engagement in interventions intended to reduce HIV transmissions. Telephone-administered motivational interviewing may be a feasible and efficacious intervention for this population.

Methods:

This randomized controlled trial will test the efficacy of a 5-session telephone-administered motivational interviewing plus behavioral skills training (teleMI+BST) intervention versus a 5-session telephone-administered coping effectiveness training (teleCET) control intervention to reduce condomless sex in OPLWH. A diverse sample of 336 OPLWH will be recruited across the U.S. The primary analysis will test the efficacy of teleMI+BST to reduce occasions of non-condom protected anal and vaginal intercourse with HIV serodiscordant sex partners. Secondary analyses will examine the efficacy of teleMI+BST to reduce depressive symptoms in mildly depressed OPLWH.

Conclusion:

This is the first large-scale RCT intended to reduce HIV sexual transmission risk behavior in OPLWH and will add to the literature on secondary prevention telehealth interventions for people living with HIV.

1. Introduction

By 2020, 70% of people living with HIV/AIDS in the United States will be greater than 50 years of age [1]. Medical morbidities are common in older people living with HIV (OPLWH) [2], and many OPLWH have detectable HIV viral loads due to poor adherence to HIV antiretroviral therapies, thus increasing their HIV infectivity [3]. In addition, rates of condomless sexual intercourse with HIV-negative sexual partners range from 18% to 37% [4,5], leading to increased risk of HIV transmission in non-virally suppressed persons, in the absence of sexual partners’ use of pre-exposure prophylaxis (PrEP). To date, behavioral interventions aimed at reducing HIV transmission have focused largely on younger HIV-positive persons and those at greatest risk of acquiring HIV (e.g., intravenous drug users, men who have sex with men, commercial sex workers, individuals in low resource settings). These interventions have not explicitly addressed the sexual health needs of OPLWH, such as age-related biological and libidinal changes, the use of erectile dysfunction medications or hormone replacement therapies, negotiating sexual safety in the context of long-term romantic partnerships, or perceptions of sexual unattractiveness in long-term survivors of HIV [6,7]. As such, existing HIV secondary prevention interventions may not be an optimal fit and, in turn, may be less efficacious for OPLWH.

Many OPLWH face barriers to engaging in face-to-face secondary prevention services because of HIV- and age-related stigma, comorbid health conditions that complicate travel to medical and social service appointments, or geographic isolation [8,9]. Secondary prevention interventions delivered over cellular and landline telephones are acceptable to PLWH [10–12], and have shown preliminary efficacy to reduce HIV sexual transmission behaviors [10], increase intervention engagement and evidence low attrition rates [12], and may be a viable and cost-effective way to circumvent barriers to care. However, to date, no large scale randomized controlled trials (RCTs) of telehealth interventions targeting HIV transmission behaviors in OPLWH have been conducted.

The primary purpose of this RCT is to compare the efficacy of a telephone-administered behavioral counseling intervention targeting condom use in OPLWH versus an attention equivalent telephone-administered coping intervention to reduce occasions of condomless sex with HIV serodiscordant partners.

2. Conceptual Model

HIV transmission is multiply determined. Figure 1 shows that interrelated intrapersonal [13–18], interpersonal [19–23], and socio- and geo-political factors [24–26] influence engagement in unprotected anal and vaginal intercourse (UAVI). Preventative measures for HIV transmission have included antiretroviral treatment as prevention [27], sexually transmitted infection symptom mitigation [28], harm reduction interventions that encourage alternative sexual acts with lower HIV transmission risk [29], and PrEP use [30], all of which help reduce the spread of HIV. However, behavioral interventions that increase condom use remain a key component of the HIV prevention armamentarium [31]. This RCT tests an intrapersonal pathway to UAVI engagement, as depicted in Figure 1 with a solid arrow.

Figure 1.

Conceptual model of HIV transmission. HIV = human immunodeficiency virus, PrEP = pre-exposure prophylaxis, STI = sexually transmitted infection, UAVI = unprotected anal or vaginal intercourse.

3. Methods

3.1. Study Design

This two-arm parallel-groups randomized controlled trial tests the efficacy of telephone-administered motivational interviewing plus behavioral skills training (teleMI+BST) to reduce UAVI with HIV serodiscordant partners. To determine whether teleMI+BST reduces UAVI to a greater extent than simple contact with a helping professional in the context of telephone counseling, the study employs an attention-equivalent comparator of telephone coping effectiveness training (teleCET), which targets stressors in participants’ lives instead of sexual behavior. TeleCET has been shown to establish a therapeutic working relationship between counselors and persons living with HIV (PLWH) [32] but does not reduce UAVI [33]. Figure 2 depicts this RCT’s study flow, including anticipated rates of enrollment and follow-up. This RCT has been registered at ClincalTrials.gov (Identifier: NCT03004170).

Figure 2.

Study flow chart. UVAI = unprotected anal and vaginal intercourse with serodiscordant partners, TeleMI+BST = telephone-administered motivational interviewing plus behavioral skills training, TeleCET = telephone-administered coping effectiveness training. Anticipated assessment follow-up of 80% retention.

3.2. Study Interventions

The active and control interventions both consist of 5 telephone counseling sessions (4 initial sessions and 1 booster session) and are administered by “real world” bachelors- and masters-trained social workers, nurses, and chemical dependency counselors who work in community agencies that provide HIV medical and social service case management to PLWH. All counseling sessions are delivered one-on-one over the telephone, last 40–50 minutes, and are digitally audio-recorded to facilitate monitoring of intervention fidelity and to prevent therapist drift. Consistent with Motivational Enhancement Therapy (MET) [34] and MET-adapted interventions that target UAVI in PLWH [35], the 4 teleMI+BST initial sessions take place in weeks 3, 4, 8, and 12 following participant enrollment. This allows sufficient intervention intensity at the outset to build rapport and identify participant goals, while providing sufficient time between sessions 2 through 4 for participants to enact behavior change plans and trouble-shoot barriers to goal attainment with therapists in the latter sessions. The teleMI+BST booster session will take place in Week 24, approximately 3 months following completion of Session 4. The addition of a booster session to MI+BST interventions that target UAVI in PLWH has been shown to sustain intervention effects over 12 months of follow-up [36]. Participants in both intervention conditions will receive the same number of sessions, including the booster session, over the same time interval.

The Information-Motivation-Behavioral Skills (IMB) Model [37] provides the theoretical framework for behavior change mechanisms of teleMI+BST. The IMB Model posits that knowledge about condom use practices, condom use motivation, and acquisition and application of requisite condom use skills lead to engagement in condom-protected sex. The teleCET intervention is based on the Lazarus and Folkman Transactional Model of Stress and Coping [38] and uses cognitive-behavioral principles to: (1) appraise stressor severity, (2) develop problem- and emotion-focused coping skills, (3) determine the match between coping strategies and stressor controllability, and (4) optimize coping through use of social support resources. Components of teleMI+BST and teleCET interventions are outlined in the tables below.

For the teleMI+BST intervention, emphasis is placed on building clients’ intrinsic motivation to change before facilitating the development of change plans for condom use practices. This approach is taken because requiring individuals to engage in change planning before they are ready to change a behavior sometimes leads to worse target behavior outcomes [39–40]. As such, discussion of change plans may occur later than Session 3, depending on participants’ readiness to reduce UAVI.

To clearly differentiate the teleCET and teleMI+BST interventions, teleCET therapists work in concert with clients to identify a therapeutic focus that is unrelated to UAVI. The coping skills taught in the teleCET intervention focus on the client-identified stressor.

3.3. Treatment fidelity

Study therapists administer both teleMI+BST and teleCET, thereby removing potential “Condition x Therapist” confounds [41]. Extensive training, ongoing supervision, and fidelity coding procedures are employed to maximize therapist adherence to treatment protocols and reduce therapist crossover effects between treatments. Training of study therapists includes a 5-day in-person workshop on MI+BST, CET, delivering treatments over the telephone, and study intervention manuals. To qualify as a study therapist, each trained individual completes 4 mock teleMI+BST sessions and 4 mock teleCET sessions. Mock sessions are digitally audio recorded and coded for motivational interviewing fidelity using the Motivational Interviewing Treatment Integrity code (MITI) [42]. The MITI is designed to assess specific motivational interviewing-consistent therapist speech as well as consistency with the overall motivational interviewing style, assessed via random 20-minute segments of audio-recorded sessions, evaluated by trained raters. For the teleCET condition, MITI coding is used to ensure distinctness from the motivational interviewing condition, and to ensure that non-motivational interviewing therapeutic approaches were used [43].

At the start of the trial, all therapy sessions are reviewed immediately following session completion, and feedback provided to therapists as needed, to ensure treatment fidelity. Once each therapist has completed 4 teleMI+BST sessions and 4 teleCET sessions, a random 10% of sessions are coded for treatment fidelity. MITI reports are provided to the therapists’ supervisor and used in monthly in-person group supervision meetings to ensure treatment fidelity is maintained throughout the trial.

3.4. Recruitment and eligibility assessment of participants

This RCT recruits a geographically-diverse sample of OPLWH, comprising men who have sex with men, women, and men who have sex with women from urban and rural areas across the U.S. Traditionally underserved persons in rural areas and those who self-identify as African-American or Latinx are oversampled because of disproportionate HIV burden in these groups [44]. Participants across the depression symptom severity spectrum (no depression, mild depression, moderate-to-severe depression) will be recruited to test secondary effects of the intervention to reduce depressive symptoms. We will regularly monitor enrollment across the depression symptom severity spectrum and will weight sampling as needed across the depression strata to enroll near equal numbers of OPLWH with no, mild, or moderate-to-severe depression. Participants are community dwelling OPLWH recruited from community AIDS service organizations, infectious disease clinics, social service agencies that serve people living with HIV across the U.S., and via online dating sites frequented by some OPLWH (e.g., Grindr, SCRUFF).

Recruitment materials direct interested persons to a weblink where they may complete the eligibility screening survey. An option to be screened via telephone interview is also provided. Study inclusion criteria include: (1) English speaking, (2) access to landline or cellular telephone, (3) at least 50 years of age, and (4) engagement in one or more occasions of UAVI in the past 3 months with an HIV-negative or unknown HIV serostatus partner. Participants who meet exclusion criteria are: (1) those with active suicidal ideation with intent or plan, or suicidal behavior in the past year, as determined by the Patient Health Questionnaire 9-item Depression measure (PHQ-9) and follow-up questions for persons endorsing suicidal ideation on the PHQ-9, or (2) those with undetectable HIV viral load who engage in UAVI within a monogamous relationship with an HIV-negative or unknown serostatus partner, as risk for HIV transmission in this circumstance is near zero [45].

3.5. Baseline and follow-up assessments

Eligible participants complete 1-hour surveys and brief telephone interviews at baseline, 3-, 6-, 9-, and 12-month follow-up. Study assessors are blind to treatment allocation. Participants are mailed $40 for completing each assessment but are not paid for participation in counseling sessions.

3.6. Randomization

During informed consent procedures, participants are told the study focuses on romantic relationships and stress coping for OPLWH and that they will be randomized to receive 5 telephone counseling sessions that focus on either relationships or stress coping. Participants are not blind to intervention condition but are blind to the study’s primary and secondary outcomes and overall study purpose.

Following completion of the baseline assessment, participants are randomized to one of the two study arms using permuted block randomization with varying length blocks generated via a randomization sequence. Opaque envelopes numbered 1 through 336 (the targeted sample size for this RCT) contain a slip of paper denoting either the teleMI+BST or teleCET condition. Following completion of the baseline assessment, the assessor opens the next envelope in the sequence, revealing the condition to which the participant has been randomly assigned.

3.7. Outcome measures

3.7.1. Primary outcome

This RCT’s primary outcome is the number of UAVI acts in the past 3 months, assessed via telephone interview using the validated Timeline Followback (TLFB) method [46]. TLFB utilizes a calendar prompt to aid in participants’ retrospective recall of sexual behavior, including type of sexual act (oral, vaginal, and anal), partner HIV serostatus, and condom use.

3.7.2. Secondary outcome

OPLWH are 5 times more likely to be diagnosed with a depressive disorder than HIV-negative older adults [47]. In addition, teleMI+BST interventions may have secondary effects of reducing depressive symptoms among mildly depressed OPLWH [48]. Given high rates of depression in OPLWH [47], and research linking mild depression with UAVI [13], secondary outcomes explore whether teleMI+BST will reduce depression severity, as assessed with the PHQ-9, among mildly depressed participants in this sample.

3.7.3. Covariates

Baseline and follow-up surveys assess mental health and sexual risk reduction service utilization using a modified version of the HIV Cost and Services Utilization Survey [49–50]. In addition, the TLFB measures alcohol and other drug use prior to or during each sex act, sexual partners’ use of PrEP, and whether sexual partners were main or casual. To evaluate if participants have a positive view of the counseling sessions and their therapist, the Working Alliance Inventory, Short-Revised form [51] is administered as part of the 3- and 6-month surveys.

3.8. Statistical analysis

3.8.1. Sample size

The sample size of 336 participants (168 per intervention arm) was calculated using α=0.05, repeated measures correlations of 0.30 for the primary outcome variable, and 20% study attrition (i.e., 66 participants lost to follow-up), for which a sample of 270 participants without attrition could detect a Time x Condition interaction effect of 0.10 (based on Cohen’s f statistic) with approximately 0.90 power.

3.8.2. Planned analyses

Preliminary data indicate baseline UAVI counts tend to follow a negative binomial distribution [10]. For this study, zero-inflated negative binomial (ZINB) mixed effects regression models will test intervention efficacy [52]. Models will specify random effects for participants to account for repeated measures observations modeling the Time (3-, 6-, 9-, and 12-month follow-up) by Treatment Condition (teleMI+BST vs teleCET) interaction. The primary outcome analysis will control for baseline values of the outcome variable, and time-varying covariates of alcohol/drug use during sex, sexual partner status (main versus casual), mental health/risk reduction service utilization, PrEP use by sexual partners, and therapeutic alliance as assessed by the Working Alliance Inventory.

Preliminary data indicate that teleMI+BST may be efficacious at reducing depressive symptomatology among mildly depressed persons [48]. To test this hypothesis, linear mixed effects regression models will compare the efficacy of teleMI+BST to teleCET to reduce depression, as measured by the PHQ-9, among the approximately 90 mildly depressed participants retained through 12-month follow-up in this study. Models will control for baseline PHQ-9 values, therapeutic alliance, and time-varying mental health/risk reduction service utilization and model the Time x Treatment Condition interaction.

3.8.3. Sensitivity analyses for missing data

We will compare baseline data between participants with and without missing follow-up data to be used in multiple imputation procedures [53]. Sensitivity analyses will compare parameter values from analyses with completers only, multiple imputation procedures described above, and pattern-mixture approaches [54–55], consistent with recommendations by the National Research Council Panel on Handling Missing Data in Clinical Trials [56].

4. Discussion

This RCT aims to reduce UAVI among OPLWH with HIV serodiscordant sexual partners through a brief, 5-session telephone-administered behavioral treatment. A review of 6 RCTs of motivational interviewing to reduce UAVI in younger and middle-age PLWH found that, across all RCTs, motivational interviewing reduced UAVI [36]. The review also reported preliminary efficacy from a pilot RCT of telephone-administered motivational interviewing for OPLWH [10]. Most interventions in the review included a cognitive and/or behavioral skills training component, and interventions that offered one or more “booster” sessions maintained effects through 12-month follow-up. Notable limitations of the reviewed trials include absence of clinically-relevant attention-equivalent control groups (i.e., most used “usual care” or general health education controls), and exclusive use of advanced-degree therapists who are less commonly employed in HIV service agencies in which many interventions would be implemented. An additional limitation of the reviewed research is that, with the exception of a single pilot RCT, no studies addressed the sexual health needs of OPLWH. This study thus addresses gaps in the field by conducting the first large-scale RCT to specifically address the sexual health needs of OPLWH, employing an attention-equivalent control condition rather than usual care, and using “real world” community HIV social service providers to deliver the interventions.

This RCT also has a unique methodology seldom used in behavioral clinical trials. One threat to validity in behavioral clinical trials is termed “therapist effects” [57–58]—i.e., a good therapist may have greater influence on outcomes than a poor therapist. This is particularly problematic in trials that assign therapists to deliver a single type of intervention, as observed intervention effects may be due to the quality of the therapist delivering the intervention rather than the intervention itself. To mitigate this threat to validity, RCTs may have therapists deliver all interventions in a trial to ensure that therapist effects are spread equally across interventions. This approach, however, can lead to “drift” in which therapists begin to blend the treatments, offering a hybrid version of treatment to all participants. This RCT employs the latter approach but institutes extensive training, ongoing fidelity monitoring, and supervision, with the explicit goal of having therapists objectively differentiate the active intervention, teleMI+BST, from the control intervention, teleCET. In addition, preliminary data suggest that a strong therapeutic alliance (i.e., the extent to which participants have positive impressions about the counseling and therapist) is achieved in both MI+BST and CET [10, 32]. Thus, common threats to validity in behavioral trials—therapist effects and therapeutic alliance—are mitigated through design and analysis in this clinical trial.

RCTs can be placed on a continuum from highly controlled efficacy studies that maximize internal validity of findings to pragmatic effectiveness trials that aim to maximize external validity. Louden and colleagues [59] characterized the Pragmatic-Exploratory Continuum Indicator Summary-2 (PRECIS-2) Wheel. This tool allows researchers to rate 9 categories of study design on an internal- to external validity continuum. Figure 3 characterizes the extent to which this study maximizes internal versus external validity across the 9 PRECIS-2 domains. Each spoke on the wheel represents a scale from 1 (explanatory, high internal validity) to 5 (pragmatic, high external validity). Although this RCT is characterized as an efficacy trial, several characteristics align more closely with pragmatic trial designs. First, few participant exclusion criteria were employed, resulting in a sample that is representative of the population of people who would receive this intervention in practice. Second, practicing HIV social service providers employed at a community HIV organization deliver both study interventions, as would be the case in usual practice. Third, no effort, other than reminder phone calls and text messages, is made to ensure participants adhere to the intervention, mirroring effort that would occur in practice and as contrasted with the considerable effort that is made to ensure participants received the full active intervention or comparator in efficacy trials. Finally, planned data analyses will use all available data and intent-to-treat procedures, rather than per-protocol analyses with restricted data. This RCT is an example of how trials can exist in the middle of the efficacy-effectiveness continuum to help balance the need for both internal and external validity.

Figure 3.

Pragmatic-Explanatory Continuum Indicator Summary-2 (PRECIS-2) Wheel of Randomized Controlled Trial. Adapted from Loudon and colleagues (2015) and applied to Project BRIDGE.

5. Conclusion

Nearly 40 years into the HIV epidemic, many OPLWH continue to engage in UAVI with HIV-negative partners. This clinical trial’s goal of developing a telephone-administered, individually tailored secondary prevention intervention for OPLWH informed by motivational interviewing—a highly flexible, client-centered, evidence-based, and widely used form of therapy—can meet OPLWH where they currently are in terms of their sexual relationships and health needs. This study will determine whether teleMI+BST can help reduce UAVI and depression and enhance healthier sexual behaviors in this understudied population.

Table 1.

Outline of session content for the teleMI+BST intervention

Session 1	Session 2	Session 3	Session 4	Booster Session
Week 3	Week 4	Week 8	Week 12	Week 24
• Rapport building; engage client in treatment  ○ Explore current and past sexual relationships  ○ Ascertain impact of aging on sexuality • Focus treatment:  ○ Identify condom use practices  ○ Assess attitudes and beliefs about condom use	• Evoke change talk  ○ Identify relationship goals  ○ Develop discrepancy between goals and condomless sex  ○ Process condom use ambivalence  ○ Identify behavior change goals  ○ Provide information with clients’ permission	• Revisit behavior change goals • Planning:  ○ Develop behavior change plan  ○ With clients’ permission, teach and discuss condom use behavioral skills germane to older people living with HIV and clients’ specific needs	• Review progress toward behavior change plan implementation • Help clients modify change plans as needed	• Revisit clients’ behavior change plans • Troubleshoot barriers to plan implementation or strengthen clients’ commitment to use condoms

Table 2.

Outline of session content for the teleCET control

Session 1	Session 2	Session 3	Session 4	Booster Session
Week 3	Week 4	Week 8	Week 12	Week 24
• Rapport building • Discuss clients’ feelings and concerns about aging with HIV • Identify a prominent life stressor that will be the focus of treatment	• Differentiate between problem- and emotion-focused coping • Teach clients to match stressor changeability with appropriate coping strategies	• Discuss how to obtain and maintain social support that can facilitate coping efforts • Identify support networks, evaluate support needs	• Review elements of sessions 1–3 • Further evaluate support needs • Discuss coping skill implementation strategies	• Assess client progress in coping with identified stressors • Review key concepts • Troubleshoot barriers to effective coping

Highlights:

• This is the first large-scale prevention trial for older people living with HIV.

• Novel methods employed enhance fidelity to treatment delivery.

• Features of this RCT blend efficacy and effectiveness designs.