ART Adherence and Prevention Support By Interactive Voice Response (IVR) Calls in India

BACKGROUND: Mobile phones are rapidly approaching 100% penetration globally and present opportunities for disruptive innovations in providing support for antiretroviral adherence, factors associated with non-adherence (depression, substance abuse, sexual risks), and quality of life. Text-message ART reminder interventions have demonstrated efficacy recently, but focused on reminders only and text-messaging is not universally acceptable. 

METHODOLOGY: Mixed-methods formative research, comprised of a PLH CAB, one-month pre-post pilot study (n=46), and post-pilot focus groups, was completed in August 2013 to inform an IVR intervention and RCT for PLH in Kolkata, India. The RCT (n=400), launching in November 2013, consists of baseline and three follow-up assessments over six months with medical chart abstraction and weekly IVR assessment calls. The intervention condition (n=200) receives two IVR calls daily timed to ART dose schedules, compared to standard care (n=200).  Eligibility includes currently taking India’s 1^st-line ART (twice-a-day dosing) for at least six months and reporting at least one missed ART dose in prior six months. IVR messages, in Bengali and Hindi, focus on nutrition, hygiene, mental health (i.e., cognitive-behavioral strategies), stress, social support, entitlements, adherence, and monthly ART appointment reminders. Opt-in IVR modules specifically mention HIV/AIDS, sexually transmitted infections (STI) and risk behaviors, and substance abuse.

RESULTS:  Formative results demonstrated that: 1) Text-messaging is not feasible in this context due to low literacy (39% illiterate) and text-messaging experience, while IVR is common; 2) depression, hopelessness, and stigma are significant concerns; 3) Nutrition/diet, mental health, and social support messages were particularly helpful and engaging; 4) General messages were preferred to protect HIV-status confidentiality and sensitivities around children answering calls, but PLH living alone or who disclosed to household members preferred HIV- and sex-specific messages; and 5) IVR calls functioned as ART reminders even if calls could not be answered. Pre-post pilot results (no control; n=46, 80% women, 60% sex workers) found significant reductions in self-report non-adherence (p = 0.016).  Regression models demonstrated that depression is significantly associated with non-adherence (p < 0.01).

CONCLUSIONS: Engaging and tailored message content is likely critical to sustained participation in phone messaging interventions, which are likely to be cost-effective compared to in-person modalities and for their massive scalability.