Impact of Short Message Service and Peer Navigation on Linkage to Care and Antiretroviral Therapy Initiation in South Africa

"[F]indings indicate that peer navigation coupled with SMS reminders can improve linkage to HIV care and ART [antiretroviral therapy] initiation as well as retention in HIV care in South Africa..., where scalable interventions are sorely needed to meet the 95-95-95 goals."

In South Africa, where there were 7.8 million people with HIV (PWH) in 2020, improving early HIV diagnosis, linkage to care, treatment initiation, retention in care, and adherence to antiretroviral therapy (ART) could bring substantial gains in HIV prevention. Among feasible strategies with the potential to improve care engagement outcomes, short message service (SMS or text messaging) and personal support interventions have demonstrated promise. I-Care was a cluster randomised trial conducted in primary care facilities in North West Province, South Africa. The primary study outcome was retention in HIV care; this analysis includes secondary outcomes: linkage to care and ART initiation.

The trial was conducted in the Bojanala Platinum District of North West Province, where HIV prevalence in the adult population is estimated at 22.7% and ART uptake is lower than the national average, with approximately 61% of PWH aware of their status on ART. The I-CARE interventions were as follows:
 

• The SMS strategy included three types of messages delivered from an automated system that determined messaging based on sex, clinical/ART status, and clinic visit dates. Behavioural messages were sent every 2 weeks and addressed engagement in care (e.g., blood tests every 6 months will help you keep track of your health. Ask your nurse when your next blood test should be); medication adherence or awareness (e.g., Be smart! Take your meds with you to work or travel.); and healthy living (e.g., Reduce alcohol consumption). Finally, every other week, participants received two-way check-in messages.
• Participants in the SMS + peer navigation (PN) condition also received appointment reminders and automated bi-weekly behavioural messages identical to those in the SMS-only condition. However, check-in messages were personalised and unscripted and were sent by peer navigators: HIV-positive adults in care who worked with participants to develop strategies for overcoming barriers to care through monthly in-person meetings and text/phone check-ins, with additional contacts as needed. Navigators completed training prior to commencing direct one-on-one patient contact. The training consisted of 1 week of intensive didactic and skills-building materials related to rapport building, listening skills, identification of barriers to care and prevention, setting and monitoring of behaviour change goals, and ethics.
• Standard of care (SOC) participants did not receive messages or services.

Eighteen primary care clinics were randomised to automated SMS (n = 7), automated and tailored SMS + PN (n = 7), or SOC (n = 4). Recently HIV diagnosed adults (n = 752) were recruited from October 2014 to April 2015. Those not previously linked to care (n = 352) contributed data to this analysis.

Overall, SMS (n = 132) and SMS + PN (n = 133) participants linked at 1.28 [95% confidence interval (CI): 1.01-1.61] and 1.60 (95% CI: 1.29-1.99) times the rate of SOC participants (n = 87), respectively. SMS + PN significantly improved time to ART initiation among non-pregnant women (hazards ratio: 1.68; 95% CI: 1.25-2.25) and men (hazards ratio: 1.83; 95% CI: 1.03-3.26) as compared with SOC.

Thus, these data demonstrate that overall both the SMS and SMS + PN interventions improved the time to linkage to care and that the SMS + PN intervention hastened the time to ART initiation. When stratified by sex, men experienced a doubling in linkage rates in the SMS + PN arm compared with men in SOC, and close to a doubling of ART initiation rates as well. Among women, effects were most apparent in the non-pregnant population, where both interventions demonstrated significant improvements in time to linkage to care, and SMS + PN participants experienced more rapid ART initiation. Effects were attenuated when including pregnant women in the analyses, likely given their rapid triage into care for prevention of mother-to-child transmission across study arms. The data indicate that both interventions merit consideration for scale-up, though sex-specific programming may be warranted.

Reflecting on the findings, the researchers note that, for both men and women, SMS may be a good first-step option for those who only require appointment reminders or have fewer barriers to care whereas the peer navigation approach is likely more adept at addressing the needs of people who face significant barriers and require emotional support and encouragement. For example, men may require more support to enter care: Men's concern about HIV-related stigma and inadequate social support have been found to be key barriers affecting treatment initiation. aving peer support through the early to mid-stages of HIV care engagement may be important for those who perceive stigma associated with clinic attendance and require additional social support for linkage to care. Ongoing peer mentorship that normalises treatment uptake and adherence may also be a strategy to engage men.

The researchers conclude that both SMS and PN interventions should be considered candidates for integration into national programmes.

Full list of authors, with institutional affiliations: Sheri A. Lippman, University of California, San Francisco; Julia de Kadt, International Training and Education Center for Health (I-TECH); Mary J. Ratlhagana, I-TECH; Emily Agnew, University of California, San Francisco; Hailey Gilmore, University of California, San Francisco; Jeri Sumitani, I-TECH; Jessica Grignon, I-TECH and University of Washington; Sarah A. Gutin, University of California, San Francisco; Starley B. Shade, University of California, San Francisco; Jennifer M. Gilvydis, I-TECH; John Tumbo, Sefako Makgatho Health Sciences University; Scott Barnhart, University of Washington; Wayne T. Steward, University of California, San Francisco