Methods: We conducted in-depth interviews with a convenience sample of 28 providers (nurses and doctors) and 11 patients in 17 government clinics and one tertiary hospital between June 2015 and November 2016 in Klerksdorp (North West province), South Africa. Two analysts reviewed all transcripts and developed a codebook. Inter-coder reliability was assessed to be 0.75, and codes were discussed and the codebook was revised. Analysts coded all transcripts and carried out a thematic analysis using a social-ecological framework.
Results: Individual level barriers to seeking HIV care included demographic factors such as gender (women were more open to seeking and receiving treatment than men), denial and fear, and financial difficulties. Dyadic barriers included sexual partners’ differing willingness to seek and remain on treatment, and fears over disclosing HIV status to sexual partners. Community level barriers included stigma, with patients fearing mistreatment and discrimination by health workers and neighbors and friends. Health system barriers included resource shortages at government clinics, and absence of a robust system to process and communicate laboratory results and monitor whether patients returned for follow-up visits.
According to providers, perceived benefits of the mHealth system were reduced workload for clinicians and reduced wait time for patients, potential for expanded uses of the system beyond HIV care, and patient empowerment. Providers raised concerns about the confidentiality of delivering test results via cell phone and breaking potentially distressing results through text messages. They also expressed doubts over whether patients would be able to understand their test results without a clinician’s interpretation, and doubts as to the proportion of their community with regular access to cell phones. Compared to providers, patients did not view confidentiality or cell phone access to be as large of a concern. Patients were open to receiving test results via SMS but also indicated that they would like a face-to-face explanation of their results.
Conclusion: Using a social-ecological framework, we found that barriers to entering and remaining in the HIV care continuum were present at the individual, dyadic, community, and health system levels. Given the magnitude of the HIV epidemic in South Africa, no single intervention can adequately address all these barriers and a multi-pronged approach is necessary to improve retention in care. However, through interviews with providers and patients, we found that an mHealth program that provides test results directly to patients outside of the clinical setting has the potential to address several key barriers, particularly at the health system level, that make it difficult for patients to initiate and remain engaged in care.
For patients who do not require treatment changes, receiving results through cell phone could preclude the need for unnecessary trips to the clinic, saving them long wait times, lost wages, transportation costs and exposure to other sick patients. For patients with an actionable result, receiving their test results through cell phone could encourage them to return to the clinic and be seen by a clinician as soon as possible, therefore reducing delays in initiation or adjustment of ART. This would result in a reduced workload for clinicians. Providers also highlighted the potential of the mHealth system to empower patients as an important benefit. Receiving their test results to their phone without a trip to the clinic could allow patients can take a more active role in understanding and acting on their health status. Ultimately, such a system could be expanded to provide a variety of test results and health information, leading to improved integration with other health services such as those for cancer, heart disease, or diabetes.
However, some important concerns remain, such as the ethics surrounding privacy and disseminating potentially distressing results through text messages and patient comprehension of results. These findings recommend that an mHealth intervention address confidentiality issues, for example, by implementing a secure password-enabled system such as USSD. Success of an mHealth intervention would rely on patients’ understanding of the messages sent, and therefore messages would need to be carefully crafted and tested. Rather than completely replacing face-to-face communication, an mHealth system should serve as a complement to it, leading to more meaningful interactions between patients and providers. These considerations will need to be addressed for an mHealth system to be successfully integrated with the clinic workflow. This formative research supports delivering HIV laboratory results directly to patients through a secure mHealth system. This type of mHealth intervention will capture those patients who currently do not return for their laboratory results, and may empower patients to know their status and return for treatment, improving engagement and retention in the HIV care continuum.