“Your Path”: Using Artificial Intelligence (AI) to support HIV health decision making

Organization: Health Economics and Epidemiology Research Office (HE2RO) and Audere 

 

 

Study objectives:

In this innovative study, Indlela in partnership with Audere through funding granted by the South African Medical Research Council, will test whether the “Your Path” AI app integrated with a customized and localized large language model (LLM) chat interface, could mediate health decision-making after participants conduct a simulated HIV self-test. We hypothesize that using an always-available two-way AI-powered communication tool could limit a potential pain point for clients who wish to interact with healthcare providers on their schedule and at their pace. 

 

 

The specific objectives of the study are:

  • Objective 1: Assess intention to engage with healthcare services after a simulated HIV self-test and interaction with the “Your Path” app, and evaluate the usability, acceptability and feasibility of the app for self-testing among community members and care recipients.

 

  • Objective 2: Assess the acceptability, appropriateness, and feasibility of healthcare providers reviewing interactions available from the “Your Path” app for post HIV self-testing counselling.

 

 

Intervention description: B-Hub community members and care recipients will  interact with the “Your Path” app following a simulated HIV self-test. A sub-sample of these participants will be selected for semi-structured in-depth interviews to explore the barriers and motivators of app usage. Additionally, healthcare providers will  review the “Your Path” AI app interaction scripts, script summaries, and participate in semi-structured in-depth interviews to share their perspectives on clients using the app for post-HIV self-test counselling. 

 

 

Evaluation description: Utilising a mixed-methods approach, quantitative and qualitative methods will be used to assess the outcomes for this study.  

 

 

Study outcomes: 

Primary outcome:

  • Client user intention to access health services after a simulated HIV self-test.

 Secondary outcomes: 

  • Client user insights on usability, acceptability, and feasibility of the app for self-testing.
  • Healthcare provider perspectives on the acceptability, appropriateness, and feasibility of the app for post HIV self-testing counselling.

   

 

Study site: The study will be conducted in Johannesburg, South Africa.   

 

 

Target population: We will recruit and enroll participants from Indlela’s B-Hub (care recipients, community members and healthcare providers) who meet the following inclusion criteria: 

 

Inclusion criteria for community members and care recipients 

  • ≥18 years old
  • Willing to provide written informed consent
  • Previous participant in the “Your Choice” study (for half of the participants)
  • Self-reported HIV status
  • Comfortable consenting and speaking in English
  • Able and willing to travel to the study site

 

 Exclusion criteria

  • Participants who are not willing or available to travel to the study site to take part in this sub-study

 

Inclusion criteria for the healthcare providers

  • ≥18 years old
  • Willing to provide written informed consent
  • Comfortable consenting and speaking in English
  • Willing and able to participate in an in-depth interview 

 

 Exclusion criteria:

  • Participants who are not willing or available to take part in this sub-study

 

 

Sample size & analysis: We will enrol 100 community members and care recipients to participate in objective 1. In addition, a convenience sample of 25 participants from the 100 who are successfully enrolled in the “Your Path” study and who engaged with the “Your Path” app will be selected to participate in semi-structured in-depth interviews. Up to 25 healthcare providers will be enrolled to participate in objective 2.

 

Surveys will be analysed using descriptive statistics and regression analyses. Recordings of in-depth interviews will be transcribed and analysed thematically with deductive and inductive approaches to identify key patterns using NVIVO software.  

 

 

Duration: 12 months