Research Topic and Rationale

I want to explore the feasibility of using individual patient data (IPD) and characteristics of people living with HIV (PLHIV) on antiretroviral therapy (ART)—specifically those successfully traced and categorized as lost to follow-up (LTFU)—to develop a predictive analysis model. The model will estimate the risk of four key outcomes: attrition (loss to follow-up), silent transfers (continuing care at a different facility), patients still in care, and mortality.

Retention in ART programs is critical for reducing HIV transmission and improving patient outcomes. However, retention rates remain suboptimal, with approximately 74% retained at 24 months and 60% at 60 months (Fox et al., 2018). Successfully tracing PLHIV classified as LTFU is essential for:

• Monitoring the success of ART programs, as retention in care and mortality serve as key indicators.

• Improving patient outcomes by re-engaging individuals in care.

Rationale

• Retention of patients on ART is beneficial to reducing transmission to uninfected partners, leading WHO to remove CD4 threshold to initiate patients onto ART, but this can only have significance if retention is improved (Fox et al., 2018). Retention is at about 74% at 24 onths and 60% at 60 months.

  
  

• Successfully tracing PLHIV LTFU is important for monitoring the success of ART programmes (retention in care and mortality are some indicators of success) and bringing patients back into care to improve their health status (Ballif et al., 2022).

  
  

• Studies that have only analysed PLHIV who have remained in care are underestimating the mortality of all patients who have initiated ART (Ballif et al., 2022). Estimates of mortality based on patients in care and under observation through single facility data system will underestimate mortality at the level of the program, that is, among all patients who started therapy (Chammartin et al., 2018) and [8, 9, 52]. This hampers the evaluation of programmes and comparisons between programmes, settings and countries (Ballif et al., 2022) and [22].

  
  

• Retention estimates are faulty because they are based on clinical/facility cohorts (usually well resourced clinics and cannot track patients LTFU) (Fox et al., 2018).

• Loose definitions of LTFU and reactive tracing mean that many PLHIV (29% [16] in (Ballif et al., 2022)) defined as lost to follow-up remained lost despite tracing. In (Ballif et al., 2022), the health status of 41% of PLHIV LTFU remained unknown.

  
  

• Tracing success is hampered by the delay between LTFU and tracing, this should be prioritized (Ballif et al., 2022) and [16].

  
  

• Silent transfers are a barrier to programme evaluation and accounts for the most misclassified patient outcomes (Ballif et al., 2022) (Ballif et al., 2022)and [26].