HIV: Current Research
Open Access

Abstract

Aggregating Loss to Follow-Up Behavior in People Living With HIV on ART: A Cluster Analysis Using Unsupervised Machine Learning Algorithm in R

Amobi Andrew Onovo*, Abiye Kalaiwo, Christopher Obanubi

Background: This study aimed to aggregate Loss to Follow Up (LTFU) behavior in People Living with HIV (PLHIV) into clusters to examine and describe PLHIV clusters havings similar characteristics and patterns according to their risk profile.

Objectives/methods: This was a retrospective, cross-sectional study that randomly reviewed 11,589 records of LTFU adult patients initiated on first-line ART from 313 USAID/PEPFAR-supported HIV clinics spread across 5 of Nigeria’s 6 geographical regions between July 1, 2008, and June 30, 2020. LTFU, was defined for PLHIV on ART as >28 days without an encounter since the last scheduled ART refill appointment. Using the Minkowski method and ward. D2 clustering technique for unsupervised machine learning algorithm "agglomerative hierarchical clustering" in R, we identified 6 clusters associated with patients LTFU behavior.

Results: Within the review period, 497,620 patients were ever enrolled on ART. 324,225 (65.2%) remained on treatment, 101,716 (20.4%) had an LTFU event captured, 36,021 (7.2%) were transferred out to other facilities, 25,633 (5.2%) died and 10,025 (2.0%) self-terminated treatment. Majority (66.7%) of the clusters consist of female LTFUs. LTFU doubled steadily by age among adolescents (15-19 years), but as age increased above 40-years the rate of LTFU decreased. High rate of LTFU was reflective of shorter duration on ART. Viral load test was low, with only half (50.0%) of the clusters having a documented viral load test result.

Conclusion: LTFU rates in HIV-positive patients receiving ART in our clinical sites have varied by the duration of ART, with rates declining in recent years.

Published Date: 2025-01-23; Received Date: 2023-08-17