A Systematic Review of Digital Interventions to Improve ART Adherence among Youth Living with HIV in sub-Saharan Africa

Table 1

Characteristics of included studies.

Author, year	Country	Study design	Sample size (N)	Age range (years)	Intervention type	Intervention length (weeks)	Adherence measurement method	Main adherence outcomes	Risk of bias	Strength of evidence
Linnemayr et al. 2017 [13]	Uganda	RCT	332	15–22	Weekly SMS (1-way) and weekly SMS with response option (2-way)	48	Medication event monitoring system (MEMS)	1-Way SMS: 95% CI 0.77–1.14,  = 0.27; 2-way SMS: 95% CI 0.75–1.12,  = 0.15	Low	N/A
Ivanova et al. 2019 [19]	Kenya	Pretest-posttest	90	15–24	ELIMIKA digital platform supporting blog posts, discussion, health care provider Q&A, stories contest, and private messaging	13	Perceived importance, confidence, and intentions associated with maintaining adherence pre-intervention versus post-intervention (multiple items assessed but values for “taking all drugs as prescribed by doctor” presented here)	Importance:  = 0.75; Confidence:  = 0.58; Intentions:  = 0.50	N/A	Fair
MacCarthy et al. 2020 [20]	Uganda	RCT	179	15–24	Weekly SMS with individual adherence level from the previous week (T1) and weekly SMS with individual plus peer adherence levels from the previous week (T2)	36	Wisepill smartpill box device	T1 95% CI: -9.9, 2.3; T2 95% CI: -3.0, 7.9	Unclear	N/A
Mavhu et al. 2020 [21]	Zimbabwe	RCT	500	13–19	SMS reminders, phone calls, monthly support groups, home visits, and clinic-based counseling	96	VL (virological failure defined by a VL ≥ 1000 copies/μL)	Adjusted prevalence ratio: 0.58 (95% CI 0.36–0.94,  = 0.03)	Low	N/A
Dulli et al. 2020 [4]	Nigeria	RCT	349	15–24	SMART Connections intervention utilizing private Facebook groups to improve social support and HIV-related knowledge	22	Self-report (ACTG) at 6–9 months post-enrollment	= 0.57	High	N/A
Abiodun et al. 2021 [6]	Nigeria	RCT	212	15–19	SMS reminders for follow-up appointments 48 and 24 hours before the visit date and daily interactive ART adherence SMS reminders requiring participants to respond with “1” if they found the reminder acceptable or “2” if not	20	VL, pill counts, self-report (VAS, ACTG)	VL:  = 0.044; logVL:  = 0.001	Low	N/A

RCT, randomized controlled trial; VL, viral load; ACTG, AIDS Clinical Trials Group adherence questionnaire; VAS, Visual Analog Scale; T1, treatment 1; T2, treatment 2.