Background
There is a dearth of high-quality evidence from digital health interventions in routine program settings in low- and middle-income countries. We previously conducted a randomized controlled trial (RCT) in Zimbabwe, demonstrating that 2-way texting (2wT) was safe and effective for follow-up after adult voluntary medical male circumcision (VMMC).
Objective
To demonstrate the replicability of 2wT, we conducted a larger RCT in both urban and rural VMMC settings in South Africa to determine whether 2wT improves adverse event (AE) ascertainment and, therefore, the quality of follow-up after VMMC while reducing health care workers’ workload.
Methods
A prospective, unblinded, noninferiority RCT was conducted among adult participants who underwent VMMC with cell phones randomized in a 1:1 ratio between 2wT and control (routine care) in North West and Gauteng provinces. The 2wT participants responded to a daily SMS text message with in-person follow-up only if desired or an AE was suspected. The control group was requested to make in-person visits on postoperative days 2 and 7 as per national VMMC guidelines. All participants were asked to return on postoperative day 14 for study-specific review. Safety (cumulative AEs ≤day 14 visit) and workload (number of in-person follow-up visits) were compared. Differences in cumulative AEs were calculated between groups. Noninferiority was prespecified with a margin of −0.25%. The Manning score method was used to calculate 95% CIs.
Results
The study was conducted between June 7, 2021, and February 21, 2022. In total, 1084 men were enrolled (2wT: n=547, 50.5%, control: n=537, 49.5%), with near-equal proportions of rural and urban participants. Cumulative AEs were identified in 2.3% (95% CI 1.3-4.1) of 2wT participants and 1.0% (95% CI 0.4-2.3) of control participants, demonstrating noninferiority (1-sided 95% CI −0.09 to ∞). Among the 2wT participants, 11 AEs (9 moderate and 2 severe) were identified, compared with 5 AEs (all moderate) among the control participants—a nonsignificant difference in AE rates (P=.13). The 2wT participants attended 0.22 visits, and the control participants attended 1.34 visits—a significant reduction in follow-up visit workload (P<.001). The 2wT approach reduced unnecessary postoperative visits by 84.8%. Daily response rates ranged from 86% on day 3 to 74% on day 13. Among the 2wT participants, 94% (514/547) responded to ≥1 daily SMS text messages over 13 days.
Conclusions
Across rural and urban contexts in South Africa, 2wT was noninferior to routine in-person visits for AE ascertainment, demonstrating 2wT safety. The 2wT approach also significantly reduced the follow-up visit workload, improving efficiency. These results strongly suggest that 2wT provides quality VMMC follow-up and should be adopted at scale. Adaptation of the 2wT telehealth approach to other acute follow-up care contexts could extend these gains beyond VMMC.
Trial Registration
ClinicalTrials.gov NCT04327271; https://www.clinicaltrials.gov/ct2/show/NCT04327271