Electronic Pillbox-enabled Self-administered Therapy Versus Standard Directly Observed Therapy for Tuberculosis Medication Adherence and Treatment Outcomes in Ethiopia: a Multicenter Randomized Controlled Trial

Abstract

Background

The prolonged and complex nature of anti-tuberculosis regimens contributes to suboptimal medication adherence, leading to poor treatment outcomes and drug resistance. Trials evaluating the effectiveness of digital adherence technologies are urgently needed. This study aimed to evaluate the effectiveness of pillbox-enabled self-administered therapy (SAT) compared to standard DOT on adherence to tuberculosis medication and treatment outcomes.

Methods

In this multicenter, randomized controlled trial, adults diagnosed with new or previously treated, bacteriologically-confirmed, drug-sensitive pulmonary tuberculosis and eligible to commence anti-tuberculosis therapy were enrolled from 10 healthcare facilities across Ethiopia. Participants were allocated in a 1:1 ratio to either receive a 15-day supply of tuberculosis medication dispensed with an evriMED500® digital medication event reminder and monitor (MERM) device for self-administration and return every 15 days, or to undergo standard DOT. The MERM device integrates an electronic module and medication container, serving to record adherence, securely store medication, emit audible and visual alarms onboard to prompt patients for timely intake and refills, and facilitates healthcare providers in downloading data for monitoring adherence closely. Both groups were monitored throughout the standard two-month intensive treatment phase. The primary endpoints, analyzed following the intention-to-treat (ITT) principle, included were individual-level percentage adherence during the two-month intensive phase, and sputum smear conversion. Secondary endpoints were a negative IsoScreen urine isoniazid test, adverse treatment outcomes, and self-reported adherence. ClinicalTrials.gov: NCT04216420.

Results

A total of 337 patients underwent eligibility screening, with 114 randomly assigned and included in the final analysis (57 in the control group and 57 in the intervention group). Adherence to tuberculosis medication showed comparable rates between the intervention arm (geometric mean percentage [GM%] 99.01%, geometric standard deviation [GSD] 1.02) and the control arm (GM% 98.97%, GSD 1.04), falling within the predefined margin for non-inferiority [mean ratio (MR) 1.00 (95% CI 0.99-1.01); p=0.954]. Urine isoniazid testing was conducted on 443 (97%) samples obtained from 114 participants, revealing that 13 participants yielded at least one negative result. A negative test was more prevalent among the control group compared to the intervention group (p=0.008). There were no significant difference observed regarding smear conversion, adverse treatment outcomes, and self-reported non-adherence.

Discussion

In this randomized controlled trial involving patients with drug-susceptible pulmonary tuberculosis, self-administered therapy facilitated by the MERM device demonstrated treatment adherence comparable to the standard in-person DOT. Adherence to treatment in the intervention arm proved non-inferior when compared to the control arm, suggesting the feasibility of monitoring participants throughout their treatment without requiring daily visits to a healthcare facility.

Conclusion

In addressing patient barriers linked to DOT, tuberculosis medical programs should explore options such as MERM.