Effect of Digital Medication Event Reminder and Monitor-Observed Therapy vs Standard Directly Observed Therapy on Health-Related Quality of Life and Catastrophic Costs in Patients With Tuberculosis: A Secondary Analysis of a Randomized Clinical Trial

Key Points

Question  Does a digital medication event reminder monitor–observed therapy for patients with tuberculosis result in higher health-related quality of life compared with the standard directly observed therapy?

Findings  In this secondary analysis of a randomized clinical trial involving 109 adults with pulmonary tuberculosis, health-related quality of life was significantly better in patients who used the digital medication event reminder monitor system compared with those receiving standard directly observed therapy. Patients using the digital therapy also had significantly lower rates of catastrophic costs.

Meaning  These results suggest that digital health technologies may improve quality of life and reduce catastrophic costs among patients with pulmonary tuberculosis, in particular those who face structural barriers to standard therapy.

Importance  Little is known about whether digital adherence technologies are economical for patients with tuberculosis (TB) in resource-constrained settings.

Objective  To test the hypothesis that for patients with TB, a digital medication event reminder monitor (MERM)-observed therapy provides higher health-related quality of life (HRQoL) and lower catastrophic costs compared with standard directly observed therapy (DOT).

Design, Setting, and Participants  This study was a secondary analysis of a randomized, 2-arm, open-label trial conducted in 10 health care facilities in Ethiopia. Eligible participants were adults with new or previously treated, bacteriologically confirmed, drug-sensitive pulmonary TB who were eligible to start first-line anti-TB therapy. Participants were enrolled between June 2, 2020, and June 15, 2021, with the last participant completing follow-up on August 15, 2021.

Interventions  Participants were randomly assigned (1:1) to receive a 15-day TB medication supply dispensed with a MERM device to self-administer and return every 15 days (intervention arm) or the standard in-person DOT (control arm). Both groups were observed throughout the standard 2-month intensive treatment phase.

Main Outcomes and Measures  Prespecified secondary end points of the original trial were HRQoL using the EuroQoL 5-dimension 5-level (EQ-5D-5L) tool and catastrophic costs, direct (out-of-pocket) and indirect (guardian and coping) costs from the individual patient perspective using the World Health Organization’s Tool to Estimate Patient Costs, and common factors associated with lower HRQoL and higher catastrophic costs.

Results  Among 337 patients screened for eligibility, 114 were randomly assigned, and 109 were included in the final complete-case intention-to-treat analysis (57 control and 52 intervention participants). The mean (SD) age was 33.1 (11.1) years; 72 participants (66.1%) were men, and 15 (13.9%) had HIV coinfection. EQ-5D-5L overall median (IQR) index value was 0.964 (0.907-1). The median (IQR) value was significantly higher in intervention (1 [0.974-1]) vs control (.908 [0.891-0.964]) (P < .001). EQ-5D-5L minimum and maximum health state utility values in intervention were 0.906 and 1 vs 0.832 and 1 in control. Patients’ overall median (IQR) postdiagnosis cost was Ethiopian birr (ETB) 80 (ETB 16-ETB 480) (US $1.53). The median cost was significantly lower in intervention (ETB 24 [ETB 16-ETB 48]) vs control (ETB 432 [ETB 210-ETB 1980]) (P < .001), with median possible cost savings of ETB 336 (ETB 156-ETB 1339) (US $6.44) vs the control arm. Overall, 42 participants (38.5%; 95% CI, 29.4%-48.3%) faced catastrophic costs, and this was significantly lower in the intervention group (11 participants [21.2%]; 95% CI, 11.1%-34.7%) vs control (31 participants [54.4%]; 95% CI, 40.7%-67.6%) (P < .001). Trial arm was the single most important factor in low HRQoL (adjusted risk ratio [ARR], 1.49; 95% CI, 1.35-1.65; P < .001), while trial arm (ARR, 2.55; 95% CI, 1.58-4.13; P < .001), occupation (ARR, 2.58; 95% CI, 1.68-3.97; P < .001), number of cohabitants (ARR, 0.64; 95% CI, 0.43-0.95; P = .03), and smoking (ARR, 2.71; 95% CI, 1.01-7.28; P = .048) were the most important factors in catastrophic cost.

Conclusions and Relevance  In patients with TB, MERM-observed therapy was associated with higher HRQoL and lower catastrophic costs compared with standard DOT. Patient-centered digital health technologies could have the potential overcoming structural barriers to anti-TB therapy.

Trial Registration  ClinicalTrials.gov Identifier: NCT04216420

Tuberculosis (TB) remains a disease of poverty that disproportionately affects the world's most economically vulnerable populations. Although highly efficacious, current anti-TB treatments are very costly for patients and their households as they have to cover costs associated with facility-based medication pick up, imposing significant challenges to treatment adherence and resulting in treatment failure, drug resistance, and continuous disease transmission.^1,2 Current anti-TB treatments last at least 6 months for drug-susceptible TB and 20 months for drug-resistant TB, with patients swallowing medications at a health care facility under directly observed therapy (DOT) throughout the intensive phase in most high-burden countries. This leads patients and their households to be overburdened with structural barriers to therapy adherence, including catastrophic costs and income losses resulting from transportation, food, and accommodation for in-person DOT that they are incapable of avoiding, which contradicts the World Health Organization (WHO)’s End TB strategy to shift the percentage of TB-affected families facing catastrophic costs to 0%. Efforts are currently under way to develop short-course anti-TB regimens^3,4 and cost-effective digital adherence technologies (DATs) as alternatives to DOT.^5,6

An effectiveness-implementation type 2 hybrid, randomized controlled trial was designed to assess whether a digital medication event reminder monitor (MERM)-observed self-administered therapy is effective for patients with TB compared with the standard in-person DOT in a high-burden, resource-constrained context like in Ethiopia. In the protocol, the co-primary end points were the level of treatment adherence (individual-level percentage adherence over the 2-month intensive phase measured by adherence records compiled from a MERM device vs DOT records) and sputum conversion (participant with sputum smear converted following the standard 2-month intensive phase treatment). The trial protocol⁷ and a systematic review in support of the trial⁸ have been published previously. This manuscript describes prespecified secondary outcomes of the trial: health-related quality of life (HRQoL), catastrophic costs, and postdiagnostic costs from an individual patient’s perspective. The MERM device (Wisepill Technologies) holds an electronic module and a medication container to record adherence to treatment, store medication, emit audible alerts and color-coded visual light cues (ie, green, yellow, and red) to remind patients to swallow and refill their medication, and enable clinicians to monitor adherence digitally.⁹ The device is described further in the trial protocol. Some studies have demonstrated improved TB care with the use of electronic medication event monitors,^10-13 but results were inconsistent and only a few followed a randomized trial design.

Generating evidence on the effects of such digital health technologies on HRQoL and patient cost of treatment within vastly different country contexts is equally important to ensure patients leverage the benefits. The WHO Digital Health Strategy 2020-2025 considered cost-effectiveness as one of the major arguments for the appropriate use of digital health.¹⁴ Studies carried out in sub-Saharan African countries, including Eritrea,¹⁵ Nigeria,¹⁶ South Africa,¹⁷ Zimbabwe,¹⁸ and Ethiopia¹⁹ reported a strong inverse association between TB and patient HRQoL. However, studies that have described HRQoL and the economic benefits of the MERM device for patients with TB are scarce, and to the best of our knowledge, no study has been carried out in sub-Saharan Africa and no randomized controlled studies have been documented. Thus, this study aimed to test the hypothesis of a prespecified secondary outcome that MERM-observed self-administered therapy improves the HRQoL and reduces catastrophic costs for patients with TB compared with the standard DOT.

This study describes prespecified secondary outcomes of a multicenter, 2-arm, randomized, attention-controlled, effectiveness-implementation type 2 hybrid trial in 10 health care facilities in Ethiopia. The trial is registered with ClinicalTrials.gov (NCT04216420), and a full description of the study was provided elsewhere (Supplement 1).⁷ Reporting of the trial follows the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

The protocol was approved by the Ethiopian Food and Drug Authority, the Ethiopian National Research Ethics Review Committee, the institutional review board of the College of Health Sciences at Addis Ababa University, and the Ethical Clearance Committee of the Ababa Health Bureau. All participants provided written informed consent. A standard operating procedure was developed for each activity and placed at each study site following the review and approval by the implementing clinicians.

Eligible potential patients were adults aged 18 years or older; had new or previously treated, bacteriologically confirmed drug-sensitive pulmonary TB; were eligible to start the standard WHO-approved fixed-dose combination of 6-month first-line anti-TB medication; and from the outpatient setting. Patients were ineligible if they had known drug-resistant TB or if they had a concurrent health condition that precluded informed consent or safely participating in the study procedures.

Consented participants were randomly assigned (1:1) to either MERM-enabled self-administered therapy (intervention arm) or DOT standard care (control arm) using a computer-generated random number sequence developed by a trial expert who did not participate in patient recruitment. Permuted block randomization method was used to randomly allocate participants and maintain a balance of the number of participants. The study investigators who were responsible for assessing study outcomes and writing the report were masked to group allocation until the manuscript was completed. Because of the scope of the trial, participants and the other study staff were not masked to group allocation. Stratification was not needed for key variables. A statistician masked to group allocation performed the analyses.

Treatment follow-up was made by the full-time clinicians in the TB clinic following a moderate on-site orientation, and the MERM software was set up on computers that had already been in use in TB clinics or similar facilities to understand the sustainability of the intervention in a broad sense. For both arms, baseline information, including demographic, socioeconomic, behavioral, and social characteristics were collected using a semi-structured questionnaire. Both arms were followed throughout the intensive treatment phase lasting 2 months for drug-susceptible TB. Treatment was based on the WHO-recommended 2-month fixed-dose combination of first-line anti-TB drugs delivered as a single daily dose, ie, 2RHZE (150 mg of rifampicin [R], 75 mg of isoniazid [H], 400 mg of pyrazinamide [Z], and 275 mg of ethambutol [E]).

Participants in the intervention arm were informed on how to use the MERM device and given an instructional leaflet with patient-friendly explanatory graphics prepared in the national language that outlined the procedures. Thereafter, participants received a 15-day TB medication supply (HRZE fixed-dose combination therapy of 15 doses) in the MERM device to self-administer. Participants returned to the clinic every 15 days, at which point a clinician counted any remaining tablets in the pillbox and connected the MERM module with the computer. Along with the participant, the clinician downloaded pill-taking data from the device to the computer and reviewed the event reports over the previous 15 days. Any missed event where no ingestion occurred over a particular prescribed ingestion period in the event report was evaluated against any remaining tablets in the pillbox and discussed further with the participant for confirmation.

Participants in the control arm were managed according to the standard DOT practice, where they visited the health care facility each day throughout the 2-month intensive phase to swallow their daily dose of RHZE with direct observation by TB clinicians. The continuation phase (4 months) followed the standard DOT practice for both arms.

HRQoL was measured at the end of the intensive phase using the EuroQoL 5-Dimension 5-Level Questionnaire (EQ-5D-5L),²⁰ which was administered by trained study staff. The tool measures health across 5 domains: mobility, self-care, ability to do usual activities, pain or discomfort, and anxiety or depression. Each domain has 5 levels of response—no problems, slight problems, moderate problems, severe problems, and extreme problems or inability—that provide a descriptive profile used to generate a health state utility value. Health state index scores generally range from less than 0 (where zero is the value of a health state equivalent to death; negative values representing values worse than death based on patient perception) to 1 (the value of full health), with higher scores indicating higher health utility states.²¹ Patient responses were converted into a health state, a 5-digit number generated by concatenating each of the 5 responses to provide a single index value score that determines the health state (eg, full health would be coded 11111). Health state index scores were calculated from individual health profiles using the Ethiopia value set.²²

Individual patient cost of treatment was measured at the end of the intensive phase using the Tool to Estimate Patients’ Costs developed by the WHO, KNCV Tuberculosis Foundation, and the Japan Anti-Tuberculosis Association,²³ which was adapted for the trial and administered by trained study staff. The tool is designed to assess how TB affects the welfare of households and individuals by estimating the costs that patients with TB and their households incur before and during diagnosis and during treatment of TB. The tool subdivides the direct and indirect medical costs incurred into 3 stages: prediagnostic (before diagnosis), diagnostic (during diagnosis or pretreatment), and treatment (during treatment) costs. In general, there were no fixed or marginal costs required from patients. Because the prediagnostic and diagnostic costs were similar for the 2 arms, data were collected on costs during treatment with a focus on costs related to anti-TB drug pick-up, guardian costs, and coping costs over the 2-month intensive phase. The anti-TB drug pick-up cost determined costs for travel, food and accommodation, frequency of travel, length of stay, and distance from the facility. The guardian cost captured data to determine if participants were going with guardians and if so, the reason and frequency. The coping costs were costs to meet daily living requirements despite the extra expenditures or loss of income, which could include selling assets, borrowing money, availing loans, incurring another job to cover the costs, or removing children from school. Costs for anti-TB medications were not included as these were fully covered by the Ethiopian government for all patients with TB.

The main objective of this prespecified secondary analysis was to evaluate the association between MERM-observed therapy and HRQoL, where the HRQoL was measured and calculated for each participant by arm using the EQ-5D-5L score ranging from 0 to 1, with a higher score designating better HRQoL. The secondary outcomes were: (1) catastrophic cost (patients with overall TB treatment cost exceeding or equivalent to 20% of their income over the 2-month treatment period); (2) postdiagnostic cost from an individual patient’s perspective, including cumulative direct costs (out-of-pocket costs related to anti-TB drug pick-up) and indirect costs (guardian and coping costs) over the 2-month intensive phase, and (3) factors contributing to lower HRQoL (having at least 1 health problem) and higher catastrophic costs.

Analysis was based on the complete-case intention-to-treat (ITT) principle. Descriptive statistics including frequency and percentage were used to describe the health state of the study participants. Multiple bar charts with cross-tabulation were used to illustrate distributions of health states by study arms. χ² and Fisher exact tests were employed to compare the 5 EQ-5D-5L health domains by study arms. Both Kolmogorov-Smirnov and Shapiro–Wilk tests revealed that the distribution of EQ-5D-5L index value (ie, the health state utility) was not normal. Median values were used to summarize EQ-5D-5L index value. A nonparametric Mann-Whitney U test was employed to compare the difference in EQ-5D-5L index value among study arms. A log-binomial model was used to identify risk factors for lower HRQoL, which was having at least 1 health problem.

The overall TB treatment cost was estimated by considering costs related to anti-TB drug pick-up, guardian costs, and coping costs over the 2-month intensive phase. Both Kolmogorov-Smirnov and Shapiro-Wilk tests revealed that the distribution of overall TB treatment cost was non-normal, therefore, median values with IQR were used to summarize costs. A nonparametric Mann-Whitney U test was employed to compare cost differences among study arms. The proportion of study participants who faced catastrophic costs at a cut-off point of 20% was estimated. A cross-tabulation was employed to evaluate the distribution of catastrophic cost over the study arms and a χ² test was used to test the association between catastrophic cost and study arms.

The sample size was calculated considering a 2-sided type I error of 5%, a power of 80%, 20% attrition rate, 20% noninferiority margin, a delta of 0.63, and a continuous outcome of percentage adherence over the 2-month intensive phase, with a standard deviation of 41% and 79% of average adherence. The results yielded a sample size of 57 in each arm for a total of 114 participants.

Recruitment took place between June 2, 2020, and June 15, 2021, with the last participant completing follow-up on August 15, 2021. Three hundred thirty-seven patients were screened for eligibility and 114 of these were included and randomly assigned into the trial 1:1 allocation ratio. In the final complete-case ITT analysis, 109 consented participants were included, 52 from the intervention arm and 57 from the control arms (Figure). The imbalance resulted from 4 participants being transferred and 1 lost to follow-up from the intervention arm. Imputations were not considered as the percentage of missing data was below 5%.

Of the 109 participants analyzed, the mean (SD) age was 33.1 (11.1) years; 72 participants (66.1%) were men, 6 (5.5%) were currently homeless, 68 (62.4%) lived in a house with a single bedroom, and 18 (17.4%) smoked cigarettes (Table 1). Eleven (10.1%) were re-treatment cases and had completed their previous treatment, and 15 (13.9%) had HIV coinfection, of whom 11 (73.3%) were receiving antiretrovirals. Laboratory diagnostic tools for pretreatment confirmation of TB were a MTB/RIF (mycobacterium tuberculosis/resistance to rifampicin) assay (68 participants [62.4%]) and acid-fast bacillus smear microscopy (41 participants [37.6%]). Mean (SD) monthly income was US $67.50 ($71.20; range, $0-$333.30). Baseline characteristics and HIV status were balanced between the 2 groups.

Within the overall sample, none of the study participants had severe or extreme problems in any of the 5 dimensions of the EQ-5D-5L (Table 2). The highest percentage of participants had no problems (level 1) in the EQ-5D-5L dimensions of mobility (86 [78.9%]), self-care (88 [80.7%]), usual activities (67 [61.5%]), pain or discomfort (66 [60.6%]), and anxiety or depression (69 [63.3%]). The highest proportion of problems in the EQ-5D-5L dimensions was in pain or discomfort (43 [39.4%]) and the lowest was in self-care (21 [19.3%]).

EQ-5D-5L scores were significantly different between the 2 arms, which favored the intervention arm across all 5 dimensions (Table 3). The proportion without problems in the EQ-5D-5L dimensions (level 1) was higher among the intervention arm compared with the control arm (eg, pain or discomfort: 46 of 52 participants [88.5%] vs 20 of 56 [35.1%]; P < .001) and the proportion of participants with problems was higher among the control arm (eg, level 2 anxiety or depression: 32 of 56 [56.1%] vs 8 of 52 [15.4%]; P < .001).

Overall, 39 of 109 participants (35.8%) had full health (ie, a health state of 11111 based on EQ-5D-5L value sets); of whom, 36 of 39 (92.3%) were in the intervention arm. Comparing by arm, 3 of 57 (5.3%) of the control vs 36 of 52 (69.0%) of the intervention arms had full health.

The index value for the EQ-5D-5L ranged from 0.832 for health state 22322 to 1 for health state 11111. The maximum index value following full health was 0.976 for the health state 12111. The overall median (IQR) index value of the study participants was 0.964 (0.907-1). The minimum and maximum index values for the control arm were 0.832 for health state 22322 and 1 for health state 11111, respectively. For the intervention arm, these were 0.906 for health state 11222 and 1 for health state 11111, respectively. The EQ-5D-5L median (IQR) index value was significantly higher in the intervention arm (1 [0.974-1]) compared with the control arm (.908 [0.891-0.964]) (P < .001).

A log-binomial model was applied to determine the effect of sociodemographic, behavioral, and health characteristics of participants on HRQoL. In the simple regression model with a threshold of P < .20, 6 variables, namely trial arm, age, residency, smoking, khat (a stimulant) use, and previous TB treatment were potential candidates. In the multiple regression analysis, the trial arm was significantly associated with lower HRQoL. The risk of having lower HRQoL in the control arm was 1.49 times (95% CI, 1.35-1.65; P < .001) more likely compared with the intervention arm (eTable in Supplement 2).

The minimum and maximum overall TB treatment costs in the control arm were zero (for those treated in nearby facilities) and Ethiopian birr ETB 16 200 (US $310.34). For the intervention arm, these were zero and ETB 460 (US $8.81), respectively. Patients’ overall median postdiagnosis cost was ETB 80 (ETB 16-ETB 480) (US $1.53). The median (IQR) overall TB treatment cost was significantly higher in the control arm (ETB 432 [ETB 210-ETB 1980]) compared with the intervention arm (ETB 24 [ETB 16-ETB 48]). There was a statistically significant difference in the overall TB treatment cost between the 2 arms. The median (IQR) overall TB treatment cost in the control arm, assuming that they followed the procedure in the intervention arm, would have been ETB 32 (ETB 14-ETB 164). This means that if participants in the control arm had followed the intervention procedure, they would have saved a median 92.6% (93.3%-91.7%) (ETB 32 [ETB 14-ETB 164] vs ETB 432 [ETB 210-ETB 1980]) of their actual expenditure, with an estimated median saving of ETB 336 (ETB 156-ETB 1339) (P < .001).

Of the 109 total participants, 42 (38.5%; 95% CI, 29.4%-48.3%) faced catastrophic costs due to TB treatment, with a significantly higher proportion in the control arm (31 participants [54.4%; 95% CI, 40.7%- 67.6%]) compared with the intervention arm (11 participants [21.2%; 95% CI, 11.1%-34.7%]) (P < .001). None of the study participants reported using coping mechanisms. There was a statistically significant association between catastrophic cost and the study arm.

A log-binomial model was applied to identify risk factors of catastrophic cost. In multiple regression analysis, trial arm (adjusted risk ratio [ARR], 2.55; 95% CI, 1.58-4.13; P < .001), occupation (ARR, 2.58; 95% CI, 1.68-3.97; P < .001), number of cohabitants (ARR, 0.64; 95% CI, 0.43-0.95; P = .03), and smoking (ARR, 2.71; 95% CI, 1.01-7.28; P = .048) were statistically significant factors for patients facing catastrophic costs (Table 4). Thus, the risk of having catastrophic costs in the control arm was 2.55 times (ARR, 2.55; 95% CI, 1.58-4.13; P < .001) more likely compared with the intervention arm.

To our knowledge, this is the first randomized clinical trial investigating the effects of MERM-enabled self-administered therapy on HRQoL and costs for patients with TB compared with the standard DOT. Consistent with our hypothesis, patients with TB who received a 15-day TB medication supply in the MERM device to self-administer and return every 15 days (intervention arm) had a significantly higher HRQoL and lower catastrophic and total costs compared with those who visited health care facilities each day to swallow their daily dose with direct observation by clinicians (control arm).

This study provides strong evidence of the HRQoL benefits of MERM in adults with new or previously treated, bacteriologically confirmed, drug-sensitive pulmonary TB who were eligible to start first-line anti-TB therapy. The EQ-5D-5L median index value was significantly higher in the intervention arm compared with the control arm, and the EQ-5D-5L health state utility value was higher for the intervention arm compared with the control arm, with patients in the control arm being more likely to have a lower HRQoL (ie, at least 1 health problem). The risk for lower HRQoL was 49% likely higher in DOT control arm compared with the MERM intervention arm. This indicates that this intervention has the potential to maintain the economic, psychological, and social well-being of patients with TB as compared with the standard DOT. The main explanation is that, unlike in-person daily DOT, patients who used the MERM device visited the health care facility every 15 days and this significantly reduced vulnerability of patients to the underlying barriers including costs for travel, food, and accommodation for daily in-person DOT. Studies demonstrate that quality of life is an important indicator of the effectiveness of a TB treatment strategy.^24,25 TB is not simply confined to the sphere of biomedicine, but also physical, mental, and social sufferings that the current TB treatment strategy exacerbates.^26-28 Daily in-person DOT is reported by patients as a tiresome and demanding procedure,^29-32 and the problem is severe in challenging conditions that impair population mobility, such as the onset of the COVID-19 pandemic.^33-35

In this study, the overall EQ-5D-5L median index value was equivalent to a previous score for patients with HIV in Ethiopia.³⁶ None of the study participants had severe or extreme problems in any of the 5 dimensions. The highest proportion of problems in the EQ-5D-5L dimensions was in pain or discomfort and the lowest was in self-care. In the modified regression done to determine the effect of sociodemographic, behavioral, and health characteristics of participants on HRQoL, only the trial arm was found to be a statistically significant factor.

In this study, the cost of treatment was measured to assess how TB may affect the welfare of households and individuals by estimating the costs that patients with TB and their households incur during treatment, with a focus on costs related to anti-TB drug pick-up, guardian costs, and coping costs over the 2-month intensive phase. The mean monthly income of study participants was US $67.50, which was lower than the national average of US $183.50.³⁷ The median cost was significantly lower in the intervention arm compared with the control arm. If patients in the control arm followed the intervention procedure, they would have the possibility of saving about 92.6% of their current expenditure on TB treatment. Previous studies done in different countries reported a lower patient cost of TB treatment with the use of digital technology–observed therapy compared with DOT under various cost consequences,^38-41 which is in agreement with the current study, while none of the previous studies reported on cost-effectiveness relevant to a MERM device.

Similarly, a significantly lower proportion of patients with TB in the intervention arm faced catastrophic costs as compared with those in the control arm at a 20% cut-off point. Trial arm, occupation, number of cohabitants, and smoking cigarettes were the most important factors in estimating catastrophic cost. The financial burden of treatment remains a critical issue for patients with TB preventing them from being retained in care,^42-44 which alternative mechanisms including the use of digital health may alleviate.

The present study had some limitations. The overall sample size was modest. This study was carried out in Addis Ababa, Ethiopia, and this limits the generalizability of our results. The WHO’s tool to estimate patient costs had the potential for recall bias and omits costs related to other types of TB-related morbidity, such as pain and nausea medications. The EQ-5D-5L was completed once after treatment using a cross-sectional design and thus did not show pre-post treatment changes in participant health status. Despite these limitations, this study clearly compared the effects of MERM-observed self-administered therapy on HRQoL and costs for patients with TB vs the standard DOT in Ethiopia, one of the countries with the highest burden of TB but poorly represented in such clinical trials.

In this study of patients with TB, MERM device-observed self-administered therapy was associated with higher HRQoL and lower catastrophic costs compared with the standard DOT. Patient-centered digital health technologies could have the potential overcoming structural barriers to anti-TB therapy.

Accepted for Publication: July 22, 2022.

Published: September 15, 2022. doi:10.1001/jamanetworkopen.2022.30509

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Manyazewal T et al. JAMA Network Open.

Corresponding Author: Tsegahun Manyazewal, PhD, Center for Innovative Drug Development and Therapeutic Trials for Africa, College of Health Sciences, Addis Ababa University, PO Box 9086, Addis Ababa 9086, Ethiopia (tsegahunm@gmail.com; tsegahun.manyazewal@aau.edu.et).

Author Contributions: Dr Manyazewal had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Manyazewal, Fekadu, Holland, Marconi.

Acquisition, analysis, or interpretation of data: Manyazewal, Woldeamanuel, Holland, Marconi.

Drafting of the manuscript: Manyazewal, Holland.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Manyazewal, Marconi.

Obtained funding: Manyazewal.

Administrative, technical, or material support: Manyazewal, Woldeamanuel, Fekadu, Holland.

Supervision: Woldeamanuel, Fekadu, Marconi.

Conflict of Interest Disclosures: Dr Marconi reported receiving grants from the National Institutes of Health; he reported receiving consultation fees from Eli Lilly, Bayer, Gilead Sciences, and Viiv outside the submitted work. No other disclosures were reported.

Funding/Support: The work was supported in part by the Fogarty International Center and National Institute of Allergy and Infectious Diseases of the US National Institutes of Health (grant No. D43TW009127) and the Emory Center for AIDS Research (grant No. P30 AI050409).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Emory Center for AIDS Research.

Data Sharing Statement: See Supplement 3.

Additional Contributions: The authors gratefully acknowledge the support received throughout the study period from Prof Henry M. Blumberg, MD, of the Emory University School of Medicine. The authors thank Dr Kidist Bobosha, PhD, of the Armauer Hansen Research Institute, Ethiopia, and Alison Hoover, MPH, of the Emory University School of Medicine for their major support with study coordination. The authors thank the contributions of Dr Wondwossen Amogne, MD, PhD, Dr Getnet Yimer, MD, PhD, and Dr Merga Belina, PhD (Data Safety Monitoring Board); Oumer Fuad, MSc (trial coordination); Tewodros Getinet, MSc, and Haileleuel Bisrat, MSc (for randomization, database design, and statistical expertise); Alemwork Derib, Beyene Lizino, Daniel Urgessa, Hawi Wendimsisnew, Lidiya Belihu, Mahlet Dagne, Naod Girma, Nesira Kedir, Tsegaye Gelan, and Zelalem Abiye (healthcare provision and data collection); Hiwet Eyob, Bezawit Dinberu, Leslie Sorensen, Getahun Alemu, Tesfaye Asefa, Yodit Zegeye, and Dereje Asrat (supply and laboratory testing coordination).