Tuberculosis (TB) is the leading infectious cause of death globally. India has the largest TB epidemic, accounting for one-quarter of cases and one-third of TB deaths worldwide. Nonadherence to medications is a central challenge in TB care leading to increased death, disease recurrence, and drug resistance. Despite its importance, detecting nonadherence in routine care is challenging, as current measurement approaches are inaccurate, not person-centered, or ineffective at improving outcomes. Early and accurate detection of nonadherence may serve as an entry point for differentiated care, in which people with TB at risk for poor outcomes can be given intensified interventions. Urine isoniazid testing is a validated, low-cost, point-of-care, and direct adherence measure that may be predictive of TB outcomes and therefore serve as an ideal triage test to enable differentiated care. However, to integrate urine testing into routine care, research is needed to: (1) understand how the test performs at scheduled clinic visits, (2) gain a rich understanding of root causes of nonadherence to better leverage urine test results, and (3) identify barriers and facilitators to implementation. In this study, the investigators propose conducting a 900 participant prospective cohort study with translational research involving clinical, behavioral, and implementation science to facilitate integration of urine isoniazid testing into India's national TB program. The investigators' central hypothesis is that urine testing can be integrated into routine care to facilitate early and accurate identification of people with TB who are likely to suffer poor outcomes, including death and TB recurrence. In Aim 1, the investigators will assess the accuracy of urine test results assessed at scheduled clinic visits in comparison to those assessed at unannounced home visits. In Aim 2, the investigators will assess the relationship between nonadherence detected by urine testing and subsequent unfavorable TB outcomes of death, loss to follow-up, treatment failure, and post-treatment TB recurrence.. This study proposal aims to develop an innovative but pragmatic strategy for early identification of TB medication nonadherence that is feasible in low- and middle-income countries with a high TB burden.

Tuberculosis (TB) is the leading infectious cause of death globally. India has the largest TB epidemic, accounting for one-quarter of cases and one-third of global TB deaths. Nonadherence to TB medications is associated with increased death, TB recurrence, and drug resistance. Specifically, missing >10% of doses is associated with six times greater risk of poor TB outcomes.2 The association between nonadherence to TB medications and emergence of drug resistance is one of the most concerning from a public health perspective. For example, one study found that nonadherence to TB therapy was associated with independent elevated risk of acquiring drug resistance (adjusted odds ratio 19.7, 95%CI 1.7-234). Another study found that, among people being treated for multidrug-resistant (MDR) TB, each additional month in which a patient failed to take 80% of their prescribed medication doses was associated with elevated risk of developing extensively drug-resistant (XDR) TB (adjusted hazard ratio 1.2, 95% CI 1.01-1.4).

Despite its importance, detecting nonadherence in routine care is challenging, especially as TB programs in low- and middle-income countries (LMICs) have moved away from directly observed therapy. Our research shows indirect measures-e.g., cellphone-based technology, pill counts, and patient-reported measures-miss detecting >30% of people with TB who are nonadherent by urine isoniazid testing.

Urine isoniazid testing is a low-cost, validated, direct point-of-care adherence measure that is predictive of TB outcomes. In a 650-participant cohort study in India, one unannounced home visit was conducted for each participant during TB therapy to conduct urine testing. That study found: (1) nonadherence was common by urine testing, with 18% of participants having missed more than 1 dose, and (2) negative urine test results were associated with 3.9 increased odds of death and 3.8 increased odds of loss to follow-up from treatment.

While urine testing has benefits, further research is needed to evaluate its uses under TB programs setting First, urine testing may perform differently in clinic as compared to unannounced home visits, as people with TB may modify adherence behavior in anticipation of scheduled clinic visits. Research is needed to understand how well testing at clinic visits represents adherence and predicts TB outcomes. Second, research is needed to understand reasons for nonadherence (i.e., negative urine test results) to inform personalized interventions for differentiated care. Third, implementation science research is needed to evaluate the use of urine testing in routine care in a manner that could lead to sustained use and impact.

The investigators propose conducting translational research involving clinical, behavioural, and implementation science to evaluate the use of urine isoniazid testing in the TB program. The investigators' central hypothesis is that urine testing could be useful within the routine care to facilitate early and accurate identification of people with TB who are likely to suffer poor outcomes, including death and TB recurrence. If this hypothesis is correct, the urine assay could serve as a triage test for developing differentiated care strategies, in which people at risk for poor outcomes are identified early and given intensified care

The investigators have the following specific aims:

Aim 1. To assess the agreement between urine test results collected at prescheduled clinic in comparison to unannounced home visits.

Summary of this aim: The investigators will enroll a cohort of 900 people starting 6-month drug-susceptible TB therapy. For each participant, to enable comparison of clinic and home samples, the investigators will conduct 1 urine test at a scheduled clinic visit and 1 urine test at an unannounced home visit during each of the intensive (first 2 months), early continuation (second 2 months), and late continuation (last 2 months) treatment phases.

Aim 2. To assess the relationship between nonadherence by urine testing and subsequent TB outcomes.

Summary of this aim: The investigators will follow the Aim 1 cohort participants through the completion of TB treatment (which usually takes 6 months) and then for 12 months after treatment and assess the following unfavorable outcomes: death, treatment failure, loss to follow-up, and TB recurrence. The investigators will assess the association between nonadherence by urine testing (separately for clinic and home samples) and the composite of these unfavorable outcomes (primary outcome) and with each unfavorable outcome individually (secondary outcomes).

Inclusion Criteria:

1. 18+ years of age,
2. presumed or confirmed drug-susceptible pulmonary TB (with or without a previous TB history), and
3. taking an isoniazid-containing TB regimen.

Exclusion Criteria:

1. inability to provide informed consent, or
2. drug-resistant TB diagnosis or switch to another non-isoniazid-containing regimen (Note that this second exclusion criteria may result in retrospective exclusion from the study).

Tuberculosis (TB) adherence and outcomes quantitative dataset: This will comprise one quantitative dataset of de-identified individual study participant data from 900 people with TB on treatment, including all quantitative data for the Aim 1 and Aim 2 cohort study. These data will be captured and stored as a Microsoft Excel CSV (comma-separated delimited text) file that can be uploaded into most statistical programs for analysis. As this dataset will be de-identified, individual-level data will be shared publicly through the NIAID Clinical Trials Data Repository (AccessClinicalData@NIAID).

Data will be made available as soon as possible or at the time of associated publication, whichever comes first.

The NIAID Clinical Trials Data Repository provides metadata, persistent identifiers (i.e., digital object identifier), and long-term access. This repository is supported by the National Institute of Allergy and Infectious Diseases and datasets are available under a data access request process.