This is a multicenter, prospective, cluster randomized, noninferiority clinical trial comparing the protective efficacy and safety of the 1H₃P₃ regimen with those of the 3HR regimen for TPT in adolescents and adults with LTBI. Prior to trial initiation, close contacts of pulmonary TB index cases within school environments will undergo active screening through the TB-YOUTH Screening program, a cross-sectional study (ClinicalTrials.gov NCT06033807), under the auspices of the National Tuberculosis Control Program in China. After excluding active TB, eligible LTBI participants will be grouped into clusters by their corresponding index cases. The clusters will then be randomized in a 1:1 ratio to receive either the 1H₃P₃ or the 3HR regimen for TPT. Medication adherence and adverse events (AEs) will be monitored at week 2, month 1, and month 2 for the 1H₃P₃ group, and at week 2 and then monthly until month 4 for the 3HR group after treatment initiation. Participants will undergo follow-up clinical assessments every 6 months for 2 years following randomization. The preventive efficacy of each study arm against active TB will be evaluated at the conclusion of the follow-up period.

Eligible participants, clustered by epidemiologically linked index cases, will undergo 1:1 randomization through an interactive web response system into 2 arms: the intervention arm will receive a 12-dose 1H₃P₃ regimen, and the active control arm will receive a 90-dose 3HR regimen. Intervention discontinuation should be considered under the following circumstances: (1) at any time upon request by the participant and/or their guardian, (2) major protocol deviation, or (3) an AE deemed by the safety review committee to result in an unfavorable risk-benefit profile for the participant. Interruption of one or more drugs of the assigned regimen consecutively for ≥1 month will be considered permanent discontinuation. Table 1 shows the detailed dosage and administration protocol of study drugs in each regimen. Any dose adjustment is strictly prohibited and will be considered a protocol deviation. All medications will be provided free of charge. TPT other than the assigned intervention is prohibited during the study.

All screened contacts will receive comprehensive TB education (eg, on-site education class, pamphlets, and digital resources) during initial and follow-up screenings. Participants enrolled in TPT will receive a standardized medication diary to log each administration (time, dose, and any AEs). Investigators will review entries prior to dispensing subsequent doses. A patient education booklet summarizing key TPT information (eg, frequently asked questions derived from the pilot study) will be distributed with the diary. Participants within clusters will be encouraged to select a peer coordinator to facilitate group medication adherence and foster a peer-supported compliance network.

Participants in both study groups will undergo prospective monitoring comprising two phases: (1) on-treatment assessment (from enrollment to 1 month after treatment completion) and (2) posttreatment assessment (up to 24 months after randomization). On-treatment assessments will involve standardized clinical evaluations, including physical signs, laboratory testing, medication adherence (pill counts and self-reporting), and safety monitoring, and will be scheduled at week 2, month 1, and month 2 for the 1H₃P₃ arm, and at week 2, then monthly until month 4 for the 3HR group. On the basis of recommendations from the independent protocol review board, and to comprehensively evaluate the impact of interventions on quality of life, a protocol amendment was implemented after approximately 2300 participants had been enrolled. Following ethics committee approval, the EuroQol 5-Dimensions 3-level questionnaire (version 6.0, December 2018) was incorporated as an exploratory end point. This questionnaire will be administered at every on-treatment assessment for participants enrolled after the amendment’s effective date. Posttreatment assessment will focus on active TB surveillance as well as delayed AE monitoring and will occur every 6 months from randomization until study exit. At each visit, participants will receive a standardized symptom screening for active TB, including persistent cough (≥2 weeks), fever, night sweats, weight loss, hemoptysis, or other TB-associated symptoms. Participants with positive symptom screens will undergo confirmatory sputum culture and molecular diagnostic testing. The end point adjudication committee will determine the final active TB diagnosis per predefined criteria.

The schedule of this study is summarized in Table 2.

The primary objective is to demonstrate noninferiority of the ultrashort course of the 1H₃P₃ TPT regimen compared with the standard 3HR regimen in preventing active TB, as measured by the 24-month cumulative incidence of active TB following randomization. The secondary objectives include: (1) to compare safety, patient-reported quality of life, and treatment discontinuation rates attributable to AEs between the regimens; (2) to evaluate treatment adherence using protocol-defined completion criteria; (3) to assess acquired M tuberculosis drug resistance among participants who develop active TB after TPT; and (4) to establish a contemporary nonintervention cohort for exploratory comparison, delineating the natural progression to active TB among individuals with high-risk LTBI in whom active TB has been ruled out.

The primary end point is a composite of bacteriologically confirmed or clinically diagnosed active TB within 24 months after randomization. Bacteriologically confirmed active TB refers to M tuberculosis identification via ≥1 sputum culture and/or molecular assay (eg, GeneXpert MTB-RIF) with detectable bacterial load. Clinically diagnosed TB combines objective clinical symptoms (≥2 of the following: cough ≥2 weeks, fever, night sweats, weight loss >10%, or hemoptysis) and imaging findings highly suggestive of active TB (eg, cavity or nodular lesions on chest CT), after exclusion of alternative diagnoses. Secondary end points include severity-graded AEs, treatment discontinuation (defined as any permanent cessation prior to regimen completion), treatment completion (defined as completing ≥90% of scheduled doses within 110% of the planned treatment duration), all-cause mortality, and acquired drug resistance determined by drug susceptibility testing or targeted molecular assays of M tuberculosis isolates from active TB cases. Exploratory analysis will use the nonintervention cohort to establish the background incidence of active TB, against which the risk reduction of each TPT regimen can be calculated. Patient-reported outcomes (PROs) will be calculated using health state utility values and EuroQol visual analog scale scores.

Enrolled LTBI participants linked to the same index case will be defined as a cluster [14,15], with an assumed average cluster size of 4 participants. Cluster-level randomization will be triggered when ≥1 eligible participant is enrolled within a cluster. Randomization sequences will be pregenerated by the study statistician using SAS (version 9.4; SAS Institute Inc) and will not be accessible to any personnel involved in enrollment, assignment, or intervention delivery. Designated personnel will perform centralized randomization via the interactive web response system. Randomization results will be relayed to the respective site investigators. This is an open-label study, and all participants within the same cluster will receive identical group assignments. To address intercenter heterogeneity, a 2-stage randomization strategy will be applied: in the first stage, stratification by study site will balance baseline characteristics across centers; in the second stage, permuted block randomization (block size=4) within each site will ensure immediate group balance during participant enrollment.

On the basis of surveillance data from China’s National Tuberculosis Control Program, we estimate that the 2-year cumulative active TB incidence among LTBI participants will be 7% [16,17]. Both the 3HR and 1H₃P₃ regimens are assumed to have 70% protective efficacy (ie, reducing the 2-year cumulative incidence of active TB into 2.1%).

The intracluster correlation coefficient (ICC) was incorporated into calculations. A systematic review reported ICC values of 0.028 (range 0.0005‐0.21) for school-based clusters [15], while the National Institutes of Health Pragmatic Trials Collaboratory suggested a range of 0.01 to 0.05 [18]. Therefore, we adopted a conservative ICC estimate of 0.05 for this study.

Using these parameters and a prespecified noninferiority margin of a 20% relative reduction in protective efficacy, a total of 396 clusters (n=1584 participants) per treatment arm are required to achieve 80% power at a 1-sided α=.05. This sample size is calculated to satisfy the clinically meaningful noninferiority margin, accounting for the cluster randomized design (ICC=0.05, average cluster size=4). Accounting for a 10% attrition rate to mitigate the impact of loss to follow-up on statistical power and outcome validity, each arm will enroll 1760 participants.

The trial workflow adheres to the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) statement [19]. Site investigators and staff from each center will undergo centralized training to standardize protocols for participant screening, intervention delivery, follow-up procedures, end point adjudication, and AE documentation and management. Each participant will be assigned a unique anonymized identifier, accessible only to authorized personnel for monitoring purposes. Data will be prospectively captured on paper-based case report forms (CRFs) by on-site investigators or clinical research coordinators, followed by single-entry migration into a customized electronic data capture system. The electronic data capture platform enforces role-based access control to ensure confidentiality. Data integrity audits will be conducted monthly by the data management team. Upon trial completion, the central coordinating office will execute database locks, after which only 2 primary investigators and 1 independent statistician will retain access to the final analytic dataset.

For the primary end point assessment, investigators will perform symptom screening (eg, cough, fever, night sweats, weight loss, and hemoptysis) and physical examination to identify suspected active TB cases at each visit. Participants meeting clinical criteria will undergo confirmatory testing, including chest imaging (preferably chest CT), sputum culture, and sputum molecular testing (eg, GeneXpert MTB-RIF). All suspected cases will undergo centralized adjudication by an independent TB expert panel to confirm the diagnosis.

For secondary end points assessment, AEs will be documented in a dedicated CRF section, capturing event type, onset date, severity (graded according to Division of AIDS criteria), presumed causality, management, and resolution status. Medication adherence will be tracked through participant interviews, drug diary review, and pill counts; discrepancies will be resolved by prioritizing measured residual drug quantities. Survival status will be verified at each visit. Postdropout mortality ascertained through external sources will trigger outcome reclassification from “loss to follow-up” to “deceased.” Drug resistance profiling via molecular assays will be performed exclusively in bacteriologically confirmed active TB cases.

Statistical analyses will be performed using SAS and R (R Foundation for Statistical Computing). Multiple imputation will be conducted to handle missing data, with additional inverse probability weighting applied for the nonintervention cohort to adjust for potential attrition bias. Continuous variables will be summarized as mean (SD) or median (IQR), as appropriate. Categorical variables will be summarized as counts and percentages.

The primary analysis will assess the noninferiority of the 1H₃P₃ regimen compared to the 3HR regimen in preventing active TB over the 24-month follow-up period, and the primary end point is the composite of bacteriologically confirmed or clinically diagnosed active TB. To account for the cluster randomized design and the inherent correlation within clusters, we will use a Cox proportional hazards regression model with shared frailty terms and exact likelihood estimation. Noninferiority of the 1H₃P₃ regimen relative to the 3HR regimen will be concluded if the analysis demonstrates that the 1H₃P₃ regimen does not experience a ≥20% relative reduction in protective efficacy for preventing active TB over the 24-month follow-up period. To enhance precision and address potential baseline imbalances despite randomization, the model will adjust for key prespecified covariates, such as age, sex, BMI, characteristics of the index case (smear or cavitation status), and cluster size. Outcomes will be assessed via 2 prespecified analyses with explicit definitions: the modified intention-to-treat population includes all randomized participants who received at least 1 dose of the study medication, analyzed by the assigned treatment group regardless of subsequent protocol deviations or treatment discontinuation; the per-protocol population includes all randomized participants who completed ≥90% of scheduled doses within 110% of the planned treatment duration without major protocol deviations (including use of prohibited concomitant anti-TB medications, enrollment despite exclusion criteria, or unblinding to outcome assessors). Prespecified subgroup analyses will examine treatment effect heterogeneity by age (13-17 years vs ≥18 years) and index case features (smear or cavitation status), tested through interaction terms in Cox models (interaction P<.10).

For secondary end points, all AEs, serious AEs (SAEs), and treatment discontinuations due to AEs will be documented. Group differences in AE proportions will be analyzed using the chi-square test or Fisher exact test (2-sided α=.05). Treatment completion is defined as consuming ≥90% of planned doses within 110% of the planned treatment duration (≥11 doses within 6 weeks for the 1H₃P₃ group and ≥81 doses within 16 weeks for the 3HR group). The 2-year cumulative incidence of active TB in the nonintervention cohort will be calculated using the Kaplan-Meier method, with 95% CI estimated. The protective efficacy of the 1H₃P₃ and 3HR regimens relative to the nonintervention cohort will be quantified as the risk reduction rate: 1–(intervention group incidence/nonintervention group incidence). The Fisher exact test will be used to compare incidence differences among the 3 cohorts.

All analyses concerning the PRO end point will be explicitly restricted to the subgroup of participants who complete the EuroQol 5-Dimensions 3-level questionnaire. The baseline characteristics of this PRO subgroup will be compared between the 2 arms to confirm that randomization balance was maintained. Any interpretation of PRO findings will explicitly acknowledge their exploratory nature and the fact that they are derived from a subset of the total study population. Trends in PRO metrics over time will be visualized using line graphs, with mean values and 95% CIs plotted to illustrate group-level changes. A mixed-effects model for repeated measures will be used to compare PRO differences between the 1H₃P₃ and 3HR groups across time points, with fixed effects including treatment group, assessment time point, treatment group×time point interaction, age, sex, BMI, and index case cavitation status, and a random participant-specific intercept to account for within-subject correlation from repeated assessments.

AE severity will be graded per the standardized Division of AIDS table (version 2.1) [20]. A safety review committee composed of 3 independent infectious disease experts will oversee safety monitoring and conduct periodic reviews of safety data. All AEs will be systematically assessed for their causal relationship to study medications and categorized as related (adverse drug reaction), unrelated, or indeterminate (insufficient evidence for classification). AEs identified during the study will be promptly documented in the dedicated AE module of the CRF. SAEs necessitate immediate notification: study investigators or site personnel must report SAEs to the local ethics committee and study safety board within 24 hours of detection, followed by comprehensive documentation including severity, causality, and potential corrective actions.

All participating sites are regionally designated medical centers for TB, staffed by experienced clinical teams. Prior to study initiation, all personnel will complete mandatory protocol training, including certification in intervention delivery, data collection standards, and compliance monitoring. Dedicated project managers will provide oversight of drug accountability procedures and on-site auditing to ensure protocol adherence and perform trial conduct monitoring quarterly.

Medication adherence is one of the key assessments for this study. Thus, adherence monitoring uses a 3-strategy verification framework: direct participant reporting during follow-up visits, cross-checking of self-administered medication diaries, and quantitative pill counts of returned drug containers. Within each cluster, peer coordinators will conduct weekly adherence reminders and provide real-time troubleshooting for medication-related challenges.

Participants will be identified and recruited through active screening of close contacts associated with school-based TB index cases. Prior to enrollment, informed consent will be mandatorily obtained, ensuring adequate time for eligible participants and, if needed, their guardians to review study details and voluntarily decide on participation. For individuals under the age of 18 years or those lacking decision-making capacity, written consent will be obtained from their guardians or representatives as well. To standardize ethical practices, the research team will conduct targeted training sessions for all participating health personnel, covering protocols for ethical consent procedures and strategies to minimize participant stress. All staff are required to deliver compassionate support throughout the study and provide psychological counseling if needed. For those who experience harm from study participation, free treatment and appropriate financial compensation will be offered. The study has been reviewed and approved by the ethics committee of Huashan Hospital, Fudan University (2023M-020), and will undergo annual continuing review. Protocol amendments require rereview and approval before implementation. Measures will be taken to protect participant privacy and to ensure confidentiality throughout the study. First, all identifiable personal information (eg, name and phone number) will be separated from other study data once collected, then replaced with unique anonymized identifiers. Second, study data will be stored on an encrypted server that is accessible only to core staff. Third, when obtaining informed consent, participants will be explicitly informed about data anonymization, secured storage, the scope of data use, and the prohibition against data sharing with third parties without participants’ permission. Last, participants will be informed of their rights to withdraw from the study at any time and to request deletion of their data. No identifiable features will appear in any published materials, and all reported results will be aggregated and deidentified.