Study setting
The sampling frame consisted of operational districts (ODs) without ACF interventions currently, and in the preceding six months from the project implementation date. Using data from the Demographic Epidemiological Model of Cambodia developed in collaboration with the National Center for Tuberculosis and Leprosy Control (CENAT)[5] and other health and demographic data from Cambodia[24, 25], eight ODs (Fig. 1) have been purposively selected to include ODs with high and low incidence to maintain as much balance as possible. The selection was also based on the number of health centers to increase comparability and generalizability of study findings across the groups.
Eligibility criteria
In this study, an individual will be defined as presumptive TB if he/she exhibits any of the following symptoms[26]:
- Pulmonary TB: A cough more than two weeks and at least one general symptom
- Extra-pulmonary TB: Presence of symptoms, depending on the location of TB (e.g., enlarged lymph node, swollen backbone, swollen articulation, etc.), and at least one general symptom
- General symptoms: Fever, night sweats for more than two weeks, or unintentional weight loss (>5% reduction in usual body weight over the last six to 12 months)[27]
Presumptive TB cases will be referred to the health centers for TB screening and diagnosis in the intervention arms and self-presented to the health centers in the control arm. We will include the aggregated number of cases diagnosed and notified from all arms regardless of age. However, we will only recruit people aged 18 years or over with TB (all-forms) for the baseline and follow-up survey and assessment of TB treatment outcomes.
Who will take informed consent?
Data collectors (project staff from KHANA, CENAT, CATA, and/or health center staff) will be trained to explain the study to the participants, obtain written consent for the surveys (Additional File 2), and administer the questionnaire. The data collectors will brief participants on the nature and purpose of the study and potential risks and highlight that information will be kept confidential. Consent will be taken by the data collectors in a private space at the respective OD after randomization. For participants who cannot read or write, verbal consent will be taken, and it will be noted by the consent taker in the consent form.
Additional consent provisions for collection and use of participant data and biological specimens
No biological samples are collected for research purposes; hence no additional consent will be sought from the participants. Biological samples will be collected as part of standard care (e.g., TB diagnosis and case management) defined in the national TB guideline under the purview of the national TB program.
Interventions
Explanation for the choice of comparators
Passive case finding (PCF)
PCF strategy is a default setup in the national health system. PCF is a patient-initiated pathway that relies on the self-presentation of presumptive TB cases to the health centers to be diagnosed with TB. Upon screening by health center staff, sputum samples from presumptive TB cases belonging to the KAP will be collected for examination by GeneXpert® or direct smear microscopy if GeneXpert® system is not available. CXR will be performed at the referral hospital if the sputum examination result is negative. TB diagnoses will be made by clinicians based on clinical, radiological, and microbiological evidence. For presumptive TB who presents with cough less than two weeks, CXR will be performed at the referral hospital, and sputum samples will be collected for GeneXpert® testing if radiological abnormalities are detected.[23] For presumptive TB cases that are not KAP, smear examination will be performed if presented with a cough lasting for two or more weeks. Those presented to the health centers with cough less than two weeks will be referred to the referral hospital for CXR, and cases with any radiological abnormalities will undergo smear examination.[23] Newly diagnosed people with TB will be treated and managed according to the national TB guideline. We will follow-up all people with TB diagnosed during the study period for six months from the treatment initiation.
This is a patient-initiated pathway to TB diagnosis involving: (1) a person with active TB experiencing symptoms that he or she recognizes as serious; (2) the person having access to and seeking care, and presenting spontaneously at an appropriate health facility; (3) a health worker correctly assessing that the person fulfills the criteria for suspected TB; and (4) the successful use of a diagnostic algorithm with sufficient sensitivity and specificity to diagnose TB.
Intervention description
Intervention 1: ACF with the seed-and-recruit model
In the intervention clusters, a lay counselor appointed to a health center will act as focal point for identifying potential seeds (TB survivors and their family members, and other key informants in the community such as moto-taxi drivers and grocery store owners), screening presumptive TB cases using a screening questionnaire, and training seeds and recruiters to find other presumptive TB cases in the community. Each lay counselor will be tasked to identify at least five seeds per health center to identify presumptive TB cases (recruits) in their community. These recruits will be screened for TB at the health centers and linked to care if TB was diagnosed. Among the new diagnoses, those who are eligible (Fig. 2) will be trained to become recruiters to find and refer other presumptive TB cases to the health centers in a snowball approach. The research team will work with staff at the health centers to facilitate screening and enrollment of recruits who are diagnosed with TB to care. Eligibility criteria for lay counselors, seeds, and recruiters are illustrated in Additional File 1.
All presumptive TB cases referred by seeds and recruiters will be screened by lay counselors using a symptoms screening questionnaire[17]. The screening for symptoms compatible with TB includes a cough lasting for at least two weeks, hemoptysis, fever, weight loss, and night sweats.[17] Presumptive TB cases and symptoms screening questionnaire will be assessed by a nurse practitioner in the health center. Those who are presumed to have TB will be asked to submit sputum samples at the health center for acid-fast bacilli examination. Sputum samples of KAPs will be assessed using GeneXpert® system at the referral hospital laboratory. For non-KAPs, direct smear microscopy will be undertaken at the nearest laboratory. If the sputum examination results are negative, further assessments and diagnosis of TB will be made on clinical and radiological grounds by clinicians. TB screening and treatment will be provided at no cost to the patient, in accordance with the policies of the NTP. Cases suspected of extrapulmonary TB (presence of local and general symptoms[23]) will be referred to the OD referral hospital for further assessment and diagnosis. We will follow-up all people with TB referred by seeds for six months from the treatment initiation.
Intervention 2: ACF targeting household and neighborhood contacts
In this arm (Fig. 3), community health volunteers will recruit household contacts of people with TB and TB survivors diagnosed in the preceding two years. Immediate neighbors (10 nearest households) of the index cases (people with TB) who are symptomatic will also be invited by the community health volunteers to the screening session. Other KAPs for TB and presumptive TB cases in the community encountered by the community health volunteers will also be invited to the screening session.
The one-off screening session will be held at all the health centers in the selected ODs on specific days. Presumptive TB cases will be screened for TB symptoms on-site, and CXR will be taken. Sputum samples from presumptive TB cases exhibiting features suggestive of TB (on CXR and/or presence of TB symptoms) will be collected for GeneXpert® testing. Test results will be communicated to the newly diagnosed people with TB, and they will be referred to the health centers for treatment and follow-up. Should the health center of their choice do not fall within the selected sites, follow-up (for follow-up survey) will be conducted via phone calls. Cases suspected for extrapulmonary TB will be either diagnosed on-site or referred to the referral hospital for further assessment if needed. We will follow-up on all cases found in this arm for six months from the start of TB treatment.
Intervention 3: ACF targeting the older population (people aged above 55) using mobile screening units
In the intervention clusters (Fig. 4), an outreach team will conduct training and sensitization of the target population prior to the screening session. We will set up a screening site at a designated time, day, and place. Communities in the districts will be informed of the schedule of a one-off mass screening session before the screening day. Each person who visits the screening session will be screened and surveyed at registration by trained staff. Information on their age, demographics, and presence of TB symptoms will be collected. A CXR will then be performed on-site for all persons exhibiting TB symptoms, and all individuals aged ≥ 55 regardless of symptoms. CXR will be assessed by trained CXR readers on-site. When features suggestive of TB are found, sputum samples will be collected for GeneXpert® testing on-site as well. Test results will be communicated to the participants immediately or via phone calls, and people with TB will be referred for treatment and follow-up at the health center where screening is conducted or a health center of their choice. TB treatment and care at the health center will be provided according to the guidelines of the NTP. Should the health center of their choice do not fall within the selected sites, follow-up will be conducted via phone calls. Cases suspected for extrapulmonary TB will be referred to the referral hospital for further assessment and diagnosis. A follow-up will be conducted six months from treatment initiation. The standard operating protocol of mobile screening sessions depicts that the intervention will only be conducted in the selected sites once in 1–3 years.
Criteria for discontinuing or modifying allocated interventions
Since the interventions implemented in this study are part of standard procedures endorsed by the national TB program in Cambodia, the activities will not adversely affect the welfare of the participants and pose no more than minimal risk. Therefore, we do not foresee discontinuation or modification of allocated interventions. In this study, participation in the activities are voluntary and the participants can withdraw from the study at any point in time.
Strategies to improve adherence to interventions
In this study, the implementing organizations will design and ensure that the interventions (TB case finding) protocols are strictly adhered. The medical procedures and laboratory tests (e.g., GeneXpert, chest x-ray, and sputum microscopy) conducted in this study are in accordance with the national TB guideline. In addition, the National University of Singapore that is not an implementing organization will conduct regular field visits and interim analyses to ensure adherence to study protocol.
Relevant concomitant care permitted or prohibited during the trial
No concomitant care would be specifically permitted or prohibited during the trial due to the pragmatic nature of the study. Interventions will be conducted at the community level according to standard operating procedures in a ‘real-world’ setting.
Provisions for post-trial care
Post-trial care will be provided in accordance with the standard care as defined in the national TB guideline under the purview of the national TB program.
Outcomes
We will measure the following endpoints in all four arms. The primary endpoint will be the 1) case notification rate in the intervention and control ODs during the study period. We define a TB case to be either bacteriologically confirmed (smear-positive identified by smear microscopy, culture or WHO-approved rapid diagnostics), smear-negative TB (two sputum samples tested negative and clinically diagnosed by a clinician – with abnormal CXR or no improvement in response to a course of broad-spectrum antibiotics) or extrapulmonary TB (presence of local symptoms such as enlarged lymph nodes, joint swelling, and Gibbus deformity and presence of general symptoms such as cough, fever, weight loss, and night sweats).[23] This will be ascertained from the results of the screening and routine case notification data reported to the NTP. We will also determine 2) additionality, comparing the yield in each arm with its respective historical baseline (same period in the preceding year) and 3) cumulative yield over the project implementation period. Secondary endpoints will include 4) the number of people needed to screen to detect one TB case, 5) direct and indirect costs per TB case notified, and 6) the treatment outcome of all people with TB in this study. These data will be collected in the aggregated form at the health center level or as reported by the project implementation team in each arm, except cost data, which will be collected from study participants in the baseline and follow-up survey.
Participant timeline
| Study period |
Timepoint | Enrolment | Allocation | Post-allocation |
‘19 | ‘19 | ‘19 | ‘20 | ‘20 | ‘20 | ‘20 | ‘21 | ‘21 |
Q3 | Q3 | Q4 | Q1 | Q2 | Q3 | Q4 | Q1 | Q2 |
Enrolment (cluster level) | | | | | | | | | |
Assessment of cluster | X | | | | | | | | |
Randomization of cluster | X | | | | | | | | |
Allocation of intervention | | X | | | | | | | |
Interventions | | | | | | | | | |
Active case finding with a seed-and-recruit model | | | X | X | X | X | X | | |
Active case finding targeting household and neighborhood contacts | | | X | X | X | X | X | | |
Active case finding targeting people aged 55 and above using mobile screening units | | | X | X | X | X | X | | |
Passive case finding | | | X | X | X | X | X | | |
Assessments (cluster level) | | | | | | | | | |
TB case notification by selected clusters | | | X | X | X | X | X | | |
TB cases reported by each arm | | | X | X | X | X | X | | |
Individuals screened for TB | | | X | X | X | X | X | | |
Treatment outcomes | | | | | X | X | X | X | X |
Costs data | | | X | X | X | X | X | X | X |
Assessments (individual level) | | | | | | | | | |
Enrolment | | | X | X | X | X | X | | |
Baseline survey* | | | X | X | X | X | X | | |
Follow-up survey† | | | | | X | X | X | X | X |
Sample size
Simulations were conducted to estimate the power of the trial to detect a statistically significant difference between the intervention arms and the control arm for the primary outcome of the trial. To develop a basis for these simulations, we examined the ranges and variability of effect sizes from previously collected data that included operations data from Cambodia Anti-Tuberculosis Association (CATA), KHANA, CENAT, and the NTP annual TB report 2018. We defined effect size as the percentage increase in the number of cases notified to the NTP because of the intervention (baseline: PCF). KHANA, CATA, and CENAT have implemented the respective interventions (different strategies of ACF) at different sites in Cambodia between 2017 and 2018. The percentage of estimated additional cases notified to the NTP by KHANA, CATA, and CENAT during the intervention period was (1) 98%, 110%, and 30%, respectively. For this scenario, the power was over 90%. We also considered three more sets of conservative effect sizes: (2) assuming all intervention arms have an effect size of 75%, (3) assuming all intervention arms have an effect size of 50%, and (4) assuming all intervention arms have an effect size of 25%. There was a similar level of power for a more cautious estimate (scenario (2) - a 75% improvement in case notification). In scenario (3) and (4), the power was maintained above 70% that a difference between intervention and control arms could be detected.
Recruitment
The trial will be conducted at the selected operational districts, and all communities and individuals who are eligible will be invited to participate in the activities.
Assignment of interventions: allocation
Sequence generation
The unit of randomization will be the ODs (clusters). From the list of underserved ODs in Cambodia provided by CENAT, we will randomly allocate four high TB incidence ODs and four low TB incidence ODs to the intervention and control arms matched by the population size and the number of health centers.
Concealment mechanism
Randomization will be performed by a person who will not be otherwise engaged in the project and be blinded to the identity of the districts. Project coordinators and residents in the selected ODs will not be masked to the intervention. All case finding activities share the same objective – to find TB cases – and they will be done without reference to the intervention group.
Implementation
The population living in the selected OD will be included in the respective cluster. There will be four arms (Fig. 1), and two clusters will be randomly allocated to each arm. As all interventions have been endorsed by the CENAT and conducted in Cambodia, participants will not be notified if they are part of the intervention or the control group.
Assignment of interventions: Blinding
Who will be blinded
The data analysts will be masked to intervention allocation and will only analyze de-identified data.
Procedure for unblinding if needed
Not applicable.
Data collection and management
Plans for assessment and collection of outcomes
For the primary endpoints, we will collect case notification data from the intervention implementers (KHANA, CENAT, CATA), the provincial health department, and health centers (Table 1). For the secondary endpoints, data on treatment outcomes will be collected from the health centers. To assess the number needed to screen, we will collect screening data and the number of TB cases diagnosed from intervention implementers, the provincial health department, and health centers. Data on costs – direct and indirect medical costs,[28, 29] intervention, diagnostics, and medications – will be collected from the intervention implementers and the baseline and follow-up survey (Table 1). In the survey, we will document the baseline characteristics of new diagnoses found in all arms using a questionnaire. Information on age, gender, education level, occupation, number of family members and housing, smoking, alcohol use, drug use, TB history, and the presence of other co-morbidities will be collected. We will adapt and contextualize the knowledge, attitude, and practice survey instrument tools developed by the World Health Organization to collect information on TB knowledge, attitudes, and care-seeking behavior towards TB.[30] A short follow-up survey will be conducted at the end of the TB treatment period for each person with TB to collect information on missed doses, side effects, and other TB treatment-related costs incurred.
Table 1
Endpoints and data collection
Study endpoints | Description | Data source |
Primary endpoints |
Case notification rates (cases notified per 10000 population per year) | Numerator: The number of cases notified by the selected districts | Case notification data from CENAT |
Denominator: Total population in the OD | Population statistics (Ministry of Planning/Department of Statistics/CENAT) |
Additionality (additional number of cases reported compared to historical baseline) | The number of cases notified by the selected districts | Case notification data from CENAT |
Historical data (cases notified in the preceding 3 years) | Historical case notification data from CENAT |
Cumulative yield (cases diagnosed per 1000 screened) | Numerator: The number of cases reported by each arm | Program data In the control arm, data to be collected from the health centers monthly |
Denominator: Total number of individuals screened | Program data Provincial health department laboratory data to determine the number of people screened at health centers in the control arm |
Secondary endpoints |
Treatment outcomes | TB treatment outcomes of all new patients 6 months after treatment initiation | Health centers |
Number needed to screen | Numerator Total number of individuals screened | Program data Provincial health department laboratory data to determine the number of people screened at health centers in the control arm |
Denominator: The number of cases reported by each arm | Program data Control arm: data to be collected from the health centers monthly |
Cost per TB case diagnosed/notified | Direct and indirect medical costs while seeking care for TB Staff and intervention costs Diagnostics and medication costs | Health care seeking costs: Data to be collected in the baseline and follow-up survey Intervention costs: Data to be collected from KHANA, CENAT, and CATA DALY: Data to be extracted from WHO global burden of disease studies and other existing literature |
Incremental cost-effectiveness ratio per DALY averted |
OD, operational district; CENAT, National Center for Tuberculosis and Leprosy Control; TB, Tuberculosis; CATA, Cambodia Anti-Tuberculosis Association; DALY, Disability-adjusted life year; WHO, World Health Organization |
Data collectors (project staff from KHANA, CENAT, CATA, and/or health center staff) will be trained to explain the study to the participants, obtain written consent for the surveys (Additional File 2), and administer the questionnaire. The data collectors will brief participants on the nature and purpose of the study and potential risks and highlight that information will be kept confidential. Consent will be taken by the data collectors in a private space at the respective OD after randomization. For participants who cannot read or write, verbal consent will be taken, and it will be noted by the consent taker in the consent form.
Data collectors will administer the questionnaire (both questionnaire 1 – baseline survey and 2 – follow-up survey) on-site using a tablet-administered questionnaire. The baseline survey will take between 30 and 40 minutes to complete, and the follow-up survey will take approximately 30 minutes to complete. Surveys may be conducted via phone call for participants who are unable to visit the health center for interviews. Refusals will be excluded from this study, but consent will be sought to record demographic characteristics (participants would only be completing the “Socio-Demographics” section of the questionnaire if they have signed the Consent Form indicating consent to participate in the study, but later choose not to complete the questionnaire). Participants will be reimbursed USD 2 for their participation at baseline (questionnaire 1) and an exit interview (questionnaire 2) at the end of the six-months follow-up.
Data coding, quality control, and data entry will be done following established procedures at KHANA and Saw Swee Hock School of Public Health, National University of Singapore. A database will be developed for data entry in KOBO toolbox[31], which will include a built-in range, and within and between variable consistency checks. The program will also run error-checking applications and produce reports of inconsistencies to be checked daily. The database will be exported into excel to check for consistency and to clean before data analyses. All data will be entered into the database within one month of the collection. For baseline data, a pre-coded questionnaire will be used to minimize data coding errors. All data forms and questionnaires will be checked for errors by the field supervisors, and necessary corrections will be made during data entry.
Plans to promote participant retention and complete follow-up
Refusals will be excluded from this study, but consent will be sought to record demographic characteristics (participants would only be completing the “Socio-Demographics” section of the questionnaire if they have signed the Consent Form indicating consent to participate in the study, but later choose not to complete the questionnaire).
Data management
Data coding, quality control, and data entry will be done following established procedures at KHANA and Saw Swee Hock School of Public Health, National University of Singapore. A database will be developed for data entry in KOBO toolbox[31], which will include a built-in range, and within and between variable consistency checks. The program will also run error-checking applications and produce reports of inconsistencies to be checked daily. The database will be exported into excel to check for consistency and to clean before data analyses. All data will be entered into the database within one month of the collection. For baseline data, a pre-coded questionnaire will be used to minimize data coding errors. All data forms and questionnaires will be checked for errors by the field supervisors, and necessary corrections will be made during data entry.
Confidentiality
Data collectors (project staff from KHANA, CENAT, CATA, and/or health center staff) will be trained to explain the study to the participants, obtain written consent for the surveys (Additional File 2), and administer the questionnaire. The data collectors will brief participants on the nature and purpose of the study and potential risks and highlight that information will be kept confidential. Consent will be taken by the data collectors in a private space at the respective OD after randomization. For participants who cannot read or write, verbal consent will be taken, and it will be noted by the consent taker in the consent form.
Plans for collection, laboratory evaluation and storage of biological specimens for genetic or molecular analysis in this trial/future use
Not applicable
Statistical methods
Statistical methods for primary and secondary outcomes
For the primary endpoints, we will compute TB case notification at the cluster level by dividing the total number of TB cases (all forms) notified by the selected OD by the total population in each OD. Case notification rates will be presented as TB cases notified per 10,000 population per year. The difference in the total number of TB cases notified during the intervention period and the number of TB cases notified in the preceding three years will be computed to determine additionality. We will calculate the cumulative yield of TB cases by dividing the number of cases notified in each arm by the total number of people screened for TB. Cumulative yield will be presented as TB cases notified per 1000 individuals screened.
For the secondary endpoints, we will evaluate the treatment outcome of TB cases detected in this study according to the following categories – treatment completed (without evidence of failure), treatment failed (sputum smear remains positive at treatment-month 5 or later), died (death due to any reasons before and during treatment), lost to follow-up (people newly diagnosed with TB who did not initiate treatment or treatment course that was disrupted for two or more successive months), not evaluated (no treatment outcome data available), cured (sputum smear converted from positive to negative in the final month of treatment and on at least one earlier assessment. The number needed to screen to detect one TB case in each arm will be calculated by dividing the total number of persons screened by the number of TB cases detected, i.e., the reciprocal of the absolute cumulative yield. Direct and indirect medical costs that will incur before and during health-seeking will be measured based upon the baseline and follow-up questionnaire. Costs for identification of presumptive TB cases and engagement with KAPs for screening, including training and meetings before initiating ACF activities, will be categorized separately from the costs for diagnosis using radiological/bacteriological means and management of people with TB. Incremental cost-effectiveness ratio (ICER) will be estimated by dividing the difference in total costs between trial arms by the difference in the total number of TB cases detected. ICER per disability-adjusted life year averted will also be considered as a benchmark to determine if the interventions are cost-effective.
We will analyze the primary and secondary endpoints by intention to treat, and the analyses will be conducted in two stages. First, we will analyze and present primary and secondary endpoints using the cluster summary method. The baseline characteristics of all arms at the cluster level and individual-level characteristics at baseline and end-line will be compared using student’s t-test and hierarchical or mixed-effect models to estimate the ratio of these means. As the ODs have several health centers, we will adopt a linear mixed-effects model adjusting for baseline measurements to account for clustering in the data. Individual-level data will also be incorporated in multivariable regression models to account for potential confounders differing by the trial arm and improve the precision of the risk ratios by accounting for inter-cluster variations. The impact of TB case finding strategies on treatment outcomes will be evaluated using survival analysis to compare treatment success rates across the intervention and control arms. The study will be reported according to the CONSORT[32] statement for the reporting of cluster randomized controlled trials, the TIDieR[33] checklist for intervention description and replication, the SPIRIT[34] statement, and PRECIS-2[35] for pragmatic trials (Additional File 3–7). All statistical analyses will be conducted using STATA 14 (Stata Corp LP, Texas, United States of America) and R (R Foundation for Statistical Computing, Vienna, Austria).
Interim analyses
Interim analyses will be conducted by the data analyst who is otherwise masked to intervention allocation.
Methods for additional analyses (e.g. subgroup analyses)
The baseline characteristics of all arms at the cluster level and individual-level characteristics at baseline and end-line will be compared using student’s t-test and hierarchical or mixed-effect models to estimate the ratio of these means. As the ODs have several health centers, we will adopt a linear mixed-effects model adjusting for baseline measurements to account for clustering in the data. Individual-level data will also be incorporated in multivariable regression models to account for potential confounders differing by the trial arm and improve the precision of the risk ratios by accounting for inter-cluster variations.
Methods in analysis to handle protocol non-adherence and any statistical methods to handle missing data
We will adopt intention-to-treat analyses, and imputation methods will be considered in the event of missing data.
Plans to give access to the full protocol, participant level-data and statistical code
Information from the full protocol will be published in a peer-reviewed journal.
Oversight and monitoring
Composition of the coordinating centre and trial steering committee
The coordinating center is based in the KHANA Center for Population Health Research. The steering committee comprises of the representatives from KHANA, the Cambodia Anti-Tuberculosis Association, the National Center for Tuberculosis and Leprosy Control, and the Saw Swee Hock School of Public Health National University of Singapore. Data will be managed by the team at KHANA, with inputs from the other two implementing organizations. The research team from the Saw Swee Hock School of Public Health National University of Singapore will be responsible for the analysis of the data from the trial. The steering committee will adjudicate the endpoints and develop the findings dissemination plan in consultation with the data monitoring committee.
Composition of the data monitoring committee, its role and reporting structure
The data monitoring committee (DMC) comprises of the representatives from KHANA Center for Population Health Research and the Saw Swee Hock School of Public Health National University of Singapore. The DMC will be formed independent from the funders. The DMC will be responsible for the independent assessment of the validity and integrity of the cluster randomized controlled trial. Interim analyses will be presented to the DMC. The DMC will report trial progress and the results of the trial steering committee. The DMC will meet twice a year or more if needed.
Adverse event reporting and harms
There are no major perceivable risks in this study. The potential psychological distress because of the questions that the study poses are minimal. Participation is purely voluntary, and the decision to not participate in this study or terminate their participation at any point during the study will not have any negative consequences. All participants and those who refused to partake in the survey will not be deprived of the standard care that they ought to receive for TB. As the interventions that we are evaluating are operations that have been implemented and endorsed by the national TB program, the risks have been thoroughly deliberated. The interventions involve only models of care to reach people with TB. TB screening, diagnosis, and treatment algorithms included in these models are per standard protocol in Cambodia.
Frequency and plans for auditing trial conduct
The research team will conduct at least two monitoring visits to each OD during the one-year implementation period. Field supervisors will also conduct at least four monitoring visits (e.g., every quarter for KHANA arms where the intervention is spread over one year and at every screening session conducted by CATA and CENAT during the implementation period, respectively). During monitoring visits, we will review consent forms, completeness of the data collection forms, and compliance with the trial protocol. Any anomalies in recruitment and data will be followed up by the field staff. Interim analyses will be conducted by the data analyst who is otherwise masked to intervention allocation. Data, recruitment processes, and interim results will be monitored by the research team and reported to the DMC.
Plans for communicating important protocol amendments to relevant parties (e.g. trial participants, ethical committees)
Amendments to the protocol will be duly communicated to all relevant parties including the national TB program and all partners, investigators, ethics board in Cambodia and Singapore, trial registry, and journals.
Dissemination plans
We will share the findings from this study in results dissemination meetings with key stakeholders and peer-reviewed journals. Findings from this trial could also potentially inform ACF strategies in other countries with a high TB burden.