Two-phase study to optimise diagnosis and treatment for children and adolescents with tuberculosis in Namibia: knowledge, attitudes and practices of healthcare workers before and after the rollout of updated child and adolescent tuberculosis clinical guidelines

Abstract

Introduction

Child and adolescent tuberculosis (CATB) poses a significant global burden with an estimated 1.25 million diagnoses and 191 000 deaths annually, disproportionately higher than adult rates. Over half of CATB cases remain unreported, and burden increases are expected following the USA’s tuberculosis (TB) aid withdrawal, particularly affecting high-burden countries including Namibia. Namibian children under the age of 15 years account for 10% of notified TB cases, with under-reporting suggesting higher actual morbidity and mortality. Through a multistakeholder process in 2024, Namibia’s National Tuberculosis and Leprosy Programme (NTLP) updated its National Clinical Guidelines for CATB management. We aim to (1) assess baseline levels of CATB diagnosis and management knowledge and (2) evaluate post-training knowledge, attitudes and practices (KAP) with updated guidelines to identify gaps informing targeted interventions and policy refinement.

Methods and analysis

We plan to conduct a two-phase evaluation of healthcare workers’ (HCWs) KAP regarding CATB diagnosis and management across Namibia. A collaborative team from Dartmouth, the University of Namibia (UNAM) and NTLP developed, piloted and refined a standardised assessment tool. 256 HCWs involved in all TB care levels are selected at three facilities per district across 16 districts stratified by TB burden (high, medium and low). Baseline data (KAP1) were collected before national training on updated guidelines, with follow-up assessment (KAP2) planned at 6 months post-training using the same procedures on a secure Research Electronic Data Capture (REDCap) platform. Descriptive analyses and regression models will summarise baseline KAP, while McNemar tests and paired t-tests will evaluate changes over time for participants who completed both KAP assessments.

Ethics and dissemination

Ethics approval was obtained from the UNAM and the Ministry of Health and Social Services (MoHSS). Written informed consent is obtained from all participants, protecting confidentiality and using secure data storage. Findings will be disseminated through Namibian MoHSS reports, conference presentations, peer-reviewed publications and policy briefs.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Comprehensive national coverage across diverse districts and healthcare worker cadres using a standardised, pilot-tested tool.

• Pre–post training design enables evaluation of guideline uptake.

• Participant replacement at follow-up and the 6-month follow-up limits individual-level comparisons and assessment of long-term retention and sustained practice change.

• Self-reported practices may introduce reporting bias.

• Pre–post design without a control group cannot distinguish training effects from secular trends or general clinical experience gained over the 6-month period.

Introduction

Globally, child and adolescent tuberculosis (CATB) represents a significant burden resulting in 1.25 million incident cases and 191 000 deaths for children aged 0–14 years annually.¹ While this represents 12% of overall tuberculosis (TB) cases, child and adolescent deaths represent 16% of overall TB deaths, evidencing a disproportionately higher mortality rate in reported cases in this age group. Furthermore, an estimated 52% of CATB diagnoses for patients under the age of 5 years remain unreported,² over 27% higher than the estimated 25% of adult patients with unreported TB.³

Because the USA withdrew bilateral TB aid in 2025, projections of CATB burden are expected to rise with an estimated 22% increase in patients and 31% increase in deaths over the next decade. This burden is projected to disproportionately impact high-TB burdened countries, comprising approximately 95% of cases and deaths.⁴

In Namibia, one of the WHO-designated high TB-burden countries, children and adolescents under the age of 15 years represent 10% of the total reported TB cases. In 2023, 914 children and adolescents were diagnosed with TB, including regional CATB proportions as high as 18% in the district Khomas.⁵ The same year, an estimated 2911 TB-related deaths occurred which translates to a rate of 98 per 100 000 population. Although complete age-disaggregated data on TB deaths were not available, an under-reporting estimate of 33% suggests CATB morbidity and mortality are likely higher than reported rates.^{5 6} Data from neighbouring South Africa reveal persistently high mortality rates among children (6.8%)⁷; in one study, regression models showed that children aged 0–4 years and adolescents aged 10–19 years are at increased risk of TB-related death.⁸

The Namibian National Tuberculosis and Leprosy Programme (NTLP) developed updated National Clinical Guidelines for Management of TB in Children and Adolescents through a multistakeholder, consultative 9-month process to adapt the 2022 WHO consolidated guidelines and operational handbook to the Namibian context. Nationwide training of trainer sessions began in November 2025. Immediately preceding the rollout of the training cascade, we conducted a baseline knowledge, attitudes and practice (KAP) survey of HCWs (KAP1) with the goals to guide revisions to the training curricula and inform content emphasis, and determine baseline knowledge of healthcare workers (HCWs) for CATB management and treatment.

To identify uptake efficacy of the updated training, we will conduct a second KAP assessment (KAP2) to the same HCWs who participated in KAP1 6 months after the rollout of the training cascade. In figure 1, the study flow is illustrated. The difference between baseline and post training identified in KAP1 and KAP2 will show the efficacy of training and be used to guide targeted follow-up training and/or clinical supervision activities.

Figure 1. Study timeline. Intended study activities from April 2025 through September 2026 are shown. KAP1 includes protocol development (April–September 2025), IRB submission and approval (September–November 2025) and pilot and data collection (November–December 2025). The new guidelines will be implemented from December 2025 through June 2026. KAP2 includes refining the KAP2 assessment tool and staff retraining (May–June 2026), data collection (June–July 2026) and data cleaning and analysis (July–September 2026). MoHSS reporting includes a progress update submitted in November 2025, an initial report to be submitted in June 2026 and a final report to be submitted in September 2026. IRB, institutional review board; KAP, knowledge, attitudes and practices; MoHSS, Ministry of Health and Social Services.

Methods and analysis

Following the development of the updated child and adolescent TB guidelines, the Ministry of Health and Social Services (MoHSS) guidelines team invited researchers from the Geisel School of Medicine at Dartmouth, the University of Namibia (UNAM) School of Medicine and the NTLP to design this study. Two community-based non-governmental organisations (NGOs) with field staff who have both TB care and research experience and nationwide coverage were engaged to facilitate assessment distribution.

Assessment tool

A standardised survey was developed collaboratively and iteratively to assess the KAPs of HCWs regarding TB diagnosis and treatment for children and adolescents under the age of 15 years (CA). External experts were also consulted and their feedback incorporated. The assessment tool was piloted in October 2025 among 11 HCWs across various settings: 5 in two urban clinics in Windhoek, 2 in a periurban clinic in Rehoboth and 4 in a rural clinic outside Windhoek. Results of the pilot assessment informed modifications of the assessment tool to maximise clarity, eligibility and brevity. The final assessment tool was created and administered in English, the sole official language of Namibia (see online supplemental material 1).

Patient and public involvement

The target population for this study is frontline HCWs who diagnose, treat and manage children and adolescents with presumptive or confirmed TB. Therefore, to ensure the survey tool would be responsive to both provider training needs and the TB care needs of paediatric and adolescent patients in Namibia, we sought input from a broad range of providers and community-based health workers during its development. This included feedback from colleagues at our two community-based NGOs, Regional Directors and Managers and rural and urban HCWs at non-survey sites during the piloting phase.

Study findings will be disseminated to the contributing parties of MoHSS/NTLP, Regional Directors and Managers and participating health facility directors and interviewees. We anticipate producing one-page summaries of key findings and recommendations, though we will engage the contributors listed above in determining the most appropriate dissemination formats and channels.

Study site and participant selection

Using NTLP aggregated CATB notifications between 2022 and 2024, 16 districts were selected for investigation. These districts were selected to represent the full range of TB burden of high-burden areas such as Gobabis (18.1%) and low-burden districts such as Tsandi (2.3%) in order to identify regional differences. TB burden was determined through absolute CA notifications and per cent of total notifications that were among CA. Practical accessibility to sites was also considered, including driving distances between major city centres and assessment sites, enabling multiple site assessments to occur over the course of a single trip. Within each selected district, three facilities were then chosen based on accessibility and their ability to fulfil the required facility-type and healthcare cadre requirements; where multiple facilities met these criteria, the closest accessible facilities were selected.

For each of the 16 assessment districts, three health facilities were selected, each including a hospital, health centre and primary healthcare clinics, covering the range of potential HCW cadres involved in TB care and prevention with a focus on those with clinical care decision-making authority. The organisation and scope of TB service delivery across primary, intermediate and hospital levels are summarised in table 1.

Table 1. Tuberculosis service delivery across levels of care in Namibia.

Healthcare facility type	Core functions	Enhanced capacity and support
Primary healthcare clinics	Routine outpatient TB screening, specimen collection and transport for testing, diagnosis, treatment initiation, TPT, follow-up and adherence monitoring	In remote settings, selected clinics diagnose and manage both drug-susceptible and DR-TB.
Health centres	Routine outpatient TB screening, specimen collection and transport for testing, diagnosis, treatment initiation, TPT, follow-up and adherence monitoring	Selected centres provide expanded capacity to manage drug-susceptible and DR-TB, particularly in underserved areas.
Hospitals (district, intermediate, specialised)	Comprehensive TB screening, diagnosis, treatment initiation, TPT	Tiered referral system: district hospitals provide first-level inpatient and outpatient TB care; intermediate hospitals manage more complex cases and complications; specialised (referral) hospitals provide advanced care for severe or complicated TB, including DR-TB. Hospitals are supported by on-site GeneXpert Ultra testing, centralised culture and DST and CAD for chest X-ray at selected facilities.

CAD, computer-aided detection; DR-TB, drug-resistant TB; DST, drug susceptibility testing; TB, tuberculosis; TPT, TB preventive therapy.

At each participating facility, the study team worked with facility leadership to identify eligible HCWs available during the data-collection period. Facility managers and nurse in-charges assisted in coordinating access to eligible staff members. Participants were selected through facility-based recruitment from among available eligible HCWs at each site, adhering to the cadre-level targets. HCWs identified through this process were invited to participate in the survey; if the selected participant declined, the next available eligible participant was invited. This approach was applied consistently across all participating facilities.

Some health facilities did not have sufficient HCWs on site during assessment times to meet the HCW target; in such cases, the deficit number of assessments were conducted at the next health facility selected for study inclusion within the district. The same selected districts and respective health facilities will be investigated for both KAP1 and KAP2, ideally allowing for the same HCWs to be selected for assessment. Selected districts and health facilities are shown in table 2.

Table 2. Selected health facilities by district and relevant characteristics that determined study inclusion.

District	U15/%U15 notifications (2022–2024)	Distance from Windhoek (hours)	Facility name and target number of HCW interviewees
Eenhana	35/6.4	7.25	Eenhana District Hospital - 6 Oshandi Clinic - 5 Onangolo Clinic - 5
Gobabis	434/18.1	2.5	Gobabis District Hospital - 6 Epukiro Post 3 Clinic - 5 Otjinene Health Centre - 5
Grootfontein	176/20.8	4.5	Grootfontein District Hospital - 6 Mangetti Dune Health Centre - 5 Tsumkwe Clinic - 5
Keetmanshoop	100/10.4	4.75	Keetmanshoop District Hospital - 6 Bethanie Health Centre - 5 Keetmanshoop/Tseiblaagte Clinic - 5
Mariental	78/8.8	2.75	Mariental District Hospital - 6 Maltahohe Clinic - 5 Kalkrand Clinic - 5
Okahao	6/2.5	7.5	Okahao District Hospital - 6 Otamanzi Clinic - 5 Indira Gandhi Health Centre - 5
Omuthiya	7/2.4	5.5	Omuthiya District Hospital - 6 Onanke Clinic - 5 Onkumbala Clinic - 5
Oshikuku	24/4.6	7.5	Oshikuku St. Martin District Hospital - 6 Okalongo Health Centre - 5 Onaanda Clinic - 5
Otjiwarongo	33/8.9	2.25	Otjiwarongo District Hospital - 6 Orwetoveni Clinic - 5 Otavi Health Centre - 5
Outapi	49/7.4	7.75	Okahao District Hospital - 6 Oshaala Clinic - 5 Mahenene Health Centre - 5
Outjo	28/10.3	3	Outjo District Hospital - 6 Thomas Adam Sheya Clinic - 5 Kamanjab Health Centre - 5
Rehoboth	48/5.6	1	St. Mary’s Rehoboth District Hospital - 6 Rehoboth Health Centre - 5 Klein Aub Clinic - 5
Rundu	174/18.1	7	Rundu Intermediate Hospital - 6 Rundu Clinic - 5 Mashare Clinic - 5
Tsandi	4/2.3	7.5	Tsandi District Hospital - 6 Onesi Health Centre - 5 Omakange Clinic - 5
Walvis Bay	161/11.2	4	Walvis Bay District Hospital - 6 Kuisebmund Health Centre - 5 Narraville Clinic - 5
Windhoek	523/12.7	0	Katutura Intermediate Hospital - 6 Okuryangava Clinic - 5 Katutura Health Centre - 5
Total: 16 districts			256 HCW interviews across 48 facilities

HCW, healthcare worker; U15, Number of TB patients under 15 Years of Age; %U15, Proportion of Patients Under 15 Years of Age out of Total Number of TB Patients.

The target sample of 256 HCWs was determined pragmatically to ensure representation across all three district rurality types, three levels of care and major HCW cadres involved in CATB management. 16 participants were targeted per district (approximately 6 from the district hospital, 5 from a health centre and 5 from a primary healthcare clinic). A formal a priori sample size calculation was not performed because the primary objective was to obtain a logistically feasible, resource efficient and nationally representative baseline and follow-up assessment across implementation sites, rather than to test a predefined hypothesis.

Data collection

KAP1

The 24 field-staff involved in data collection participated in a 3-day in-person training prior to starting data collection. The training curriculum outlined a general overview of paediatric TB epidemiology, transmission, diagnosis and treatment; key updates to the new guidelines; and familiarisation to both the assessment tool and interview process, including environmental simulation and use of the electronic tablets for data entry.

The MoHSS and UNAM ethics approval were communicated to regional management teams, and a virtual meeting was held to provide an overview of the study protocol, methodology and timelines. This information was cascaded to district and health facility management teams. Field teams conducted courtesy meetings with regional and district management teams in addition to sharing hard copies of the ethics approvals. Within the three-person field teams, the field staff supervisors oversaw data collection and reviewed the data at the end of each day to validate for completeness.

Data were entered electronically at the time of the interview using a secure, password-protected tablet. The survey interface on the tablet was created using the latest version Research Electronic Data Capture (REDCap) application, a secure, web-based software platform designed to support data capture for research studies, which is also password protected. When technical difficulties with our electronic data collection system required use of hard copies of the data, these were stored in secure, lockable cabinets or lockable portable cases.

Following completion of KAP1, training with the updated guidelines was planned nationwide including across the 16 assessment districts. KAP2 data collection will be scheduled. Minor edits will be made to create the KAP2 assessment tool during this timeframe.

KAP2

At least 6 months after rollout of the updated CATB guidelines to allow HCWs to integrate the knowledge and practices in the updated guidelines, KAP2 will be conducted in the same 16 districts and health facilities to reassess HCWs’ knowledge, attitudes and practices using the same standardised tool and procedures as KAP1. Field staff will complete a 3-day refresher training. Teams will make every effort to re-interview the same HCWs who participated in KAP1 to permit paired longitudinal analyses. If an original participant is unavailable, a replacement from the same cadre within the same facility will be recruited to maintain the planned sample size and representativeness of the cross-sectional assessment. All responses will be uploaded daily to the secure REDCap server following supervisor validation of completeness.

Analysis

Statistical tests

Descriptive analyses of KAP1 data will be conducted using Stata V.18.0 SE (Standard Edition). Data will be aggregated using REDCap and Microsoft Excel and summarised according to cadre, facility type and district/region. Survey items were classified a priori into the three KAP domains: knowledge, attitudes and practices. Knowledge items were further categorised into clinical presentation, diagnosis, treatment and TB infection/preventive therapy (TPT/latent TB infection) subdomains based on their alignment with recommendations in the updated CATB Clinical Guidelines (see online supplemental material 2). Attitude and practice items were similarly categorised according to the clinical construct being assessed (see online supplemental material 3). The prespecified classification of individual survey items, scoring rules and definitions of guideline-concordant responses are also provided in the respective supplemental materials. The KAP domains, subdomains, item classifications, scoring algorithms and thresholds for interpretation were all defined before completion of the primary analyses and will be applied consistently to both KAP1 and KAP2 assessments.

Several survey items are measured using Likert-scale response options and will be analysed as ordinal outcomes. For descriptive analyses, item-level response distributions will be summarised, including the proportion of participants providing responses considered correct or guideline-concordant. These proportions will be categorised using predefined Bloom’s cut-off criteria, with satisfactory performance defined as ≥80% of participants providing correct/guideline-concordant responses and unsatisfactory performance defined as <80%. A secondary classification will categorise item-level performance as high/good (80–100%), moderate/neutral (60–79%) or low/poor (<60%).

For inferential analyses, the original Likert-scale responses will be retained as ordinal outcomes. Associations between HCW cadre, facility type and location and individual survey item responses will be assessed using ordered logistic regression models. Separate univariable models will be fitted for each covariate of interest (HCW cadre, facility type and location) to account for the ordinal structure of the response categories.

For subdomain-level analyses, composite scores will be constructed by summing correct response selections for Likert-scale items within each predefined subdomain. In constructing composite scores, each item within a subdomain is equally weighted, with responses reflecting the prespecified desirable or correct response assigned one point and incorrect responses, or ‘Not Sure’ responses assigned zero points. This scoring approach was selected to measure the extent to which respondents demonstrated the target KAP captured by each subdomain. For knowledge items, ‘Not sure’ responses will be treated as not demonstrating the measured knowledge construct because they indicate uncertainty or lack of demonstrated knowledge. For attitude and practice items, responses will be scored according to the prespecified construct being measured, with non-endorsing responses assigned zero points. The total domain score will be calculated by summing individual item scores, with the maximum obtainable score determined by the number of included items within each domain.

These composite scores will be treated as approximately continuous and analysed using univariable linear regression models following the same covariates. Residual diagnostic plots will be examined to assess linearity, homoscedasticity, normality of residuals and influential observations using standardised residuals, leverage values and Cook’s distance. If substantial violations of model assumptions are identified, appropriate sensitivity analyses or alternative modelling approaches will be considered, and results will be interpreted accordingly.

Non-Likert-scale survey items will be analysed according to prespecified response structures. For items with objectively defined correct responses, correctness will be determined according to guideline concordance (see online supplemental materials 2 and 3). Select-all-that-apply questions will be scored according to predefined response criteria, with each response option evaluated independently rather than assigning a single overall correct/incorrect classification.

Statistical significance will be assessed using a two-sided alpha level of 0.05. For ordered logistic regression models, results will be presented as ORs with corresponding 95% CIs. Results will be interpreted as evidence of associations between healthcare facility type, location or cadre, and KAP outcomes. Specifically, statistically significant associations indicate differences in the distribution of Likert-scale responses or in composite scores across levels of the explanatory variables, providing descriptive evidence of heterogeneity in KAP-related responses across healthcare facility settings, geographical regions, and cadre groups.

Gaps in knowledge and practice deviations will be identified based on responses misaligned with recommendations in the updated CATB Clinical Guidelines. KAP1 and KAP2 data will be compared between groups using the same individual and subdomain classifications as well as composite scoring systems from the initial analysis. A McNemar test will be used to test for paired significance for categorical variables, and paired t-tests for continuous variables across cadre, facility type and district/region. To correct for participant replacement, participants not retained for KAP2 will be excluded from the dataset analysed in the paired analysis, although still included in the unpaired cross-sectional pre-training and post-training analysis. Statistical significance will similarly be considered met if the p value is<0.05.

Interpretation of statistics

Gap closure will be interpreted based on improvement in specific KAP domains identified as deficient at baseline rather than on a prespecified absolute magnitude of change. Item-level performance will be assessed using predefined Bloom’s satisfactory cut-off criteria; items or domains below this threshold will be considered areas requiring additional attention and targeted improvement in guideline updates. For paired comparisons, statistically significant changes identified by McNemar tests for categorical outcomes and paired t-tests for continuous composite scores (two-sided α=0.05) will be interpreted as evidence of improvement or decline following guideline rollout. Notably, statistical significance will be considered alongside the absolute performance thresholds, such that statistically significant improvements that remain below satisfactory Bloom’s criteria will continue to be considered implementation gaps requiring further targeted updates. Conversely, domains achieving satisfactory guideline concordance after rollout will be considered areas of successful training.

Ethics and dissemination

Approvals

This study was approved by the University of Namibia Decentralized Ethics Committee (Reference Number: SOM44-2025) and the Republic of Namibia MoHSS Research Office (Ref: 22/4/2/3). Dartmouth College’s Committee for the Protection of Human Subjects (CPHS) also reviewed the study protocol and determined this activity met the definition of programme evaluation and qualified as an exempt activity according to guidance (CPHS#: STUDY00033493). Written consent to participate, record the interview and publish anonymised responses was and will be obtained before data collection. This study complies with the Declaration of Helsinki for medical research involving human participants.

Informed participant consent

Each interviewee was provided a one-page brochure describing the study and written informed consent obtained prior to starting the interview. All individuals 18+ are eligible for inclusion in the study. Eligible participants provided consent for themselves. Anyone who did not consent was excluded from the study.

Participant privacy and protection

We strictly adhered to ethical guidelines to safeguard the privacy and confidentiality of our research participants. This involves obtaining informed consent, anonymising sensitive information and securely storing data with restricted access. Dissemination of results will consist of de-identified data. Hard copies of participant data are securely safeguarded. Electronic files are password protected. HREC-H (Human Research Ethics Committee for Health), MoHSS and Dartmouth have access to the data records.

Dissemination

The results will primarily be summarised in two reports submitted to the Namibian MoHSS: one presenting the KAP1 data, and the other presenting the KAP2 data along with a comparative analysis between KAP1 and KAP2. We will also present our findings at local and international meetings, publish our findings in peer-reviewed open access journals and work with the NTLP to develop and publish policy briefs on CATB management.

Discussion

Improved global CATB care is urgently needed, along with improved strategies for clinician training effectiveness. This two-phase assessment in Namibia was designed to generate nationally representative evidence on the KAP of HCWs responsible for CATB care. By establishing a clear baseline prior to rollout of the updated national guidelines, the study will identify priority gaps in diagnostic skills, disease management and familiarity with the new recommendations. Results of the KAP1 will be compiled in a report for the Namibian MoHSS and provide recommendations for informing refinement of the national training materials and enhancing the efficacy of the training cascade. Similar studies have shown that uptake of revised practice guidelines can be measured following national rollout and targeted training developed to address any gaps identified.⁹ Additionally, a systematic review and meta-analysis from 2023 found that training provided to health professionals and volunteers significantly improved the detection of patient with TB.¹⁰

A major strength of this study is its comprehensive national coverage across diverse healthcare settings. By including 16 districts selected to represent a range of TB burdens and 48 facilities spanning hospitals, health centres and primary healthcare clinics, this study captures variation across the settings where CATB care is delivered in Namibia. Inclusion of multiple HCW cadres involved in TB diagnosis, treatment, prevention and follow-up further strengthens the ability of this assessment to identify implementation gaps relevant to different clinical environments. Additionally, the use of a standardised, collaboratively developed and pilot-tested assessment tool provides a systematic approach for evaluating guideline-related KAPs across geographically and operationally diverse settings. The pre–post implementation design provides an opportunity to assess changes following national rollout of updated CATB guidelines and generate actionable information for future training, supervision and refinement.

The study design also has several limitations. First, a formal sample size calculation was not performed because the target sample size was determined based on implementation and feasibility considerations, including achieving broad categories. Therefore, while the sample was designed to maximise national representativeness, the study may have limited power to detect small differences between KAP1 and KAP2 or within subgroup analyses. The study was designed primarily as a strategy evaluation rather than as a study requiring a predefined effect size.

Second, the pre–post design does not include a control or comparison group. Therefore, observed changes between KAP1 and KAP2 cannot be attributed solely to the guideline training intervention and may also reflect secular trends or general clinical experience gained during the follow-up period. Findings will therefore be interpreted as changes associated with guideline implementation rather than causal estimates of efficacy.

Third, participant replacement at KAP2 may limit individual-level comparisons. Although every effort will be made to reassess the same HCWs who participated in KAP1, some participants may be unavailable at follow-up, requiring replacement participants from the same cadre and facility. As a result, paired analyses will include only participants with complete KAP1 and KAP2 data, while secondary cross-sectional analyses including replacement participants will provide complementary information regarding overall guideline uptake but will not permit individual-level assessment of change. Additionally, the 6-month follow-up period allows assessment of early guideline uptake but may not capture long-term knowledge retention, sustained practice change or continued adherence to recommendations.

Fourth, some study outcomes rely on self-reported practices, which may introduce reporting bias if reported practices differ from routine clinical behaviours. Future validation studies incorporating clinical observations, chart review or programmatic indicators would provide additional assessment of implementation fidelity and real-world practice change. Additionally, the survey was administered only in English, Namibia’s official language and the language of national clinical guidelines. This approach may disadvantage some HCWs, particularly HCWs working in rural settings where English proficiency may vary.

Finally, HCW cadre, facility type and location may be correlated, and certain cadres may be concentrated within specific facility types. However, the purpose of these analyses is not to estimate independent predictors of KAP outcomes, but rather to describe variation in guideline-related KAP across predefined implementation-relevant strata. These strata were selected because they correspond to the operational categories through which training, supervision and future guideline implementation activities are organised. Therefore, results will be interpreted as descriptive associations that identify groups or settings where targeted implementation support may be warranted rather than as adjusted estimates of causal effects.

Given the large number of item-level and subdomain-level analyses, there is a potential concern regarding interpretation of multiple comparisons. However, these analyses are not considered parallel co-primary analyses. Instead, each item-level and subdomain-level analysis represents an assessment of a specific reinforced or new component of the updated CATB guidelines and corresponding training objectives. Individual survey items are designed to evaluate whether training successfully addressed specific KAP competencies; therefore, results will be interpreted independently according to the targeted guideline content assessed. Because these analyses represent evaluations of distinct implementation objectives rather than multiple comparisons of a single hypothesis, adjustment for multiple testing will not be performed. Findings will instead be used to identify specific areas of successful guideline uptake and areas requiring additional training or implementation support.

Despite these limitations, the combined results will guide targeted follow-up interventions to strengthen adherence to the updated guidelines. In addition to compiled reports for the Namibian MoHSS and government stakeholders, findings will be disseminated through conference abstracts and a peer-reviewed manuscript, providing a model and results to inform uptake of future guideline updates across varying settings. Together, these outputs are expected to strengthen the quality of CATB care nationwide and contribute to regional and global efforts to reduce morbidity and mortality among children and adolescents affected by TB. A follow-up validation study is planned to assess the implementation fidelity and real-world impact of the updated guidelines across diverse healthcare settings. This work is ongoing, and sustained funding is critical to support continued progress in this neglected area of global health, where CATB remains underprioritised despite its significant burden.