Introduction
In 2024, of 10.7 million people estimated to have developed tuberculosis (TB) according to the WHO, 11% (1.2 million) were children. Only 49% of children with TB were notified to WHO by National TB programmes (NTPs) mainly due to under diagnosis.1 Challenges in diagnosing childhood TB include the paucibacilliary nature of pulmonary TB in children resulting in low sensitivity of existing microbiological diagnostic tests and the difficulty of obtaining respiratory samples in young children.2 3 In addition, the centralisation of childhood TB diagnostic services in several high TB incidence and low- and middle-income countries (LMICs) reduces access to TB care, and thereby contributes to the gap of childhood TB diagnosis.4
In 2022, to improve diagnosis and access to treatment for childhood TB, WHO recommended (1) decentralising childhood diagnosis services to low-level peripheral health facilities and (2) improving the quality of the diagnosis approach and specifically the microbiological testing component by using low-complexity automated nucleic acid amplification tests (LC-aNAAT) like the Xpert MTB/RIF Ultra (Ultra; Cepheid, Sunnyvale, California, USA) on stool samples and nasopharyngeal aspirates (NPAs).5 This automated GeneXpert platform reduces sample manipulation time and biosafety risks and thereby increases the possibility of using the Xpert assay at low levels of healthcare.6 The single module GeneXpert Edge (G1), a cheaper, heat and dust resistant, more user-friendly and battery-operated platform could be introduced at primary health centre (PHC) levels where there is a lower number of tests to perform daily.7 Literature is limited on its use and placement at PHCs or at bedside. Two proof of concept studies reported good feasibility of using G1 at community level by study staff to provide on-site testing too difficult to reach people and at home by means of mobile transportation (boats) and in-home testing of household TB contacts where it eliminated dependence on clinic and referral based testing systems.8 9 In PHCs without laboratory and thus laboratory technicians, nurses could be trained on how to operate the G1 platform and to perform the Ultra assay. Another use for G1 and reason justifying training of nurses to perform Ultra would be to make the Xpert testing available in paediatric wards to shorten time to diagnosis for highly vulnerable children outside laboratory opening hours.
Apart from simple point-of-care testing such as urine-lipoarabinomannan test, there is no literature on task-shifting of microbiological TB diagnostic tests from laboratory technicians to nurses.10 11 However, previous experiences in other fields of TB-related healthcare have shown that task-shifting/sharing from workers in specialised or higher cadres to healthcare workers (HCWs) from a different background could address human resource challenges, address unmet health needs, improve healthcare quality in terms of service delivery and accessibility after short trainings with support supervisions in health services.12–15
To address the unmet need to improve access to childhood TB microbiological diagnosis, the TB-Speed Decentralisation study implemented Ultra testing on NPA by either nurses or laboratory technicians in children with presumptive TB at PHC level.16 Similarly, the TB-Speed Pneumonia study that evaluated systematic TB detection in children admitted for severe pneumonia implemented in-wards Ultra testing on NPA by nurses to reduce time to treatment decision.2
In this study, we compared the feasibility of Ultra testing on NPA in terms of uptake, performance and turnaround times (TATs) between nurses and laboratory technicians, and we assessed the nurses’ perception and experience of Ultra testing on NPA within the TB-Speed Decentralisation and Pneumonia studies.
Facts Only
* In 2024, 10.7 million people were estimated to have developed tuberculosis according to the WHO.
* 11% of these cases (1.2 million) involved children.
* Only 49% of children with TB were notified to the WHO by National TB programs, mainly due to under-diagnosis.
* Challenges in diagnosing childhood TB include the paucibacillary nature of pulmonary TB and difficulty obtaining respiratory samples.
* WHO recommended decentralizing childhood diagnosis services and improving diagnostic quality using low-complexity automated nucleic acid amplification tests (LC-aNAAT) like Xpert MTB/RIF Ultra on stool samples and NPAs in 2022.
* The GeneXpert Edge (G1) platform is a cheaper, heat/dust-resistant, user-friendly device proposed for use at PHC levels.
* Proof of concept studies showed feasibility using G1 for on-site testing by study staff and home testing of household contacts.
* Nurses could be trained to operate the G1 platform and perform Ultra assays in PHCs without laboratory staff.
* The TB-Speed Decentralisation study implemented Ultra testing on NPA by nurses at the PHC level.
* The TB-Speed Pneumonia study implemented Ultra testing on NPA by nurses in inpatient wards.
Executive Summary
The development of tuberculosis in 2024 involved 10.7 million people, with 1.2 million being children. Only 49% of children with tuberculosis were reported to the WHO by National TB programs, primarily due to under-diagnosis. Challenges in diagnosing childhood TB include the paucibacillary nature of pulmonary TB in children, which reduces the sensitivity of existing microbiological tests, and difficulties obtaining respiratory samples in young children. Centralization of childhood TB diagnostic services in many low- and middle-income countries limits access to care and contributes to diagnostic gaps.
To address these issues, the WHO recommended decentralizing childhood diagnosis services and improving diagnostic quality by using low-complexity automated nucleic acid amplification tests (LC-aNAAT), such as Xpert MTB/RIF Ultra, on stool samples and nasopharyngeal aspirates (NPAs). This automated approach reduces sample handling time and biosafety risks. A cheaper, portable platform, GeneXpert Edge (G1), was proposed for use at primary health centers (PHCs). Proof of concept studies showed feasibility in using G1 for on-site testing in community settings and testing household contacts, bypassing traditional clinic referral systems. Furthermore, there is an opportunity to task-shift microbiological diagnostic tests from laboratory technicians to nurses, supported by training, as evidenced by the TB-Speed Decentralisation and Pneumonia studies which used Ultra testing on NPAs by nurses.
Full Take
The narrative presents a clear trajectory from a systemic failure in diagnosing childhood TB to proposed technological and structural interventions aimed at improving access. The initial facts highlight that diagnostic limitations—stemming from the nature of the disease and centralized service delivery—create significant gaps in care for children. The proposed solutions, centered on decentralization and automation (LC-aNAAT/G1), suggest a shift in where and by whom testing occurs.
The move towards task-shifting diagnostics from specialized staff to frontline health workers like nurses is a crucial nexus point. This moves beyond mere technical capability toward addressing human resource constraints and service delivery accessibility, suggesting that the implementation feasibility rests as much on contextual training and organizational acceptance as it does on the technology itself. The comparison between the decentralized and pneumonia studies shows an intent to test the operational efficacy of these interventions across different settings.
The central tension lies between the promise of high-tech, point-of-care testing and the practical realities of training health workers and establishing sustainable decentralized systems in resource-limited settings. A critical question arises regarding whether the implementation of technology like G1 or the task-shifting of complex molecular tests will translate into sustained, equitable care, or if it risks creating a new layer of dependency on supervision rather than true autonomy. What are the long-term implications for the quality assurance chain when testing shifts from centralized labs to varied frontline settings? What metrics beyond uptake and turnaround time are necessary to assess the impact on clinical outcomes for vulnerable populations?