Contribution of Xpert® MTB/RIF to pediatric TB diagnosis in Mozambique

Dear Editor, GeneXpert testing was introduced in Mozambique in 2011. In 2013, the National TB Control Program (NTP) adopted Xpertw MTB/RIF (Xpert; Cepheid, Sunnyvale, CA, USA) as the initial diagnostic test for TB in children and people living with HIV due to its higher sensitivity (compared to microscopy) and ability to identify rifampin (RIF) resistance.1,2 By 2018, there were 109 GeneXpert machines in use, and testing volume increased from 36,927 tests in 2015 to 163,530 in 2018.3 Sample referral networks were developed to increase access for facilities without a machine on site. During this scale-up of Xpert, the number of annual pediatric TB cases reported nearly doubled from 6,672 in 2015 to 12,522 cases in 2018. The number of bacteriologically confirmed pediatric cases also more than doubled from 677 in 2015 to 1,543 in 2018. However the percentage of cases of bacteriologically confirmed only rose from 10.1% to 12.3%, and only 60 children were treated for multidrug-resistant TB (MDR-TB) in 2018.3–5 Simultaneous to the expansion in Xpert use, there was a national training program on a new pediatric diagnostic algorithm that facilitated clinical diagnoses.6 This effort likely explains the large increase in overall cases, but it was unclear how Xpert expansion contributed to bacteriologic confirmation and MDR-TB treatment in children. The NTP does not receive routine programmatic data indicating the type of test done for confirmed cases, and the national GxAlert system does not provide reliable age-disaggregated information to quantify pediatric testing. Due to persistent concerns about pediatric TB incidence and case detection rates despite recent progress, we conducted a programmatic evaluation of laboratory diagnosis to estimate the contribution of Xpert to confirmed pediatric drug-susceptible and MDR-TB cases, and to evaluate age-appropriate sampling methods.7–9 Data from laboratory registers from January 2017 to December 2018 were collected from three of the 11 provinces in the country (Cabo Delgado, Gaza, Maputo Province) and compared to aggregate provincial treatment data for the same time period. All children aged 0–14 years with Xpert results registered in laboratories with a GeneXpert machine were included. This programmatic review had a nonresearch exemption from the National Bioethics Committee, Maputo, Mozambique (Ref 409/CNBS/ 19). No patient-identifying information was collected and informed consent was not required. A total of 4,488 children were included. The median age was 8 years and the distribution of patients in the 0–4, 5–9, and 10–14-year age groups was 26.9%, 33.3% and 39.8%, respectively; 2,321 (51.7%) were female. Only 44 (1.0%) patients had documentation of previous TB treatment. The majority of specimens were respiratory, collected using spontaneous sputum production (n1⁄4 3,961, 88.3%), followed by sputum induction (n 1⁄4 259, 5.8%) and gastric aspiration (n 1⁄4 191, 4.3%). Extrapulmonary specimens, including pleural and cerebral spinal fluid, represented 1.7% (n1⁄4 77) of all samples. In children aged 0–4 years, respiratory specimens were obtained using sputum induction and gastric lavage in 13.5% and 14.6% of cases, respectively. Xpert results were classified as MTB-detected, MTB-not detected, and invalid in 266 (5.9%), 4,125 (91.9%) and 97 (2.2%) patients, respectively. Positivity rates were 5.0%, 5.2% and 7.1% in the 0–4, 5–9, and 10–14-year age groups, respectively. Positivity by specimen type was 6.6%, 1.9% and 1.6% for spontaneous expectorated sputum, induced sputum, and gastric aspirate specimens, respectively. Of the 266 Xpert MTB-detected samples, 11 (4.1%) were RIF-resistant, 4 (1.5%) were RIF-indeterminate, and 251 (94.4%) were RIF resistance not detected. During the 2-year time period of the evaluation, the NTP reported 4,039 cases of TB (all forms) among children aged 0–14 years from the three provinces, including 1,161 bacteriologically confirmed cases and 26 cases of presumptive or confirmed MDRTB.3 The 266 Xpert MTB-detected cases identified from the review of laboratory registers represent 6.6% of the 4,039 cases and 22.9% of the 1,161 bacteriologically confirmed cases reported. The 11 Xpert RIF-resistant cases represent 42.3% of the 26 MDR-TB cases notified (Table).3,10 In 2018, MDR-TB represented only 0.5% of pediatric cases reported by the NTP, and pediatric cases constituted only 5% of total cases of MDR-TB, whereas 13.4% of all reported TB cases were in children.3 In this evaluation, Xpert RIF-resistant tests accounted for 42.3% of reported pediatric MDR-TB cases, and we suggest that expanded access to Xpert should be prioritized to improve MDR-TB diagnosis