Decentralisation and integration of paediatric tuberculosis services to primary healthcare facilities as an approach to optimise management in Cameroon and Kenya: a descriptive cohort study.

Introduction: Tuberculosis (TB) remains a major cause of morbidity and mortality for children less than 5 years. Diagnosis and treatment of children with active TB is often centralised in district hospital settings due to poor public health infrastructure and lack of diagnostic capabilities in primary healthcare (PHC) facilities. This analysis aims to evaluate TB case detection and treatment outcomes by comparing district hospitals and PHC facilities. Methods: To increase paediatric TB case detection, an intervention was designed to decentralise and integrate paediatric TB diagnosis and management into PHC facilities. Between May 2019 and March 2021, we screened and enrolled children under age 5 years with presumptive TB at different entry points in 32 study facilities in Kenya and Cameroon. TB services were described by level of care. Fischer's exact test was used to determine significance when comparing categorical variables and a Wilcoxon-Mann-Whitney test was used to test for significance of continuous variables. Results: A total of 610 children were enrolled; 481 (79%) had received services at the district hospitals and 129 (21%) at the PHC facilities. The median age was 15.4 (IQR: 6.1-36.0) months; 59% were children below age 2 years, 53% were male and 5% were HIV coinfected. A total of 74 (12%) children were diagnosed with TB, 19 (15%) in PHC and 55 (11%) in district facilities, 11 (15%) with a bacteriological confirmation. The time from TB symptom onset to TB diagnosis was significantly shorter in the PHC (1.0 month (IQR: 0.1-2.1)) than district hospitals, (1.8 months (0.8-4.4), p=0.043). The proportion of lost to follow-up was higher in district hospitals (15.8%) than in PHC (1.8%, p=0.05). Mortality (overall 1.4%) did not significantly differ by setting. Discussion: Decentralisation of TB screening and diagnosis at the PHC level was feasible and significantly shortened the time from the onset of symptoms to TB diagnosis. Trial registration number: NCT03862261.