Long term outcome following pulmonary TB or lower respiratory tract illness in children: a prospective South African study

Abstract

Background Pulmonary tuberculosis (PTB) and lower respiratory tract infection (LRTI) are common in children in low- and middle-income countries. While the long-term impairment following PTB is well recognised in adults, there is limited data in children. Additionally, there is a lack of comparative studies assessing these outcomes relative to other (non-TB) causes of LRTI. To longitudinally assess the long-term impact of PTB and non-PTB LRTI on lung function, including children under 5 years of age, utilising both oscillometry and spirometry. Methods A prospective study of South African children with suspected PTB from 3 September 2015 to 18 September 2024 was done. Microbiological testing included 2 induced sputa for liquid culture and Xpert Ultra. Participants were classified as: confirmed TB (microbiologically positive); unconfirmed TB (clinical diagnosis); unlikely TB (microbiology negative, not clinically diagnosed with PTB and improvement without TB treatment at 3 month follow up). Lung function was measured at enrolment and at follow-up visits: 1 and 3 months for all participants; 6, 12 and 18 months for confirmed or unconfirmed TB. Measures included pre- and post-bronchodilator (BDR) oscillometry in all and spirometry in those 5 years or older. Linear mixed effects models were used to explore the association of PTB or LRTI with lung function z-scores over time. Findings Of 276 children included, 112 (41%) had confirmed TB, 113 (41%) unconfirmed TB and 51 (18%) unlikely TB. The median (IQR) age was 8.3 (5.7–11.1) years; 51% were female. Children with confirmed TB had lower pre-BDR resistance at 6 Hz z-score (β = −0.53, 95% CI −0.95, −0.11); lower post BDR z-scores for FEV 1 (−0.55, 95% CI −1.08, −0.01) and FVC (−0.59, 95% CI −1.14, −0.04) compared to unlikely TB at 3 months. Lower post-BD FEV 1 (−0.31, 95% CI −0.762, 0.00) and FEV 1 /FVC ratio (−0.33, 95% CI −0.64, −0.33) persisted through 18 months in children with confirmed compared to unconfirmed PTB. Abnormal spirometry was common across all groups, with a predominant restrictive pattern. Interpretation Children with PTB demonstrated significant lung function impairment, predominantly restrictive, that persisted over time. These findings highlight the importance of ongoing respiratory evaluation and stronger preventative interventions to mitigate long-term pulmonary sequelae in paediatric PTB. Funding National Institutes of Health, USA (RO1HD058971), Regional Prospective Observational Research in TB consortium (TB-REPORT), South African Medical Research Council (SA-MRC) and EDCTP2 programme (TMA2020CDF-3200). S Chaya is funded by Carnegie Deal 3 program at UCT and the SA National Research Foundation; L Githinji was funded by the Foundation Botnar supported by the European Union, HJ Zar is supported by the SA-MRC.