Article Text
PDF
Abstract
Background Despite tuberculosis (TB) being a curable disease, current guidelines fail to account for the long-term outcomes of post-tuberculosis lung disease—a cause of global morbidity despite successful completion of effective treatment. Our systematic review aimed to synthesise the available evidence on the lung function outcomes of childhood pulmonary tuberculosis (PTB).
Methods PubMed, ISI Web of Science, Cochrane Library and ProQuest databases were searched for English-only studies without time restriction (latest search date 22 March 2023). Inclusion criteria were (1) patients who had TB with pulmonary involvement at age ≤18 years; (2) pulmonary function tests (PFTs) performed on patients after treatment completion; and (3) observational studies, including cohort and cross-sectional studies. We adhered to the recommendations of the Cochrane Collaboration and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses.
Results From 8040 records, 5 studies were included (involving n=567 children), with spirometry measures from 4 studies included in the meta-analyses. The effect sizes of childhood TB on forced expiratory volume in the first second and forced vital capacity z-scores were estimated to be −1.53 (95% CI −2.65, –0.41; p=0.007) and −1.93 (95% CI −3.35, –0.50; p=0.008), respectively.
Discussion The small number of included studies reflects this under-researched area, relative to the global burden of TB. Nevertheless, as childhood PTB impacts future lung function, PFTs (such as spirometry) should be considered a routine test when evaluating the long-term lung health of children beyond their completion of TB treatment.
PROSPERO registration number CRD42021250172.
- Paediatrics
- Respiratory Medicine
- Child Development
- Child Health
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Tuberculosis (TB) is a treatable disease, but despite resolution of the infection, lung function deficits associated with post-tuberculosis lung disease (PTLD) can persist.
While this is well appreciated in adults, the extent and severity of PTLD in children are not well characterised.
This area of work is important because of the potential long-term impacts of PTLD on children’s lung health and development.
WHAT THIS STUDY ADDS
Our meta-analyses showed that childhood TB causes significant decline in at least two spirometry parameters despite high levels of between-study heterogeneity.
The effect sizes of childhood TB on forced expiratory volume in the first second (FEV1) and forced vital capacity (FVC) z-scores were clinically significant.
While direct comparison with published adult TB studies was not possible, this study suggests that childhood TB results in PTLD.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This study supports incorporation of routine pulmonary function tests into the follow-up of children with history of TB, allowing for early detection and management of PTLD.
Introduction
Tuberculosis (TB) is an airborne disease caused by Mycobacterium tuberculosis. Inhalation of the bacterium into the airways can result in TB infection. Pulmonary disease is established when the host’s innate immune response is unable to eliminate the bacterium.1 In 2021, an estimated 10.6 million people fell ill from TB globally, and children under 15 years old accounted for 11% of this burden.2 Childhood TB causes a spectrum of clinical presentations, most commonly pulmonary disease. Irrespective of organ involvement, obtaining bacteriological confirmation for infants and young children still proves challenging. Age is the key determinant of disease progression, with risk of progression to pulmonary tuberculosis (PTB) about 30%–40% when primary infection occurs in infants under a year old.3 While improving the diagnosis and management of childhood TB is important,4 children with prior PTB can experience detrimental changes irrespective of successful completion of treatment.5 There is a significant knowledge gap in the occurrence and severity of post-tuberculosis lung disease (PTLD) in children.6 Adult PTLD is better described, including post-TB bronchiectasis7 and lung function changes.8 One study reported decline in mean forced expiratory volume in the first second (FEV1) and forced vital capacity (FVC) z-scores by −1.07 and −0.91 on treatment completion, and −0.91 and −0.64, respectively, 3 years post-treatment.9
Specific data on childhood PTLD are required as early-life lung injuries from respiratory infections and pneumonia cause deficits during children’s peak lung growth and development.10 It is possible that childhood PTB is more detrimental to future lung function, compared with acquiring PTB as TB-naïve adults. A recent review recommended the evaluation of PTLD using objective tests for early detection of post-TB pulmonary changes irrespective of symptoms.11 These could promote initiation of treatments to prevent irreversible lung function decline, reduce healthcare costs, and alleviate burden to patients, their families and healthcare systems.
Given the absence of a systematic review evaluating the effects of childhood TB on pulmonary function test (PFT) outcomes, we undertook this review and meta-analysis aiming to synthesise available evidence regarding the effects of childhood PTB on future lung function.
Methods
This systematic review was registered in PROSPERO with identification number CRD42021250172. Deviations from the registered protocol were (1) redefining the primary outcome as spirometry measurements and the secondary outcome as non-spirometry measures of lung function; and (2) reduced number of searched databases due to record duplication. Study findings were reported in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analyses guidelines and checklist.12
Literature search
Eligible studies were searched from PubMed, Cochrane Library and ISI Web of Science databases up to 22 March 2023 (online supplemental appendix 1). Grey literature searches were performed on ProQuest database, followed by manual citation searching of included studies. No publications were excluded based on publication date.
Supplemental material
Eligibility criteria
Included studies fulfilled the following inclusion criteria : (1) patients with TB with pulmonary involvement at age ≤18 years; (2) PFTs performed on patients after treatment completion; and (3) observational studies, including cohort and cross-sectional studies. The exclusion criteria were (1) mixed-population studies which did not report the ≤18 years subgroup separately, (2) TB studies without pulmonary involvement, (3) evidence of non-standard anti-TB treatment regimens, (4) did not perform PFTs post-treatment or (5) reviews and case studies.
Studies from all countries and settings were included. Studies were included regardless of bacteriological confirmation, unreported treatment regimens or timing of PFTs. Studies with other concurrent disease as primary domain were included providing that PFT measures were sufficiently reported for inclusion in the analysis.
Outcomes
The primary outcomes were spirometry measures. The secondary outcomes were measurements from non-spirometry PFTs. There were no limits to the timing of PFTs after completion of TB treatment.
Data extraction and quality assessment
Screening and eligibility assessment was performed by two reviewers independently (YLL and AFT). The references of eligible studies were assessed to ensure inclusion of relevant studies. No automation tools were used throughout the review process. For eligible studies, data extraction was performed according to a standardised collection form (online supplemental appendix 2). Using the Newcastle-Ottawa Scale, both reviewers independently assessed the studies’ risk of bias and certainty assessments, with consensus achieved through discussion.13
Statistical analysis
All statistical analyses were done using R for Windows (V.4.2.2).14 The median and IQR were normalised to give mean and SD, where the distance between Q1 to median was equal to median to Q3.15 Spirometry results presented as percentage of predicted values were converted to z-scores using the ‘rspiro’ package,16 based on the Global Lung Function Initiative (GLI)-2012 equation,17 accounting for North East Asian ethnicity, mean age of 11.9 years, a male to female ratio of 53.5:46.5 and a median height for age based on the 2017 Korean National Growth Charts.18 Studies which reported primary outcomes were included in the meta-analyses, and effect sizes were calculated using Hedges’ g and presented with 95% CI.19 Secondary outcomes not included in the meta-analyses were presented separately.
We used random-effects models (DerSimonian-Laird method) to estimate overall effect due to variable data with significant heterogeneity. Between-study heterogeneity was assessed using the I2 statistic, with values >75% representing considerable heterogeneity. Meta-analyses were performed using the ‘metafor’ package.20 Sensitivity analysis was not performed as substantial interstudy differences rendered statistical approaches meaningless.
Results
Search results
The screening process is detailed in figure 1. After removing duplicates, 8040 records were screened, from which we reviewed 34 full-text articles and finally included 5 studies.21–25 The characteristics of the excluded studies are summarised in online supplemental appendix 3.
PRISMA diagram showing identification, screening and inclusion of studies. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.
Characteristics of included studies
The included studies were studies conducted in urban or periurban settings of TB-endemic countries of Africa, except for one retrospective review study in South Korea, a country with upper-moderate TB incidence.22 A total of 567 children with history of PTB were included; the median number of children in the studies was 68 (range: 42–305).21–25 Key characteristics are reported in table 1, and details of quality scores are shown in online supplemental appendix 4. One study performed non-spirometry PFTs after early-life TB25 and was thus excluded from the meta-analyses. Granular details of the studies included in the meta-analyses are provided in online supplemental appendix 5.
Included studies and their characteristics
Diagnosis and treatment of PTB
Bacteriological confirmation of TB varied between studies, ranging from 13.7%25 to 58.5%.22 GeneXpert MTB/RIF was used to rule out active infection pre-spirometry.21 26 Only one study reported treatment regimen for drug-susceptible TB according to national guidelines, modified for drug resistance25; other studies21–24 did not report treatment regimen details.
Pulmonary function tests
The time between treatment completion to PFTs ranged between 6 and 24 months in the three most recent studies23–25; two earlier studies21 22 did not specify this duration. Three studies21–23 reported performing spirometry according to the American Thoracic Society/European Respiratory Society 2005 standards,27 and one24 according to the 2019 update.28 Three studies21 23 24 performed bronchodilator responsiveness testing.
For the effects of childhood TB on FEV1 and FVC z-scores, meta-analyses were possible and these are presented in forest plots, with pooled effect size estimates of −1.53 (95% CI −2.65, –0.41; p=0.007; figure 2) and −1.93 (95% CI −3.35, –0.50; p=0.008; figure 3).
Forest plot of effect sizes of childhood pulmonary tuberculosis on FEV1 z-scores. DL, DerSimonian and Laird method; FEV1, forced expiratory volume in the first second; TB, tuberculosis.
Forest plot of effect sizes of childhood pulmonary tuberculosis on FVC z-scores. DL, DerSimonian and Laird method; FVC, forced vital capacity; TB, tuberculosis.
Meta-analysis was not possible for FEV1:FVC ratios presented in the included studies, thus summarised in table 2 instead. Only one study performed non-spirometry PFTs. Association coefficients between childhood PTB occurring between 1 and 4 years of age with measurements taken at 5 years are reported and presented in online supplemental appendix 6.25 Only one study reported lung function patterns directly attributed to PTB. Spirometry taken at 19.2 (IQR: 10.2–44.4) months after TB diagnosis showed 61.5% (32/52) with normal lung function, 36.5% (19/52) with restriction and 2% (1/52) with obstructive patterns at a median age of 8.9 (IQR: 7.2–11.2) years.24
Summary of studies reporting FEV1:FVC ratios in any manner
Heterogeneity, sensitivity and bias
Due to the small number of studies, sensitivity analysis was not performed. Significant between-study heterogeneity was observed in the meta-analyses for FEV1 and FVC effect sizes at I2=89.11% (p<0.0001) and I2=91.63% (p<0.0001), respectively, as indicated by the blue diamonds in figure 2 and figure 3. The study22 which reported PTB–bronchiectasis overlap had the largest effect size on both FEV1 and FVC z-scores. The exclusion of this study22 from the meta-analysis for FEV1 led to a significant reduction in heterogeneity, as shown by the red diamond in figure 2, but not observed in the meta-analysis for FVC.
Publication bias was judged as unlikely as the included studies were observational, funded by research grants and unlikely to be influenced by industry-based sponsorship or agenda. It was notable that three included studies reported primarily on non-TB diseases, thus less likely to be affected by reporting bias in terms of PTB-related outcomes, at the cost of being less comprehensive in reporting PTB-related details.21–23 One included study had significant information bias as summary statistics were not reported numerically, necessitating pixel counting of error bars from interval plots to approximate the data dispersion within the PTB subgroup.22
Discussion
Interpretation of results
The small number of included studies highlights under-representation of childhood TB globally.2 The overall direction of effects of PTB on lung function was negative, that is, reduced lung function in both meta-analyses of FEV1 and FVC. These findings align with our current understanding of PTLD in adults,29 lending further support to the validity of our approach. While pooled effect sizes appear to be significant, high I2 values indicate substantial between-study heterogeneity, which is a key limitation in our study. This suggests a research gap in quantifying the impact of PTB during childhood on lung function outcomes, particularly in high-prevalence settings.
Of the included studies, three had primary diseases21–23 that were not PTB, which were reasonable to include as HIV coinfection is a significant comorbidity30 and bronchiectasis is a well-established sequela of PTB.6 One study24 evaluated health-related quality of life post-PTB, suggestive of recent paradigm shifts to better evaluate PTLD. As only two studies reported spirometry performed at >24 months23 and >6 months24 after treatment completion, the actual effect of spirometry timing was indeterminate due to variability of the included studies. A prospective cohort of adult TB survivors did show greater deterioration in FEV1 and FVC values 3 years after treatment completion9 compared with the first year post-treatment.29
Due to a low quality score, contextual interpretation for one included study22 is presented here. As the GLI-2012 equation17 for North East Asian ethnicity was developed using only subjects aged ≥16 years, the spirometry z-scores for young children in this study22 were calculated using extrapolation, which may have inadvertently inflated the effect sizes. This inference was partially supported by a validation study31 which found that South Korean females aged 7–8 years have mean FEV1 and FVC z-scores lower than the GLI-2012 predictions by −0.23 (95% CI −0.31, –0.15) and −0.26 (95% CI −0.36, –0.16), respectively, suggesting the actual lung function for this subgroup is slightly below established baseline. A secondary analysis which excluded this outlier study22 from the meta-analyses yielded an alternative random-effects model for FEV1 based on three studies with reduced statistical heterogeneity (red diamond; figure 2). As pooled effect sizes regardless of exclusion remained below −0.8, the overall interpretation was that childhood TB exerts a large effect32 on FEV1. Removal of this study22 did not appreciably change the pooled effect size estimate nor the I2 statistic for FVC (not shown). It is noteworthy that one study23 reported more significant HIV-associated decline in FVC than FEV1; the combinatory effect of PTB within an all-HIV cohort gave a greater change in FVC relative to baseline and subsequently a larger standardised effect size as compared with FEV1. This was partly supported by another study10 which found early childhood respiratory infections had a marginally greater effect on FVC than FEV1, raising plausibility that HIV coinfection is a clinical contributor to heterogeneity observed in figure 3.
Limitations of evidence and review process
The included evidence had limitations inherent to the population, nature of disease and outcome measures. The WHO classifies childhood TB as diagnosed in children <15 years old, leading to bias in age stratification at study design level.2 Adolescents aged ≥15 years are classified as adults, inadvertently excluding evidence encompassing full age range of childhood. Bacteriological confirmation of TB was relatively low (range: 13.7%–59.5%); thus, misclassification bias among children with clinically diagnosed TB was possible.33 Numerical SD values were not reported in one study22, requiring manual pixel-counting based on published figures, measurement errors arising from this step may be propagated when Hedges’ g was calculated. One study23 reported z-score changes as association coefficients, thus necessitating Fisher’s z-transformation,34 resulting in CIs that were much smaller than other studies21 22 24 included in the meta-analyses. Thus, pooled effect sizes should be interpreted with awareness of our approach used.
Clinical and policy implications
To the best of our knowledge, this is the first meta-analysis to investigate the effects of childhood PTB on lung function decline. Our findings suggest that childhood PTB is associated with overall decreases in subsequent FEV1 and FVC z-scores. Concurrent bronchiectasis exerted the greatest additive negative impact on spirometry parameters compared with HIV coinfection or TB on its own.22 Childhood TB and resultant PTLD remain understudied within paediatric populations despite clear association with lung function decline, further compounded by underdiagnosis and subsequent failure to treat.2 6
WHO-defined outcomes of TB treatment include cured or treatment completed positive outcomes, and negative outcomes of lost to follow-up, treatment failure or death.35 These outcome indicators are based primarily on bacteriological clearance and treatment compliance, with post-TB sequelae and residual respiratory impairment unaccounted for. The most recent roadmap for ending TB in children and adolescents does not address the fact that post-TB disabilities and PTLD do occur beyond completion of treatment.36
In the first consensus-based set of clinical standards for PTLD,37 the foremost standard recommends clinical, functional and subjective evaluation of every patient completing TB treatment for PTLD, with considerations for paediatric care, including selection of age-appropriate PFTs and quality-of-life questionnaires. The second and third standards called for evaluating patients with PTLD for pulmonary rehabilitation (PR) and the organisation of PR programmes with health settings and individual patients’ needs in mind. While not routinely done for children and thus far unreported for childhood PTB, individualised PR programmes have been attempted for paediatric asthma38 and could be adjusted to younger patients in high-TB settings.
Objective lung function measurements allow for prompt initiation of PR37 or other adjunctive therapies39 to prevent late-life onset of respiratory diseases such as bronchiectasis, asthma or chronic obstructive pulmonary disease. As performing spirometry on young children can be challenging, non-spirometry PFTs should be considered for children below certain ages and others on a case-by-case basis. At least one study has explored oscillometry for children above 2 years, alongside spirometry for those above 4 years of age.40 Subsequent findings may address the evidence gap for performing scheduled PFTs as part of national TB programmes or routine post-TB pulmonary health surveillance,11 especially in low-income to middle-income countries with significant disease burden.
Our findings suggest that spirometry or other PFTs should be performed as routine follow-up of children beyond TB treatment completion to monitor their lung function and diagnose any impairments promptly. Lung health monitoring enables appropriate and timely interventions to reduce the frequency and severity of PTLD beyond treatment completion.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Ethics statements
Patient consent for publication
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors YLL collected and analysed the data, and wrote and revised the manuscript; YLL accepts full responsibility for the finished work and/or the conduct of the study, had access to the data, and controlled the decision to publish. YLL is guarantor. AFT collected and analysed the data, and reviewed the manuscript. SY critically reviewed the data analysis and interpretation, and reviewed the manuscript. TWY conceptualised and designed the study, and provided supervision. AC conceptualised the study, critically reviewed the manuscript and provided supervision. CL reviewed the initial analysis, critically reviewed the manuscript and provided supervision.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
Linked Articles
- Highlights from this issue