The prevalence of tuberculosis infection in India: A systematic review and meta-analysis

Tuberculosis (TB) caused by Mycobacterium tuberculosis (MTB) is a well known and major public health challenge globally, and is the leading cause of death from a single infectious agent1. According to the Global TB Report 20221, an estimated 10.6 million incident cases of TB were reported in 2021. The World Health Organization (WHO) END TB strategy aims to reduce 95 per cent of TB deaths and 90 per cent of new incident cases of TB by 20352. The global TB targets for reductions in disease burden of TB can only be achieved if diagnostic, prevention and treatment services of TB are strengthened3,4. A state of persistent immune response due to stimulation by MTB antigens without any evidence of clinically manifested active TB is defined as TB infection (TBI). TBI can serve as the precursor for the development of full-blown TB disease, particularly among the immunocompromised5. This constitutes a large reservoir of individuals with TBI, and thus, the management of TBI is crucial for global efforts to curb TB burden, particulaly in high TB burden countries such as India.

Geographically, out of the six high-burden countries from south-east Asia region, India accounts for 28 per cent of the global TB burden6. It also has the highest TBI burden globally6. According to the National TB prevalence survey, 2021, the crude prevalence of TBI among individuals >15 yr was 31.3 per cent [95% confidence interval (CI); 30.8-31.9]7. Around 5-10 per cent of those with TBI reportedly develop clinically active TB disease8. A single active case of TB may infect several other individuals before receiving any anti-tubercular treatment due to delayed diagnosis, hence creating a perpetual reservoir of TB-infected individuals9. Since TB-infected individuals constitute a perennial source of risk for progression towards active disease, prevention of active TB disease by treating TB-infected individuals and interrupting the chain of transmission is the chief component of the WHO End TB strategy2,8.

The prevention of TB disease by the treatment of TBI is largely undervalued but remains as an important component of the National Strategic Plan 2017-25 for Ending TB in India by 2025, five year ahead of the sustainable development goals10. The lancet commission on TB mentions that the diagnosis and treatment strategies to end TB would be ineffective unless TB preventive treatment (TPT) is included in the comprehensive strategy11. There is a need to improve the implementation of proven interventions such as effective new regimens for TPT and ensure their efficient and rapid scale-up12. This applies to finding high-risk groups and initiating TPT. Thus, it is essential to estimate the burden of TBI among various states and groups in India, which has programmatic implications. Against this background, this review aimed to estimate the burden of TBI in the Indian population stratified according to geographies and sociodemographic profiles.

Material & Methods

Protocol and inclusion criteria: A systematic review of various studies examining the TBI among Indians was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. All primary studies conducted on Indian residents reporting TBI irrespective of the tests used to measure were included. We considered the use of any available test for identifying TBI including interferon gamma release assay or tuberculin skin test (TST) or T SPOT, TB or C-TB or Cy-TB among the participants. Reviews, case reports, editorials, opinion pieces, study protocols, conference abstracts, posters, thesis, reports or any unpublished material were excluded. TBI for the present study was defined as the immune response to MTB antigen without clinical evidence of active TB.

Databases, study selection and data extraction: We conducted electronic searches of four databases namely: MEDLINE, EMBASE, CINHAL and Scopus limiting the search from January 1, 2013 to December 31, 2022. A comprehensive search was done using the keywords and Medical Subject Headings (MeSH) terminology for Tuberculosis including, TB, Latent Tuberculosis Infection, sub-clinical Tuberculosis, inactive Tuberculosis, Tuberculosis Infection, LTBI, Pulmonary Tuberculosis, Koch’s Tuberculosis, Extra-pulmonary Tuberculosis, Latent TB, Tuberculin Skin test, Mantoux test, TST, Purified protein derivative, Interferon gamma release assay, skin test, IGRA, Enzyme-linked immunospot assay, Quantiferon, Quantiferon-TB Gold, T.SPOT TB, C-TB and Cy-TB. The complete search strategy per database consulted is presented as Appendix I in the Supplementary material.

Two authors independently searched from all four databases mentioned. Reference checking, hand searching of citations and reference lists were done to identify potentially missing literature.

All citations retrieved from the electronic searches were first imported to EndNote where duplicate entries were removed, the resulting entries were then uploaded to the Rayyan software, 201613. Two independent researchers then screened the titles and abstracts of the retrieved studies to identify all the articles that could be eligible for inclusion. Uncertainty or disagreement was settled by a third reviewer’s consensus decision. Two researchers conducted the full-text screening of the previously identified articles, and disagreement was settled by a third reviewer.

The information on study characteristics and the results of the included studies were extracted using a standardized data extraction form. Information on authors, publication, country, study design, study area, sample size, publication year, study year, sociodemographic factors, proportion of TBI and the test used was extracted. In addition, information for assessing the risk of bias was extracted. When the data were insufficient or missing or full text was not available, corresponding authors of the original articles were contacted via e-mail to provide relevant information.

Assessment quality and the risk of bias: Methodological quality and the risk of bias among the included studies were assessed by two independent researchers using the Joanna Briggs Institute (JBI) Critical Appraisal tools designed for use in systematic reviews – set of questionnaires used to asses cross-sectional, cohort and randomized controlled trials13,14. Studies were then graded, and according to the score obtained, these were classified as ‘low’, ‘moderate’ or ‘high’ risk of bias.

Data synthesis and analysis: The studies were categorized according to the study design, and then, the study characteristics using percentages and frequencies for categorical variables and standard deviations or median for continuous variables were summarized. The pooled prevalence of TBI was extracted using STATA at 95 per cent confidence interval (CI) to account for small variability between studies. The prevalence of TBI based on geographical location and age group was estimated. Studies were weighted using the random effect model. As two tests (TST and IGRA) were commonly used for estimating TBI prevalence, pooled prevalence was estimated for studies using TST as well as IGRA separately using the random effect model. Sub-group meta-analysis was carried out based on the demography and prevalence of active TBI in India. Effect sizes were expressed as odds ratio for dichotomous data and as weighted mean difference for the continuous data. All effect estimates were expressed using 95 per cent CI. We pooled data using the random effects model (DerSimonian and Laird) in the MetaXL software.

Results

Study characteristics: Initially, 10,521 studies were retrieved. Of these 519 duplicates were excluded and 9996 studies were eligible for screening. Of these, 9836 studies were further excluded after reviewing the title and abstract leaving 160 studies for full text review. Of these, the full text of 55 studies was unavailable. Additionally, 10 studies were included from cross references. So, in all 115 studies met the inclusion criteria for a full-text review. Following the full-text review, 77 studies were finally included for investigation. The PRISMA flow diagram is presented in Figure 1.

Close

Fig. 1

PRISMA flow diagram. PRISMA, preferred Reporting Items for Systematic Reviews and Meta-Analysis

Export to PPT

Out of the 77 articles, 31 were cohort and 46 were cross-sectional studies. According to the zonal divisions of India15, most of the research was conducted in the southern part (15,813 i.e. 40.8% of patients) followed by the northern region (12,527 i.e.; 32.3% of patients), western region (3424 i.e. 8.8% of patients), central region (2296 i.e. 5.9% of patients) and eastern (174 i.e. 0.4% of patients) region of India. TST was used for diagnosing TBI in majority of the studies (70/77), IGRA was used in 10 per cent (7/77) studies and both TST & IGRA were used in 61 per cent (47/77) studies. A total of 59 studies considered TST positivity when reading recorded more than 10 mm induration and 11 studies considered it as more than 5 mm. A total of 38,767 individuals were studied across the included articles ranging from 33 to 2351 individuals. The Table provides an overview of the main characteristics of the included studies. As IGRA is a more specific diagnostic test for TBI compared to TST5, the pooled prevalence of TBI as depicted below was estimated using IGRA data as available. In the absence of the availability of IGRA data, TST results were considered for estimation. Furthermore, the pooled prevalence of TBI in India was estimated based on the community-based cohort studies only since hospital-based and cross-sectional studies generally, have inherent selection biases.

Prevalence of tuberculosis infection based on community-based cohort studies: As per the random-effects model, the pooled community-based cohort study-based prevalence of TBI was found to be 40.8 per cent (95% CI: 29.5-52.6%, Q=1648.9, P<0.0001, I²=99%) (Fig 2). After excluding the risk groups, the pooled prevalence of TBI for community-based general population group was found to be 36 per cent (95% CI: 28-45%, Q=17.38, P<0.0001, I²=83%). Among the general population, TBI prevalence among adults (≥15 yr) was 35 per cent (95% CI: 24-46%, Q=13.93, P<0.0001, I²=86%). Due to the remarkable heterogeneity (P<0.00001, I²=99%), a sensitivity analysis was performed. After excluding the studies causing the heterogeneity, the pooled TBI prevalence was found to be 38 per cent (95% CI: 29-46%, Q=10.72, P=0.02, I²=72%).

Close

Fig. 2

Community-based cohort studies TBI prevalence in India. TBI, tuberculosis infection

Export to PPT

Prevalence of tuberculosis infection based on community-based cohort study according to the geography, age, gender and year-wise distribution: According to the available data, the prevalence of TBI based on community-based cohort studies in Delhi was 68 per cent (95% CI: 46-87%), in Tamil Nadu 42 per cent (95% CI: 24-61%) and in Maharashtra 26 per cent (95% CI: 16-36%). The pooled prevalence of TBI in urban areas was 37 per cent (95% CI: 16-60%), in rural areas 27 per cent (95%CI: 11-48%) and in tribal areas 33 per cent (95% CI: 20-47%). Among the paediatric population, the pooled prevalence of TBI was estimated as 33 per cent (95% CI: 24-42%) in under 5-yr-old children and in those aged 6-14 yr, it was 40 per cent (95%CI: 30-51%), respectively. In adult population, ranging from 15-45 yr of age, the TBI prevalence was 52 per cent (95% CI: 39-69%) and among older adults (above 45 yr), it was estimated as 62 per cent (95% CI: 50-74%). The estimated pooled prevalence of TBI was 41 per cent (95% CI: 19-65%) among the male population and in female population, it was 31 per cent (95% CI: 09-59%). A steady trend of TBI over the years was observed according to the data reported by various studies during 2013-2022.

Quality and risk of bias: Majority of the studies (76) were considered having a low risk of bias according to the JBI critical assessment score (>70%). Only one study was considered to have a moderate risk of bias (score 50-69%); Appendix I, Supplementary Table I and II. No articles were excluded based on the quality assessment.

• Q1. Were the two groups similar and recruited from the same population?

• Q2. Were the exposures measured similarly to assign people to both exposed and unexposed groups?

• Q3. Was the exposure measured in a valid and reliable way?

• Q4. Were the confounding factors identified?

• Q5. Were strategies to deal with confounding factors stated?

• Q6. Were the groups/participants free of the outcome at the start of the study (or at the moment of exposure)?

• Q7. Were the outcomes measured in a valid and reliable way?

• Q8. Was the follow up time reported and sufficient to be long enough for outcomes to occur?

• Q9. Was follow up complete, and if not, were the reasons to loss to follow up described and explored?

• Q10. Were strategies to address incomplete follow up utilized?

• Q11. Was appropriate statistical analysis used?

Publication bias:

Discussion

The present review appraised available evidence on the burden of TBI among people residing in India based on the results of IGRA as well as TST, which are the tests being used to diagnose and target individuals eligible for TPT. Data were gathered from over 38,767 IGRA and TST results covering the different regions of India. Based on the community-based cohort studies, it was observed that TBI was prevalent in more than one-third of India’s population and it increased with advancing age.

Various researchers previously reported a similar prevalence of TBI in India, using Bayesian model. Woodruff et al15 and Collins et al93 estimated 31.9 per cent and 33.9 per cent prevalence among Indians, respectively whereas Houben and Dodd94 by mathematical modeling presented a 31 per cent estimate of TBI in the south-east Asian region94. The national TB prevalence survey in 2019-2021 also recorded a crude TBI prevalence of 31 per cent7. Considered together, all these evidences indicate that India has a huge reservoir of TBI, a considerable proportion of whom may progress to TB disease. This review noted a rising TBI prevalence with age, especially among adults and older adults. High prevalence among such groups is an impediment to the national TB elimination efforts. Chong et al95 in their mathematical model, pertaining to countries with intermediate TB prevalence, suggested that screening and providing TPT to 20-40 per cent of the elderly could result in 50 per cent overall decline of TB incidence.

We noted a steady trend of TBI maintained over the last decade suggesting the persistence of TB incidence. TBI treatment rates among the high-risk groups in India was sub-optimal (12%)6. The World Health Organization (WHO) recommends the prevention and treatment as key interventions to achieve, the End TB targets. End TB targets2 India’s National Strategic plan (2017-25) rests on prevent, detect and treat as the core pillars for the elimination of TB10. The treatment of active TB reduces the prevalence of infectious TB, eventually cutting down the transmission and lowering the incidence4, whereas treatment of TBI prevents the progression of latent infection to disease and directly reduces the incidence3. Evidence suggests the treatment of TBI and active TB act synergistically to reduce TB incidence96,97. Hence, when the treatment levels of active TB are high (95%)6, a similar payoff could accrue by stepping up the treatment of TBI. Thus, scaling up of the TPT will hasten the lowering of TB incidence and to achieve the EndTB targets in India98.

The current review recorded remarkable regional differences in TBI prevalence. A high burden was found in those areas with prevalent active TB (Delhi, Tamil Nadu, etc.) implying that TBI probably has a high conversion rate to active TB disease7. TBI case-finding has high implementation cost99. However, in the absence of TBI case-finding, the number of infectious individuals will grow demanding increased need for active case-finding and case-holding efforts99. At present, India’s National TB Elimination Programme (NTEP) recommends active case-finding and case-holding efforts for areas with high TB prevalence leading to a longer implementation period100. Considering, both active as well as TBI case-finding having similar cost-sensitivities, TBI case-finding has a shorter implementation period. TBI case-finding relies on the use of diagnosis and treatment of TBI99. Evidence suggests that there is no ideal way of testing for and diagnosing TBI101-103. The two main tests in use are TST and IGRA, both of which measure the immunological response to MTB antigens101. TST, although inexpensive, is logistically challenging, produces variable results depending on the type of purified protein derivative (PPD) used and shortage of quality-assured tuberculin posing an impediment globally104,105. On the other hand, IGRA has high specificity, demands single facility visit; however, it is costly and there are issues related to the reproducibility of the results102. Recently, validity of both the tests has been questioned with evidence, suggesting only 10 per cent patients who harbour viable MTB organisms capable of causing disease showing immunoreactivity to these tests12,106. Further, issues related to testing hesitancy challenges the implementation of widespread screening107,108. Growing experience from clinical studies and implementation in the field suggests newer regimens such as three-month weekly rifapentine and isoniazid (3HP) regimen and one-month daily rifapentine and isoniazid (1HP) regimens, both of these appear to have a similar efficacy as the daily isoniazid dosing-based preventive therapy but with a better safety profile, higher acceptance and completion rate3. Thus, faced with operational issues related to the TBI diagnostics and considering the consequences of missing positive cases leading to increase in number of infectious individuals, increased efforts for active case-finding and high case-holding efforts in high prevalence areas and with recent 3HP implementation in the programme7,100, ‘no test, treat only approach’ after ruling out active TB needs to be contemplated109,110.

The present review is a comprehensive review that reports the prevalence of TBI in India, wherein an extensive systematic literature search with hand-searching of references was performed with a multidisciplinary team. The pooled prevalence was estimated only for community-based cohort studies. This increased the validity and offered a deeper and comprehensive understanding of the burden of TBI among different groups as well as the general population. However, the evidence reflected all the parts of India except the eastern region thus limiting the comprehensive depiction of the findings across India. There were also some limitations. For TST, studies were not separated based on the strength of PPD or use of standardization of PPD used in the studies. Furthermore, there was a lack of uniformity in the diagnosis of TBI, especially while using TST method as some studies reported more than 5 mm as TBI positivity irrespective of the immunocompromised patient inclusion. Further, there is a possibility of poor detection of true heterogeneity, especially when small number of studies are analyzed using Cochran’s Q, Higgins and Thompson’s I² test.

Implications and way-forward: Prioritization of initiation of TPT is required for the regions with a high prevalence of TBI. This is important as the chances of missing positive patients are high owing to the less predictive value of diagnostic tests and their high operational costs. A more robust approach is needed with consideration of ‘No test, treat only’ approach after ruling out active TB for specific high disease burden geographies in comparison to the current strategy of active TB case-finding and case-holding approach for such regions. Research is further needed to identify the prevalence of TBI among individuals with multimorbidity (presence of two or more chronic conditions in one person), e.g. diabetic with cardiovascular disease or rheumatoid arthritis. Further evidence needs to be generated related to TBI among migrants, prisoners and residents of mental health homes. Prioritization of States with high TB disease burden and TBI prevalence for community-based screening to rule out active TB and implement TPT policy at the population levels is needed. As the treatment of TBI is a pre-requisite for achieving TB elimination goals, the evidence presented in this review will pave way for a more strengthened programmatic management of TBI in India.

Overall, this review demonstrated a high prevalence of TBI, commensurate with active TB prevalence suggesting conversion of TBI to active TB-disease. A high burden among people residing in the northern and southern regions was observed, indicating the need for the TBI country-specific strategies targeting population-level interventions.