Tuberculosis in the elderly population: Findings from a State-level TB prevalence survey (2022) from India

Materials & Methods

This cross-sectional study was conducted by the department of Clinical Research, ICMR-NIRT, Chennai, Tamil Nadu after obtaining the ethical clearance from the Institute Ethics Committee.

Tamil Nadu is a Southern Indian State with a population of nearly 80 million people. The present survey covered 32 districts in the State of Tamil Nadu to understand the prevalence of TB. The survey was conducted between February 2021 and July 2022, and the detailed methods have been provided elsewhere^13-15. Participants who had a cough for more than two wk, fever for more than two wk, haemoptysis, chest pain for more than one month, weight loss, loss of appetite, expectoration and night sweat, past history of TB, and any abnormality in chest X-ray (CXR) were eligible for sputum examination. A trained medical officer read CXR in the field, and it was also read by a panel of two pulmonologists at the district level. In case of discrepancies, a third expert reading was performed. CXR were transmitted through the internet via a dedicated local server at the field to the central server at NIRT. All the participants were offered a CXR. Participants with symptoms and/or CXR suggestive of TB were offered confirmatory microbiological diagnostics tests. The Xpert, microscopy, and liquid cultures were offered upfront for diagnosis. One spot sputum was subjected to Cartridge-Based Nucleic Acid Amplification Test (CBNAAT) at the field, and the second early morning sputum was transferred to a nearby pre-identified quality-assured lab for smear and liquid culture. Appropriate linkage and treatment support were offered to the participants diagnosed with TB with the help of the national programme. Quality assurance was ensured by a group of units named ‘Central Project Monitoring Unit (CPMU)’. In the training conducted by CPMU, manpower, frequent field monitoring, data checking, and GPS-based supervision of the flow of participants and their recruitment in the survey were included.

Results

There were 1,30,932 (91.6%) participants who underwent screening with either of the methods. The screening coverage is given in figure. Of those screened by either of the methods, 20,086 were eligible for sputum collection, and in this, the first sample was tested for 18,654 and a second sample was tested for 18,255 participants. Of the 1,30,932 screened individuals, 16,555 (12.64%) were ≥60 yr old, and 1,14,377 (87.36%) were <60 yr of age. Among the elderly participants, 74 (0.45%) were diagnosed with MCPTB, and among the adults under 60 years of age, 170 (0.15%) were diagnosed with MCPTB (Figure). The characteristics of the participants are presented in table I.

Close

Figure.

Consort diagram of TB among elderly population. TB, tuberculosis.

Export to PPT

The crude prevalence of MCPTB per 100,000 population among elderly participants was 447 (95% CI: 351,561), and for adults <60 yr, it was 130 (95% CI: 111, 151). The adjusted prevalence of MCPTB per 100,000 population among elderly participants was 482 (95% CI: 3,85,578), and for adults <60 yr it was 166 (95% CI: 137-195). The adjusted prevalence ratio (APR) of MCPTB in the elderly population was 2.99 (95% CI: 2.25-3.98, P<0.0001).

In the analysis of the elderly subgroup (Table II), the aPR of MCPTB among elderly people was significantly higher in the male sex, under nutrition, smoking, and previous episode of TB. Males had a 2.54 times higher chance of having MCPTB when compared to females (aPR: 2.54; 95% CI: 1.41-4.57). Similarly, participants with under-nutrition (BMI <18.5) were 3.53 times more likely to have MCPTB (aPR:3.53; 95% CI: 1.65,7.54). Self-reported smokers were 1.94 times more likely to have MCPTB when compared to non-smokers (aPR: 1.94; 95% CI: 1.02-3.71), and the elderly with a previous episode of TB had a 2.26 times higher chance of having MCPTB than others (aPR: 2.26; 95% CI: 0.92-5.51). However, a statistically significant association among the elderly was not established in other risk factors like geography, self-reported diabetes, alcohol use status, and falling below the poverty line.

The NNS for the general population was 537 (95% CI: 473-611); however, in the elderly group, this was only 224 (95% CI: 178-285). Among the elderly, NNS was very low at 56 (95% CI: 26-152) for the participants with previous episode of TB: for smokers it was 75 (95% CI: 52 - 112), among those with alcohol use history NNS was 78 (95% CI: 55 - 114), in males it was 133 (95% CI: 103-175) and in individuals with under-nutrition (BMI <18.50) NNS was 172 (95% CI: 136-222) (Table III).

Discussion

Our study revealed five significant findings. Firstly, the prevalence of MCPTB among elderly people in south India was 447 (95% CI: 351-561) per 1,00,000 population, which was comparatively lower than the finding from the ‘National TB Prevalence Survey of India (NTBPS)’ (588; 95% CI 516-660 in ≥55 yr). However, the confidence intervals were overlapping in both surveys. India has not witnessed any other large sample size surveys other than NTBPS. A single-site study among the low socio-economic elderly group in northern India among elderly in 2016 recorded a prevalence of 2300 per 1,00,000. However, this is incomparable with the national surveys¹⁷. These findings reveal the higher prevalence of TB among the elderly and the magnitude of the problem in India. It was also noted that the MCPTB prevalence among the elderly was three times higher than that in adults <60 yr of age (aPR-2.99). It is therefore understood that the rapidly growing elderly population in India would be of greater concern if TB among the elderly is not brought under control⁴.

We conducted an additional subgroup analysis among the elderly to better understand the risk factors associated with TB. Male sex, under-nutrition, and smoking were the factors among the elderly having an association with TB. Similar findings have been recorded by other researchers¹⁸. The second notable finding was that elderly males were 2.5 times more likely to have TB, which was slightly less than the general population (3 to 4.5 times)^2,19. Thirdly, the undernourished elderly individuals were 3.5 times more likely to develop TB. However, in the general population prevalence ratio can be as high as six times¹⁹. Malnutrition is the primary risk factor for TB in India, contributing significantly to both the incidence and prevalence of the disease.¹⁰ The fourth finding was that self-reported elderly smokers had a two-fold higher chance of having TB. However, in the general population, this can be as high as five times¹⁹. The above-mentioned risk factors are similar to those found in the general population; however, the strength of association was less among the elderly because old age was a risk factor for TB.

The fifth finding, potentially relevant for the National Health Programme, was the NNS, which was only 224 for the elderly compared to 537 for the general population. Additionally, among the elderly individuals who were males, who had already been affected by TB, reported smoking, used alcohol, and were undernourished, the NNS was even much lower. This will aid in establishing targeted screening of high-risk groups when planning active community screenings or opportunistic screenings at healthcare facilities.

The elderly are the population with many risk factors for TB and, hence, highly susceptible to TB. They present with atypical and nonspecific symptoms and presents with atypical chest radiographic presentations, and collecting sputum for diagnosis also becomes difficult. Late diagnosis, associated comorbidities, and poor adherence would lead to poor and unfavourable outcomes²⁰. Therefore, active community screening and opportunistic hospital screening should be implemented and enhanced to ensure early diagnosis and prompt treatment for the elderly. Understanding the need for health care of the elderly, India has come up with a ‘National Programme of Health care of Elderly’ by which various packages of services are offered to the elderly at different levels in health care. One among them is conducting a weekly ‘geriatric clinic’ at the primary health centre level²¹. Tapping this opportunity would help in diagnosing elderly patients with TB, and the finding of our study would help support such newer initiatives. Without having evidence at the subgroup level and applying in the programme through specific measures, attaining the End TB-2035 target of 95 per cent reduction in mortality and 90 per cent reduction in incidence would remain unachievable. In terms of mortality, India is supposed to reach the target of 14 per lakh in this year (2025), six per lakh in 2030 and three per lakh in 2035; however, India has reached 22 per lakh in 2020¹⁰. Sriram et al¹² identified that TB mortality was 10 times higher in the elderly group compared to that in the younger population and Kolappan et al²² identified that TB mortality was more than three times among the older population²². These studies underline the importance of screening the elderly population with TB, without which the end-TB targets would not be achievable. Our study provides the data for better TB screening through the public health care system. The strengths of this study were that it was carried out with a large sample size and followed a well-designed standard operating procedure. The survey staff received thorough training, and the survey was closely supervised. However, one limitation of this study was that self-reported information on factors such as alcohol consumption, smoking, and diabetes history was used, which may have introduced some biases.