Epidemiology and outcomes of drug-resistant tuberculosis cases notified in a low-resource district in Kerala, India 2017–2021 – A 5-year retrospective analysis

INTRODUCTION

Tuberculosis (TB), caused by the bacillus Mycobacterium tuberculosis, is the second leading infectious disease cause of death after COVID-19 globally, with an estimated 10.6 million cases and 1.6 million deaths in 2021.^[1] Drug-resistant TB (DR-TB) is often more challenging to treat, with poorer patient outcomes than drug-sensitive TB. DR-TB patients face high economic and social costs, and only one in three can access quality care globally.^[2] Globally, in 2021, 71% of the microbiologically confirmed pulmonary TB patients were tested for at least Rifampicin (Rif) resistance, with 0.45 million drug-resistant cases detected.^[2] India contributes 26% of the total TB burden and 27% of Rif resistance globally, with an estimated DR-TB incidence of 0.9/million population.^[3] A recent situation analysis conducted in Kasaragod – one of the low-resource settings in Kerala state, India, revealed a higher notification of DR-TB cases, 0.2/million, compared to the state average of 0.1/million population. The district also reports a 3 times higher proportion of DR-TB cases out of total notified TB cases than the state. Hence, it was pertinent to understand the epidemiology and outcomes of DR-TB cases notified in the district. We analyzed the National Tuberculosis Elimination Program (NTEP) data in the Kasaragod district with the objectives of describing the DR-TB cases notified in the Kasaragod district from January 2017 to December 2021 by time, place, person, and by treatment outcomes and to determine the factors associated with unfavorable treatment outcomes in DR-TB patients.

MATERIALS AND METHODS

We conducted programmatic data analysis of NTEP data in the Kasaragod district of Kerala, India. Kasaragod is the northernmost district of Kerala sharing the border with Karnataka state with a population of 1.5 million and three TB treatment units (TU), the decentralized units for programmatic management of TB cases in the district. An analytical cross-sectional study design was used to describe the DR-TB cases notified in the district for 5 years from January 2017 to December 2021 and to determine the factors associated with unfavorable treatment outcomes. We included all the 128 DR-TB patients notified in the Kasaragod district during the period for our analysis.

We extracted data from the “Nikshay portal” (the web-enabled case-based monitoring application for TB patients as part of India’s NTEP), DR-TB treatment cards, and the DRTB registers kept at the district TB center in Kasaragod. The variables on sociodemographic characteristics, anthropometry, DR-TB diagnosis details, clinical features, and outcome were entered into the structured data abstraction form in Epicollect5. We used the standard definitions used by the NTEP program for classifying the treatment outcomes.^[4] For comparing the groups, we classified outcomes as favorable treatment outcomes, including cured and treatment completed, and unfavorable treatment outcomes, including treatment failure, loss to follow-up, and death. A pulmonary TB patient with bacteriologically confirmed TB at the beginning of treatment who completed treatment as recommended by the national policy with evidence of bacteriological response (two consecutive cultures negative) and no evidence of treatment failed (treatment regimen needs to be terminated or permanently changed to a new regimen option or treatment strategy) is considered cured. A patient who completed treatment as recommended by the national policy whose outcome does not meet the definition of a cure or whose treatment failed is considered treatment completed. To ensure the data quality, the data were triangulated using the Nikshay portal, DR-TB treatment cards, and DR-TB registers for all the patients. We used “Nikshay ID” as the unique identifier variable during the triangulation process.

The data were analyzed using STATA version 17 (StataCorp. 2021. Stata Statistical Software: Release 17. College Station, TX: StataCorp LLC), and maps were created using QGIS version 3.22.6 (QGIS.org, 2022. QGIS Geographic Information System. QGIS Association. http://www.qgis.org). The burden of DR-TB is expressed as DR-TB notification per 100,000 population and the proportion (%) of DR-TB out of total TB notification in the district. This burden is described by year, local self-government area (the local administrative setup), age, gender, and income level. All DR-TB patients with outcomes evaluated were included in the analytical study. Bivariate odds ratio with 95% confidence intervals (CIs) was estimated to examine the association between risk factors of interest, such as sociodemographic factors, DR-TB type and treatment, comorbidities, and clinical and laboratory features at the time of diagnosis and unfavorable treatment outcomes. We performed multiple logistic regression to determine the independent risk factors of getting unfavorable treatment outcomes. A P < 0.20 in the bivariate analysis, plausible biological association, or evidence of significant association as documented in earlier studies were the criteria for including as covariates to the logistic regression model. CIs of 95% were reported for each odds ratio. All reported P-values are two tailed, and statistical association is considered significant at a P < 0.05.

Ethical clearance was not required as this study was a secondary data analysis of the programmatic data available with the district health system. Permission to utilize the data was obtained from the Director of Health Services, Kerala, The District Medical Officer (Health) Kasaragod, and the District TB Officer, Kasaragod. The entire patient details were coded and anonymized before the analysis.

RESULTS

128 DR-TB patients were notified in the Kasaragod district from 2017 to 2021. Among them, 47 (36.7%) were monoresistant, 75 (58.6%) were multidrug-resistant (MDR), and 6 (4.7%) were extensively drug-resistant (XDR) TB.

Time distribution of DR-TB cases

TB notification shows a decreasing trend in Kasaragod across 2017–2021, and DR-TB notification is increasing with the highest notification rates in 2019. Figure 1 shows the TB and DR-TB notification rates in Kasaragod.

Close

Figure 1:

Tuberculosis (TB) and drug-resistant (DR)-TB notification rates of Kasaragod, Kerala, 2017–2021.

Export to PPT

The proportion of DR-TB cases out of notified TB cases in Kasaragod increased from 1.8% in 2017 to 3.3% in 2021, with the highest notification of 5.4% in 2019. The proportion of new patients notified rose from 17.6% in 2017 to 61.5% in 2021, and previously treated notifications decreased from 82.4% in 2017 to 38.5% in 2021. Monoresistance cases were first detected in 2018. The proportion of monoresistant patients rose from 10.5% in 2018 to 58% in 2021, and MDR and XDR cases dropped from 88% and 12% in 2017 to 42% and 0% in 2021, respectively.

Place distribution of DR-TB cases

Figure 2 shows the distribution of DR-TB cases by Local Self Government (LSG) area, which is the local administrative units in the state of Kerala. Seven out of 41 LSG units had not notified any DR-TB cases from 2017 to 2021. Across the 5 years, the DR-TB notification per 100,000 population ranged from 0 to 20, and the proportion of DR-TB cases out of total notified TB cases ranged from 0% to 33%. The LSG units nearing the state administrative boundary with the Karnataka state and those adjacent to the coastal belt had higher notifications of DR-TB cases over the years. To understand the DR-TB notification in the context of drug sensitivity testing (DST), we identified the DST rates for the three TB units from 2017 to 2021. Over 5 years, from 2017 to 2021, the DST rate augmented from <30% to 85% across all three TB units.

Close

Figure 2:

Drug-resistant tuberculosis (DR-TB) notification by local self-government units, Kasaragod, 2017–2021.

Export to PPT

DISCUSSION

The notification rates of DR-TB are higher in Kasaragod compared to the state average. TB notifications per 100,000 population in Kasaragod district decreased from 2017 to 2021, and the highest dip in notifications was noted in the year 2020, which can be related to the COVID-19 pandemic, which disrupted the TB diagnostic and treatment services. This notification drop is noted globally, with India contributing 41% of the total reduction in notifications.^[2] The DR-TB notification per 100,000 population is higher than the state average of 1.7 and much lower than the national average of 9. The DST rate in Kasaragod is showing a steady increase across the three TUs of the district, reaching more than 90% of the microbiologically confirmed cases, which may be the reason for the increasing DR-TB notification. The higher proportion of DR-TB among notified TB cases in Kasaragod may be due to the lower TB notification, especially from the private sector hospitals bordering the Karnataka state, which does not report to the routine surveillance system. These notification rates suggest that the district may be “missing cases” that need urgent attention. The findings of the National Drug Resistance Survey released in 2018 also estimated that only 44% of the estimated MDR-TB cases were diagnosed in India.^[5] Kasaragod district started CBNAAT for Rif resistance for all microbiologically confirmed TB cases from 2017 onward leading to increased DR-TB notifications and primary drug resistance notifications. The line probe assay for INH and Rif resistance was started for all M. tuberculosis cases from 2018 onward, and hence, monoresistance cases are increasing from 2018 to 2021. Due to the Kerala TB elimination mission and the availability of CBNAAT facilities in all districts from 2017, the entire program received a new focus, and notification rates were improving for DR-TB across the state.^[6]

From the spatial distribution of the cases across the district, coastal areas and areas sharing boundaries with the Dakshin Kannada district of Karnataka consistently report more DR-TB cases throughout the 5 years. This is consistent with studies conducted in Jodhpur, India, and Ethiopia.^[7,8] Further spatio-epidemiological studies are required to explore the clustering of cases in administrative and coastal borders.

The incidence of TB is higher in the 15–45 age group in India.^[9] Our study shows the highest notification of DR TB in the 45–59 age group, followed by the above 60 age group. Males and those living below the poverty line have higher DR-TB notifications per previous evidence from India.^[7,10,11]

The overall favorable treatment outcome among evaluated cases was 75.4%, and the unfavorable treatment outcome was 24.6%. Several studies across the globe have demonstrated the treatment success rate of DR-TB from 25% to 72 % in various settings.^[12-14] The reported mortality for different regimen-wise DR-TB cohorts initiated treatment in 2019 ranged from 5% to 21% and our study reports an overall mortality of 14%.^[3]

Our study has identified age of more than 45 years, living below the poverty line, being admitted more than twice during treatment, and BMI at diagnosis in the underweight category as significant predictors for unfavorable treatment outcomes. Many studies conducted in similar settings have identified an age of more than 45 years as a significant predictor for unfavorable treatment outcomes.^[15,16] Poor socioeconomic status is a proven risk factor for poor treatment outcomes among TB patients.^[17,18] Our study also identified poor economic status as a risk factor for poor outcomes and the existing financial and nutritional support programs, which reached almost 100% of the patients in the district, to be continued. A recent individual participant data meta-analysis has identified that underweight patients had 1.7 times the odds of unfavorable treatment outcomes compared to patients with normal weight.^[19] Other studies conducted in India and Vietnam have also identified unfavorable treatment outcomes for underweight patients (BMI <18.5 kg/m²) at diagnosis.^[13,20] BMI at diagnosis can be used to predict outcomes, and serial monitoring is needed to identify treatment prognosis.

CONCLUSION

DR-TB notifications have increased in the Kasaragod district from 2017 to 2021, with a high burden among males aged 45 years and above. The favorable treatment outcome is better than the national and state average. Regular monitoring and follow-up of MDR patients with low incomes, above 45 years, and underweight may improve the final treatment outcomes.