**Combined Tuberculosis and Diabetes Mellitus Screening and Assessment of Glycaemic Control among Household Contacts of Tuberculosis Patients in Yangon, Myanmar**

#### **Nyi-Nyi Zayar 1,2 , Rassamee Sangthong <sup>1</sup> , Saw Saw <sup>3</sup> , Si Thu Aung <sup>4</sup> and Virasakdi Chongsuvivatwong 1,\***


Received: 4 May 2020; Accepted: 19 June 2020; Published: 29 June 2020

**Abstract:** Background: This study aimed to identify the prevalence of diabetes mellitus (DM) and tuberculosis (TB) among household contacts of index TB patients in Yangon, Myanmar. Method: Household contacts were approached at their home. Chest X-ray and capillary blood glucose tests were offered based on World Health Organization and American Diabetes Association guidelines. Crude prevalence and odds ratios of DM and TB among household contacts of TB patients with and without DM were calculated. Results: The overall prevalence of DM and TB among household contacts were (14.0%, 95% CI: 10.6–18.4) and (5%, 95% CI: 3.2–7.6), respectively. More than 25% of DM cases and almost 95% of TB cases among household contacts were newly diagnosed. Almost 64% of known DM cases among household contacts had poor glycaemic control. The risk of getting DM among household contacts of TB patients with DM was significantly higher (OR—2.13, 95% CI: 1.10–4.12) than those of TB patients without DM. There was no difference in prevalence of TB among household contacts of TB patients with and without DM. Conclusion: Significant proportions of the undetected and uncontrolled DM among household contacts of index TB patients indicate a strong need for DM screening and intervention in this TB–DM dual high-risk population.

**Keywords:** screening; tuberculosis; diabetes mellitus; contact investigation; TB-DM

### **1. Introduction**

Globally, approximately 15% of tuberculosis (TB) patients are comorbid with diabetes mellitus (DM) [1]. Meanwhile, patients with DM have almost a three times higher risk of developing TB than the general population [2]. It is estimated that stopping the increasing prevalence of DM could prevent 6 million TB cases and 1.1 million TB deaths in 13 countries over 20 years [3]. Myanmar is ranked 10th among high TB burden countries, with an estimated annual TB incidence of 181,000 (95% CI: 119,000–256,000) [4]. The country is also ranked 9th among countries with the highest incidence of TB with DM comorbidity [5]. The TB incidence in the Yangon Region was 506/100,000 of the population, which was significantly higher than the global average 132/100,000 [6,7]. The prevalence of DM in the Yangon urban area was 12.1% [8], which was also higher than the global and regional averages [9]. It was found that the adjusted hazard ratio of TB was 2.09 among patients with DM over three years [10]. Therefore, the burden of TB is predicted to increase in the future, with an increased prevalence of DM in the Yangon Region.

It is known that the risks of getting TB and DM are higher among household contacts and family members. The prevalence of TB among household contacts was 4.5% in low and middle income countries [11]. Additionally, close blood relatives and spouses of patients with DM have more than a two times higher risk of developing DM due to sharing genetic susceptibility and/or sharing lifestyles, including physical inactivity, dietary habits associated with BMI, and central obesity [12–14]. Previous studies reported that DM is more prevalent among patients with TB compared to the general population [15]. Therefore, the prevalence of both TB and DM might be increased among household contacts of TB patients because of the high prevalence of DM among patients with TB.

In Myanmar, household contact investigation of TB, implemented since 2011, is an effective screening method for undiagnosed TB among household contacts of patients with TB [16]. Screening of DM was launched in 2018, according to a WHO recommendation, among patients with TB aged 40 years and above in 32 townships in 23 districts of Myanmar [17]. However, there is no guidance on screening of DM among household contacts of index TB patients. Without appropriate intervention, the burden of future TB might be higher among household contacts with DM living in the same household with patients with TB in Yangon. In addition, the risk of TB is especially higher among patients with poorly controlled DM who are exposed to patients with active TB [18]. Therefore, integration of DM screening and glycaemic monitoring of DM in household contact investigations could uncover not only hidden TB patients but also the undiagnosed DM and poor glycaemic control DM patients who are at high risk of developing TB. Comparing the TB and DM prevalence among household contacts of TB index cases with and without DM allows us to examine whether the two groups had different risks and different levels of needs in screening. This study intended to examine the prevalence of DM and of TB among household contacts of patients with TB and compare this prevalence based on the DM status of index TB patients. The proportion of poor glycaemic control of known cases of DM among these contacts was also assessed.

#### **2. Materials and Methods**

#### *2.1. Study Design and Setting*

A cross-sectional study was conducted in the Insein and North Okkalapa townships in the Yangon urban area from March to December 2018. These areas were selected for two reasons. First, they had the highest TB case notification rates in the Yangon region, with 387 and 248 per 100,000 population, respectively. Second, Yangon Region had the highest DM prevalence (12.1%) in Myanmar [8].

More than 80% of patients with TB living in the study townships are registered in the Township TB clinics and take regular treatment, while others are registered in clinics supported by local and international non-government organizations [19]. Existing routine household contact investigation for TB in these areas is carried out by basic health staff at the township level. The screening of signs and symptoms of TB followed by sputum smear and chest radiography (CXR) to confirm the diagnosis and a gene Xpert test to detect multidrug resistant TB (MDR-TB) were done in Township TB clinics. The findings from this study may give a clue to the situation on a larger scale throughout the country.

#### *2.2. Study Sample*

The study included newly diagnosed sputum smear positive index TB patients, aged ≥ 25 years, registered in township TB clinics. The study recruited only patients with newly diagnosed TB who had never received TB treatment to avoid hyperglycaemia due to the TB treatment [20]. Pregnant women, HIV positive TB patients, MDR-TB patients and those who had no family members were excluded. The exclusion criteria were based on some medical conditions that may affect blood sugar or risk of having DM. Pregnant women may have gestational DM [21]. Consistently, people living with HIV and

receiving treatment for HIV may have increased risk of type 2 DM [22]. Furthermore, patients with DM have lower risk of developing MDR-TB compared to drug sensitive TB [2,23]. Therefore, the study excluded them to maintain homogeneity among the index cases. Family members living in the same households with an index TB patient for at least 3 months before having diagnosis of TB were invited for a contact investigation of TB. Among them, household contacts aged < 25 years were excluded for DM screening because the prevalence of DM increases steadily from age 25 years [24].

#### *2.3. Sample Size Calculation*

The sample size was calculated to test that the prevalence of TB among household contacts with DM (28%) was higher than that without DM (13%) [25]. At least 259 household contacts of index TB patients without DM and 65 household contacts of index TB patients with DM were required.

#### *2.4. Data Collection Tools*

A set of structured questionnaires was modified from previous research on household contact investigation in Myanmar [25] and a manual for WHO STEPwise approach to surveillance of non-communicable diseases and their risk factors [26].

The questionnaire included (1) Sociodemographic characteristics, such as age, gender, education, employment status and daily income per household member. Daily income per household member was converted from Myanmar kyats into United States dollars (USD) and cut off values for income level were 1.9 and 3.1 USD/day based on the poverty level defined by the World Bank [27]. (2) Signs and symptoms of TB, including evening rise in temperature, cough, coughing up blood, breathlessness, chest pain and weight loss. (3) Previous history of TB and DM (history of TB and/or DM diagnosed by a medical doctor). (4) Risk factors of TB (closeness with index TB patients including sharing the same room, sleeping in the same bed and taking care of index TB patients) and risk factors of DM (family history of DM, low physical activity: defined as <3 days of vigorous-level activity (e.g., carrying heavy loads, running, etc.,) of at least 20 min/week, or <5 days of moderate-level activity (e.g., walking very briskly, performing domestic chores, etc.,) using standard metabolic equivalents [28].

A physical examination including body mass index (BMI) was calculated using weight and height (kg/m<sup>2</sup> ) and classified < 18.5 kg/m<sup>2</sup> as underweight, 18.5–24.9 as normal, 25.0–29.9 kg/m<sup>2</sup> as overweight and ≥30 kg/m<sup>2</sup> as obese [29], and central obesity defined for those with waist circumference ≥ 94 cm in men and ≥80 cm in women [30], were also done during household visits.

#### *2.5. Diagnosis of DM in Index TB Patients and Household Contacts*

Both index TB patients and their household contacts were asked about their previous history of DM diagnosed by a medical doctor. The diagnosis was used as a confirmation of DM. They were also tested for their glycaemic control by fasting capillary blood glucose (FBG). Those without history of DM were tested by random capillary blood glucose (RBG) followed by FBG on the following day for TB index patients, and within 2 weeks for household contacts at township TB clinics. In case a contact did not revisit the clinics within 2 weeks, researchers then visited them at their home within 4 weeks to perform an FBG test. Any single positive test result was repeated on a separate day. All blood tests were done by an Accu-Chek® Performa glucometer [31].

Based on the American Diabetes Association [32], DM was defined as subjects with known DM or newly diagnosed DM with RBG ≥ 200 mg/dl and FBG ≥ 126 mg/dl (or) RBG ≥ 200 mg/dl for two times on separate days (or) FBG ≥ 126 mg/dl for two times on separate days. Poor glycaemic control was defined as FBG ≥ 130 mg/dl among known cases of DM [33].

All participants were duly informed of their results. New patients with DM were referred to DM clinics for further management.

#### *2.6. Diagnosis of TB among Household Contacts*

Household contacts were firstly screened by (1) asking about signs and symptoms of TB and (2) taking CXR at township TB clinics.

For household contacts with at least one signs and symptoms of TB (or) abnormal CXR suggested of TB, smear microscopy test was done by collecting one spot sputum specimen at the time of visit to TB clinic and next-day early morning sputum specimen. Gene Xpert test was done for those who had at least one positive sputum smear microscopy result.

For those who could not produce sputum, a 2-week trial of broad-spectrum antibiotics was prescribed. Disappearance of abnormalities on CXR would indicate non-TB. Otherwise, the subject was classified as TB. All contacts who were deemed to have positive TB were invited to register in the township TB clinics, where treatment would be provided according to the National TB treatment guideline.

Confirmation of TB could be made by 2 criteria. First, bacteriological confirmation using either a sputum smear or GeneXpert tests [34]. Second, a clinical diagnosis for those who did not meet the first criteria but had abnormal CXR. All contacts with newly diagnosed TB were asked to receive a full course of treatment [34].

#### *2.7. Data Collection Procedure*

Eligible patients registered in each township's TB clinic were invited to participate in the study. After informed consent, a face-to-face interview and a blood investigation for DM were done. Participants were then asked for their permission to visit their home and screen their household contacts for TB and DM.

#### *2.8. Data Analysis*

Epidata version 3.1 was used for data entry and R software was used for data analysis. The prevalence of TB or DM among household contacts was calculated by dividing the number of contacts with TB or DM by the number of household contacts screened for TB or DM. Chi-square test and Fisher's exact test were used to test differences between the prevalence of TB or DM among household contacts of TB patients with and without DM.

Multivariate logistic regression models were done to determine the odds ratios of getting DM and TB among household contacts based on the DM status of index TB patients, adjusted for covariates identified in previous studies [14,35–37]. Covariates included age, gender, socioeconomic status (SES), low physical activity, BMI, central obesity for risk of DM, age, gender, SES and closeness to index TB patient for risk of TB. These covariates were added cumulatively into each model. The model with the lowest Akaike's information criterion (AIC) value indicated the best fitting model. All *p* values < 0.05 were taken as statistically significant.

#### *2.9. Ethical Approval*

Ethical approval was obtained by the Ethics Review Committee at Prince of Songkla University (33/2017) and the Ethics Review Committee at the Department of Medical Research, Myanmar (023/2018).

#### **3. Results**

#### *3.1. Investigation for DM among Index TB Patients*

#### Prevalence of DM among Index TB Patients

Figure 1 shows a flow chart of the study. A total of 235 index TB patients were registered in the township TB clinics during the study period. Among them, 16 were excluded for various reasons. A total of 219 patients were invited and 193 agreed to participate (88% response rate). Among the 193

index TB patients, the prevalence with 95% confidence interval (CI) of known DM, newly diagnosed DM and overall DM were 24.9% (18.9–31.5), 7.8% (4.4–12.5) and 32.6% (26.1–39.7), respectively. Eight (12.7%) patients were aged between 29 and 39 years.

≥ ≥ ≥ ≥ **Figure 1.** Flow chart of recruiting and investigation for diabetes mellitus (DM) among newly diagnosed index tuberculosis (TB) patients. MDR-TB—Multidrug resistant TB; RBG—Random capillary blood glucose; FBG—Fasting capillary blood glucose; Known DM—Patients with previous history of DM diagnosed by a medical doctor; New DM—Newly diagnosed DM patients (RBG ≥ 200 mg/dl & FBG ≥ 126 mg/dl (or) RBG ≥ 200 mg/dl for two times on separate days (or) FBG ≥ 126 mg/dl for two times on separate days).

#### *3.2. Prevalence and Glycaemic Control of DM among Household Contacts*

Figure 2 shows a flow chart of the 347 household contacts aged 25 years and above who were investigated for DM. Among them, 33 had history of diagnosed with DM. Among the 314 household contacts without a history of DM and invited to have an FBG test in township TB clinics, 139 (44.3%) attended. The remaining 156 contacts (49.7%) had their FBG tested at their home during the household revisit.

In total, 328 contacts completed all DM investigations. Among these, 33 (10.1%) were known to have DM and 13 (4.0%) were newly diagnosed, resulting in an overall prevalence of DM among household contacts of 14.0%. Most (94%) of the contacts with DM were aged 40 years and above. Among the 33 known cases, 21 (63.6%) had poor glycaemic control.

≥ ≥ ≥ ≥ ≥ **Figure 2.** Flow chart for investigation for diabetes mellitus (DM) among household contacts. \* Capillary blood test—Only fasting capillary blood glucose (FBG) test was done for known DM; and both random capillary blood glucose (RBG) test and FBG test were done for contacts without history of DM; Known DM—Household contacts with previous history of DM diagnosed by a health care personnel; Uncontrolled DM—FBG ≥ 130 mg/dl among known DM; New DM—Newly diagnosed DM (RBG ≥ 200 mg/dl & FBG ≥ 126 mg/dl (or) RBG ≥ 200 mg/dl for two times on separate days (or) FBG ≥ 126 mg/dl for two times on separate days).

#### *3.3. Prevalence of TB among Household Contacts*

Figure 3 shows a flow chart of the 553 household contacts who were investigated for TB. One household contacts among them was known case of TB. All household contacts, except the known case of TB, were invited to township TB clinics to have CXR regardless of having signs and symptoms of TB. Among them, 249 (45.1%) complied after the 1st home visit and 190 (34.4%) complied after the 2nd visit.

Overall, 439 (79.4%) completed all TB investigation. One (0.2%) was a known case and 21 (4.8%) were newly diagnosed, resulting in an overall TB prevalence of 5.0% among household contacts. MDR-TB was not detected.

#### *3.4. Comparing Prevalence of DM and TB among Household Contacts of Index TB Patients with and without DM*

Table 1 compares the prevalence of DM and TB among household contacts of TB patients with and without DM. The prevalence of DM, including known and newly diagnosed cases, was higher among household contacts of index TB with DM, although the difference was statistically significant only for the overall prevalence of DM. The prevalence of TB among household contacts, however, was comparable between the index TB with DM and without DM groups.

**Figure 3.** Flow chart for investigation for tuberculosis (TB) among household contacts.



N/A–Not applicable; \* including both existing DM and household contacts who completed both RBG and FBG tests; CI—confidence interval; † Fisher's exact test.

#### *3.5. Number of Patients with DM among Household Contacts* ≥ *25 Years Old*

Among household contacts aged ≥ 25 years without TB, 32 (12.1%) were diagnosed with DM, including all newly diagnosed cases.

Among household contacts aged ≥ 25 years with TB, one case was found to have existing DM with poor glycaemic control.

#### *3.6. Odds Ratio of Getting DM and TB among Household Contacts Based on DM Status of Index TB Patients*

Table 2 shows the odds ratios of obtaining DM and TB among household contacts, based on the DM status of index TB patients adjusted for covariates. From models D1 to D7, household contacts of index TB patients with DM had significantly higher risks of getting DM than household contacts of index TB patients without DM. Among these models, D2 had the lowest AIC. Based on this model, household contacts of index TB–DM patients had 2.13 times higher risk of getting DM than household contacts of index TB patients without DM.

From models T1 to T5, the risk of getting TB was lower in household contacts of index TB patients with DM, although the risk was not significantly different.


**Table 2.** Odds of getting DM and TB among household contacts based on the DM status of index TB patients in combination with various covariates.

OR—odds ratio; CI—confidence interval; AIC—Akaike's information criterion; \* *p* value < 0.05; SES—Socioeconomic status including formal education, employment and daily income per household member; Closeness with TB patients—sharing same room, sleeping in same bed and taking care of index TB patients.

#### **4. Discussion**

In this study, comorbidity with DM among 32.6% of index TB patients highlighted the increased double burden of TB and DM in Yangon. The adherence to DM and TB screening among household contacts was increased with repeated home visits. Among household contacts, the prevalence of DM and TB were 14.0 and 5.0%, respectively, and more than a quarter of DM and 95% of TB were newly diagnosed. Among household contacts with existing DM, almost 63.6% had poor glycaemic control. The risk of DM among household contacts of index TB patients with DM was two times higher than those of TB patients without DM. There was no difference in the prevalence of TB among household contacts of both index TB patient groups.

The adherence to having an FBG test in a TB clinic after the 1st home visit was 44.3% and an additional 49.7% were tested in their homes during the 2nd revisit. Therefore, testing FBG during household visits increased the screening coverage of household contacts. The adherence of household contacts to TB investigation in clinics after the 1st home visit was similar to a study conducted in South Africa (45.1%) [38]. The adherence rate in our study was increased up to 79.5% after the 2nd visit. Therefore, repeated home visits achieved more than 30% adherence to TB investigation.

Among index TB patients, the prevalence of DM, 32.6% (95% CI: 26.1–39.7) was significantly higher than that in the general population 12.1% (95% CI: 8.4–17.0) [8]. After adjustment for age, the risk of DM among index TB patients was 2.8 times greater than the risk of DM in the general population. This finding is similar to the results of a meta-analysis, where the risk of DM was found to

be three times higher in TB patients [2]. In the current study, the prevalence of newly diagnosed DM among 29 to 39 year old TB patients was almost 13% of all newly diagnosed DM cases. The current Myanmar guideline on screening for DM confined to those aged ≥ 40 years for TB patients would miss nearly 13% of the DM cases among TB patients in the Yangon Region.

Among household contacts who were aged 25 years or older, 14.0% (95% CI: 10.6–18.4) had DM, which was slightly higher than the 12.1% (95% CI: 8.4–17.0) prevalence in the general population [8]. After adjustment for age, no significant difference was found between the risk of DM among household contacts of TB patients and the risk of DM in the general population. Around 94% of patients with DM among household contacts were aged 40 years or over, which was similar with a previous Indian study, where household contacts aged 35 years and above comprised almost 93% of total DM patients [39]. Household contacts of TB–DM patients were more likely to have DM than those of TB patients without DM. This could be explained by genetic linkage similarity in risk behaviours among subjects in the same household.

Our contact investigation could identify around 4.0% of new DM. In a survey in the Yangon urban area, 7.6% of the participants were identified as new DM based on a one time FBG test [8]. In our study, confirmation of DM was defined based on results of both RBG and FBG tests. Therefore, the specificity of the test is higher than that of either test [40]. The prevalence of DM among household contacts without TB was 12.1%, including all newly diagnosed DM patients. These newly diagnosed DM patients comprised almost 25% of all DM patients among household contacts. Although integration of DM screening in household contacts of TB patients increases the case detection rate of a relatively small percentage of hidden DM cases in a community, newly detected DM cases are more at risk to TB than those detected by other screening methods.

Our overall prevalence of TB among household contacts (5.0%) was similar with a systematic review of contact investigation done in low and middle-income countries [11]. The prevalence was higher than studies done in China (3.8%) estimated by the same method [41], but lower than the 13.8% found in a study done in Mandalay region, Myanmar [25]. The difference in prevalence of TB among household contacts can be explained by different TB backgrounds in different populations, and the lifestyles and living conditions of people in each area [42]. There was no significant difference in prevalence of TB among household contacts of index TB patients with and without DM. This result is similar to a previous one reported from India [39].

In the current study, nearly two-thirds of known DM cases among household contacts had poor glycaemic control. The corresponding percentage was 54% among the general population but the cut off value for poor glycaemic control was FBG ≥ 126 mg/dl, which was lower than FBG ≥ 130 mg/dl in our study [8]. A previous study conducted in the US found that uncontrolled DM patients continued to have a significantly elevated risk of TB infection (OR, 2.6; 95% CI, 1.5–4.6) compared to nondiabetics [43]. A study conducted in Taiwan reported that DM patients with poor glycaemic control had a significantly higher hazard of TB than those without DM during a median follow-up time of 4.6 years [18]. In the current study, screening of DM was done among household contacts of recently diagnosed TB patients and only one TB case was identified among household contacts with poorly controlled DM. Therefore, these DM patients need good follow up to improve their DM conditions and reduce the risk of developing TB in the future.

#### **5. Limitations**

Our screening of DM was based on the finger prick method due to limited resources. The prevalence of DM may be different had the standard glucose concentrations in plasma been used [31,44]. Screening using CXR without sputum culture among those whose sputum could not be obtained might have led to an overestimated prevalence of TB among our contacts.

#### **6. Conclusions**

Despite these limitations, our findings indicate that screening of DM during household visits increased feasibility and repeated household visits improved the adherence of TB screening using CXR among household contacts. In a high TB and DM co-prevalent area, DM screening and glycaemic control assessment for TB household contacts aged 40 years or over should be integrated with routine TB screening programmes.

**Author Contributions:** N.-N.Z. completed this article as part of his Ph.D. thesis project supervised by V.C. and R.S. All authors contributed equally to this manuscript. N.-N.Z., V.C. and R.S. designed this study. N.-N.Z. collected and analysed the data, and prepared the first draft. S.S. and S.T.A. supervised the data collection. V.C. and R.S. supervised the data analysis. V.C. and R.S. reviewed the literature, contributed to interpretation of the results and reviewed the final manuscript. All authors gave equal contribution on revising and preparing the final version. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was a part of the PhD/Master thesis of the first author to fulfil the requirement for the TB/MDR-TB research training program at Epidemiology Unit, Prince of Songkla University under the support of Fogarty International Center, National Institutes of Health—Grant number D43TW009522. Additional funding also supported from the Department of Medical Research, Myanmar and the Graduate School Scholarship from Prince of Songkla University.

**Acknowledgments:** We would like to acknowledge all the respondents, research assistants, Township Tuberculosis Coordinators, Township Health Officers, and Basic Health Staff in research townships and responsible persons in Yangon Regional Tuberculosis Program, Department of Public Health, Myanmar.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Tropical Medicine and Infectious Disease*

#### *Perspective*

### **An Opportunity to END TB: Using the Sustainable Development Goals for Action on Socio-Economic Determinants of TB in High Burden Countries in WHO South-East Asia and the Western Pacific Regions**

**Srinath Satyanarayana 1,2,\* , Pruthu Thekkur 1,2, Ajay M. V. Kumar 1,2,3 , Yan Lin 2,4 , Riitta A. Dlodlo <sup>2</sup> , Mohammed Khogali <sup>5</sup> , Rony Zachariah <sup>5</sup> and Anthony David Harries 2,6**


Received: 12 May 2020; Accepted: 12 June 2020; Published: 18 June 2020

**Abstract:** The progress towards ending tuberculosis (TB) by 2035 is less than expected in 11 high TB burden countries in the World Health Organization South-East Asia and Western Pacific regions. Along with enhancing measures aimed at achieving universal access to quality-assured diagnosis, treatment and prevention services, massive efforts are needed to mitigate the prevalence of health-related risk factors, preferably through broader actions on the determinants of the "exposure-infection-disease-adverse outcome" spectrum. The aim of this manuscript is to describe the major socio-economic determinants of TB and to discuss how there are opportunities to address these determinants in an englobing manner under the United Nations Sustainable Development Goals (SDGs) framework. The national TB programs must identify stakeholders working on the other SDGs, develop mechanisms to collaborate with them and facilitate action on social-economic determinants in high TB burden geographical areas. Research (to determine the optimal mechanisms and impact of such collaborations) must be an integral part of this effort. We call upon stakeholders involved in achieving the SDGs and End TB targets to recognize that all goals are highly interlinked, and they need to combine and complement each other's efforts to end TB and the determinants behind this disease.

**Keywords:** tuberculosis; End TB; sustainable development goals; South-East Asia; Western Pacific Region; national TB program; socio-economic determinants

#### **1. Introduction**

Tuberculosis (TB), caused by the bacteria *Mycobacterium tuberculosis* (Mtb), is one of the top 10 leading causes of death world-wide and the leading cause of death from a single infectious agent [1]. Mtb spreads from person to person through airborne droplet nuclei. When a person with active TB of the lungs or throat, coughs or sneezes, droplets containing Mtb are expelled into the air and inhalation of this contaminated air may cause TB infection [2]. Once infected, about 5–15% of the people develop

active TB disease in their lifetime, with the risk of developing the disease being highest in the first two years of infection [3,4]. About 1.7 billion people or 23% of the world's population are estimated to be infected with Mtb, of which 55.5 million (0.8% of the world's population) are estimated to be recently infected and at high risk of progression to active TB [5]. The aim of this manuscript is to describe the major socio-economic determinants of TB and to discuss how there are opportunities to address these determinants in an englobing manner.

#### **2. TB Burden: Globally and in the WHO South-East Asia and Western Pacific Regions**

Globally, about 10 million people developed TB in 2018, with this number being relatively stable in recent years. The annual TB incidence rate ranged from less than 5 people per 100,000 population to more than 500 people per 100,000 population, with the global average of 130 people per 100,000 population. In 2018, an estimated 1.2 million HIV-negative people and an additional 251,000 HIV-positive people died from TB. Although these numbers are large, the number of deaths among HIV-negative people has reduced by 27% from 1.7 million in 2000 to 1.2 million in 2018. Similarly, the number of deaths among HIV-positive people has reduced by 60%, from 620,000 in 2000 to 251,000 in 2018.

Geographically, in 2018, the TB burden varied across the six regions of the World Health Organization (WHO): this comprised 44% in the South-East Asia region [6], 24% in the Africa region, 18% in the Western Pacific region, 8% in the Eastern Mediterranean region, and 3% each in the Americas and Europe. Globally, 30 high TB burden countries accounted for 87% of the world's patients, with 11 of these being in the South-East Asia and Western Pacific regions (Table 1). Five countries in the South-East Asia and Western Pacific regions accounted for more than half of the global total: India—27%, China—9%, Indonesia—8%, the Philippines—6%, and Bangladesh—4%. Even within countries, the distribution of the TB burden is highly heterogeneous [7].


**Table 1.** List of high TB burden countries in the WHO South-East Asia and Western Pacific Regions.

Source: WHO Global TB Report 2019 [6].

#### **3. End TB Strategy**

In 2014, to rid mankind of the enormous burden of TB, the 67th World Health Assembly adopted a resolution to make the world free of TB by the year 2035. WHO's "End TB Strategy" provides a holistic overview of this resolution and has four principles and three pillars [8]. The three high-level target indicators of the End TB Strategy are reductions in TB deaths by 95%, reductions in the TB incidence rate by 90% and the percentage of TB patients and their households experiencing catastrophic costs being maintained at zero. These indicators and targets are relevant to all countries, with interim milestones to be achieved by 2020, 2025 and 2030.

#### *Progress towards the 2020 Milestones of the End TB Strategy*

By the end of 2019, at the global level, most of the WHO regions and many high TB burden countries were not on track to reach the End TB Strategy's 2020 milestones (20% reduction in TB incidence rate, 35% reduction in the number of TB deaths and reduction in the households experiencing catastrophic costs to 0%). The reduction in the cumulative global TB incidence rate between 2015 and 2018 was only 6.3%, and the reduction in the total number of TB deaths between 2015 and 2018 was 11% [6]. Table 2 shows the reduction in the TB incidence rate, TB mortality, and catastrophic costs in the 11 high TB burden countries in the Asia Pacific region (which comprises South-East Asia and the Western Pacific). However, with this rate of decline in incidence and mortality, and with the data on catastrophic costs unavailable, it is unlikely that any of the countries in the Asia Pacific region will be able to reach all the End TB Strategy's 2020 milestones.


**Table 2.** High TB burden countries in the WHO South-East Asia and Western Pacific Regions and the relative change in crucial indicators between 2015 and 2018.

Source: Point estimates from WHO Global TB Report 2019 [6]; NA = Data Not available; \* negative sign indicates an increase in the TB incidence rate or number of TB deaths.

#### **4. Factors Influencing the Risk of Exposure to Mycobacterium tuberculosis (Mtb), Infection, the Progression from Infection to Disease and Adverse Treatment Outcomes**

There are two important aspects to understanding the TB epidemiology. First, Mtb infection in humans results in a spectrum of clinical presentations. As mentioned earlier, most infections are subclinical and asymptomatic, with Mtb replication contained by the host immunity—a condition called latent TB infection (LTBI)—and only a small subset of infected individuals presenting with symptomatic, active TB disease. Even within and between these two states, there is a wide ranging spectrum of Mtb bacterial load, immune responses, pathologies and clinical presentations [9]. Second, like all other infectious diseases, the risk of infection and disease is dependent on the characteristics and interaction of the bacteria, the human host and the environment [7]. A good understanding of these factors and their unique complex interactions—at both the population level and the individual level—is crucial for designing the intervention strategies to mitigate the TB burden.

The factors that influence the risk of exposure to Mtb, infection, the progression of infection to disease, and adverse treatment outcomes (such as death) are shown in Box 1 [10,11]. The factor that is essential for TB infection and disease is close contact with a person with a person with infectious TB disease; the greater the closeness, bacterial load and duration of contact, the higher the chances of infection. Other factors such as age, sex, tobacco use, alcohol use, malnutrition, human immunodeficiency virus (HIV) infection, diabetes mellitus and silicosis increase the risk of infection, the progression from infection to disease and adverse TB treatment outcomes, and are therefore called major health-related risk factors. Factors such as poverty, socio-economic and/or

gender inequality, food and/or job insecurity and weak health systems affect not only all aspects of the "exposure-infection-disease-adverse outcome" spectrum, but also several aspects of the health of populations in general, and are therefore called the critical underlying 'determinants' or the 'root causes' of TB. While age and sex are not modifiable, all the other factors listed in Box 1 can be modified by human interventions. Table 3 shows the health-related risk factors and the corresponding lifetime increase in the risk of TB disease and the 'population attributable fraction (PAF)' of these factors [12–18].

**Table 3.** Health-related risk factors and the corresponding lifetime increase in the risk of TB disease and their corresponding population attributable fraction.


\* Source: Global TB report 2019 [6]); \*\* these are approximates and vary widely across countries; \*\*\* PAF indicates the proportion of all cases of a particular disease in a population that is attributable to a specific exposure and is estimated based on the relative risk and the prevalence of the risk factor in the population [18].

**Box 1.** Factors influencing the "exposure-infection-disease-adverse outcome" spectrum of TB.

#### **The essential factor for TB infection and TB disease**

• Close contact with a person with infectious TB disease

#### **Major health-related risk factors (factors that increase the chances of infection, disease, adverse TB treatment outcomes)**

• Age, sex, tobacco use, alcohol abuse, malnutrition, HIV infection, diabetes mellitus, exposure to indoor air pollution, silicosis, intake of immunosuppressive drugs/medications (e.g., tumor necrosis factor-alpha (TNF) antagonists, corticosteroids)

#### **Major underlying determinants**

• Poverty, socio-economic and gender inequality, overcrowding, food and job insecurity, weak health systems

#### *4.1. Relationship between Various Health-Related Risk factors and Major Underlying Socio-Economic Determinants of TB*

In the past, the trends in TB incidence rates in high-income countries clearly show that efforts towards improving the socio-economic status, living conditions and nutritional status (as was seen before and soon after the world wars) resulted in the rapid decline in the TB burden, and, deterioration in these conditions (during times of war) increased the TB incidence rates. Both in high and low-income countries, TB predominantly affects people of lower socio-economic status [19,20].

Most of the risk factors for TB are associated with poverty, socio-economic and gender inequalities, and living conditions [21]. Malnutrition, poor housing/living conditions, and overcrowding are direct markers of poverty [22]. People from lower socio-economic groups are more likely to live and/or work in overcrowded settings, experience greater food insecurity, have lower levels of awareness about healthy behavior, and are less likely to have access to quality health care services [23]. They are also more likely to come into contact with people with active TB disease. The prevalence of tobacco use, alcohol use, HIV, and diabetes mellitus is relatively higher in people of low socio-economic status groups in various settings [24–28].

The Multisectoral Accountability Framework to accelerate progress to end TB (MAF-TB) by 2035 developed by the WHO in 2019 [29] urges governments to address a wide range of socio-economic determinants through collaborations and partnerships. Although the primary responsibility to pursue public health in all policies rests with different ministries within governments, the national TB programs, as champions and implementers of the TB care and prevention services in the countries, should take the lead in developing partnerships and support the implementation of the multisectoral accountability framework, both through advocacy and by helping to address the social conditions of patients and their families.

#### *4.2. Need for Action on Risk Factors and Major Determinants of TB*

TB (as described earlier) is a multifactorial disease, and the achievement of the End TB Strategy milestones requires universal access to quality-assured diagnosis, treatment, and prevention services promptly. For this, it is necessary to strengthen the national TB programs, by ensuring adequate resources for deploying latest WHO-endorsed rapid TB diagnostics and drug susceptibility testing (DST) facilities, the provision of appropriate treatment services for drug-susceptible and drug-resistant TB, preventive treatment for high risk individuals (people living with HIV and household and other close contacts of TB patients), and the implementation of infection control measures in all health facilities. While these are necessary, it is being increasingly recognized that End TB targets are ambitious and unlikely to be achieved by these measures alone [10,11,30]. Massive efforts are needed to mitigate the prevalence of health-related risk factors, preferably through broader actions on the determinants of the "exposure-infection-disease-adverse outcome" spectrum, such as health system strengthening, poverty alleviation, addressing socio-economic and gender inequality, limiting job loss and food insecurity, improving housing quality and reducing overcrowding. An increase in the health-related risk factors of TB or the worsening of the determinants of TB (as is currently happening due to the socio-economic consequences of the SARS-CoV-2 pandemic [31]) can harm the global progress made towards ending TB.

#### **5. Framework for Action on Social Determinants of Tuberculosis: Sustainable Development Goals**

The sustainable development goals (SDGs) [32] are a list of 17 global goals, which outline the vision, principles, and commitments of all United Nations member countries to a fairer and more sustainable world, now and in the future. The SDGs, launched in 2015 by the United Nations General Assembly, and intended to be achieved by the year 2030, are part of UN Resolution 70/1, the 2030 Agenda. The 17 SDGs are mentioned in Box 2.

**Box 2.** The Sustainable Development Goals (2015 to 2030) \*.


The seventeen SDGs each have a list of targets that are measured with indicators [33]. These goals necessitate collaboration and the alignment of all actions to secure a fair, healthy, and prosperous future for everyone on this planet—Earth. These SDGs provide a framework for action on the determinants of TB disease. Action on the determinants of TB through the SDGs framework will also mean endorsing the socio-ecological model of health which outlines that disease prevention and its mitigation may require action at five levels: individual, interpersonal, organisational, community and public policy [34] and addressing every TB determinant may require actions at these five levels.

#### *5.1. Sustainable Development Goals that Are Likely to Have a Significant Impact on the Burden of TB*

#### 5.1.1. SDG Goal 1: No Poverty

As discussed above, poverty is a significant determinant of several aspects of health including TB [22,35]. Globally, more than 90% of TB cases and deaths occur in low and middle income countries [36]. There is an inverse correlation between a country's gross domestic product (GDP) per capita and TB incidence rates [21]. Therefore, efforts to reduce poverty will have a substantial impact on the TB burden. Despite great progress made globally in reducing poverty levels, some countries in the WHO South-East Asia and Western Pacific Regions, such as India (~21%) and Papua New Guinea (38%), have high reported levels of people living below the international poverty line (SDG Indicator 1.1.1 in Table (SDG Indicator 2.1.1 in Table 4).

#### 5.1.2. SDG Goal 2: Zero Hunger

Hunger leads to undernutrition, which is one of the significant determinants of TB [37]. It is estimated that in India (the highest TB burden country in the world), where the prevalence of undernutrition is high, nearly 50% of TB cases are attributable to undernutrition [38,39]. Undernutrition is also ubiquitous in all high TB burden countries in Asia and the Pacific region (SDG Indicator 2.1.1 in Table 4). Therefore, ending hunger and improving the nutritional status of populations can dramatically reduce the burden of TB.

#### 5.1.3. SDG Goal 3: Good Health and Well Being

One of the key targets of this goal (Target 8) is to achieve universal health coverage (UHC) by 2030 [40]. UHC means that everyone can access and receive sufficient and quality health services that they need without suffering financial hardship. There are two key indicators to monitor progress. They are: a) UHC service coverage index (SCI)—SDG Indicator 3.8.1 (Table 5), and b) the percentage of the population experiencing expenditures on health care that are relatively large in relation to the household expenditures or income—SDG Indicator 3.8.2 (Table 5). The achievement of UHC and improved patient-centered TB care will have a direct effect on the reach and delivery of quality TB services and on catastrophic costs incurred by TB patients and their families [40]. Apart from this, the SDG goal also has indicators for dramatically reducing the prevalence of HIV, tobacco and alcohol use, and diabetes mellitus, all of which will have a considerable impact on reducing the TB burden.

#### 5.1.4. SDG Goal 4: Quality Education

Quality education typically leads to better and secure jobs, more money and higher purchasing power, resulting in better access to quality health care (including for TB) [41]. Higher earnings also allow people to afford homes in safer neighbor hoods, as well as consume healthier diets. Incorporating health within the ambit of 'quality education' builds knowledge, skills, and positive attitudes about health and all other determinants of health, which can directly or indirectly have a considerable impact on efforts to End TB [42]. Education also improves the ability to identify the symptoms suggestive of TB and seek timely care for diagnosis and treatment of TB [43], thus, limiting the delays in the diagnosis of TB and the community spread of the disease.

#### 5.1.5. SDG Goal 5: Gender Equality

TB can affect either gender. In almost all countries, the notification rates of TB are higher in males than in females. Although more men than women develop TB disease and die from it, TB is nevertheless a leading infectious cause of death among women. Higher tuberculosis notification rates in men partly reflect epidemiological and biological differences in exposure, risk of infection, and progression from infection to disease [44]. Despite this, in several countries, gender inequality, socio-economic, and cultural factors act as barriers to accessing health care among women. These may lead to the under-detection and under-notification of TB in women. The stigma and discrimination associated with TB and certain co-morbidities, such as HIV infection, adversely affect women more than men, often leaving them in a more vulnerable position [45,46]. It is also widely believed that the medical care-seeking behavior of men and women with TB is mostly determined by how they and those around them perceive the symptoms, regard the diagnosis, accept the treatment, and complete it. Gender may influence each of these components and affect the early detection of the disease and its outcome [47]. Studies that have assessed gender differences have shown that, on average, women are either undiagnosed, or diagnosed late in the course of TB disease when compared to men [48,49]. Promoting gender equality in all spheres of life helps to mitigate some of these issues and contribute towards ending TB.

#### 5.1.6. SDG Goal 6: Clean Water and Sanitation

Access to clean water and sanitation is essential to reduce illness, malnutrition, poor physical and cognitive development, and death due to water-borne diseases [50]. In countries of the Asia-Pacific region, the proportion with access to clean water ranges from 16.7% in rural Cambodia to 92.3% in urban China. The proportion with access to safe sanitation ranges from 5.1% in rural Democratic People's Republic of Korea to 83.7% in urban China. The provision of clean water and sanitation can affect the TB burden by reducing infections and improving nutritional status.

#### 5.1.7. SDG Goal 7: Affordable and Clean Energy

There is evidence of an association between indoor air pollution (such as that caused by the burning of solid fuels for cooking at homes), outdoor ambient air pollution and TB infection and TB disease [51–54]. Depending on the prevalence of indoor air pollution, the fraction of TB cases attributable to indoor air pollution varies across countries. The proportion of the population using clean fuels in the South-East Asia and Western Pacific regions ranges from 11% in DPR Korea to 74% in Thailand (Table 4; indicator 7.1.2). Interventions such as clean cook stoves to reduce the adverse effects of indoor air pollution merit rigorous evaluation [50], particularly in high TB burden countries in Asia and the Pacific, where the prevalence of both indoor air pollution and TB is high. Clean energy is also expected to reduce outdoor air pollution and improve ambient air quality, which can reduce the risk of TB [55].

#### 5.1.8. SDG Goal 8: Decent Work and Economic Growth

TB predominantly affects people in the economically productive age-groups [56]. Apart from providing stable and regular job opportunities for the economically productive age groups, provision for early diagnosis and treatment of TB at workplaces, making workplaces safe by reducing the chances of airborne transmission of infections, adopting favorable policies towards social/job security in case of diseases and reducing occupational diseases like silicosis, will help in reducing the TB burden and improving the economic productivity of the workforce [57]. Stable/formal employment also increases access to employer-sponsored social health insurance programs and paid sick leaves, both of which are known to be associated with the reduced risk of occupational diseases and all-cause mortality [58–60].

#### 5.1.9. SDG Goal 10: Reduced Inequalities

Empowering and promoting social, economic and political inclusion of all, irrespective of age, sex, sexual orientation, disability, race, ethnicity, origin, religion or economic or another status will help in mitigating the effects of socio-economic disparities—disparities that are key drivers of all the risk factors of TB [61].

8 19

#### 5.1.10. SDG Goal 11: Sustainable Cities and Communities

Rapid industrialization, urbanization, and migration—dominant occurrences in most developing countries in the Asia Pacific region—can create ideal conditions for infectious diseases (including TB) to flourish [62,63], unless accompanied by proper urban planning, social reforms, environmental protection, adequate housing, transportation, and a well-coordinated and robust health system.

#### 5.1.11. SDG Goal 13: Climate Action

Climate change that manifests itself in the form of higher variations in temperatures and rainfall is known to have a substantial effect on several aspects of human health and behaviour (such as crowding, migration, changes in food habits), either directly or through several intermediaries, resulting in an increase in the burden of infectious diseases including TB [64,65].

The linkages between the various SDGs and TB are pictorially depicted in Figure 1.

**Figure 1.** The possible direct and indirect linkages between SGDs 1, 2, 3, 4, 5, 6, 7, 8, 10, 11, 13 and reduction in tuberculosis burden. The arrows indicate the probable direction of action, with the bi-directional arrow indicating that effects in both the directions are possible. The relationships/pathways shown in this figure are 'indicative' only and not 'definitive'. The size of the square boxes in the figure have been arrived at using the best fit function, and therefore the varying sizes or shapes of the text boxes/circles in the figure do not carry any special significance. SDGs 1, 4, 5, 8, 10 are grouped together, to indicate the author's view that they are interdependent and can act alone or in a combined manner to influence the other aspects in the pathway.

*Trop. Med. Infect. Dis.***2020**,*5*, 101


**Table 4.**Current levels of TB relevant SDG 1, 2, 7, 8, 10, 11 indicators in high TB burden countries in the WHO South-East Asia and Western Pacific Regions (in 2018).

 PPP = Purchasing Power Parity; NA = Not available; Source of data: [66–68]; \* Gini index = measure of income distribution across income percentiles in a population. It ranges from 0% to 100%, with 0% representing perfect equality and 100% representing perfect inequality.

*Trop. Med. Infect. Dis.***2020**,*5*, 101


**Table 5.**Estimated prevalence of TB relevant SDG-3 indicators in high TB burden countries in the WHO South-East Asia and Western Pacific Regions (in 2018).

> NA=Not available; Source of data: [66–68].

#### **6. Role of National TB Programs in Accelerating the Progress towards Achieving SDGs**

Together, TB and poverty form a vicious cycle: TB decreases people's capacity to work and adds to treatment expenses. This, in turn, exacerbates their poverty. Poor people also go hungry and live in close, unhygienic quarters, where TB and its risk factors flourish. Progress in ending TB will accelerate the progress on SDG goal 1, and through it, progress on other related SDGs [36].

According to the WHO's "A Multisectoral Accountability Framework to accelerate progress to end Tuberculosis by 2030 (MAF-TB)" [29], the following are recommended actions for national TB programs:


WHO has identified the following fourteen SDG indicators as relevant to TB. These include

	- (1) SDG 1 (No poverty)—Indicator 1.1.1: Proportion of the population living below the international poverty line
	- (2) SDG 1 (No poverty)—Indicator 1.3.1: Proportion of the population covered by social protection floors or systems
	- (3) SDG 2 (Zero hunger)—Indicator 2.1.1: Prevalence of undernourishment
	- (4) SDG 7 (Affordable and clean energy)—Indicator 7.1.2: Proportion of the population with primary reliance on clean fuels and technology
	- (5) SDG 8 (Decent work and economic growth)—Indicator: 8.1.1: Gross domestic product (GDP) per capita
	- (6) SDG 10 (Reduced inequalities)—Indicator: 10.1.1: Gini index for income inequality
	- (7) SDG 11 (Sustainable Cities and Communities)—Indicator: 11.1.1: Proportion of the urban population living in slums

The levels of these seven indicators under SDG 1, 2, 7, 8, 10 and 11 at the end of 2018 in the 11 high TB burden countries in WHO South-East Asia and Western Pacific Regions are given in Table 4.

	- (1) SDG Indicator 3.8.1: Coverage of essential health services (UHC) measure with an UHC index based essential health services and ranges between 0 and 100
	- (2) SDG Indicator 3.8.2: Proportion of the population with large household expenditures on health, as a share of total household expenditure or income
	- (3) SDG Indicator 3.c.1: Current health expenditure per capita in current international dollars
	- (4) SDG Indicator 3.3.1: Prevalence of HIV
	- (5) SDG Indicator 3.a.1: Prevalence of smoking
	- (6) SDG Indicator 3.4.1: Prevalence of diabetes
	- (7) SDG Indicator 3.5.2: Prevalence of alcohol use disorder

The levels of these seven within SDG-3 indicators at the end of 2018 is in the 11 high TB burden countries in WHO South-East Asia, and Western Pacific Regions is given in Table 5.

National TB programs, Ministries of Health and Governments in the 11 high TB burden countries must realize that interventions beyond diagnosis and treatment of TB and LTBI are needed to reduce the risk factors and determinants of TB. TB programs must take a lead and list the stakeholders in their countries who are working on the other SDGs and equip themselves with resources and the necessary skillsets to engage with them. Since TB is not homogenously distributed within the country, there will be geographical areas/communities with high TB burdens. As a first step, TB programs should share details/information of high TB burden geographical areas, assess the socio-economic determinants that are locally relevant/prevalent, and facilitate concentrated action on those socio-economic determinants as a priority in these areas wherever possible. The list of sample interventions that can be undertaken to reduce the prevalence of the socio-economic determinants of TB (based on needs assessment) are given in Table 6. (In this table, using India as an example, we have highlighted public programs [69–82] that can be galvanized to address the determinants at the local level.)

The stakeholders/partners who can be involved in carrying out the interventions could be respective government departments (who have the mandate and jurisdiction to perform the intervention), non-governmental/community based organizations, private sector, developmental partners etc., who may share the common vision of improving the lives of the people. Identifying suitable partners will help the national TB programs, in facilitating their interventions in such geographic high TB burden areas. This holistic approach may contribute towards ending TB in such communities or geographic areas in a sustainable manner. Demonstration projects on how to operationalize intersectoral coordination and build partnerships (at the local level) are urgently needed to generate evidence and to show that the impact of such initiatives goes beyond the simple sum of its immediate returns. Research—to find out the optimal mechanisms for collaboration and to measure the impact of such collaborations—must also be undertaken, so that the lessons learnt are disseminated widely. Apart from this, national TB programs must also find themselves represented in all in-country developmental committees and agendas of the government, so that ending TB is seen, not just as a responsibility of the health sector, but seen as a necessity for human development in all spheres of life. International technical and funding agencies/organizations, like WHO, the Global Fund, etc., should advocate for progress on broader SDGs and provide technical and financial assistance on this aspect to TB programs. **Table 6.** Sample policies/interventions that can be implemented at the community level in high TB burden geographical areas within a country by the National TB program and its partners to address determinants of tuberculosis.


#### **Table 6.** *Cont.*


### **7. Conclusions**

Achieving End TB targets with the current pace of progress is highly challenging and un-realistic if the thrust is mainly medical and focusing only on 'diagnosis and treatment' of TB and LTBI, without addressing the underlying determinants of TB. While the strengthening of national TB programs under the framework of universal health coverage is quintessential for accelerating the progress towards End TB targets, the SDG framework provides an excellent opportunity for acting on several determinants and risk factors of TB. All stakeholders, be it government ministries, non-governmental organizations or private sectors involved in achieving SDGs and the End TB targets must recognize that most of their goals are strongly interlinked. Failure to acknowledge this fact may result in ineffective and inappropriate actions and a delay in the achievement of both SDG goals and End TB targets.

**Author Contributions:** S.S. and P.T. wrote the first draft which was critically reviewed by A.M.V.K., Y.L., R.A.D., M.K., R.Z. and A.D.H. All authors contributed to subsequent drafts and agreed upon and approved the final version. All authors have read and agreed to the published version of the manuscript.

**Funding:** The research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Disclaimer:** The views expressed in this document are those of the authors and may not necessarily reflect those of their affiliated institutions.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Tropical Medicine and Infectious Disease*

#### *Article*

### **Yield of Systematic Longitudinal Screening of Household Contacts of Pre-Extensively Drug Resistant (PreXDR) and Extensively Drug Resistant (XDR) Tuberculosis Patients in Mumbai, India**

**Roma Haresh Paryani <sup>1</sup> , Vivek Gupta <sup>2</sup> , Pramila Singh <sup>1</sup> , Madhur Verma <sup>3</sup> , Sabira Sheikh <sup>1</sup> , Reeta Yadav <sup>1</sup> , Homa Mansoor <sup>1</sup> , Stobdan Kalon <sup>1</sup> , Sriram Selvaraju <sup>4</sup> , Mrinalini Das 1,\* , Chinmay Laxmeshwar <sup>1</sup> , Gabriella Ferlazzo <sup>5</sup> and Petros Isaakidis <sup>5</sup>**


Received: 13 April 2020; Accepted: 6 May 2020; Published: 15 May 2020

**Abstract:** While risk of tuberculosis (TB) is high among household contacts (HHCs) of pre-extensively drug resistant (pre-XDR) TB and XDR-TB, data on yield of systematic longitudinal screening are lacking. We aim to describe the yield of systematic longitudinal TB contact tracing among HHCs of patients with pre-XDR-TB and XDR-TB. At the Médecins Sans Frontières (MSF) clinic, Mumbai, India a cohort comprising 518 HHCs of 109 pre-XDR and XDR index cases was enrolled between January 2016 and June 2018. Regular HHC follow-ups were done till one year post treatment of index cases. Of 518 HHCs, 23 had TB (21 on TB treatment and two newly diagnosed) at the time of first visit. Of the rest, 19% HHCs had no follow-ups. Fourteen (3.5%) TB cases were identified among 400 HHCs; incidence rate: 2072/100,000 person-years (95% CI: 1227–3499). The overall yield of household contact tracing was 3% (16/518). Of 14 who were diagnosed with TB during follow-up, six had drug susceptible TB (DSTB); six had pre-XDR-TB and one had XDR-TB. Five of fourteen cases had resistance patterns concordant with their index case. In view of the high incidence of TB among HHCs of pre-XDR and XDR-TB cases, follow-up of HHCs for at least the duration of index cases' treatment should be considered.

**Keywords:** incidence; tuberculosis; household contacts tracing; operational research

#### **1. Introduction**

Tuberculosis (TB) is a long-standing global health issue, disproportionately affecting low- and middle-income countries. India is a country with a high TB burden [1]. Drug-resistant tuberculosis (DR-TB) is also reported to have an increasing trend in the country. Pre-extensively drug resistant (pre-XDR-TB) is the presence of resistance to rifampicin and isoniazid along with resistance to either one of the fluoroquinolones (ofloxacin, levofloxacin or moxifloxacin) or second-line injectables (amikacin, capreomycin or kanamycin). In case the TB bacilli have resistance to both fluoroquinolones and second-line injectables, along with rifampicin and isoniazid, they are classified as extensively drug resistant (XDR). In India, the estimated incidence of TB in India was 2,700,000 [2]. According to the drug-resistance survey report, in India, 28% patients with TB have resistance to any anti-tubercular drugs and 6.2% have multi-drug resistant TB (MDR-TB); among MDR-TB, 21.8% have pre-XDR-TB and 1.3% have XDR-TB [3]. Some studies from the country have reported a higher prevalence of XDR-TB [4–6].

Systematic reviews suggest that contact investigations are an important intervention in early identification of active TB cases, thus minimizing TB transmission [7–9]. Fox and colleagues have reported active TB among 3.4% contacts of patients with multi-drug-resistant or extensively drug-resistant TB in low-and middle-income settings [9]. In Pakistan, baseline contact investigation of index cases with MDR-TB identified MDR-TB in 17.4% and DS-TB in 4.2% of close contacts [10]. A longitudinal study of household contact (HHC) among newly diagnosed TB patients in Vietnam reported incident TB in 0.5% of contacts at 12-month follow-up and in 0.3% of contacts at 24-month follow-up, while the prevalence at baseline evaluation was 0.5% [11]. In China, four-year follow-up of close contacts of bacteria-positive TB index cases reported a high risk of incident TB (333/100,000 person years) [12].

To add to the evidence around longitudinal contact tracing in addition to baseline contact tracing among drug-resistant TB cases, the present study aimed to determine the yield of systematic longitudinal household contact tracing among patients with pre-XDR-TB and XDR-TB initiated on treatment at Médecins Sans Frontières (MSF) Clinic, Mumbai. Secondary objectives included (1) assessment of profile of index cases and HHCs developing TB and (2) drug resistance patterns in TB diagnosed among HHCs.

#### **2. Materials and Methods**

#### *2.1. Study Design*

This was a retrospective cohort study using routine program data.

#### *2.2. Setting*

The study was done in Mumbai, which is among the most populous cities of India. Located in the western part of India in the state of Maharashtra, it has a population density of 21,000/km<sup>2</sup> . The city has high rates of migration and is home to a large slum population. Nearly 9895 patients with DR-TB were diagnosed in Maharashtra, and TB case notifications were nearly 209,642 in 2018 [2,12–14].

The MSF clinic in Mumbai started offering treatment to patients with pre-XDR and XDR-TB in 2012 [15]. Eligibility criteria for enrollment are based on resistance profile and treatment history. Individualized treatment is offered on an outpatient basis based on drug sensitivity patterns. Bedaquiline and Delamanid are provided to patients who require it. All patients receive free-of-charge services, which include counseling, diagnosis, treatment, social enabler (dry ration), travel support and medical reimbursement for co-morbid conditions as per need. Treatment is provided using an ambulatory model of care described elsewhere [15].

#### 2.2.1. Household Contact (HHC) Investigation

Since 2013, the clinic has been performing investigation of HHCs of registered pre-XDR and XDR-TB patients. HHCs of DR-TB cases were members of the household regularly living with the patients registered for care in the MSF Clinic (hereafter mentioned as index cases). At enrollment of a new index case, a trained nurse visited their homes, a list of HHCs was prepared, and details were noted in a standardized form. All the HHCs are examined for signs and symptoms of TB and following risk factors: pregnancy, diabetes mellitus, HIV, alcohol use disorder, drug use, previous history of TB, age >55 years and immune-suppressive therapy.

Adults and adolescents HHCs over 15 years of age with risk factors and all pediatric (aged <15 years) HHCs are invited to visit the MSF clinic for a detailed screening. Those who visit the clinic are assessed using chest X-ray (CXR) and physical examination. If they have symptoms suggestive of TB or CXR abnormality, sputum examination or gastric lavage for smear, cartridge based nucleic acid amplification test (CBNAAT)/GeneXpert and culture and drug susceptibility testing (C-DST) are done. HHCs that are diagnosed with active TB are initiated on treatment as per their TB drug susceptibility pattern.

#### 2.2.2. Follow-Up of HHCs

HHCs aged >15 years and not having any high-risk factors are followed every six months in their homes. Those who have risk factors are followed at three-monthly intervals during the first year and then six-monthly at home. All children <15 years are followed up three-monthly at the MSF clinic. At each follow-up visit, screening of contacts is done for presumptive TB, defined as any of the symptoms/signs: (a) cough >2 weeks, (b) fever >2 weeks, (c) significant weight loss, (d) any abnormality in chest radiograph. Contacts with presumptive TB undergo detailed clinical, radiological and laboratory evaluations. All laboratory investigations are done in an accredited private laboratory contracted by MSF. In case a contact does not report for follow-up at the clinic, they are contacted telephonically and encouraged to visit the clinic. The follow-up continues for the entire duration of TB treatment of the index case and one year after treatment completion. In case of the death of the index case, follow-up of HHCs is still offered for the next one year.

#### *2.3. Study Population and Participants*

For the present analysis, we included all HHCs of index cases with pre-XDR and XDR-TB registered for treatment between January 2016 and June 2018. Index cases who did not have bacteriologically confirmed pre-XDR or XDR-TB, and those who refused treatment, were excluded.

#### *2.4. Data Variables and Sources, Data Analysis*

Data for each index case and HHCs is maintained in EMR software (Bahmni™), individual care record files, and in Microsoft Excel™ spreadsheets. Data for this study were extracted from these sources. Data was analyzed using EpiData Analysis (version 2.2.2.178 EpiData Association, Odense, Denmark) and Stata version 15.1 (StataCorp, College Station, TX). The numbers of HHCs assessed at each stage of follow-up were noted to understand the contact investigation cascade. Survival analysis was used to analyze the incidence rate of incident TB overall, by age, sex and time from follow-up initiation. HHCs were censored in case of loss to follow-up (LTFU), completion of follow-up period, or on 30th April 2019, whichever was earlier. The date of last visit was taken as the date of censoring. Per-capita floor area of the house was calculated, and overcrowding was considered present in case the per-capita area was less than 50 square feet. In all analyses, p < 0.05 was considered statistically significant. Wherever appropriate, 95% confidence intervals (95% CI) are being reported.

#### *2.5. Operational Definitions*

(1) Index case: patients with pre-XDR or XDR-TB registered for treatment in MSF Clinic.

(2) Incident TB: any HHC diagnosed as suffering from TB after one month of baseline contact evaluation.

(3) Lost to follow-up (LTFU): HHCs not evaluated during a scheduled visit and any time during the subsequent 6-month period.

#### *2.6. Ethics*

Ethics approval was obtained from the Ethics Advisory Group of the International Union Against Tuberculosis and Lung Disease, Paris, France (EAG No. 96/18). The study met the criteria for a posteriori analysis of routinely collected clinical data and did not require MSF Ethics Review Board full review. It was conducted with permission of the medical director, Operational Centre Brussels, MSF.

#### **3. Results**

The clinic enrolled 129 cases of pre-XDR and XDR-TB during the study period; among these, we excluded 11, and nine refused treatment later, yielding 109 eligible index cases (Figure 1). There were 58 (53.4%) cases with XDR-TB; nine (8.3%) were under 15 years of age, and 60 (55.1%) were female (Table 1). Two index cases were from the same household. These 109 index cases had 530 HHCs, among whom 518 (97.7%) underwent baseline evaluation. Among 518 HHCs, 23 were reported to have TB at the time of first visit (21 already on treatment, two newly diagnosed). The rest of the 495 HHCs were eligible for follow-up.

**Figure 1.** Assembly and follow-up of cohort comprising household contacts of index cases with pre-XDR and XDR tuberculosis registered at the MSF Clinic, Mumbai between Jan 2016 and June 2018. TB = Tuberculosis; DSTB = Drug-sensitive TB; DRTB = Drug-resistant TB; pre-XDR = Pre-extensively drug-resistant; XDR-TB = Extensively drug-resistant tuberculosis; PTB = Pulmonary TB; ATT = Anti-tubercular therapy; LTFU = Loss to follow-up. While 114 HHCs missed clinic evaluation, we considered them as eligible for follow-up since they were at risk of incident tuberculosis.


**Table 1.** Characteristics of index cases with pre-XDR and XDR tuberculosis registered at the MSF clinic, Mumbai, India, January 2016 to June 2018 (N = 109). IQR: Inter-quartile range.

#### *3.1. TB Diagnosis among HHCs*

Among the 495 HHCs eligible for follow-up, 95 (19.2%) did not complete the first follow-up visit (at 3 or 6 months, based on age and risk profile) and were early LTFUs. For the remaining 400 HHCs, median follow-up of 18.9 months [Interquartile range (IQR): 14.0–25.8] and cumulative follow-up of 675 person-years (py) was achieved. During the follow-up period, we identified 14 (3.5%) new TB cases (Figure 1). The median time to TB diagnosis among HHCs was 18 months (IQR: 6–21 months). The incidence rate of TB was 2072 (95% CI: 1227–3499) per 100,000 person-years over the entire follow-up period.

Of 14 HHCs with incident TB, ten (69.2%) incident cases were 15 years of age or older, and eight (61.5%) were female (Table 2). Six cases occurred in siblings of index cases and three in children. None of the index cases or HHC characteristics were associated with incident disease in univariate analysis. **Table 2.** Profile and factors associated with incident TB among household contacts of patients with pre-XDR and XDR tuberculosis in Mumbai, India, January 2016–June 2018 (N = 400). \* Column percentage; \*\* Row percentage; N(%) were compared using the Chi-square test or Fisher's exact test; †: Median (Inter-quartile range), compared using the Wilcoxon rank-sum test; Pre-XDR TB: Pre-extensive drug resistant TB; XDR-TB: Extensive drug-resistant TB; PTB = Pulmonary TB; EPTB = Extra-Pulmonary Tuberculosis; AFB = Acid fast bacilli.


#### *3.2. Overall Yield of TB in HHCs*

The overall yield among 518 identified HHCs was 3%. A total of 16 patients were diagnosed with TB, which included two patients who were diagnosed with TB on first day of screening and 14 who were identified during follow-up.

#### *3.3. Drug Resistance Patterns*

Among 37 identified TB cases among HHCs, the TB resistance profiles of 23 patients who were reported with TB on the first day of screening (including 21 on TB treatment and two newly diagnosed) were as follows: three had DSTB, seven had MDR TB, five had pre-XDR, and eight had XDR-TB, respectively. Among 14 cases found during follow-up, six had DSTB, six had pre-XDR-TB, one had XDR-TB and a TB resistance profile could not be ascertained for one. Five (38.5%) of these 14 cases were concordant with index case (4 Pre-XDR-TB and 1 XDR-TB).

#### **4. Discussion**

The overall yield was 3% during the longitudinal household contact tracing for TB in patients diagnosed with pre-XDR and XDR-TB receiving care in MSF Clinic in Mumbai, India. Through systematic longitudinal evaluation of HHCs, the study reported a higher incidence of TB disease than the previously reported study from India [2].

The yield of TB reported in our study is comparable to findings of systematic reviews [6–8]. However, the systematic review by Morrison et al. [6] also included studies where TB index cases were bacteriologically and clinically confirmed patients, while our study included only biologically confirmed cases as index cases.

Our findings are in line with those of Leung and colleagues, who conducted follow-up of HHCs of MDR-TB and reported that presence of XDR-TB significantly increased the odds of identifying a prevalent TB as well as hazard of incident case by almost five times after 18 months [16]. However, they did not have pre-XDR cases in their cohort. The observed incidence is higher than in China, Gambia and Peru [17–19], though it matches results from a Vietnam active case finding setting [10].

We observed discordant TB resistance profiled in HHCs compared to index cases; about half of the cases were DSTB, while the index cases were either pre-XDR-TB or XDR-TB [7]. These may hint towards community transmission of DSTB in a high burden setting.

#### *4.1. Limitations*

The study could not differentiate whether incident cases were the result of infection within the household or community acquired, since we did not conduct molecular fingerprinting. Molecular fingerprinting would have assisted in identifying the transmission pathways. We also did not evaluate for latent TB infection. Given the high incidence of TB in the program setting, it is likely that transmission would have occurred outside households, as is suggested by recent literature [20]. We observed that 44.2% of HHCs did not come for baseline clinic evaluation and 19.2% of HHCs had initial LTFU. A commonly reported reason for this was family members moving away from the index case. The index cases were largely slum-dwellers staying in rented accommodation, and the family members moving away could be a result of stigma, or a reaction to protect the health of family members. Finally, our findings may be representative of an urban Mumbai slum context and not directly applicable in other settings.

#### *4.2. Implications for Policy and Practice*

Under the Programmatic Management of Drug Resistant TB (PMDT) in India, it has been recommended that all close contacts of DR-TB cases be identified through contact tracing and evaluated for active TB disease. If the contact is found to be suffering from TB disease, irrespective of bacteriological confirmation, s/he should be identified as "presumptive DR-TB" and therapy be initiated based on their history of previous anti-TB treatment or TB resistance profile of index TB case. Simultaneously, two sputum samples should be transported for culture and DST. No definitive guidelines have been provided for prophylaxis among contacts with no active disease, and follow-up duration and close monitoring is suggested as the recommended action [21]. Our results suggest the need for follow-up of HHCs of patients with DR-TB after baseline evaluation for at least the duration of the index case's treatment.

#### *4.3. Implications for Future Research*

Cost-effectiveness analyses have shown the utility of active case finding among HHCs [22]. Since we are proposing a longitudinal follow-up of contacts of pre-XDR and XDR cases, future research should focus on establishing the cost effectiveness of this approach. An assessment of latent TB infection among contacts of pre-XDR and XDR cases is also recommended in view of the high incidence rates of active infection. Molecular fingerprinting and linkage analyses will clarify the epidemiology of transmission. Longer follow-up of more pre-XDR and XDR contacts is also recommended in research settings along with analyses of implementation effects of various screening preventive and infection control strategies so that their relationship with incident disease may be better understood.

#### **5. Conclusions**

We observed a high incidence of TB among HHCs of pre-XDR and XDR-TB in Mumbai, India. Systematic longitudinal TB screening of HHCs in cases with pre-XDR and XDR-TB must be strengthened, as they are at a high risk of incident disease.

**Author Contributions:** Conceptualization and protocol writing: R.H.P., V.G., M.V., C.L., M.D., P.S., S.S. (Sriram Selvaraju), P.I., G.F.; Data collection: R.H.P., S.S. (Sabira Sheikh), R.Y.; Data analysis: R.H.P., V.G.; Drafting the paper: R.H.P., V.G., C.L., M.D., M.V., S.S. (Sabira Sheikh); Critical review and approval for submission: R.H.P., V.G., P.S., M.V., S.S. (Sriram Selvaraju), H.M., S.K., M.D., C.L., G.F., P.I. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** This research was conducted through the Structured Operational Research and Training Initiative (SORT IT), a global partnership led by the Special Programme for Research and Training in Tropical Diseases at the World Health Organization (WHO/TDR). The model is based on a course developed jointly by the International Union Against Tuberculosis and Lung Disease (The Union) and Medécins sans Frontières (MSF/Doctors Without Borders). The specific SORT IT programme which resulted in this publication was jointly developed and implemented by: The Union South-East Asia Office, New Delhi, India; the Centre for Operational Research, The Union, Paris, France; Department of Preventive and Social Medicine, Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry, India; Department of Community Medicine and School of Public Health, Postgraduate Institute of Medical Education and Research, Chandigarh, India; Department of Community Medicine, All India Institute of Medical Sciences, Nagpur, India; Rajendra Prasad Centre for Ophthalmic Sciences, All India Institute of Medical Sciences, New Delhi, India; Department of Community Medicine, Pondicherry Institute of Medical Science, Puducherry, India; Department of Community Medicine, Kalpana Chawla Medical College, Karnal, India; National Centre of Excellence and Advance Research on Anemia Control, All India Institute of Medical Sciences, New Delhi, India; Department of Community Medicine, Sri Manakula Vinayagar Medical College and Hospital, Puducherry, India; Department of Community Medicine, Velammal Medical College Hospital and Research Institute, Madurai, India; Department of Community Medicine, Yenepoya Medical College, Mangalore, India; Karuna Trust, Bangalore, India and National Institute for Research in Tuberculosis, Chennai, India. The authors are grateful for the contribution of healthcare workers from the MSF clinic as well as the patients living with TB and their families.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Perspective*
