Predictors Of Catastrophic Costs of Tuberculosis (TB) Among Patients Co-Affected With TB-HIV And TB-Diabetes in Bhavnagar Region, Western India

Background: By the year 2030, no family with a patient of tuberculosis (TB) is supposed to incur cata-strophic costs. In India, a significant number of people suffer from TB, and many faces catastrophic costs. Our objective was to determine the predictors of catastrophic costs due to TB among co-prevalent TB-HIV and TB-diabetes patients. Methods: We conducted a cross-sectional study among 234 patients co-affected with TB-HIV and 304 patients with TB-diabetes co-prevalence in the Bhavnagar region (western part of India). TB costs was estimated using a validated questionnaire. Multivariable logistic regression was used to determine the significant predictors of catastrophic costs of TB. Results: Four percent of patients in each group incurred catastrophic costs due to TB. Female gender [aOR 6 (1.2-33)], being single [aOR 9 (1.5-52)], low socioeconomic status [aOR 7 (1.2-30)], private con-sultation for TB [aOR 9 (1.5-53)], and hospitalization in first HIV visit [aOR 19 (3-137)] were significantly predicted catastrophic costs of TB among patients co-affected with TB-HIV. Among patients with TB-diabetes co-prevalence, hospitalization in first TB visit [aOR 7 (2-29)], and private consultation for TB [aOR 7 (1.6-30)] were the significant predictors. Conclusions: Despite a lower percentage of TB-HIV/ TB-diabetes patients facing catastrophic costs, hospitalization and private care-seeking are the “modifiable” determinants of TB catastrophic costs in our study setting.


INTRODUCTION
Tuberculosis (TB) is one of the diseases on the highpriority list for elimination globally by the year 2030. For many years, TB is a leading public health problem in India. India is said to be the country reporting the highest number of cases of TB worldwide. 1 As per the India TB Report, 1.9 million cases of TB were notified in the year 2021, 0.4 million higher than the estimates released by the World Health Organization (WHO). 1,2 No TB-affected family should incur catastrophic costs -one of the targets for the elimination of TB -is to be achieved by the year 2030. 3 WHO has defined costs due to TB exceeding 20% of annual household income as catastrophic. 4 TB is a diverse disease and so is the prevalence of catastrophic costs incurred due to it in India. 5,6 Globally, the prevalence of catastrophic costs among TB patients is 43% (95% CI 34%-51%), due to drug-sensitive, drug-resistant, and HIV co-infection is 32%, 81%, and 81% respectively. 7 Globally, various studies have documented the predictors of catastrophic costs incurred due to TB. [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23] Among some recent studies, some of the predictors reported were private provider consultation for the initial visit 6 , hospitalization 16,20 , loss of job 17 , human immunodeficiency virus (HIV) co-infection 7 , lower wealth quintiles 18,20,23 , and drug-resistance. 7,16,18,23 Evidence on predictors of catastrophic costs of TB among patients with TB-HIV co-infection or TBdiabetes comorbidity is lacking. We conducted this study to determine the predictors of catastrophic costs of TB among patients with TB-HIV co-infection and TB-diabetes comorbidity.

METHODS
Study design and setting: We conducted a crosssectional study among patients with TB-HIV and TBdiabetes co-prevalence in the Bhavnagar region (western India). Bhavnagar city and district have a population of ~0.6 million and ~2.8 million respectively. The region is situated around 200 kilometres southwest of Ahmedabad city (the financial capital of Gujarat state in the western part of India). Bhavnagar reports approximately 2500 cases of TB every year. The care for patients with TB is largely decentralized, that is, patients need to make around 3-4 visits (for drug-sensitive pulmonary regimen) to a government health facility for diagnosis and follow-up. The medicines are placed nearby the homes of patients with a treatment supporter. TB health visitors pay home visits for examining the development of adverse drug reactions and provide supportive treatment for the same. Only in case of complications, do patients with TB need additional visits to health facilities.
Study population: We included 234 adult patients co-affected with TB-HIV infection and 304 adult patients with TB-diabetes co-prevalence notified during 2017-2020 in the Bhavnagar region under the public sector. We included all eligible patients meeting our inclusion criteria (universal sample) and none of the patients refused written informed consent to participate in our study.

Definitions:
Catastrophic costs: Costs were said to be catastrophic when total costs due to TB incurred by the patients exceeded 20% of the annual household income. We calculated the annual household income based on the self-reported total monthly family income at the time of the interview.
Standard of Living (SLI) index: SLI index was categorized into low vs. middle/ high SLI using a summary score calculated based on the possession of assets like the type of house, number of rooms, agricultural land, and others. 6 Data collection, variables, and analysis: We used a validated WHO questionnaire for calculating the catastrophic costs. 4 Data on socio-demographic and other clinical variables were added to the questionnaire. The outcome variable was catastrophic costs incurred due to TB (dichotomous). We considered multiple variables like age, female gender, single (vs married), sputum smear grade, years of education, site of TB, TB drug sensitivity, tobacco smoking, SLI index, first consultation at a private provider, and hospitalization on the first visit as the predictors of catastrophic costs due to TB. Uni-variable logistic regression was performed to extract the variables to be included in the multivariable logistic regression. Predictors with a p-value ≤0.2 on univariable logistic regression were included in the multivariable model. Multivariable logistic regression was performed to determine the significant predictors (p-value <0.05) of catastrophic costs of TB. Adjusted odds ratios (aOR with 95% confidence intervals CIs) were calculated using multivariable logistic regression.
Ethical considerations: The study was approved by the Ethics Committee of Government Medical College Bhavnagar. Written informed consent was obtained from all the study participants.

RESULTS
The prevalence of catastrophic costs among 234 patients co-affected with TB-HIV and 304 patients with comorbid with TB-diabetes was 4% each.

Predictors of catastrophic costs among TB-HIV
Among patients with TB-HIV co-infection, on univariable logistic regression, the variables -female gender, single (vs married), low SLI index, employed in paid work before TB, first TB consultation at a private provider, and hospitalization in first HIV visithad a p-value of <0.2 and were included in the multivariable logistic regression analysis (Table 1).
On multivariable logistic regression, female gender, single (vs. married), low (vs. middle/ high) SLI index, first TB consultation at a private clinic, and hospitalization in first HIV visit were the significant predictors of catastrophic costs of TB ( Table 2). The female gender had 6 times (95% CI 1.2-33), being single had 9 times (95% CI 1.5-52), low SLI index had 7 times (95% CI 1.2-3), private consultation for the first TB visit had 9 times (95% CI 1.5-53), and hospitalization due to HIV in the first visit had 19 times (95% CI 3-137) higher odds of catastrophic costs due to TB as compared to their counterparts among patients coaffected with TB-HIV.
Predictors of catastrophic costs among TBdiabetes: Among 304 patients co-affected with TBdiabetes, on univariable logistic regression, years of education, urban residence, first TB consultation at a private clinic, hospitalization in first TB visit, first diabetes consultation at a private clinic, and hospitalization in first diabetes clinic visit had a p-value of <0.2 and were included in the multivariable logistic regression model (Table 3).    On multivariable logistic regression, hospitalization at the first TB clinic visit, and first TB consultation at a private clinic were significant predictors of catastrophic costs of TB among patients with TBdiabetes co-prevalence (Table 4). Hospitalization in first TB clinic visit had 7 times (95% CI 2-29) and first TB consultation at a private clinic had 7 times (95% CI 1.6-30) higher odds of catastrophic costs of TB than their counterparts among patients with TBdiabetes co-occurrence.

DISCUSSION
To sum up, we found that female gender, being single, households with a low SLI index (low socioeconomic status), first TB consultation at a private clinic, and hospitalization in first HIV visit significantly predicted catastrophic costs of TB among patients co-affected with TB-HIV. Among patients with TBdiabetes co-prevalence, hospitalization in first TB visit, and first TB consultation at a private clinic were the significant predictors of catastrophic costs of TB.
Our study reported that hospitalization in the first visit due to HIV had 19 times (95% CI 3-137) higher odds of catastrophic costs among TB-HIV co-infected patients and hospitalization in the first TB visit had 7 times (95% CI 2-29) higher odds of catastrophic costs among TB-diabetes comorbid patients. Hospitalization was reported to be a significant predictor by investigators in other studies from Kenya, China, and Tanzania. 15 15,16,20,21 However, hospitalization was found not significantly associated with total costs as well as costs for drug-resistant TB in studies from Indonesia and China. 10,12 Patients lose wages and incur costs for transportation, food, and accommodation during the days of hospitalization. The costs escalate with an increase in the number of days of hospitalization and the number of accompanying members.
Among patients co-affected with TB-HIV and TBdiabetes in our study, the first TB consultation at a private clinic had 9 times (95% CI 1.5-53) and 7 times (95% CI 1.6-30) higher odds of catastrophic costs of TB respectively. Our study findings were supported by studies among patients with TB in Nigeria and Kenya -3 times (95% CI 1.5-5.9) and 1.25 times (95% CI 1.1-1.7) -higher odds of TB catastrophic costs respectively. 14,20 However, the type of facility was not a significant predictor in studies from Ghana and Indonesia. 12,18 Nearly half of the patients with TB seek a private care provider for their care in India. 24 Patients incur significantly higher costs in the private sector than government providers, especially the direct medical costs. 6 Thus, the chances of catastrophic costs among those visiting private practitioners are higher.
In our study, households with a low SLI index (low socioeconomic status) were compared with households with a middle/ high SLI index (middle/ high socioeconomic status). We established that households with a low SLI index had 7 times (95% CI 1. Similar to our study, where we found single people at 9 times (95% CI 1.5-52) higher odds, a study from China reported 'divorced/ widow' as 2 times (95% CI 1.1-4.5) more likely to incur catastrophic costs due to TB. 10 On the other hand, studies from China and Cairo found no association between marital status and catastrophic costs. 15,17 Those who are divorced or separated due to any reason might be incurring higher TB costs due to a lack of financial support from any other family member.
This is the first study determining predictors of catastrophic costs due to TB among patients co-affected with TB-HIV and TB-diabetes. The study findings can be generalized to similar semi-urban and rural settings in India.

CONCLUSIONS
We conclude that despite a lower percentage of TB-HIV/ TB-diabetes patients facing catastrophic costs, hospitalization and private care-seeking are the "modifiable" determinants of TB catastrophic costs in our study setting. Early detection, prompt treatment, outreach programs for awareness of free care in the government sector, referral of impoverished patients to the government sector by private doctors, and further expansion of the decentralized model of TB care might avert catastrophic costs due to TB in India.