Expenditure Pattern for TB Treatment among Patients Registered in an Urban Government DOTS Program in Chennai City, South India

Background. Tuberculosis (TB) patients registered in the government clinics under the DOTS (Directly Observed Treatment, Short Course) program in Chennai city catering to about 4.3 million population. Objective. To estimate the pattern and overall costs incurred by the new patients (who have never had treatment for tuberculosis or have taken antituberculosis drugs for less than one month) registered under DOTS program in the treatment of tuberculosis in Chennai city. Methodology. A cross-sectional survey among new TB patients, who had completed intensive phase of antituberculosis treatment, was done using a precoded semi-structured questionnaire between March and June 2007. Information was collected on demographic, socioeconomic characteristics and expenditure for before and during treatment. Mean costs were used for comparison. Results. Among the 300 TB patients, most economically productive age group and 186 (62%) were males. The overall estimated total costs incurred right from the onset of symptoms until treatment completion was found to be Rs. 3211 (3.8% of annual family income) under DOTS program, which is less compared to previous studies. The overall mean total cost was significantly high among male (Rs. 3270; P < 0.01), employed (Rs. 3945; P < 0.01), and extrapulmonary patients (Rs. 3915; P < 0.01). Conclusion. The study has reiterated the fact that DOTS helps in reducing out-of-pocket expenses to patients with tuberculosis and hence is a cost-effective health intervention. This cost reduction may help to increase the access to the poor people which would help in achieving universal access to TB care services.


Introduction
TB (tuberculosis) causes enormous social and economic disruption and hampers nation's development [1,2]. India accounts for one-fifth of the global TB burden, with 1.8 million developing the disease each year and of them about 800,000 are infectious. Nearly 0.4 million are dying due to TB annually which translates to two deaths every three minutes [3]. The disease is most prevalent in the age group of 15 to 54 years [4,5], which is the highly economically productive period of an individual's life with important consequences for the household when the individual falls sick with TB.
Generally, burden of TB is measured by morbidity and mortality which are key considerations [6]. However, only focusing on morbidity and mortality effects provides an incomplete picture of the adverse impact of ill health on human welfare. In particular, the economic consequences of poor health can be substantial. Health "shocks," such as unexpected increases in health expenditure, reduced functional capacity, and lost income or productivity are often a primary risk factor for impoverishment [7,8]. At a societal level, poor population health is associated with lower savings rates, lower rates of return on capital, and lower 2 Tuberculosis Research and Treatment levels of domestic and foreign investment; all of these factors can and do contribute to reductions in economic growth [9]. Measurement of these various adverse impacts provides decision-makers to take appropriate policy decisions and also to provide another dimension of justification for worth of investment in TB control [10][11][12].
Projections from earlier study conducted prior to the implementation of RNTCP (Revised National TB Control Program) in south India indicated that despite being offered free diagnosis and treatment by government, the projected out-of-pocket expenditure incurred by TB patients annually was more than US$ 3 billion [13,14]. A subsequent study in a rural area showed that for patients registered under RNTCP, the patients who returned to work early established the economic benefits to patients treated under DOTS (Directly Observed treatment, Short Course) [15]. Ten years post-RNTCP implementation, the present study was undertaken to estimate the pattern and overall cost incurred by the patients registered under RNTCP for the treatment of tuberculosis in Chennai city.

Study Setting.
Chennai city has a population of 4.34 million and is spread over 174 sq. km. The city is divided into 10 administrative zones with one tuberculosis unit (TU) in each zone. This study has been conducted so as to cover all patients, meeting the eligibility criteria, who come to each TU for DOT in one week of drug cycle.

Study Design and Period.
In this cross-sectional study, the data was collected during the period from March 2007 to June 2007.

Study Population.
New (who have never had treatment for tuberculosis or have taken antituberculosis drugs for less than one month) adult TB patients of both sexes, coming to the TU for DOT treatment and whose HIV statuses were negative were considered eligible for the study.

Tool for Data Collection.
A pretested, semi-structured, precoded questionnaire was used to collect information on demographic and socioeconomic characteristics of patients. The questionnaire also included information on expenditure for the consultation fees, investigations, medicines, travel for escort, and patient before and during treatment.

Direct Costs.
Consultation fees and money spent on investigations and drugs were classified as medical expenditure. Money spent on travel, lodging, special food, and expenditure incurred for persons accompanying the patient were classified as nonmedical expenditure.

Indirect
Costs. Indirect costs were classified as loss of wages due to illness, decreased earning ability due to illness, or long-term disability that necessitated change in type of work.

Total Cost.
Total cost includes the expenditure incurred pretreatment and during treatment under direct and indirect costs.
2.6. Data Collection. List of TB patients who met the eligibility criteria was compiled from the TB register. All the patients who met the eligibility criteria were interviewed after the completion of intensive phase of their treatment by trained investigators after obtaining written informed consent either at the TU or in their residence. Patients were informed in their mother tongue, in Tamil or a language that they understood, about the purpose of the study. Patients were told about the confidentiality of the data collected, each interview will take about half an hour to complete it and also it is their right to withdraw from the study at any time. Ethical approval for the study was obtained from the Ethics Committee of Sri Ramachandra University, Porur, Chennai. Costs data were validated throughout the interview by repeated questioning and cross checked with the prevailing rates of doctors consultation fees, costs for investigations, and market price of drugs, medical bills wherever possible. All the information was estimated for 3 periods: (1) before treatment (three months prior to start of TB treatment), (2) intensive phase of treatment (two months treatment period), and (3) continuation phase of treatment (four months after intensive phase). In calculating total treatment costs, the cost was calculated by adding costs of intensive phase and continuation phase for the whole 6 months treatment period. The costs were calculated in terms of Indian rupees and US dollars (exchange rate during the study period US $1 = Rs. 45).

Data Management.
Data were checked for errors, entered, and analyzed using the SPSS version 15.0 (SPSS Inc., Chicago, IL, USA). Average (mean and median) costs such as direct medical, direct nonmedical, indirect, and total were calculated for both before and during treatment. All these patients had taken treatment in government hospitals where the investigations and the medicines were offered free of costs. However, they had to spend for travel, and so forth, the distribution of costs is uneven and we are of the opinion that this variation is expected from all economic data such as income and expenditure, and so forth. In order to avoid any bias we have provided both mean and median values. Further we classified the costs into Nil costs, those who spent up to Rs. 500, those who spent more than 500; then we had also measured what proportion were in these categories. Student t-test was used to compare the mean difference between mean total costs. A P value < 0.05 was considered statistically significant.

Coverage.
Among the 335 patients who met the eligibility criteria, 305 (92%) were interviewed. About 30 patients Tuberculosis Research and Treatment 3 could not be interviewed because they had defaulted or migrated. Of 335, only in five patients the costs were quite high as they were out layers and inclusion of this data might distort the average hence excluded from analysis. They had gone to private providers and spent more money for TB diagnostics, test-like scanning, blood test, and so forth, this hiked the total costs and these five were excluded as outliers. Accordingly, further analysis of data relates to 300 patients only. The percentage of coverage in all the ten TUs ranged from 87% to 95%.  743, nonmedical Rs. 219) and it was higher before treatment. The indirect cost was higher during treatment period Rs. 1694 (intensive phase Rs. 949 and continuation phase Rs. 745) ( Table 2).

Costs Incurred by Patients prior to Start of Treatment (Pretreatment Costs).
Before starting treatment in RNTCP 81.7% incurred medical costs, 93% incurred direct nonmedical costs and 21% incurred indirect costs (Table 3). During intensive phase of treatment 2.3% incurred medical costs, 42% incurred direct nonmedical costs and 24.7% incurred indirect costs. In all during treatment 45% did not incur any costs (Table 4).

Proportion of Costs to the Annual Family Income.
Proportion of total costs as percentage of annual family income for all 300 patients was 3.8%; the proportion was significantly higher (7%, 4%, 2%; P = 0.000) for patients belonging to lower economic strata compared to those in higher economic strata ( Figure 1).

Discussion
This study has highlighted the patterns of expenditure of TB patients in terms of direct costs, indirect costs, and total costs for both diagnosis and treatment by the TB patients registered in government TB control program in Chennai city. The overall estimated total costs incurred right from the onset of symptoms until treatment completion were found to be Rs. 3211 under DOTS program. If these costs were extrapolated to the whole country, the cost to the country for TB treatment would be considerable. However, the encouraging finding is that the estimated mean total cost in the current series was less as against those reported in earlier studies s [9,[16][17][18]. Our results strongly suggest that DOTS program is cost saving for patients and thus lends support for universal DOTS in India. This finding also draws attention to the contribution tuberculosis control makes to alleviation of poverty by reducing the economic burden that the disease inflicts on the poor. In this study, we observed that the total costs as percentage of annual family income to the patients was 3.8%.  This is much less compared to the earlier study done in rural Tamilnadu; it was in the range from 10 to 19% and most of the study subjects (62%) had per capita income per month <Rs. 335 [11]. This could be due to the fact that this study was done in an urban area where majority of the patients were in monthly percapita income range of more than Rs. 1100.
In this study, 81.7% and 93% of patients incurred medical and nonmedical costs, respectively, before patients were actually started on treatment in the DOTS programme. In general in more than 50% of chest symptomatics the first point of contact was private health facilities [19][20][21], these patients had gone to private and spent money for diagnosis which could have been avoided if the private  healthcare providers were involved in RNTCP. Therefore, involvement of private health providers is the priority area for RNTCP [22]. Other reason may be due to low awareness among patients who spent more money for "shopping for diagnosis" [23]. advocacy, communication, and social mobilization (ACSM) activities on availability of free TB diagnosis and treatment need to be strengthened to reduce these pretreatment costs.
We also found that less proportion (24.7%) of patients incurred indirect costs as compared to the earlier studies where the indirect costs during treatment was incurred by 54% and 41% of patients, respectively [9,11]. This finding suggests that due to decentralization of DOT, patients were able to go back to their work after taking treatment under direct observation and henceforth recover in the early stage of treatment. These findings support the progress on achievement on one of the goal of RNTCP, that is, patient should not incur any money for diagnosis and treatment. Another important finding from this study was that majority did not incur any medical costs during treatment, which indicates that free medical services under government are reaching the people.
However, a considerable proportion of patients incurred nonmedical and indirect costs in the intensive phase of treatment. The total costs as percentage of annual family income in this study are much less compared to that published in study by Rajeswari et al. [13] in India and in Thailand study done prior to DOTS where the total costs as percentage of annual family income was 40% and 15%, respectively [24]. However this proportion was more for patients belonging to low economic strata as compared to high economic strata. This highlights that despite the country's propoor health DOTS implementation, they cannot protect households from all costs which necessitates appropriate support security measures for TB patients by governmental and nongovernmental organizations.
The overall mean total cost was significantly high among males (Rs. 3270, P < 0.01) and employed (Rs. 3945, P < 0.01). This is in conformity to the finding that the economically active age group (15 to 54) patients incur higher costs. The mean total cost was significantly more for extrapulmonary TB patients (Rs. 3915) as compared to pulmonary patients (Rs. 2951). This could be due to the fact that diagnosis of extrapulmonary TB is more difficult as compared to pulmonary TB because it requires tissue diagnosis such as FNAC, biopsy and investigations such as X-ray and CT scan the costs of which are higher. Program needs to provide adequate and easily accessible facilities for diagnosing extrapulmonary TB cases.
The present study, which covered all the ten zones of Chennai Corporation, is representative of the group of TB patients in an urban area. While previous studies on socioeconomic impact of tuberculosis have considered only adult sputum positive patients [13], the present study has included the all age groups and both pulmonary and extrapulmonary TB patients and has given a comprehensive picture of the costs of a cross-section of TB patients registered with the urban TB control program.
Some of the limitations identified are that there could have been recall bias in answering questions on costs. This study has not taken into consideration TB patients who take treatment in the private health sector or the patients taking DOT in rural areas.

Conclusions
Our study findings strongly suggest that RNTCP has proven to be a cost effective health intervention, with reference to reducing out-of-pocket expenses and indirect costs which indicate that making them return early to work, which in turn benefits their families and in the broader perspective contributes to the overall economic and social development of their country. Reducing out-of-pocket costs to patients may increase the access to the poor people and thus promoting the universal access of TB care services as well. Still, there is a need to provide adequate financial and social support measures to TB patients as two-thirds of TB patients registered in the programme were from the poorer sections of the community. Already attempts have been made to involve private sector, and nongovernmental organisations in the RNTCP. But this study highlights that this activity should be substantially increased. Advocacy, communication, and social mobilization (ACSM) activities on availability of free TB diagnosis and treatment also need to be strengthened to reduce these pretreatment costs.