Cost-effectiveness of TB treatments to Burmese and Laotian patients by Thai hospitals on the border with Myanmar and Laos

Purpose – The purpose of this paper is to estimate the cost-effectiveness of tuberculosis (TB) treatment and care in two Thai hospitals located on the borders with Myanmar and Laos. Design/methodology/approach – A retrospective data collection was conducted to analyze all costs relevant to TB treatment and care from Mae Sai and Chiang Sean Hospitals. The cost related to TB treatment and care and the number of successful TB treatment from January 1 to December 31, 2017 were used for the calculation. The cost-effectiveness ratio (C/E) and the incremental cost-effectiveness ratio (ICER) were the outcomes. Findings – In 2017, the total cost of the TB treatment and care program at Mae Sai Hospital was 482,728.94 baht for 57 TB patients. The cast per treated case per year was 8,468.93 baht. The C/E was 10,971.11 baht per successful TB treatment (44 successful cases). The total cost of the TB treatment and care program at Chiang Sean Hospital was 330,578.73 baht for 39 TB patients. The cost per treated case per year was 8,476.38 baht. The C/E was 22,038.58 baht per successful TB treatment (15 successful cases). The ICER was 5,246.56 baht. The Mae Sai Hospital model was more cost-effective in terms of the treatment and care provided to Burmese patients with TB than the Chiang Sean Hospital model for Laotian patients with TB. Journal of Health Research Vol. 34 No. 1, 2020 pp. 13-21 Emerald Publishing Limited e-ISSN 2586-940X p-ISSN 0857-4421 DOI 10.1108/JHR-01-2019-0011 Received 16 January 2019 Revised 20 March 2019


Introduction
Tuberculosis (TB) is one of the most serious infectious diseases and necessitates long-term treatment of at least six months duration according to the World Health Organization (WHO) recommendations [1]. TB has become a global threat, particularly in developing countries, including Thailand, which is considered to be one of the highest TB-burdened countries globally [2]. Several studies have shown that factors such as the country or health setting, economy, process of treatment, internal policy and management are all related to TB treatment outcomes [3][4][5][6]. Additionally, countries that are faced with problems of HIV/AIDS must double their efforts in their management of TB patients, especially in areas that border two or more countries, such as the Thailand-Myanmar-Republic of Laos border communities [7]. Both Thailand and Myanmar have a long border that stretches from the north to the south and is approximately 2,401 km [8]. For years, northern Thailand has been considered to be the region with the highest epidemic of HIV/AIDS and TB in the country [9]. Chiang Rai is located in the northern-most part of Thailand and borders Myanmar in the west and the Republic of Laos in the north and east [10]. In 2017, there were almost 1.5m registered Thai people in the province. However, the same year, an additional 200,000 non-Thai who were working in different locations also lived in Chiang Rai [11]. Most of the non-Thai did not meet the criteria for free access to the health care system in Thai public hospitals, including medical care for TB treatment [12].
In 2018, the Center of Disease Control No. 1, Chiang Mai Office [13] reported that the number of new TB cases and the number of relapses increased from 5,785 cases (98.6/100,000 pop) in 2017 to 6,251 cases (106.8/100,000 pop) in 2018. However, the cure rates in those respective years were 79.5 and 73.49 percent, which were lower than the WHO's goal of 90.0 percent [14]. In 2018, more than 20,000 HIV/AIDS cases were being treated with ARV, and more than 4,000 new TB cases were detected in 18 public hospitals in Chiang Rai province [15,16]. Chiang Rai province was reported to have the highest registered rate of new TB cases at 26.4 percent and the lowest cure rate at 54.05 percent [13]. People aged 60 years old represented the greatest proportion of all registered TB patients in hospitals, accounting for 20.7 percent, and 9.6 percent of HIV-positive individuals were identified as infected with TB [13].
Between 2010 and 2018, 14,024 TB cases were reported in Chiang Rai province, and 37.5 percent were identified as members of the stateless population, including individuals of Burmese and Laos nationality [17]. A large proportion of TB cases were reported by Thai hospitals along the borders of Thailand-Myanmar and Thailand-Republic of Laos [7]. Mae Sai and Chiang Sean Hospitals were reported to have the greatest proportion of new foreign TB patients in Chiang Rai province. Among these foreign TB patients, 74.3 percent were Burmese and 24.7 percent were Laos [18]. These patients were frequently listed on the Multidrug-resistant Tuberculosis register in these hospitals, particularly the Burmese and Laos TB patients [18].
All the hospitals located in border areas have very unique characteristics, which are different from those of hospitals in other areas [19]. TB patients face a number of barriers to access to a hospital, particularly those individuals who cross the border to go to Thai hospitals [12]. For instance, at Thai hospitals, foreign TB patients face difficulties with the language and medical fees, and cooperation between doctors who manage TB patients while working in two different hospitals located in different countries is also a barrier [20]. Moreover, the burdens of treatment 14 JHR 34,1 and other medical care needs must be closely monitored to maintain the quality of care for Thai citizens [21]. Currently, in Chiang Rai province, there are at least two hospitals located on the Thailand-Myanmar-Republic of Laos borders in northern Thailand that provide care for Burmese and Laos TB patients [15]. Because of the unique conditions, these two hospitals differ on many fronts in terms of how TB treatment and care is provided, such as the overall number of patients, the number of TB patients and the number of foreign TB patients. This type of information is needed to identify the optimal cost-effective program for TB treatment and care. This study aimed to assess the cost-effectiveness ratio (C/E) and the incremental costeffectiveness ratio (ICER) for TB treatment and care between these two hospitals. Ultimately, the findings of the study could be used to implement policies in hospitals located in border areas.

Study design
A retrospective study was conducted to analyze the costs and outcomes of the treatment of foreign TB patients in two Thai hospitals located on the borders between Thailand and Myanmar and Thailand and the Republic of Laos.

Study setting
Mae Sai Hospital (a 90-bed hospital) and Chiang Sean Hospital (a 60-bed hospital) were purposively selected as the study settings because these two hospitals are located in border areas in Chiang Rai province, and all the TB staff, including the directors of the hospitals, were willing to participate in the study. Mae Sai Hospital is located in Mae Sai district, west of the Chiang Rai province border with Myanmar. Chiang Sean Hospital is located in the Chiang Sean district, which is bordered by Myanmar in the west and the Republic of Laos in the north and east. The study was conducted from May 2018 to November 2018.

Study population
The study populations included Burmese TB patients who were treated at Mae Sai Hospital and Laos and Burmese TB patients who were treated at Chiang Sean Hospital between January 1 and December 31, 2017.

Sources of data and measurements
Data were collected with specifically designed sheets that were tested for validity and reliability before use. Three different cost categories were included in the analysis: capital costs, material costs and labor costs. The data on capital and material costs were collected based on the provider prospective. These data were collected from hospital accounting documents, hospital records, annual expenditure reports and interviews with hospital officers. The labor cost category (staff salary) was obtained from the accounting documents.
For the outcome measurement, all clinical data for foreign TB patients were collected from the HOSxp database in both the hospitals.

Activity-based costs
Activity-based costs were defined as the operating costs of the TB program and were assessed by developing the activity model and including all possible related domains involved in TB treatment and care. Afterward, an activity analysis was performed to identify the resources used for each activity.

Outcome measures
The outcome measure of the study was the effectiveness of the TB treatment and care program. First, we selected the number of TB cures or cases of successful treatment as the main measurement of effectiveness. The cure rate was defined as the proportion of patients who received a cured status after treatment, which was confirmed by a sputum examination that was negative for TB bacilli, divided by those who completed a full treatment regimen. Second, other outcomes included the total number of TB patients who received care at a hospital.
Activities related to the costs (Figure 1) Mae Sai Hospital had three main activities related to the TB treatment and care program: activities related to TB screening, a new TB diagnosis and treatment and care; activities regarding ongoing treatment and care of the whole cohort in a year; and activities related to hospitalized TB patients.
Chiang Sean Hospital had four main activities related to the TB treatment and care program: activities related to TB screening, a new TB diagnosis and treatment and care; activities regarding ongoing treatment and care of the whole cohort in a year; activities related to hospitalized TB patients; and village-based follow-up of TB patients.

Cost categories
All relevant costs were classified into categories such as labor, material and equipment. These categories were then classified into either direct or indirect costs. Direct costs referred to all expenses directly and exclusively related to TB treatment and care activities, such as fuel and stationery, that could not be used in other activities. Indirect costs were costs that could be reused or allocated into many activities or programs or had multiple allocable components, such as labor, buildings, X-ray machines, computers, microscopes and scales.

Cost drivers
Cost drivers were the causal factors that influenced the cost of the activity, which was determined and estimated by the nature of the usage of cost drivers regarding the resources utilized for TB treatment and care. We determined the cost drivers for treatment-related activities and supporting activities and included time in the calculation. Cost drivers were  selected based on the recommendation of selected people from a hospital, which depended on the time used and the material cost. All items were identified as the volume finally used.

Calculation of the annual costs of activities and services
Once the cost drivers were identified, the next measure was the cost allocation rate, which was calculated as the annual cost of the resources and the number of times the resource was used during a whole year (annual quantity of cost driver). Finally, we calculated the annual TB treatment and care program cost by multiplying the actual quantity of allocation based on the allocation rate of each resource for that TB treatment activity.

Outcome measures
The outcome measurement was based on the effectiveness of TB treatment and care. We selected the number of cured patients or successful treatment as the main measure of effectiveness; this is equivalent to the WHO measure of successfully treated patients. The cure rate was defined as the proportion of those patients whose cure was confirmed by sputum examination and found to be negative for TB bacilli and those who completed a full treatment regimen. Other outcomes included were the total number of TB patients who received care at Mae Sai and Chiang Sean Hospitals.

Process of cost calculation
For the calculation of labor costs, we used the monthly gross salary data of the staff base, which was obtained from the salary database and then multiplied by 12 months to convert monthly gross salaries to annual costs incurred by a hospital. Next, we divided the annual cost of labor by the annual quantity of the cost driver, which was derived from 60 min × 8 h per day × 22 days per month × 12 months (8 day-off in a month) to obtain the cost allocation rate of labor cost per minute of working for Mae Sai and Chiang Sean Hospitals. We allocated the activity cost for the TB clinic by multiplying the cost allocation rate by the actual quantity of allocation base (time spent on each activity in the TB clinic). Capital and material costs were calculated based on the use in each activity. Some purchase prices and the year of purchase for capital and materials could not be determined due to a lack of records. The estimation of the cost and the year of purchase were ascertained through interviews with the director and staff who had worked at Mae Sai and Chiang Sean Hospitals for more than ten years. The rooms or buildings used by the TB clinics were estimated for 30 years of lifetime-use; however, other lifetime-use materials were estimated at 1-20 years.
A discount rate of 5 percent was used to find all the current prices of capital and materials. Current prices of capital and materials were annualized depending on the discount rate and its lifetime-use estimate. To calculate the cost allocation rate, we divided the annual cost of capital and materials by the time in minutes that capital and materials were used in a whole year (2017) or the annual quantity of cost drivers.
We estimated the annual quantity of cost drivers (number of minutes that capital and materials were used in one year) for capital and materials based on the information obtained from the hospital director and other staff. Three criteria were considered in the study: 60 min ×24 h per day × 30 days per month × 12 months for resources that were used all the time, such as IPD and the emergency room; 60 min ×8 h per day × 22 days per month × 12 months for the components employed in office time, such as computers, scales, X-ray machines, mercury sphygmomanometers, automatic blood pressure monitoring devices and patient historical record forms; and 60 min ×3 h per day × 4 days per month × 12 months for materials and equipment used only by Mae Sai Hospital staff for services and follow-up with people in the villages every Friday afternoon.
All capital and material costs were allocated to the TB program by multiplying the cost allocated rate to each item by its annual quantity of cost drivers.
Finally, the cost-effectiveness was calculated for Mae Sai and Chiang Sean models.

Results
In 2017, the TB treatment and care program provided by Mae Sai Hospital was more cost-effective than that provided by Chiang Sean Hospital.  (Table I).
The TB treatment and care program provided by Mae Sai Hospital was approximately 2.2 times more cost-effective than the program provided by Chiang Sean Hospital. The ICER was 5,246.56 baht per successful TB treatment per year.
The majority of the cost for TB treatment and care was the capital cost, which was up to 67 percent for the TB treatment and care program at Mae Sai Hospital and 46 percent at Chiang Sean Hospital (Table II) where r is the interest rate and n is the usable life of capital asset; b Calculation of the current price of material: where C t is the current price of the asset, C o is the purchase price, r is the interest rate and t are the years since purchase Thailand and Myanmar do not have any memoranda of understanding with regard to TB case management. Therefore, most individuals diagnosed with TB at Mae Sai Hospital are asked to attend a hospital in Thailand or to live in a village close to a hospital in Thailand. This enables the TB clinic staff to closely monitor their patients and improve the successful treatment rate.
Our study findings indicate that hospitals with community or village-based TB case monitoring achieve better cost-effectiveness compared to those that do not have this service. This coincides with the findings of several studies of cost-effectiveness in TB case management [24][25][26][27].
Some limitations were found during data collection. The rotation of staff in TB clinics was one of the problems faced when collecting labor costs. Labor costs (staff salaries) were difficult to access in some hospitals.

Conclusion
The study findings show that providing TB treatment and care with standard treatment procedures (screening, ongoing treatment and hospitalized care) combined with patient follow-up in a village significantly improves treatment outcomes and the