Abstract
Background
Environmental issues related to air pollution in Southeast Asia have persisted for more than a decade, especially in Thailand. This study aims to estimate the treatment costs of respiratory diseases caused by exposure to ambient PM₂.₅ and to identify the factors that influence these costs.
Methods
This retrospective study analyzed secondary data on OPD and IPD respiratory disease treatment costs from government hospitals, along with ambient PM₂.₅ data from low-cost monitoring stations, to estimate the cost of illness across 25 districts in Chiang Mai during Thailand’s fiscal year 2023. Economic cost was estimated using the Cost-of-Illness method formula: Economic Cost Loss = Health Impact × Treatment Cost. K-means cluster analysis was used to classify estimated costs into minimum, medium, and maximum cost scenarios. Multiple linear regression was applied to identify significantly associated factors with treatment cost.
Results
Under the maximum cost scenario identified through K-means cluster analysis stratification, the total treatment cost associated with an average PM₂.₅ concentration of 42.59 µg/m³ was 460,122.58 USD, averaging 41.62 USD per case. Each 1 µg/m³ increase in PM2.5 was associated with a cost rise ranging from 403.84 to 13,159.87 USD. Non-infectious respiratory diseases incurred costs approximately two times higher than infectious ones. The estimate of maximum treatment burden for respiratory disease cases was highest in urban areas, totaling 102,878.88 USD. The urban area showed a significantly higher cost of treatment both in OPD and IPD cases (p < 0.001). Moreover, higher healthcare levels and older age were associated with higher costs in OPD cases. In IPD cases, length of hospital stay was a significant predictor.
Conclusions
Ambient PM₂.₅ exposure contributes significantly to the economic burden of respiratory diseases in polluted areas. These highlight the importance of pollution control policies and healthcare resource planning in high-risk areas.
Trial registration
not applicable.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-026-26478-2.
Keywords: PM2.5, Respiratory diseases, Treatment costs, Economic burden
Background
Environmental issues of air pollution in Southeast Asia have been of concern for more decade [1]. In Thailand, especially in the northern part, levels of fine particulate matter (PM2.5) particles rank among the highest globally, partly attributable to unavoidable cross-border pollution [2]. Thailand’s regulation remains an announcement of a standard annual concentration of PM2.5 at 15 µg/m³, which is three times higher than the World Health Organization (WHO) guideline at 5 µg/m³ [3, 4]. Chiang Mai has documented all PM2.5 concentrations beyond the maximum health alert thresholds of the air quality index for all population categories [5]. Populations residing in high-pollution regions face health risks affecting multiple organ systems, especially the respiratory systems. Elevated PM2.5 levels have been associated with increased mortality, increased incidence and exacerbation of Chronic Obstructive Pulmonary Disease (COPD) and asthma, reduced lung function, increased risk of respiratory infection, and increased risk of lung cancer [6–8]. Therefore, the problem of air pollution in northern Thailand will persist for a considerable duration. Moreover, the economic loss attributed to PM2.5-induced morbidity and mortality constitutes a significant portion of national expenses. This highlights the urgent need for strategies to mitigate health impacts and anticipate related fiscal consequences.
Thailand’s healthcare system provides universal health coverage to all citizens, encompassing approximately 69.88% of the population [9]. In accordance with UC principles, the budget is allocated as an average treatment cost per capita across various regions. Consequently, the impact of air pollution on public health means that the healthcare system endures an augmented burden from the typical injuries and illnesses prevalent in the community. Alongside the healthcare budget provided by the UC scheme, there are other related budgets, such as operational budgets from the Ministry of Public Health to support regional agencies or other budgets within healthcare facilities. Nonetheless, the allocation of these budgets for the care of individuals impacted by PM2.5 exposure requires careful management by relevant stakeholders, such as government administrators and health service providers. The costs associated with caring for individuals impacted by PM2.5 exposure might not be included in standard budget assessments, given that the effects are localized to specific regions. When stakeholders are made aware of the associated costs, it facilitates a more effective distribution of budgets. Previous research assessed the health costs related to PM2.5 exposure, which varied significantly from 152,400 USD in China to 1,250,000 USD in France based on the value of a statistical life [10, 11]. However, there is limited data regarding the estimated costs associated with treating illnesses caused by exposure to air pollution, specifically PM2.5 in Thailand. This study aims to investigate the treatment costs of respiratory diseases in relation to changes in ambient PM2.5 levels and to identify factors associated with those costs. We expect that the results of this research will provide supporting information for stakeholders affected by PM2.5 to appropriately allocate budgets related to the care of citizens suffering from PM2.5-related illnesses.
Methods
Study design
This study is a retrospective secondary-data cost-of-illness study that reviewed and analyzed secondary data on treatment costs of respiratory diseases and ambient PM2.5 levels to estimate the relative cost of illness. The study was conducted in Chiang Mai Province, Thailand, and included data from all 25 districts across the province. The data were derived from government-affiliated hospitals, covering every district and representing all levels of the public healthcare system. All treatment cost records were obtained from the centralized database of the Chiang Mai Provincial Public Health Office (PPHO). Notably, this dataset excluded data from private hospitals and medical schools. Only direct medical costs, derived from hospital charges, were included, these covered inpatient services, laboratory investigations, procedures, and medications. All costs are presented in Thai Baht (THB) and converted to United States Dollars (USD) using the average exchange rate during the study period of 32.62 THB/USD [12].
The reported data included Out-Patient Department (OPD) and In-Patient Department (IPD) respiratory disease cases recorded during the Thailand fiscal year 2023 (October 2022 – September 2023). Respiratory diseases potentially related to air pollution were identified using International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10), including: Influenza: J09–J11 Pharyngitis and Upper Airway Infections: J02, J31 Pneumonia: J13–J18 Acute Bronchitis & Bronchiolitis: J20–J21 COPD: J40–J44 Asthma: J45–J46. Respiratory diseases were classified into two categories: non-infectious diseases (COPD and asthma) and infectious diseases to align with Thailand’s the Diagnosis-Related Group (DRG) healthcare reimbursement system, supporting policy relevance.
Hospital levels were categorized based on the Office of the Permanent Secretary’s classification system [13], and re-grouped into three levels based on service capacity: Primary Level (P1–P3): Facilities that provide primary care services without inpatient departments. Secondary Level (F1–F3, M2): Community hospitals offering inpatient services, general practitioners, and some specialized care (e.g., internal medicine, pediatrics, general surgery). Tertiary Level (M1–S): Referral hospitals capable of managing complex cases, with multiple specialists and advanced treatment technologies. Administrative service areas were defined according to the Chiang Mai provincial referral system [13] and grouped into five regional nodes: Chiang Mai City, Fang, San Patong, Chom Thong, and San Sai.
Ambient PM2.5 concentration data were obtained from low-cost monitoring devices named DustBoy sensors, which were deployed at 80 monitoring stations distributed across all districts in the study area [14]. These sensors provided real-time measurements, which were subsequently averaged over the study period to generate the annual average concentration. Samae H. et al. [15] demonstrated calibrated DustBoy sensors showed strong agreement with the United States Environmental Protection Agency’s Federal Reference Method (FRM) and Federal Equivalent Method (FEM), confirming the validity and reliability of data. The air pollution periods were determined based on regional air quality monitoring data and observed seasonal patterns. Based on Sapbamrer P. et al. study [8], the high pollution period was defined as December to May and the low pollution period as June to November, using decade-long average PM₂.₅ concentrations of 42.91 ± 25.68 µg/m³ and 12.14 ± 5.20 µg/m³, respectively. These relative differences were used to define high and low pollution periods.
The integrated dataset comprising PM2.5 concentrations, hospital admission records, and categorized treatment costs was used to estimate the respiratory diseases’ economic burden attributable to PM2.5 exposure in the study area.
Statistical analysis
Cost Estimation of respiratory disease due to exposure to PM2.5
The cost estimation was based on treatment costs associated with hospital admissions for respiratory diseases related to PM2.5 exposure, identified according to relevant ICD-10 codes. The total economic cost was estimated using the Cost-of-Illness method described by Yang et al. (2019) [16] and expressed as:
 |
In this equation, Health impact denotes the health impact attributable to PM2.5 exposure, which is derived from population data and the exposure–response function described by Shi et al. (2016) [17]:
 |
In this context, population refers to the total number of people at risk who reside in Chiang Mai Province, obtained from the Chiang Mai Provincial Administrative Office [18]. E0 is the baseline incidence rate for respiratory disease [19], calculated as the proportion of hospital visits relative to the population at risk, C represents the observed average PM2.5 concentration during the study period which was 42.59 µg/m3. The variable C0 signifies the baseline PM2.5 concentration set at 5 µg/m³, as recommended by the WHO in its 2021 Air Quality Guidelines [4]. The coefficient β is the exposure–response coefficient specific to the disease outcome. As the concentration–response β varies by factors like geography, socioeconomic status, and demographics [20, 21], this study used a local β derived from Chiang Mai based on Jarernwong (2024) [19], with a value set at 0.00401 for respiratory diseases. The details of the parameters used in the estimations are provided in Supplementary Material 1. The treatment cost per 1 µg/m³ increase in PM2.5 was estimated by comparing it to the baseline PM2.5 concentration (C), with the change from baseline defined as (C − C₀) = 1. This calculation was performed using the Cost-of-Illness method as previously described.
The treatment cost refers to the average cost, derived from hospital records and disaggregated by respiratory disease type according to infectious and non-infectious. Costs are presented as the total average for Chiang Mai province and as per-person costs, calculated by dividing the total cost by the total number of cases for infectious and non-infectious diseases separately. K-means cluster analysis was employed to categorize cases into three distinct cost scenarios—minimum, medium, and maximum corresponding to varying levels of disease severity and healthcare utilization. K-means analysis equation presented as follows: [22, 23]
 |
Where 
represents the i-th observation (data point) in the dataset. In this study, 
corresponds to the total treatment cost for the i-th patient. 
is the centroid (mean value) of cluster 
, representing the average treatment cost of all patients in cluster. 
denoted cluster k — in this study, k = 3 to represent the three categories of minimum, medium and maximum case scenarios, and 
is the centroid of cluster k, calculated as:
 |
This classification captures the variability in health outcomes and associated costs attributable to PM2.5 exposure. The ArcGIS program was used to generate all figures.
Multiple linear regression analysis
Multiple linear regression was used to explore factors associated with treatment cost, following an explanatory modeling approach focused on interpreting key relationships. Variable selection was informed by an intensive literature review and included relevant factors such as age, gender, level of healthcare, service area, and pollution period. For the IPD model, the length of hospital stay was added as an additional independent variable [24–27]. Linear regression diagnostic tests included the Shapiro–Wilk test to assess the normality of residuals. A sensitivity analysis was conducted by excluding OPD cases with zero treatment costs.
Clinical trial number
Not applicable.
Result
Demographic data
A total of 10,455 OPD and 601 IPD respiratory disease cases were recorded between October 2022 and November 2023. One-third of the OPD cases were reported from the Chom Thong node area. Approximately half of the OPD visits occurred at secondary-level health care facilities (50.26%), where the most diagnosed disease was COPD, accounting for 74.01% of all OPD cases, followed by asthma (13.96%). For IPD cases, the San Patong node area reported the highest number of admissions (28.45%), followed by San Sai (21.96%) and Chiang Mai city (21.30%). The predominant causes of admission were COPD and asthma (55.24% and 33.94%, respectively). The mean age of patients was 32.04 years for OPD and 44.05 years for IPD. The average treatment cost for OPD was 17.16 USD (SD = 44.94), ranging from 0 to 1432.37 USD. For IPD cases, the mean cost was 376.47 USD (SD = 643.37), with a range of 21.46 to 10,521.43 USD. Overall case distribution and descriptive statistics are shown in Table 1. The average of annual PM2.5 was 42.59 µg/m3.
Table 1.
Demographic data of patients with respiratory disease in IPD and OPD in Chiang Mai during Thailand fiscal year 2023 (October 2022 – November 2023)
| Characteristics | OPD (n = 10,455) | IPD (n = 601) |
|---|---|---|
| n (%) | n (%) | |
| Categorical data | ||
| Sex | ||
| Male | 5223 (49.96) | 292 (48.59) |
| Female | 5232 (50.04) | 309 (51.41) |
| Service Area | ||
| Node 1: Chiang Mai City | 1972 (18.87) | 128 (21.30) |
| Node 2: Fang | 1283 (12.27) | 77 (12.81) |
| Node 3: San Sai | 2225 (21.28) | 132 (21.96) |
| Node 4: San Patong | 1509 (14.43) | 171 (28.46) |
| Node 5: Chom Thong | 3466 ( 33.15) | 93 (15.47) |
| Level of health care | ||
| Primary Health Care | 2309 (22.09) | 0 (0.00) |
| Second Health Care | 5255 (50.26) | 392 (65.22) |
| Tertiary Health Care | 2891 (27.65) | 209 (34.78) |
| Disease | ||
| Infectious Disease | ||
| Influenza (J09-11) | 751 (7.18) | 41 (6.82) |
| Pharyngitis infection (J02, J31) | 417 (3.99) | 0 (0.00) |
| Pneumonia (J13-18) | 34 (0.33) | 23 (3.83) |
| Acute bronchitis (J20-21) | 55 (0.53) | 1 (0.17) |
| Non-infectious disease | ||
| COPD (J40-44) | 7738 (74.01) | 332 (55.24) |
| Asthma (J45–46) | 1460 (13.96) | 204 (33.94) |
| Pollution period | ||
| Low pollution | 5049 (48.29) | 657 (49.92) |
| High pollution | 5406 (51.71) | 659 (50.08) |
| Continuous data | ||
| Age (Year) | ||
| Mean (SD) | 32.04 (27.32) | 44.05 (29.88) |
| Median (IQR) | 35 (55) | 53 (60) |
| Cost of treatment (USD) | ||
| Mean (SD) | 17.16 (44.94) | 376.47 (643.37) |
| Median (IQR) | 9.63 (15.48) | 197.15 (274.54) |
| Maximum | 1432.37 | 10,521.43 |
| Minimum | 0 | 21.46 |
| Length of stay (day) | ||
| Mean (SD) | N/A | 3.34 (4.0) |
| Median (IQR) | N/A | 2 (4) |
| Maximum | N/A | 39 |
| Minimum | N/A | 1 |
N/A: not applicable
The treatment cost Estimation
The treatment cost estimation of respiratory disease due to exposure to PM2.5 in Chiang Mai during the fiscal year 2023, using the maximum cost estimation scenario, showed a total cost of 460,122.58 USD, corresponding to an average cost 41.62 USD per case. For every 1 µg/m3 increase in PM2.5 levels, the estimated respiratory treatment cost across the province ranged from 403.84 to 13,159.87 USD. Non-infectious diseases accounted for a higher cost of treatment compared to infectious diseases. The overall cost estimation, under minimum to maximum scenarios, is presented in Table 2. The estimated respiratory treatment costs attributable to PM₂.₅ by district and service area in Chiang Mai are illustrated in Fig. 1, which shows that Node 1: Chiang Mai City demonstrated the highest cost burden, at 168,997.99 USD —accounting for 36.73% of the total cost in Chiang Mai across all districts. The estimated costs for each district and node, as well as the comparative economic costs of PM₂.₅-attributable infectious and non-infectious respiratory diseases across five service areas and 25 districts under the maximum scenario, are reported in Supplementary Material 2. The value of the health impact of respiratory disease is shown in Supplementary Material 3.
Table 2.
Cost Estimation in Chiang Mai Province and Per Individual Case
| Respiratory Disease | Chiang Mai province | Per person | ||||
|---|---|---|---|---|---|---|
| Minimum | Medium | Maximum | Minimum | Medium | Maximum | |
| The cost of treatment from average PM2.5 42.59 µg/m3 (USD) | ||||||
| Total | 14,119.75 | 93,951.95 | 460,122.58 | 1.28 | 8.50 | 41.62 |
| Non-infectious disease | 10,094.57 | 57,112.64 | 308,767.84 | 1.04 | 5.87 | 31.72 |
| Infectious disease | 4025.18 | 36,839.31 | 151,354.74 | 3.04 | 27.87 | 114.49 |
| The cost of treatment from every 1 µg/m3 PM2.5 increase (USD) | ||||||
| Total | 403.84 | 2687.10 | 13,159.87 | 0.04 | 0.24 | 1.19 |
| Non-infectious disease | 288.71 | 1633.47 | 8831.00 | 0.03 | 0.17 | 0.91 |
| Infectious disease | 115.12 | 1053.63 | 4328.87 | 0.09 | 0.80 | 3.27 |
Unit: USD (1 USD = 32.62 THB)
Fig. 1.
Estimated Respiratory Treatment Costs Attributable to PM2.5 by District and Service Area in Chiang Mai. This distribution map illustrates the maximum estimated treatment costs of respiratory health impacts attributable to PM2.5 exposure, including both infectious and non-infectious diseases, across 25 districts and 5 health service areas in Chiang Mai province. Color gradients represent cost variations, with darker shades indicating higher treatment costs and greater healthcare management needs. The highest treatment costs were observed in Mueang (102,878.88 USD/year), Mae Rim (66,119.12 USD/year), and Omkoi (44,556.46 USD/year) districts. Five health service nodes (Node 1–5) are also highlighted, with Node 1 exhibiting the highest total treatment cost (168,997.99 USD/year), followed by Node 4, Node 5, Node 3, and Node 2
Factors associated with treatment cost
The results of the multiple linear regression analysis assessing the association between treatment cost and multiple factors are presented in Table 3 for OPD and Table 4 for IPD.
Table 3.
Multiple linear regression analysis for the association between cost of treatment (USD) and multiple variables in out-patient respiratory disease during Thailand fiscal year 2023 (October 2022 – September 2023)
| Variables | Base case analysis | Sensitivity analysis | |||||
|---|---|---|---|---|---|---|---|
| Coefficient Estimate | 95% CI | p-value | Coefficient Estimate | 95% CI | p-value | ||
| Out-patient Respiratory case | |||||||
| Constant | 23.17 | 19.39, 26.95 | - | 22.38 | 18.43, 26.33 | - | |
| Level of Health Care | |||||||
| Primary Health Care | Ref | ||||||
| Second Health Care | 10.96 | 8.47, 13.45 | < 0.001 | 12.34 | 9.72, 14.96 | < 0.001 | |
| Tertiary Health Care | 8.93 | 6.03, 11.82 | < 0.001 | 10.00 | 6.93, 13.07 | < 0.001 | |
| Service Area | |||||||
| Node 2: Fang | -22.66 | -25.82, -19.49 | < 0.001 | -22.51 | -85.82, -19.20 | < 0.001 | |
| Node 3: San Sai | -28.34 | -31.01, -25.68 | < 0.001 | -28.05 | -30.82, -25.27 | < 0.001 | |
| Node 4: San Patong | -28.39 | -31.46, -25.31 | < 0.001 | -27.99 | -31.25, -24.74 | < 0.001 | |
| Node 5: Chom Thong | -26.14 | -28.79, -23.49 | < 0.001 | -25.32 | -28.12, -22.52 | < 0.001 | |
| Node 1: Chiang Mai City | Ref | ||||||
| Age | |||||||
| Per a one-year increase in age | 0.22 | 0.19, 0.25 | < 0.01 | 0.23 | 0.20, 0.26 | < 0.001 | |
| Sex | |||||||
| Male | Ref | ||||||
| Female | -0.32 | -1.99, 1.34 | 0.704 | -0.31 | -2.06, 1.45 | 0.733 | |
| Pollution Period | |||||||
| Low pollution season | Ref | ||||||
| High pollution season | 0.56 | -1.10, 2.22 | 0.509 | 0.47 | -1.30, 2.20 | 0.617 | |
| R2 | 0.09 | 0.09 | |||||
| Adjust R2 | 0.09 | 0.09 | |||||
Table 4.
Multiple linear regression analysis for the association between cost of treatment (USD) and multiple variables in in-patient respiratory disease during Thailand fiscal year 2023 (October 2022 – November 2023)
| Coefficient Estimate | 95% CI | p-value | ||
|---|---|---|---|---|
| In-patient Respiratory case | ||||
| Constant | 254.83 | 71.99, 443.35 | - | |
| Level of health care | ||||
| Secondary Health Care | Ref | |||
| Tertiary Health Care | 25.34 | -107.37, 158.602 | 0.705 | |
| Service Area | ||||
| Node 2: Fang | -393.22 | -523.78, -271.43 | < 0.001 | |
| Node 3: San Sai | -398.21 | -551.57, -253.72 | < 0.001 | |
| Node 4: San Patong | -343.40 | -509.15, -185.32 | < 0.001 | |
| Node 5: Chom Thong | -337.94 | -5713.09, -177.10 | < 0.001 | |
| Node 1: Chiang Mai City | Ref | |||
| Age | ||||
| Per a one year increase in age | 0.81 | -0.39, 2.03 | 0.185 | |
| Sex | ||||
| Male | Ref | |||
| Female | -10.70 | -80.50, 58.56 | 0.760 | |
| Pollution Period | ||||
| Low pollution season | Ref | |||
| High pollution season | 37.95 | -30.95, 107.70 | 0.277 | |
| Length of stay (LOS) | ||||
| Per day increase in LOS | 108.49 | 100.57, 118.83 | < 0.001 | |
| R2 | 0.58 | |||
| Adjust R2 | 0.57 | |||
For OPD, the multiple linear regression explained a significant proportion of the variation in treatment cost (R² = 0.09, adjusted R² = 0.09). The cost of treatment was significantly higher at tertiary (+ 8.93 USD; 95%CI: 6.03, 11.82) and secondary (+ 10.96 USD; 95%CI: 8.47, 13.45) healthcare levels. All peripheral areas showed significantly lower treatment costs, ranging from 22.66 to 28.39 USD, compared to the reference city area. Increasing age was positively associated with higher treatment costs of + 0.22 USD per year of age. However, no significant difference in treatment cost was observed between the high and low pollution seasons. Sensitivity analysis was conducted by excluding the zero cost of treatment, the data still showed the same significant factors associated with OPD and IPD treatment cost (R² = 0.09, adjusted R² = 0.09).
For IPD cases, the multiple linear regression explained a significant proportion of the variation in treatment cost (R² = 0.58, adjusted R² = 0.57). Treatment costs were significantly lower in peripheral areas, ranging from 343.40 to 393.22 USD, compared to the reference city area. Length of hospital stay was associated with treatment costs, increasing by 108.49 USD per day (95%CI: 100.57, 118.83). Although treatment cost was higher during the high-pollution period at about 37.95 USD (95%CI: -30.95, 107.70), this difference was not statistically significant. Other variables, including age, gender, and level of healthcare facility, did not show significant associations with treatment cost.
The Shapiro-Wilk test for normality indicated that the data were non-normality distributed in both OPD cases (W = 0.229 p < 0.01) and IPD cases (W = 0.52, p < 0.01).
Discussion
Our analysis estimated that the cost for treating respiratory diseases associated with PM2.5 ranged from 14,119.75 to 460,122.58 USD. Additionally, the estimated respiratory treatment cost across the province could increase by between 403.84 and 13,159.87 USD for every 1 µg/m³ rise in PM2.5 levels. These findings are consistent with previous studies conducted in the United States and China [28–31], which reported that rising PM2.5 concentrations were associated with increased healthcare costs and economic losses across various respiratory conditions, including COPD, asthma, and infectious diseases, for both short-term and long-term exposure periods. For instance, a study in the United States found that asthma-related treatment costs rose by 15–93 USD per 1 µg/m³ increase in monthly PM2.5 concentration [30]. Chiang Mai’s Gross Provincial Product (GPP) in 2023 was 8,507.76 million USD in total and 4,750.80 USD per capita [32]. In our study, the maximum estimated treatment cost represented approximately 0.0134% of Chiang Mai’s GPP and 0.025% of the per capita GPP in 2023. However, during the haze season, PM2.5 levels in Chiang Mai can rise to over 60 µg/m³, potentially leading to significantly higher healthcare expenses for the population. This study underscores the critical importance of air pollution control policies and the revision of Thailand’s PM2.5 air quality standards. Reducing PM2.5 levels not only decreases direct treatment costs but also contributes to reducing broader healthcare expenditures and parameters, including medication costs, social care needs, hospital admissions, and mortality rates among patients with respiratory diseases [33, 34].
Analysis of the demographic data reveals significant factors influencing expenses incurred, including level of health care, service area, and length of hospital stay. The level of health care can be explained by its correlations with the service areas, as the hospital with the highest average expenses is a provincial hospital and tertiary hospital located in the city district. The patient referral system initiates at the service unit nearest to the patient and progresses to hospitals with enhanced care capabilities based on disease severity, which may require higher procedures, investigations, and treatment might require [27]. Regarding the length of hospital stay, a previous study in Thailand, which analyzed secondary data from pneumonia patients, found that a longer hospital stay is associated with complications that occur during treatment [35]. This may also lead to increased treatment costs. From the aforementioned information, it can be seen that the treatment in the OPD setting has similar costs for patient care within the same disease group. This may be due to the initial management being conducted according to standards or the same treatment guidelines, resulting in minimal differences in medical expenses. In contrast, patient care in the IPD setting has more factors affecting treatment, including potential complications and length of stay, which are associated with increased medical costs [26, 36]. An absence of statistically significant difference in treatment costs between high and low pollution seasons contradicts our hypothesis that environmental consequences should correlate directly with PM2.5 levels. This may be attributed to the comparable volume of service visits over both periods. Data regarding service visits to public health facilities is constrained in terms of collection. Moreover, the data was derived from the “Operational Manual for Medical and Public Health Concerning Particulate Matter 2.5 Microns or Less (PM2.5),” [37], which outlines PM2.5-related illness categories by ICD-10. Although respiratory diseases—both infectious and non-communicable—are clearly classified, providers may not always recognize or record whether a condition is linked to air pollution, nor consistently identify cases using ICD-10 codes Y97 (environmental-pollution-related condition) or Z58.1 (exposure to air pollution). As a result, the current data may not fully reflect the true situation. Therefore, it may be necessary to provide feedback to relevant agencies to improve the completeness of data recording systems and to encourage the use of air-pollution–specific ICD-10 codes in cases where illness is suspected to be related to air pollution.
In terms of budget management, we found that the estimated cost from our study could be higher than the per capita average of healthcare costs under the Universal Health Coverage scheme in the same fiscal year (103.8 USD) [38] for infectious disease, and could be use 40% of budget in total cases. For all respiratory diseases on average, it could account for up to 40% of this budget. Current data suggest that areas with larger hospitals exhibit elevated resource utilization and treatment expenses compared to smaller hospitals. Budget al.location derived exclusively from standard treatment entitlements may prove inadequate, particularly in years characterized by elevated PM2.5 levels. Studies on the financial consequences of diseases use numerous analytical approaches, including the human capital method, the disease cost method, the Value Of a Statistical life (VOSL), and the Willingness To Pay (WTP) method [10]. The objective of these computations is to assess the financial worth of death risk, which pertains to the burden resulting from disease. These calculations are highly beneficial for reflecting the problems and burdens of relevant departments that need to care for those affected. However, in the details of allocating healthcare budgets to align with the population’s primary medical right, it may be necessary to use database-based information from actual public health services to directly communicate with relevant agencies that the average cost of treatment per episode may not yet al.ign with the annual budget al.located per capita. In Thailand, a recent transition involved shifting the administration of primary healthcare hospitals from the Ministry of Public Health to local administrative organization. This structural change may enhance coordination and enable more efficient budget al.location for supplementary funding by local authorities, particularly in supporting primary healthcare services during the haze season. The findings from this study will support the implementation of improved healthcare strategies during transition. Conversely, alongside the costs for medical care, the costs associated with disease prevention are equally significant in mitigating the negative impacts [39]. This entails the prevention of PM2.5 emissions, including the enactment of rules prohibiting burning during wildfire season. Policies that promote electric vehicles, educate the public on the health risks of PM2.5, and encourage behaviors to minimize PM2.5 exposure might alleviate potential health effects and may decrease the overall costs or burdens related to the PM2.5 problem [40–43]. Interestingly, we found that the minimum recorded cost does not correspond with the anticipated cost, as certain OPD costs are recorded as zero, which is implausible. Discrepancies in the hospital’s recorded cost may arise from difficulties with health insurance and schemes. Hospitals can retroactively assert charges for treating people utilizing the universal health coverage scheme from the National Health Security Office (NHSO). In these cases, the cost of treatment may initially be recorded as zero until reimburstment is processed.
This research is significant as it is one of the initial studies to analyze the financial burden of PM2.5 concerning medical treatment costs and its relationship with varying PM2.5 levels. Prior research in Thailand may lack this level of specificity. By estimating minimum to maximum cost scenarios for respiratory diseases attributable to PM2.5 exposure, this study provides a framework for forecasting treatment costs in pollution-affected areas. However, the weakness of this research lies in its retrospective descriptive study approach, utilizing secondary data that contains various ambiguities and significant differences. First, the treatment cost from our study is the average cost per visit for medical services, while the NHSO’s average healthcare budget per capita is the annual budget for average treatment per capita. In addition, the evidence suggests potential differences between actual spending and recorded expenses from zero OPD costs. Therefore, the data obtained from this study may serve as preliminary information for relevant parties to see the public health burden of PM2.5. However, its application may require additional data for further consideration in budget allocation and cannot be used solely based on the data from this study. Additionally, only direct medical treatment costs from government-affiliated hospitals were included; indirect costs and non-medical expenses were excluded, potentially underestimating the true economic burden. Secondly, there are data limitations. As mentioned earlier, the cost of treatment data we analyzed and calculated was secondary data, which had incompleteness that could have resulted from non-real-time data entry into the system, leading to missing data. Key variables that influence treatment costs, such as disease severity and patient comorbidities, were not available, which may affect the accuracy of our findings. Another limitation was the data specificity of diagnostic information. Our study results could not differentiate diseases related to PM2.5 exposure because specific causes were not identified, such as ICD10 code Z58.1 or Y97. The findings of this study could be beneficial if feedback is provided to relevant agencies to highlight system deficiencies in data collection, leading to the development of more complete medical records in the future. Furthermore, this study’s data solely encompasses governmental healthcare facilities, thereby limiting its generalizability to other facilities like private clinics and hospitals. Therefore, if agencies in these sectors intend to use this study’s data, they may need to consider that the costs incurred could differ. If data from these sectors were added, the study’s results could better and more comprehensively reflect the healthcare system. Lastly, there are the limitations of study modeling and analysis. Since the data used in this study is general data from Chiang Mai province, and the PM2.5 level measurement data we used is more comprehensive than the gold standard method in the area, it is still unable to identify individual PM2.5 exposure details, potentially leading to an ecological fallacy. Furthermore, the analysis did not include conditions with long latency periods, such as lung cancer, which necessitated long-term exposure data.
To ensure the accuracy and comprehensibility of the data for future budget evaluations, we suggest setting up systems that accurately record and store medical treatment information and establish precise terminology for noting details like diagnoses, procedures, and medical expense claims. Using identification of treatment data, such as the frequency of service received per individual, would improve understanding of incurred expenses in healthcare facilities and support better care planning for frequent patients [26]. Future studies may require further reliance on health-related data from diverse sources to evaluate the consistency and completeness of the information. Including indirect and non-medical costs—such as transportation, caregiving, and productivity loss—would offer a more complete estimate of the total economic burden. Additionally, comparing treatment costs attributable to PM₂.₅ with other causes using alternative economic analyses and studies could better quantify their relative impact. Further investigation into chronic conditions like lung cancer and the economic impacts of sustained exposure to air pollution is warranted.
Conclusions
The average PM2.5 concentration in Chiang Mai province, measured at 42.59 µg/m3, can impose significant treatment costs for respiratory diseases, reaching up to 460,122.58 USD, and costs can be increased up to 13,159.87 USD for every 1 µg/m3 increase in PM2.5 levels. The service area with higher-level healthcare hospitals area contributed substantially to the treatment burden in both OPD and IPD cases. This study emphasizes the critical need for funding strategies in polluted-affected areas, particularly during the haze season. Furthermore, it underscores the importance of implementing pollution control policies to minimize not only adverse health outcomes but also the associated treatment costs, which contribute to boarder economic loss.
Supplementary Information
Acknowledgements
The authors gratefully acknowledge the Chiang Mai Provincial Public Health Office for providing access to the secondary data used in this study.
Declaration of generative AI and AI-assisted technologies in the writing process
During the preparation of this work the authors used ChatGPT3.0 in order to improve readability and language of the work. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.
Authors’ contributions
PA.: Conceptualization, Investigation, Methodology, Formal analysis, Data curation, Validation, Writing – original draft, Writing – review and editing. KJ.: Conceptualization, Investigation, Methodology, Formal analysis, Software, Visualization, Writing – original draft, Writing – review and editing. SS.: Conceptualization, Investigation, Methodology, Resources, Writing – review and editing. JP.: Conceptualization, Investigation, Methodology, Formal analysis, Data curation, Software, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review and editing. All authors read and approved the final manuscript.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data availability
The data that support the findings of this study are available from Chiang Mai Provincial Public Health Office but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data may be available upon reasonable request and with permission from the Chiang Mai Provincial Public Health Office. For inquiries regarding access to the raw data, please contact: Environmental Health Department Chiang Mai Provincial Public Health Office Email: saraban-cmi@moph.go.th.
Declarations
Ethics approval and consent to participate
This study was reviewed and approved by the Research Ethics Committee of the Faculty of Medicine, Chiang Mai University, under the Exemption Category. The approval study number is COM-2567-0045. All procedures were performed in compliance with ethics committee guidelines. The requirement for informed consent was waived, as approved under the Exemption Category outlined in the announcement of the Research Ethics Committee of the Faculty of Medicine, Chiang Mai University dated 29 May 2020. This waiver was granted due to the use of anonymized secondary data derived from the annual public health reports provided by the Provincial Public Health Office.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Pheerasak Assavanopakun and Kannika Jarernwong contributed equally to this work.
References
- 1.Shi Y, Matsunaga T, Yamaguchi Y, Li Z, Gu X, Chen X. Long-term trends and Spatial patterns of satellite-retrieved PM(2.5) concentrations in South and Southeast Asia from 1999 to 2014. Sci Total Environ. 2018;615:177–86. [DOI] [PubMed] [Google Scholar]
- 2.Inlaung K, Chotamonsak C, Macatangay R, Surapipith V. Assessment of transboundary PM2.5 from biomass burning in Northern Thailand using the WRF-Chem model. Toxics 2024, 12(7). [DOI] [PMC free article] [PubMed]
- 3.Pollution Control Department (PCD): Pollution Control Department Announcement: Air Quality Index of Thailand, B.E. 2566; Pollution Control Department Ed.: Bangkok, Thailand. 2023.
- 4.World Health Organization (WHO). WHO global air quality guidelines: particulate matter (PM2. 5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide. Geneva: World Health Organization; 2021. [PubMed] [Google Scholar]
- 5.Sukkhum S, Lim A, Ingviya T, Saelim R. Seasonal patterns and trends of air pollution in the upper Northern Thailand from 2004 to 2018. Aerosol Air Qual Res. 2022;22(5):210318. [Google Scholar]
- 6.Kyung SY, Jeong SH. Particulate-Matter related respiratory diseases. Tuberc Respir Dis (Seoul). 2020;83(2):116–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Nhung NTT, Amini H, Schindler C, Kutlar Joss M, Dien TM, Probst-Hensch N, Perez L, Kunzli N. Short-term association between ambient air pollution and pneumonia in children: A systematic review and meta-analysis of time-series and case-crossover studies. Environ Pollut. 2017;230:1000–8. [DOI] [PubMed] [Google Scholar]
- 8.Sapbamrer P, Assavanopakun P, Panumasvivat J. Decadal trends in ambient air pollutants and their association with COPD and lung cancer in upper Northern thailand: 2013–2022. Toxics 2024, 12(5). [DOI] [PMC free article] [PubMed]
- 9.Health insurance information. service system [https://ucinfo.nhso.go.th/ucinfo
- 10.Lu ZN, Zhao M, Guo Y, Hao Y. Evaluating PM(2.5) -Related health costs in China-Evidence from 140 Chinese cities. Int J Health Plann Manage. 2022;37(4):2376–94. [DOI] [PubMed] [Google Scholar]
- 11.Alberini A, Cropper M, Krupnick A, Simon NB. Does the value of a statistical life vary with age and health status? Evidence from the US and Canada. J Environ Econ Manag. 2004;48(1):769–92. [Google Scholar]
- 12.Daily Foreign Exchange Rates. [https://www.bot.or.th/en/statistics/exchange-rate.html
- 13.Chiangmai Provincial Public Health Office. Chiang Mai Province patient transfer service system manual and Lanna zone 1. Chiang Mai; 2022.
- 14.Jarernwong K, Gheewala SH, Sampattagul S. Health impact related to ambient particulate matter exposure as a Spatial health risk map case study in Chiang Mai, Thailand. Atmosphere. 2023;14(2):261. [Google Scholar]
- 15.Samae H, Suriyawong P, Yawootti A, Phairuang W. Sate Sampattagul 5 O: Precision and Accuracy Analysis of PM₂.₅ Light-Scattering Sensor: Field and Laboratory Experiments. Atmosphere, 16(1), Article 76 2025, 16(1):76.
- 16.Yang S, Fang D, Chen B. Human health impact and economic effect for PM2. 5 exposure in typical cities. Appl Energy. 2019;249:316–25. [Google Scholar]
- 17.Shi L, Zanobetti A, Kloog I, Coull BA, Koutrakis P, Melly SJ, Schwartz JD. Low-concentration PM2. 5 and mortality: estimating acute and chronic effects in a population-based study. Environ Health Perspect. 2016;124(1):46–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Official statistics registration systems. [https://stat.bora.dopa.go.th
- 19.Jarernwong K. Extrapolation of human health damage cost from particulate matter 2.5 exposure in Chiang Mai province using grey model. Doctoral dissertation. Chiang Mai: Chiang Mai University; 2024.
- 20.Do V, Chen C, Benmarhnia T, Casey JA. Spatial heterogeneity of the respiratory health impacts of wildfire smoke PM(2.5) in California. Geohealth. 2024;8(4):e2023GH000997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kloog I, Nordio F, Zanobetti A, Coull BA, Koutrakis P, Schwartz JD. Short term effects of particle exposure on hospital admissions in the Mid-Atlantic states: a population estimate. PLoS ONE. 2014;9(2):e88578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Xu D, Tian Y. A comprehensive survey of clustering algorithms. Annals Data Sci. 2015;2(2):165–93. [Google Scholar]
- 23.Tzortzis G, Likas A. The minmax k-Means clustering algorithm. Pattern Recogn. 2014;47(7):2505–16. [Google Scholar]
- 24.Chuesakoolvanich K. Cost of hospitalizing asthma patients in a regional hospital in Thailand. Respirology. 2007;12(3):433–8. [DOI] [PubMed] [Google Scholar]
- 25.Meng Z, Hui W, Cai Y, Liu J, Wu H. The effects of DRGs-based payment compared with cost-based payment on inpatient healthcare utilization: A systematic review and meta-analysis. Health Policy. 2020;124(4):359–67. [DOI] [PubMed] [Google Scholar]
- 26.Rattanavipapong W, Wang Y, Butchon R, Kittiratchakool N, Thammatacharee J, Teerawattananon Y, Isaranuwatchai W. Retrospective secondary data analysis to identify high-cost users in inpatient department of hospitals in Thailand, a middle-income country with universal healthcare coverage. BMJ Open. 2021;11(7):e047330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wongpairin A, Lim A, Tongkumchum P, Thongpeth W, Khurram H. Assessing diagnosis-related groups based direct medical expenditures of chronic disease patients in general hospital of lower Southern Thailand. Cost Eff Resour Alloc. 2024;22(1):87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hu F, Jing L, Shi R. Analysis of related factors and disease costs of respiratory infection and environmental pollution in children. Technol Health Care. 2020;28(S1):355–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Li H, Zhai F, Ma Y, Wang Y, Gu R, Cao C, Wang L, Ge B, Wu W, Zhai C, et al. Associations of short-term exposure to air pollution with outpatient visits and treatment costs for chronic obstructive pulmonary disease in Xinxiang, China (2016–2021). Sci Total Environ. 2025;963:178438. [DOI] [PubMed] [Google Scholar]
- 30.Roy A, Sheffield P, Wong K, Trasande L. The effects of outdoor air pollutants on the costs of pediatric asthma hospitalizations in the united States, 1999 to 2007. Med Care. 2011;49(9):810–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhang P, Zhou X. Pricing air pollution: evidence from short-term exposure to air pollution on hospitalization of acute bronchitis and chronic obstructive pulmonary disease in Southwestern China. Int Health. 2022;14(6):572–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.GPP Dashboard. [https://lookerstudio.google.com/u/0/reporting/704fe6c7-aecb-4362-9062-3dc971c75982/page/p_nmxeulqptc
- 33.Pimpin L, Retat L, Fecht D, de Preux L, Sassi F, Gulliver J, Belloni A, Ferguson B, Corbould E, Jaccard A, et al. Estimating the costs of air pollution to the National health service and social care: an assessment and forecast up to 2035. PLoS Med. 2018;15(7):e1002602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Yang J, Zhang B. Air pollution and healthcare expenditure: implication for the benefit of air pollution control in China. Environ Int. 2018;120:443–55. [DOI] [PubMed] [Google Scholar]
- 35.Reechaipichitkul W, Thavornpitak Y, Sutra S. Burden of adult pneumonia in thailand: a nationwide hospital admission data 2010. J Med Assoc Thai. 2014;97(3):283–92. [PubMed] [Google Scholar]
- 36.Li L, Du T, Zhang C. The impact of air pollution on healthcare expenditure for respiratory diseases: evidence from the people’s Republic of China. Risk Manag Healthc Policy. 2020;13:1723–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Ministry of Public Health T. Operational manual for medical and public health concerning particulate matter 2.5 microns or less (PM2.5). In. Nonthaburi. Thailand: Ministry of Public Health, Thailand; 2022. [Google Scholar]
- 38.National Health Security Office. Summary of key performance achievements for fiscal year 2023 (B.E. 2566), problems and obstacles in operations, and proposed solutions. In. Thailand: National Health Security Office; 2023.
- 39.Fisk WJ, Chan WR. Health benefits and costs of filtration interventions that reduce indoor exposure to PM2.5 during wildfires. Indoor Air. 2017;27(1):191–204. [DOI] [PubMed] [Google Scholar]
- 40.Janjua S, Powell P, Atkinson R, Stovold E, Fortescue R. Individual-level interventions to reduce personal exposure to outdoor air pollution and their effects on people with long-term respiratory conditions. Cochrane Database Syst Rev. 2021;8(8):CD013441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Jin Y, Andersson H, Zhang S. Air pollution control policies in china: A retrospective and prospects. Int J Environ Res Public Health 2016, 13(12). [DOI] [PMC free article] [PubMed]
- 42.Rizza V, Torre M, Tratzi P, Fazzini P, Tomassetti L, Cozza V, Naso F, Marcozzi D, Petracchini F. Effects of deployment of electric vehicles on air quality in the urban area of Turin (Italy). J Environ Manage. 2021;297:113416. [DOI] [PubMed] [Google Scholar]
- 43.Chiang Mai Governor Office. Chiang Mai Province Announcement on legal measures for controlling open burning within the burn-control zone of Chiang Mai Province for fiscal year 2025 (B.E. 2568). In. Chiang Mai: Chiang Mai Governor Office; 2025.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from Chiang Mai Provincial Public Health Office but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data may be available upon reasonable request and with permission from the Chiang Mai Provincial Public Health Office. For inquiries regarding access to the raw data, please contact: Environmental Health Department Chiang Mai Provincial Public Health Office Email: saraban-cmi@moph.go.th.