Lifetime costs of alcohol consumption in Thailand: protocol for an incidence-based cost-of-illness study using Markov model

Abstract

Introduction

Several prevalence-based cost-of-illness (COI) studies have been conducted to estimate the economic burden of alcohol consumption borne by a particular society in a given year. Yet there are few studies examining the economic costs incurred by an individual drinker over his/her lifetime. Thus, this study aims to estimate the costs incurred by an individual drinker’s alcohol consumption over his or her lifetime in Thailand.

Methods and analysis

An incidence-based COI approach will be employed. To project individuals’ associated costs over a lifetime, a Markov modelling technique will be used. The following six alcohol-related diseases/conditions will be considered in the model: hypertension, haemorrhagic stroke, liver cirrhosis, liver cancer, alcohol use disorders and road injury. The analysis will cover both direct (ie, direct healthcare cost, costs of property damage due to road traffic accidents) and indirect costs (ie, productivity loss due to premature mortality and hospital-related absenteeism). The human capital approach will be adopted to estimate the cost of productivity loss. All costs will be presented in Thai baht, 2022.

Ethics and dissemination

The Institutional Review Board of Mahidol University, Faculty of Dentistry/Faculty of Pharmacy has confirmed that no ethical approval is required (COE.No.MU-DT/PY-IRB 2021/010.0605). Dissemination of the study findings will be carried out through peer-reviewed publications, conferences and engagement with policy-makers and public health stakeholders.

Correspondence to Dr Montarat Thavorncharoensap; [email protected]

STRENGTHS AND LIMITATIONS OF THIS STUDY

• The study will not only provide the lifetime cost of individual drinker but also the total amount that could be saved from his or her ceasing to drink alcohol at different ages over a lifetime.

• Both direct and indirect costs will be considered.

• The findings from the study will be useful for evaluating the cost-effectiveness of the policy/intervention aimed at reducing the prevalence/incidence of alcohol consumption.

• Only six of the alcohol-attributable diseases will be included in the analysis.

• Several assumptions will be inevitably required but were chosen based on expert judgement and will be further externally validated.

Introduction

Alcohol consumption is a global public health concern. Alcohol intake contributes to a wide range of health problems, including liver diseases, cardiovascular disorders, mental health disorders and various types of cancer.¹ Moreover, it is associated with a significant number of road traffic injuries, other injuries and violence.^{2 3} In 2016, alcohol use led to approximately 3 million deaths globally, representing 2.2% and 6.8% of all deaths in females and males, respectively.⁴ Additionally, it was ranked as the seventh leading risk factor for burden of disease, contributing to 6% of the total disability-adjusted life years (DALYs) for males and 1.6% for females.⁴ Alcohol consumption also leads to substantial economic burden worldwide.^{5 6}

Economic burden of alcohol consumption can be measured through a cost-of-illness (COI) study. In a COI study, the total cost incurred due to a disease or condition is estimated.^7–11 Two of the methodologies commonly used when conducting a COI study are prevalence-based and incidence-based approaches.^{8 9} A prevalence-based approach provides the total costs incurred by the disease/condition within a specified time period, typically 1 year, regardless of when the disease/condition first occurred.^{8 9} Incidence-based approaches determine the lifetime cost for a new case of disease/condition.^{8 9} The results from a prevalence-based approach can be used to raise awareness of the economic impact of disease/condition on society, while an incidence-based approach can provide an estimate of the savings that potentially accrue if a particular preventive measure is implemented. In addition, the estimates derived from an incidence-based approach can provide essential information for conducting cost-effectiveness analyses of specific preventive measures.^{8 9}

A recent systematic review of the 29 prevalence-based COI studies indicated that the economic burden of alcohol consumption represented 2.6% of the gross domestic product (GDP).⁵ Nonetheless, incidence-based COIs on alcohol consumption are scarce. Only one study estimating the lifetime cost of alcohol consumption was identified.¹² The study employed the Markov modelling technique to estimate the lifetime cost of alcohol consumption from healthcare system perspective, where only direct medical costs were included in the analysis.¹² The model comprised five possible states of alcohol consumption for an individual: hazardous drinking, harmful drinking, ex-hazardous drinking, ex-harmful drinking and death.¹² Notably, the parameters used in the model were based on a UK setting.¹²

Thailand is a middle-income country located in South-East Asia. According to the 2021 Heath Behavior of Population Survey,¹³ the prevalence of alcohol consumption among Thai adults aged 15 and over was 28.0%. In Thailand, alcohol use is the second leading risk factor for burden of disease in males, accounting for 14.0% of total DALYs in males.¹⁴ In addition, alcohol use is a significant risk factor contributing to high burden of cardiovascular disease, cancer, and road traffic accidents in males.¹⁴ In 2006, the total economic cost of alcohol consumption in Thailand was estimated at 156 105.4 million baht, representing 1.99% of the country’s GDP.¹⁵ Recently, it was estimated at 165 450.5 million baht, representing 1.02% of GDP in 2021.¹⁶ The previous incidence-based COI study in Thailand¹⁷ was conducted in 2011. In such study, only the indirect costs due to premature mortality were considered, with the analysis based on the concept that alcohol use increases the risk of mortality.¹⁷ The objective of this study is to estimate the lifetime costs of alcohol consumption in Thailand, using incidence-based COI analysis. In addition, potential economic costs saved from quitting drinking at particular ages (ie, 35, 45, 55 and 65 years old) will be estimated.

Methods and analysis

Study design

This study will adopt an incidence-based COI methodology. The analysis will be based on healthcare and societal perspectives. A Markov model will be developed to predict the economic burden of alcohol consumption among Thai drinkers over their lifespan.

Study population

A cohort of 1000 healthy Thai individuals will be entered into the Markov model. The starting age for the males is 20 years old while that for females is 23 years old. These ages were chosen as they are the ages that males and females in Thailand start drinking on average.¹³ For each gender, six different scenarios will be created depending on the individual’s alcohol use status: (1) lifetime abstainers, (2) lifetime drinkers, (3) drinkers who quit at 35 years old, (4) drinkers who quit at 45 years old, (5) drinkers who quit at 55 years old and (6) drinkers who quit at 65 years old. During the period that each individual is a drinker, he/she will be assumed to drink at the constant amount (ie, 30.1 g/day in males and 8.0 g/day in females), which is the average daily alcohol intake of Thai population¹⁸ until he/she abruptly quits at the specified age (ie, 35, 45, 55 and 65 years old).

Model structure and model assumption

Due to the limitation of epidemiological parameters, to avoid overcomplicating the model’s structure, and to keep it parsimonious, only six major alcohol-related disease/conditions will be considered in the model: hypertension, haemorrhagic stroke, liver cirrhosis, liver cancer, alcohol use disorders (AUDs) and road injury. These diseases/conditions are known to have a high economic burden in Thailand and are strongly associated with alcohol consumption.^{2 15} Additionally, these diseases/conditions were among the primary causes of death and DALYs in both Thai males and females in 2019.¹⁹ As shown in figure 1, for each cycle, every healthy individual has a probability of developing any of the six alcohol-related diseases. Nevertheless, AUDs can only be developed in those who are currently drinkers. It should be noted that an individual with AUD can also develop hypertension, haemorrhagic stroke, liver cirrhosis, hepatocellular carcinoma and road injury. Similarly, an individual with hypertension can develop haemorrhagic stroke, liver cirrhosis and hepatocellular carcinoma. In addition, liver cirrhosis can lead to hepatocellular carcinoma. As shown in the model, injuries will be classified as fatal, and non-fatal (ie, minor, severe and disability) conditions while haemorrhagic stroke will be classified as fatal and non-fatal stroke. It should be noted that individual who experiences a road injury will be allocated to one of the three possible injury states (ie, minor, severe, disability) right away rather than waiting for the next cycle. The proposed model will run until individuals die or reach the age of 100 years. The cycle length will be set to 1 year.

Figure 1. Markov model structure.

Mainly due to the availability of epidemiological parameters and to avoid overcomplicating the model’s structure, the following main assumptions will be adopted in the analysis: (1) the model will exclude those with concurrent conditions; (2) individual drinker will drink at constant amount during the specified period; (3) individuals with AUDs and hypertension are also considered at risk of developing other diseases in the model; (4) individual with cirrhosis, liver cancer, hypertension, or haemorrhagic stroke will not be considered at risk of road injuries; (5) individuals can recover only from AUDs and non-fatal injuries but not from other health states; (6) individual with non-fatal injuries will recover to the healthy state in the next cycle; (7) the annual probability of death among individuals with alcohol-attributable disease will not depend on drinking status; and (8) the risks for former drinkers will be assumed to be independent of their previous drinking duration.

Model parameters

Transitional probabilities

Transitional probabilities of developing alcohol-related diseases/conditions

The transitional probabilities of developing alcohol-related diseases/conditions by age, gender and drinking status will be calculated. Age-specific morbidity rates of all diseases except hypertension in the general population will be obtained from a database from the Burden of Disease Research Program Thailand (BOD Thailand). The rate of developing hypertension as well as the rate of developing liver cancer in patients with cirrhosis, and the rate of developing haemorrhagic stroke in patients with hypertension will be derived from a literature review. The morbidity rates among the general population will then be converted to the annual transitional probability by using formula (1)

(1) Probability1−year=1−exp(−Incidencerate1−year×1)

Given that the general population consists of current drinkers, former drinkers and abstainers, there is a need to convert the probability of developing alcohol-related diseases among the general population into the probability of developing alcohol-related diseases among the abstainers using formula (2)

(2) $$\begin{gathered} {\displaystyle \begin{array}{l}{\mathrm{P}}_{\mathrm{l}\mathrm{a}}={\mathrm{P}}_{\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{p}\mathrm{o}\mathrm{p}}/\left(\left(1-\mathrm{P}\mathrm{r}\mathrm{e}{\mathrm{v}}_{\mathrm{f}\mathrm{d}}-\mathrm{P}\mathrm{r}\mathrm{e}{\mathrm{v}}_{\mathrm{c}\mathrm{d}}\right)+\left(\mathrm{P}\mathrm{r}\mathrm{e}{\mathrm{v}}_{\mathrm{f}\mathrm{d}}\times \mathrm{R}{\mathrm{R}}_{\mathrm{f}\mathrm{d}}\right)+ \\ \left(\mathrm{P}\mathrm{r}\mathrm{e}{\mathrm{v}}_{\mathrm{c}\mathrm{d}}\times \mathrm{R}{\mathrm{R}}_{\mathrm{c}\mathrm{d}}\right)\right)\end{array}} \end{gathered}$$

where

P_la=Probability of developing the diseases in lifetime abstainers,

P_genpop=Probability of developing the diseases in general population,

Prev_fd=Prevalence of former drinkers,

Prev_cd=Prevalence of current drinkers,

RR_fd=Relative risk of developing the diseases in former drinkers, as compared with abstainers,

RR_cd=Relative risk of developing the diseases in current drinkers, as compared with abstainers.

It should be noted that the prevalence of alcohol drinking will be collected from the 2021 Heath Behavior of Population Survey¹³ while gender-specific RRs of developing the diseases will be derived from related epidemiological studies. The annual probabilities of developing the diseases among current and former drinkers will then be estimated using formulas (3) and (4)

(3) ${\displaystyle {\mathrm{P}}_{\mathrm{c}\mathrm{d}}={\mathrm{P}}_{\mathrm{l}\mathrm{a}}\times \mathrm{R}{\mathrm{R}}_{\mathrm{c}\mathrm{d}}}$

(4) ${\displaystyle {\mathrm{P}}_{\mathrm{f}\mathrm{d}}={\mathrm{P}}_{\mathrm{l}\mathrm{a}}\times \mathrm{R}{\mathrm{R}}_{\mathrm{f}\mathrm{d}}}$

where

P_la=Probability of developing the diseases in lifetime abstainers,

P_cd=Probability of developing the diseases in current drinkers,

P_fd=Probability of developing the diseases in former drinkers,

RR_cd=Relative risk of developing the diseases in current drinkers, as compared with abstainers,

RR_fd=Relative risk of developing the diseases in former drinkers, as compared with abstainers.

It is important to note that AUDs are 100% attributable to alcohol.² Therefore, the annual probabilities of developing AUDs among lifetime abstainers and former drinkers will be set at 0. The annual remission rate of AUDs will be obtained from a literature review. The annual probability of developing hypertension, haemorrhagic stroke, liver cirrhosis, liver cancer among general population will be used to estimate the probabilities among individuals without AUDs (online supplemental formula S1). Subsequently, it will be converted into the probability among patients with AUDs by multiplying the probability among patients without AUDs by the relevant RRs (online supplemental formula S2).

In our analysis, not all current drinkers but only binge drinking will increase the risk of receiving a road injury. The annual probability of experiencing road injuries among abstainers will be estimated from the annual probability of receiving a road injury in the general population (which consisted of lifetime abstainer, former drinker, current drinker with binge drinking, and current drinker without binge drinking) after considering the prevalence of binge drinking and the risk of binge drinking on the road injury (online supplemental formula S3). The probability of road injuries among binge drinkers will then be estimated based on the probability of receiving a road injury among abstainers considering for an increased risk of experiencing a road injury due to binge drinking, the annual frequency of binge drinking in binge drinkers, and the proportion of drink driving in binge drinkers (online supplemental formula S4). Information on the annual frequency of binge drinking and proportion of drink driving in binge drinker will be derived from the 2021 Heath Behavior of Population Survey.¹³ Finally, to estimate the annual probability of road injuries among current drinker, the proportion of current drinker with binge drinking will be multiplied with the annual probability of getting road injuries among binge drinker.

Transitional probabilities of death in patients with alcohol-related diseases/conditions

Transitional probability of death from each alcohol-related disease (except road injury) will be obtained from survival analysis/clinical studies, which were conducted in Thailand or comparable countries. The annual transitional probability of receiving a fatal road injury among the general population will be obtained from the BOD Thailand.

Transitional probability of death from other causes in the general population

Age-specific mortality rates of the diseases (except for hypertension) in the general population will be obtained from the BOD Thailand. For hypertension, the mortality rate will be derived from a database from the Strategy and Planning Division, Office of the Permanent Secretary, Ministry of Public Health. The annual probabilities of death from the other causes in the general population will be estimated by subtracting the sum of the probabilities of death from the alcohol-related diseases included in the model from the annual probabilities of death in the Thai population, which will be obtained from the WHO life tables.²⁰

Details of RRs and transitional probabilities that will be included in the analysis along with its sources are reported in online supplemental tables S1 and S2.

Cost parameters

Direct costs consist of direct medical costs, direct non-medical costs, and costs of property damage due to road-traffic accidents. Indirect costs consist of costs of hospital-related absenteeism and costs of premature mortality.

Annual direct medical costs of each alcohol-related disease will be derived from the claims database of the National Health Security Office (NHSO). The direct non-medical costs will be estimated using the unit costs derived from the standard cost lists for health economic evaluations in Thailand²¹ while the resource use associated with each alcohol-attributable disease (eg, annual number of outpatient and inpatient visits per patient) will be obtained from the claims database of the NHSO. Direct costs of property damage due to road traffic accidents will be derived from the annual monetary value of property damage due to road traffic accidents reported by the Royal Thai Police for 2021.²² While we assumed that individual with disability from non-fatal injuries will return to healthy state in the next cycle, it should be noted that the discounted lifetime cost of disability will be applied to account for the lifetime burden from disability. It should also be noted that the costs of individual with AUD, who subsequently develops other disease/condition, will be the summation of the cost for treating the disease/condition and the cost for treating AUDs. The same concept will also be applied to the individual who develops haemorrhagic stroke after hypertension and the individual who develops hepatocellular carcinoma from liver cirrhosis.

Human capital approach will be adopted to estimate the indirect costs. The gross national income (GNI) per capita for the year 2022²³ will be used for monetary conversion purposes. The expected life expectancy will be obtained from the WHO life table for Thailand.²⁰ Additionally, the average 20-year GNI growth rate²⁴ and the discount rate of 3%²⁵ will be applied to estimate the present value of lost earnings.

All of the costs will be reported in Thai baht at the currency rate for the year 2022. Details on the cost variables are summarised in online supplemental table S3.

Model outcomes

The primary outcomes of the study model are the lifetime costs of alcohol consumption in Thailand by gender, which will be estimated as the excess costs occurring due to alcohol-related illness for lifetime drinkers compared with lifetime abstainers. In addition, the potential cost savings that occur after quitting drinking at the ages of 35, 45, 55 and 65 by gender will also be evaluated in order to estimate the economic benefits of alcohol cessation.

Model validation

To ensure that our model provides appropriate estimates, face validity will be conducted through expert meeting to ensure the suitability of the model structure, assumptions and parameters. External validation will also be undertaken. The annual transitional probabilities of death for the six alcohol-related diseases as well as death from the other causes will be further adjusted to ensure that the life expectancy of the lifetime abstainer estimated from the model is in accordance with that of the general Thai population, derived from the WHO’s life table.²⁰

Sensitivity analysis

Both one-way and probabilistic sensitivity analyses will be performed to examine the influences of uncertainty surrounding the model parameters and assumptions on the primary outcomes. One-way sensitivity analysis will be performed by varying the input parameters, as follows: the average alcohol daily intake in Thai drinkers from the WHO (45.78 g/day in males and 16.56 g/day in females)²⁶; varying the relative risks based on their upper and lower limits of CIs identified in the literature. Also, we will set all of the non-significant risks at 1.0, assume that 50% of the binge drinkers will drive after drinking, use 0% and 6% discount rates, and use income growth rates of 1% and 5%.²⁵ Probabilistic sensitivity analyses will also be conducted by assigning probability distributions to the model parameters and running Monte Carlo simulations to evaluate the overall uncertainty. The analysis will run with 1000 iterations, and the 95% CIs of the lifetime costs will be computed.

Patient and public involvement

Patients and/or the public were not involved in the design, conduction, reporting or dissemination plans of this protocol.

Start and end dates for the study

1 February 2023–30 April 2024.

Discussion

The protocol described the incidence-based COI method to estimate the lifetime costs of alcohol consumption of an individual in Thailand. The Markov model enables the projection of health states of the alcohol-related conditions and the associated costs over a lifetime period, providing an understanding of the long-term economic impacts of alcohol consumption. Our protocol also facilitates the exploration of various scenarios (ie, lifetime drinking, quitting at specific ages), offering insights into the potential cost-effectiveness of different strategies or policies aimed at reducing alcohol consumption and mitigating alcohol-related harm.

Nevertheless, it is important to acknowledge that this protocol has certain limitations. First, the accuracy of the estimates relies on the quality of the input parameters. Given the lack of local epidemiological parameters, efforts to select appropriate parameters for studies in other settings are warranted. However, sensitivity analyses will be conducted to assess the effect of each input parameter on the findings. Second, due to the absence of the epidemiological data and to keep the model parsimonious, only six major alcohol-related diseases were included. Given that alcohol use contributes to more than 200 diseases,²⁷ this protocol might underestimate the economic burden incurred by the individual drinker. In addition, it should be noted that our model will only predict the costs incurred by drinker who starts drinking at certain age (ie, 20 years old in male and 23 years old in female) at a constant amount (ie, 30.1 g in males and 8.0 g in females) throughout his/her lifetime or for certain periods (ie, quit at age 35, 45, 55 and 65), which is an average age at start drinking and the average amount of drinking among Thai population. While the costs estimated from our study may not be directly applied to the drinker who starts drinking or quitting at different age or drinking at different amount/pattern, they could provide an approximate estimate, which could be applied with some assumptions to assessment of the impact of alcohol policy. In addition, it could serve as the useful example for conducting such analysis for drinker with different characteristics. Lastly, to balance between the natural disease’s progression, the availability of data on the relevant transitional probabilities and the modelling technique, many assumptions were inevitably required. Nevertheless, the model validation will be performed to ensure that our model provides acceptable estimates. Given the demand on the estimates and the scarcity of the studies that estimate lifetime cost of drinking, we also hope that this protocol will provide a lesson learnt for the researcher who plans to estimate the lifetime cost of an individual drinker in other settings.

Ethics and dissemination

The ethical approval for this study has been obtained from the Institutional Review Board of Mahidol University, specifically the Faculty of Dentistry and Faculty of Pharmacy (COE.No.MU-DT/PY-IRB 2021/010.0605). As per their assessment, no additional ethical approval is required for this research. The findings of the study will be disseminated via conferences, publication in a peer-reviewed academic journal and engagement with policy-makers and public health stakeholders.

The authors would like to thank Ms. Astrid Otto for conducting a comprehensive language review on this manuscript.

Ethics statements

Patient consent for publication

Not applicable.

Footnote

Contributors CL contributed to the study conception and design, data acquisition, data analysis, developed the first draft of manuscript and editing the manuscript. MT, BS and JR contributed to the study conception and design, data acquisition, data selection, data analysis, and reviewing and editing the manuscript. UC and OP contributed to the study conception and design, data analysis, and reviewing and editing the manuscript.

Funding This study is a part of PhD study at social, economic and administrative pharmacy graduate programme of CL which scholarship provided by the RGJ-Ph.D Programme under the National Research Council of Thailand, Ministry of Higher Education, Science, Research and Innovation (Grant No. PHD/0061/2561). The grantor of scholarship has no role in the design of the study; in the collection, analysis or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Competing interests None declared.

Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

References

¹ Parry CD, Patra J, Rehm J. Alcohol consumption and non-communicable diseases: epidemiology and policy implications. Addiction 2011; 106: 1718–24. doi:10.1111/j.1360-0443.2011.03605.x

² Rehm J, Gmel GE, Gmel G, et al. The relationship between different dimensions of alcohol use and the burden of disease—an update. Addiction 2017; 112: 968–1001. doi:10.1111/add.13757

³ Alpert HR, Slater ME, Yoon YH, et al. Alcohol consumption and 15 causes of fatal injuries: a systematic review and meta-analysis. Am J Prev Med 2022; 63: 286–300. doi:10.1016/j.amepre.2022.03.025

⁴ GBD 2016 Alcohol Collaborators. Alcohol use and burden for 195 countries and territories, 1990-2016: a systematic analysis for the global burden of disease study 2016. Lancet 2018; 392: 1015–35. doi:10.1016/S0140-6736(18)31310-2

⁵ Manthey J, Hassan SA, Carr S, et al. What are the economic costs to society attributable to alcohol use? A systematic review and modelling study. Pharmacoeconomics 2021; 39: 809–22. doi:10.1007/s40273-021-01031-8

⁶ Thavorncharoensap M, Teerawattananon Y, Yothasamut J, et al. The economic impact of alcohol consumption: a systematic review. Subst Abuse Treat Prev Policy 2009; 4: 20. doi:10.1186/1747-597X-4-20

⁷ Rice DP, Hodgson TA, Kopstein AN. The economic costs of illness: a replication and update. Health Care Financ Rev 1985; 7: 61–80.

⁸ Tarricone R. Cost-of-illness analysis. What room in health economics Health Policy 2006; 77: 51–63. doi:10.1016/j.healthpol.2005.07.016

⁹ Jo C. Cost-of-illness studies: concepts, scopes, and methods. Clin Mol Hepatol 2014; 20: 327–37. doi:10.3350/cmh.2014.20.4.327

¹⁰ Hodgson TA. Annual costs of illness versus lifetime costs of illness and implications of structural change. Drug Information Journal 1988; 22: 323–41. doi:10.1177/009286158802200306

¹¹ Komorowski M, Raffa J, editor. Markov models and cost effectiveness analysis: applications in medical research. In: Secondary analysis of electronic health records. Cham (CH): Springer International Publishing, 2016: 351–67.

¹² Barbosa C, Taylor B, Godfrey C, et al. Modelling lifetime QALYs and health care costs from different drinking patterns over time: a Markov model. Int J Methods Psychiatr Res 2010; 19: 97–109. doi:10.1002/mpr.306

¹³ National Statistical Office. The 2021 heath behavior of population survey. Bangkok: Statistical Forecasting Division, 2021.

¹⁴ Thailand Bod. Burden of disease attributable to risk factors in Thailand. Nonthaburi: Burden of disease, International health policy program (Thailand), 2019.

¹⁵ Thavorncharoensap M, Teerawattananon Y, Yothasamut J, et al. The economic costs of alcohol consumption in Thailand, 2006. BMC Public Health 2010; 10: 323. doi:10.1186/1471-2458-10-323

¹⁶ Luangsinsiri C, Youngkong S, Chaikledkaew U, et al. Economic costs of alcohol consumption in Thailand, 2021. Glob Health Res Policy 2023; 8: 51. doi:10.1186/s41256-023-00335-w

¹⁷ Health lntervention and Technology Assessment Program. The Evaluation of the outcomes of health promotion performance: development of the targets and indicators necessary for assessing the health promotion performance of the Thai Health Promotion Foundation using input from an incidence-based cost-of-illness analysis. Nonthaburi: The Graphico Systems, 2011.

¹⁸ Aekplakorn W, ed. The 2019-2020 Thai National health examination survey VI. Bangkok: Aksorn Graphic and Design Publishing, 2021.

¹⁹ International Health Policy Program Foundation. Thailand burden of diseases attributable to risk factors 2014. Nonthaburi: International Health Policy Program Foundation, 2018.

²⁰ World Health Organization. Life table by country: Thailand. Global Health Observatory data repository. 2020. Available: https://apps.who.int/gho/data/view.main.61640?lang=en [Accessed 20 May 2023 ].

²¹ Riewpaiboon A. Standard cost lists for health economic evaluation in Thailand. J Med Assoc Thai 2014; 97 Suppl 5: S127–34.

²² The Royal Thai Police. Statistics of road traffic accident casualty and property damaged. 2022. Available: https://ittdashboard.nso.go.th/preview.php?id_project=64 [Accessed 16 May 2023 ].

²³ Office of the national economic and social development council. Report on economic conditions in Thailand (Q1/2023). 2023. Available: https://www.nesdc.go.th/more_news.php?cid=871&filename=index [Accessed 20 May 2023 ].

²⁴ The World Bank. GDP per capita growth (annual %)-Thailand. 2022. Available: https://data.worldbank.org/indicator/NY.GDP.PCAP.KD.ZG?locations=TH [Accessed 22 May 2023 ].

²⁵ Guideline Development Working Group. Guideline for health technology assessment in Thailand updated edition: 2019. Bangkok: Health Systems Research Institute, 2021.

²⁶ World Health Organization. Alcohol, levels of consumption. 2022. Available: https://www.who.int/data/gho/data/themes/topics/indicator-groups/indicator-group-details/GHO/levels-of-consumption [Accessed 9 Jun 2023 ].

²⁷ World Health Organization. Alcohol. 2022. Available: https://www.who.int/news-room/fact-sheets/detail/alcohol [Accessed 22 Jul 2023 ].