This cross-sectional descriptive study was conducted at the MFD Clinic, IUDC, Bangkok—Thailand’s first government-operated MFD assessment centre. Two service modalities were employed: 1) clinic-based assessment, in which drivers self-presented; and 2) proactive workplace outreach, in which assessment teams deployed to transportation depots, intercity bus terminals, and taxi cooperatives. The outreach modality was adopted following observation of insufficient uptake under clinic-only provision, reflecting income-loss and scheduling barriers characteristic of this occupational group. Data were collected over the full operational lifespan of the programme (May 2020-March 2024), funded by the Road Safety Fund. The study is reported in accordance with the STROBE guidelines for cross-sectional studies [8].

Eligible participants held a current driving license for a public transport or for-hire vehicle issued by the Department of Land Transport, were aged ≥ 18 years, and provided written informed consent. No health-based exclusion criteria were applied, ensuring an unselected representation of the occupational driver population. A total of 2085 drivers with complete assessment records constituted the analytic sample; this exceeds the minimum of 1537 drivers calculated by the Krejcie and Morgan formula for a population of 10,000 (95% confidence; 2.5% precision). Drivers were classified into four vehicle categories: small for-hire vehicles (≤7 seats, including taxis), small public transport vehicles (songthaew), medium public transport vehicles (vans, ≤15 seats), and large public transport vehicles (buses and coaches, >15 seats).

2.3.1. Guideline Development

The MFD assessment guideline was developed by the DDC Injury Prevention Division in collaboration with nine Thai Royal Medical College specialty committees (neurology, cardiology, endocrinology, psychiatry, ophthalmology, otolaryngology, pulmonology, nephrology, and occupational medicine). It was adapted from the Austroads/National Transport Commission standards [4] and DVLA guidance [5], contextualised to reflect Thai diagnostic capacity, legal definitions, and operational feasibility for non-specialist clinicians. Eight disease domains were formalised: neurological disorders, cardiovascular disorders, diabetes mellitus, psychiatric disorders, substance misuse, visual disorders, renal and respiratory disorders, and miscellaneous conditions (including OSA, excessive daytime sleepiness, and medication effects).

2.3.2. Assessment Structure

The comprehensive assessment comprised six sequential parts. Part 1 captured driver and vehicle classification. Part 2 recorded sociodemographic data, driving history, self-reported chronic conditions, road crash history, and health behaviours. Part 3—the most extensive domain—covered anthropometry, blood pressure, complete blood count, urinary amphetamine screening, chest radiography, spirometry, comprehensive vision assessment (visual acuity, visual field, colour vision, dark adaptation, depth perception, and fundoscopy), audiometry, and upper-limb motor function, as well as structured OSA risk screening based on validated clinical history and anthropometric criteria (STOP-BANG criteria; snoring, felling tired, observed of OSA by other, high blood pressure, BMI ≥ 35 kg/m², Age ≥ 50, circumferential of neck 40 ≥ cm., Male gender, cut off score at 3 - 4 (moderate risk), 5 - 8 (high risk)). Part 4 comprised a standardised cognitive and psychomotor instrument assessing attention and reaction accuracy. Part 5 assessed safety attitudes and self-reported risk behaviours via a 20-item questionnaire. Part 6 administered the Thai-validated 9Q depression screening tool (scores 7 - 12 = mild; 13 - 18 = moderate; ≥19 = severe) and the 8Q suicide risk assessment instrument, with any positive response triggering immediate clinical referral.

2.3.3. Service Delivery Tiers

The operational model comprised three tiers to enable national scalability. Level A (Advanced) delivered the full protocol plus driving simulator assessment and polysomnography at a specialist centre. Level B (Basic) provided the standard clinic-based protocol excluding simulator and sleep study, operable by a traffic medicine physician and trained allied health officers. Level C (Common) offered a condensed outreach protocol deployable by a mobile team to transport workplaces, retaining history-taking, anthropometry, vision, hearing, haematology, drug screening, and mental health components while omitting radiography and spirometry.

Data were analysed using descriptive statistics. Categorical variables are expressed as frequencies and percentages; continuous variables as means (SD) or medians (IQR) according to distributional characteristics. Records with physiologically implausible values—identified by pre-specified range checks—were excluded from analysis of the affected variable, resulting in slightly varying denominators reported in each table.

The study was conducted in accordance with the Declaration of Helsinki. All data were de-identified prior to research use. Written informed consent was obtained from all participants as part of the standard MFD clinical protocol. Data were managed in accordance with Thailand’s Personal Data Protection Act B.E. 2562 (2019). The research was approved by IUDC EC 01/2024.

Of 2085 participants with complete vehicle data, the majority drove for-hire vehicles (n = 1253; 60.1%), with public transport vehicles accounting for the remainder (n = 832; 39.9%). Small vehicles (≤7 seats) predominated (n = 1475; 70.8%), followed by large vehicles (n = 492; 23.6%) and medium vehicles (n = 118; 5.7%). Most participants habitually drove short routes < 200 km (n = 1627; 78.1%) (Table 1).

Table 1. Driver and vehicle classification (n = 2085).

Note: One participant had missing vehicle classification data and was excluded from this table only.

The sample was predominantly male (n = 1830; 87.7%), with a mean age of 35.4 years (SD 14.1; median 31.5; range 18 - 84). Notably, 19.8% were aged over 60 years. Single marital status was most common (57.0%), followed by married or partnered (39.3%). Educational attainment was most frequently at the upper-secondary or vocational level (41.9%). Regarding driving experience, 46.2% had fewer than five years of commercial experience; 24.2% had more than 15 years. A self-reported chronic disease was declared by 19.4% (Table 2).

Table 2. Personal baseline characteristics (n = 2085).

**Percentages may not sum to 100% due to missing data in some sub-categories.

The most prevalent self-reported conditions relevant to driving safety were snoring (8.6%), cardiovascular disease (6.7%), and drowsiness or microsleep episodes while driving (5.8%)—a triad consistent with undiagnosed or untreated OSA. Diabetes mellitus was reported by 3.6% and asthma or chronic bronchitis by 2.4%. Critically, only 0.2% self-reported a psychiatric diagnosis—substantially lower than the screened prevalence identified in Section 3.6, indicating significant under-reporting of mental health conditions (Table 3).

Table 3. Self-reported prevalence of diseases and conditions relevant to driving safety (n = 2085).

Note: Self-reported psychiatric diagnosis rate (0.2%) contrasts markedly with the screened depression prevalence of 4.9% identified by the 9Q instrument, underscoring the inadequacy of self-declaration for mental health ascertainment in this population.

Nearly half of all participants were current smokers (46.1%), with a lifetime tobacco exposure rate of 54.0%. Habitual sleep of fewer than 7 hours per day was reported by 55.3%, and regular alcohol consumption (defined as ≥1 day/week) by 14.7%. Regular energy-drink use (≥4 times/month) was reported by 15.5%—a pattern consistent with a workforce managing chronic fatigue through stimulant behaviours. Almost half had experienced at least one road crash during their driving career (47.7%), of whom 17.2% had sustained injuries requiring hospitalisation (Table 4).

Table 4. Health behaviours and road crash history (n = 2086).

Note: Regular alcohol use defined as consumption on one or more days per week. Regular energy-drink use defined as four or more times per month. Sleep duration based on self-report of habitual daily sleep.

Mean BMI was 24.8 kg/m² (SD 4.8). Using Asian-adapted WHO criteria, 59.6% were classified as overweight or obese (BMI ≥ 23 kg/m²), including 41.0% with BMI ≥ 25 kg/m². Mean blood pressure was 124.3/69.9 mmHg (SD 19.1/13.9); elevated blood pressure (≥140/90 mmHg) was identified in 22.0% of participants (n = 439/2024 with complete BP data).

A positive OSA risk screen (loudly audible snoring by clinical history) was present in 44.5%. Composite OSA risk stratification classified 35.1% as moderate risk and 7.7% as high risk, with 57.2% at low risk. Abnormal vision findings were detected in 19.6% of participants, of whom 15.5% were already using visual aids. Abnormal complete blood count results represented the most prevalent laboratory finding (26.1%). Abnormal pulmonary function on spirometry was found in 2.9%, and abnormal chest radiograph findings in 2.6%. Hearing abnormalities were identified in 2.4%, and a positive urinary amphetamine screen in 0.8% (Table 5).

Table 5. Physical examination and functional capacity findings.

Note: Denominators vary slightly across variables due to missing or physiologically implausible values excluded before analysis; valid n is stated where it differs materially from the total sample. OSA = obstructive sleep apnoea; BP = blood pressure; BMI = body mass index. OSA composite risk score incorporates snoring, witnessed apnoeas, hypertension, BMI > 35 kg/m², age > 50 years, neck circumference > 40 cm, and male sex.

Of 2026 participants completing the emotion and concentration test, 98.8% passed and 1.2% were referred for specialist neuropsychological evaluation. The general emotional-state screen flagged 10.5% as positive. Among 2027 completing the 9Q depression instrument, 4.9% met criteria for depression of at least mild severity (mild 3.1%; moderate 0.8%; severe 1.0%)—contrasting markedly with the 0.2% self-reported psychiatric diagnosis rate. Among 2028 completing the 8Q suicide risk instrument, 2.0% demonstrated current suicide risk (low 0.9%; moderate 0.2%; severe 0.8%). All 8Q-positive participants received immediate confidential clinical referral (Table 6).

Table 6. Emotion, concentration and mental health screening results.

Note: 9Q = Thai-validated 9-item depression screening questionnaire (equivalent of PHQ-9). 8Q = Thai 8-item suicide risk assessment instrument. All participants screening positive on the 8Q were referred for immediate confidential clinical review. Denominators reflect the number completing each instrument; not all 2086 participants completed every component.

This study presents the largest systematic MFD dataset reported from Thailand and, to the authors’ knowledge, from Southeast Asia. Its central contribution is one of screening yield and burden estimation: systematic assessment detects, at scale, a convergent set of safety-relevant conditions that Thailand’s existing certification model leaves largely uncaptured. Every condition identified in this cohort is detectable through systematic assessment, treatable through existing clinical pathways, and regulatable through the commercial driver licensing framework. The question is not whether such conditions exist in the driver population—this study confirms that they do, at scale—but whether the regulatory system is designed to find and act upon them. Under Thailand’s existing certification model, it is not. This establishes the epidemiological rationale for MFD screening as a road safety intervention; it does not, and given the cross-sectional design cannot, demonstrate that screening itself reduces crashes. That effect is expected on the basis of the crash-risk associations and treatment evidence discussed below, but its confirmation requires prospective evaluation linking screening to subsequent crash outcomes.

The findings reveal a workforce carrying a convergent, multidimensional burden of conditions with direct mechanistic links to crash causation—conditions that remained undetected among drivers who had already satisfied every existing legal requirement for a commercial licence. This is not a failure of individual drivers; it is a structural failure of the regulatory framework designed to safeguard them and the travelling public. Reframing MFD assessment from a bureaucratic formality into a targeted public health intervention—with defined screening protocols, referral cascades, conditional licensing provisions, and real-time linkage to transport databases—positions it as a plausible crash prevention tool whose potential returns can be projected from existing evidence. Realising and measuring those returns, however, depends on downstream steps this dataset does not capture: effective treatment of screen-detected conditions, conditional licensing action, and longitudinal follow-up of crash outcomes.

The OSA risk screen positivity rate of 44.5% represents the single largest, most clinically actionable, and most economically consequential safety risk identified in this cohort. This figure substantially exceeds general-population OSA prevalence: contemporary meta-analytic data place the pooled prevalence of sleep-disordered breathing in the general adult population well above the low single-digit estimates historically cited [9][10], while driver-specific evidence shows a markedly higher burden still—pooled moderate-to-severe OSA prevalence reaches 37.3% in commercial drivers [11], with individual studies of Asian professional drivers reporting 28% - 78% [12]. That nearly half of an occupational driver population screened positive for OSA risk under a programme that did not previously exist underscores both the scale of the unmet clinical need and the detection value of systematic MFD assessment.

The public health significance derives from a well-characterised causal pathway. OSA produces intermittent nocturnal hypoxia and sleep fragmentation, generating excessive daytime sleepiness, impaired sustained attention, prolonged reaction times, and episodic microsleep—mechanisms directly implicated in crash causation. Meta-analytic evidence confirms that untreated OSA substantially elevates crash risk, with the mean crash-rate ratio among affected drivers estimated at approximately 1.2 to 4.9 relative to unaffected drivers [13]. The co-occurrence in this study of OSA screen positivity with self-reported microsleep while driving (5.8%), loud snoring (8.6%), and extreme obesity (BMI ≥ 35 kg/m² in 3.2%) defines a convergent, potentially synergistic risk profile of particular urgency in a workforce operating multi-passenger vehicles.

Critically, OSA is not only detectable—it is treatable. Continuous positive airway pressure (CPAP) therapy normalises crash risk among affected drivers to levels comparable to unaffected peers [14]. Despite this evidence, Thailand’s current commercial driver medical certificate does not reference sleep-disordered breathing [15]—in stark contrast to the current American Academy of Sleep Medicine guidance recommending systematic OSA screening for commercial motor vehicle operators [16], and EU Directive 2006/126/EC, which requires documented treatment for OSA-positive Group 2 licence holders [17]. The Road Safety Fund’s own economic modelling projects that systematic MFD screening—of which OSA detection is the highest-yield component—could prevent over 495 crash events annually, generating savings exceeding 1.1 billion baht per year [18]; these are modelled projections contingent on effective downstream treatment, not effects measured in the present cross-sectional sample. The case for OSA screening integration is therefore clinical, regulatory, and economic, and now warrants prospective confirmation.

Elevated blood pressure in 22.0% of participants, combined with overweight or obesity in 59.6%, defines a workforce carrying substantial population-level cardiovascular risk. In the occupational driving context, however, the primary concern is not chronic morbidity but the risk of acute incapacitating events—sudden cardiac death, myocardial infarction, hypertensive crisis, or stroke—occurring during vehicle operation. A single such event in a fully loaded intercity coach or urban bus carries mass-casualty potential categorically different from the same event in a private vehicle, making cardiovascular fitness assessment a public safety imperative rather than a routine clinical formality.

The obesity prevalence in this sample (59.6% with BMI ≥ 23 kg/m²; 41.0% with BMI ≥ 25 kg/m²) meaningfully exceeds the Thai national average of approximately 37.5% [19], consistent with systematic reviews documenting elevated metabolic risk in occupational drivers attributable to sedentary work patterns, irregular meal timing, and circadian disruption [12]. Thailand currently lacks any structured cardiovascular fitness protocol for commercial drivers—no mandatory electrocardiography thresholds, no ejection fraction criteria, and no post-event driving cessation requirements—in contrast to the detailed cardiac standards specified by Austroads [4] and the DVLA [5]. Each of these conditions is identifiable at screening, manageable through primary care, and regulatable through defined licence conditions. Inaction is itself a policy choice—and in the occupational driving context, one with measurable safety consequences for drivers, passengers, and the public alike.

Visual abnormalities in 19.6% of participants—nearly one in five—identify vision as the third most prevalent safety-relevant finding in this cohort. Given that vision accounts for approximately 90% of driving-relevant information processing [20], the detection rate from this multi-component assessment was more than three times that implied by existing certification records—demonstrating that routine licensing vision tests administered by the Department of Land Transport are insufficiently sensitive to identify the clinically relevant deficits that systematic ophthalmic assessment detects.

Critically, vision impairment is among the most readily correctable conditions in this profile. The 15.5% of participants already using visual aids confirms both that the clinical pathway exists and that correction is culturally acceptable among this occupational group. The regulatory solution is correspondingly straightforward: mandatory periodic ophthalmic review integrated within a standardised MFD assessment cycle, linked to conditional licensing provisions for drivers with correctable or progressive conditions—precisely the model implemented under Australia’s Austroads standard [4]. For Thailand’s national MFD policy framework, vision screening integration represents the highest-yield, lowest-barrier regulatory enhancement available.

The finding that 55.3% of participants reported habitual sleep of fewer than 7 hours per day—compounded by regular energy-drink consumption in 15.5%—reflects a pattern of chronic occupational fatigue constituting a simultaneous road safety hazard, occupational health failure, and regulatory gap. The neurophysiological consequences are well-quantified: sustained wakefulness of 17 - 19 hours produces cognitive impairment equivalent to a blood alcohol concentration of 0.05 - 0.10 g/dL [21], and driving after fewer than 5 hours of sleep carries crash risk comparable to operating at the legal alcohol limit [22]. Meta-analytic evidence confirms that sleepiness and fatigue significantly elevate crash risk among professional drivers (pooled OR 1.34; 95% CI 1.25 - 1.43) [23], and systematic reviews of fatigue countermeasure interventions indicate that structured screening and management programmes can reduce drowsiness-related crash risk, though effect sizes vary by intervention type and setting [24].

Current smoking in 46.1% of participants—well above the Thai national adult male smoking prevalence of roughly 35% [25]—further compounds this risk profile through promotion of OSA, nocturnal hypoxic arousal, and subjective sleepiness masking, creating a potentially synergistic interaction with the fatigue burden documented above. Thailand’s Road Transport Act B. E. 2522 [26] contains no provisions for mandatory minimum rest periods for for-hire and small public transport vehicle operators—the categories that predominate in this sample. This regulatory absence is not incidental; it is the direct legislative antecedent of the sleep insufficiency pattern documented here. Addressing it requires not only MFD screening but concurrent amendment of occupational driving-hours regulations across the full spectrum of commercial vehicle categories.

The detection of depression of any severity in 4.9% of participants and current suicide risk in 2.0%—against a self-reported psychiatric diagnosis rate of only 0.2%—represents the most policy-critical finding of this study. The 25-fold disparity between screened and self-declared prevalence reflects well-documented patterns of mental health under-reporting in occupational settings, particularly where psychiatric disclosure carries stigma and direct employment consequences [27]. Driver-specific evidence reinforces this concern: a recent systematic review of professional drivers reported depression prevalence ranging from 14.3% to 60.5%, far exceeding the 4% - 8% typically observed in corresponding general populations [28]. More fundamentally, it exposes the inadequacy of self-declaration as a mental health surveillance tool in the commercial driver certification context: a system that relies on voluntary disclosure to identify depression will, by design, fail to find it.

The road safety implications are severe. Depression impairs sustained attention, psychomotor reaction times, and risk perception, and is associated with an 8.75-fold increase in crash risk—the highest disease-specific crash risk multiplier reported in the epidemiological literature [3]. Extrapolating the 2.0% current suicide risk prevalence to Thailand’s estimated 400,000 licensed commercial vehicle drivers implies approximately 8000 occupational drivers experiencing active suicidal ideation while operating passenger vehicles on any given day. This figure reframes mental health screening not as a clinical add-on to MFD assessment, but as simultaneously a road safety imperative and a humanitarian obligation.

The barriers to implementation are institutional, not technical. Validated, culturally adapted instruments already exist: the Thai 9Q depression screening tool and the 8Q suicide risk assessment instrument require fewer than five minutes to administer and minimal staff training. Confidential referral pathways—designed to protect drivers’ employment status while connecting them to clinical support—actively reduce disclosure barriers. What is absent is not the tool, but the regulatory mandate to use it. Integration of validated mental health screening into all MFD protocols, with defined confidential referral cascades, represents one of the highest-yield and lowest-cost enhancements available to Thailand’s commercial driver certification system.

Beyond its epidemiological contribution, this study documents the operational feasibility of MFD programme implementation within a public health infrastructure characteristic of an upper-middle-income transitional economy. The programme’s architecture—nationally contextualised clinical guidelines, inter-ministerial governance, tiered service delivery, and structured workforce training—aligns with the WHO Safe System Approach’s emphasis on shared institutional responsibility and scalable design [7]. Four lessons are directly generalisable to LMIC health systems seeking a precedent for occupational driver health screening.

Formal inter-ministerial governance is non-negotiable. The Memorandum of Understanding between the Ministry of Public Health and the Ministry of Transport created shared accountability, data-sharing legitimacy, and cross-sector recruitment channels that clinic-housed programmes cannot replicate. The critical outstanding gap—real-time data linkage between the MFD health record system and the Department of Land Transport’s licence database—requires sustained political commitment and must be resolved before clinical findings can generate enforceable licensing decisions.

Guideline contextualisation is a prerequisite, not an option. Direct adoption of high-income country standards in LMIC settings consistently produces frameworks that are technically sound but operationally unimplementable. The engagement of nine Thai Royal Medical College specialty committees to adapt international criteria to Thailand’s diagnostic capacity and legal context represents a critical and transferable process step rarely documented in the published MFD literature from LMICs.

Proactive workplace outreach is the primary delivery model. Clinic-only provision generated insufficient uptake, reflecting structural barriers—income loss, shift incompatibility, and geographic distance—that characterise this occupational group. Deployment of Level C mobile assessment teams to bus depots, taxi cooperatives, and transport terminals should be the default initial strategy, with clinic-based services reserved for confirmatory investigation and complex case management.

Workforce investment is the rate-limiting factor for national replication. Structured training across traffic medicine physicians and allied health professionals—building on the Thai Traffic Medicine Association’s curriculum framework—created the human capital prerequisite for both programme operation and future scale-up, offering a model for LMICs seeking to develop MFD capacity without establishing standalone specialist infrastructure.

The findings of this study, interpreted against Thailand’s existing regulatory framework, define three tiers of policy action—regulatory, health system, and data infrastructure—each with distinct responsible agencies, implementation timelines, and measurable outcomes, presented systematically in Table 7.

At the regulatory level, legal formalisation is the foundational prerequisite. MFD findings cannot be linked to licensing decisions until the eight disease-domain criteria are enacted through the Medical Council of Thailand and promulgated in the Royal Gazette. Parallel amendment of the lifetime licence exemption under the Motor Vehicle Act B. E. 2546 and application of the Occupational Disease and Work Environment Control Act B.E. 2562 [29][30] provide complementary legislative pathways.

Table 7. Tiered policy recommendations for integrating medical fitness-to-drive assessment into Thailand’s road safety framework.

MFD = medical fitness to drive; DDC = Department of Disease Control; DLT = Department of Land Transport; MOPH = Ministry of Public Health; MOT = Ministry of Transport; RSF = Road Safety Fund. Timeline: Short = 1 - 2 years; Medium = 3 - 5 years; Long = >5 years. (RSF = Road Safety Fund, CPAP = continuous positive airway pressure).

At the health system level, the tiered service model is ready for national scale-up through existing Regional Disease Control Office infrastructure without requiring new facility construction. The highest-priority clinical enhancements are integration of structured OSA screening cascades and confidential mental health referral pathways into all MFD protocols, given the prevalence and crash-risk magnitude of these conditions documented in this cohort.

At the data and surveillance level, real-time bidirectional linkage between the Ministry of Public Health’s MFD health record system and the Department of Land Transport’s licence database is the infrastructure prerequisite for converting clinical findings into enforceable licensing decisions. A national occupational driver health registry, linked prospectively to crash records, would enable rigorous evaluation of programme impact on road traffic injury rates—the definitive outcome measure for a programme framed as a road safety intervention.

This study’s principal strengths include the largest systematic MFD sample reported from any LMIC setting (n = 2086); a multi-domain assessment protocol simultaneously evaluating physical, sensory, laboratory, functional, and mental health dimensions across six structured parts; the use of validated Thai-language instruments for depression (9Q) and suicide risk (8Q), ensuring cultural and linguistic appropriateness; a dual-modality service delivery design—combining clinic-based and workplace outreach assessment—that reduces the attendance bias limiting the generalisability of clinic-only studies; and the generation of an operationally documented, tiered service model directly informing national MFD policy development.

Limitations follow principally from the study design. The cross-sectional design supports estimation of screening yield and of the burden of safety-relevant conditions, but precludes causal inference: it can neither establish that these conditions caused the crashes recorded in participants’ histories nor demonstrate that MFD screening reduces future crashes. The reported crash history (47.7%) is retrospective and self-reported and therefore indexes co-occurring burden rather than any preventive effect of the programme; demonstrating crash reduction would require a prospective design linking screening, treatment, and subsequent crash outcomes. Further limitations include restriction to the Bangkok Metropolitan Region, limiting generalisability to provincial, rural, and border-area driver populations where health profiles and occupational exposures may differ; volunteer participation introducing potential healthy-worker and self-selection biases that may underestimate the true burden of safety-relevant conditions; reliance on self-report for behavioural and health history data, subject to recall and social-desirability bias—most consequentially for mental health and substance use; absence of confirmatory diagnoses for screen-positive OSA and mental health findings, meaning that prevalence figures represent screen-positive rather than clinically confirmed burdens; absence of a general-population comparison group, precluding direct quantification of excess risk attributable to occupational driving; and the absence of systematic medication data, a meaningful gap given the 19.4% chronic disease prevalence implying substantial pharmacological use with potential implications for driving fitness in this cohort.

Notwithstanding these limitations, the study’s large sample, comprehensive multi-domain protocol, dual-modality design, and policy-oriented analytical framework collectively support the robustness of the principal findings and their relevance to MFD programme development in Thailand and comparable LMIC settings.