A survey of 233 Thai Muslim residents from eight communities in Nakhon Si Thammarat used PLS-SEM to test a community health-governance model for COVID-19 prevention. Primary healthcare services were associated with social support, community nutrition status, and public trust. Community nutrition had the strongest direct association with public compliance, which was in turn associated with trust. The model explained 41.5% of variance in compliance and 49.0% in trust. The findings support culturally responsive system design but, as cross-sectional evidence, do not establish causal direction.
Key findings
- The model met the authors' reliability and validity criteria, with SRMR=0.082 and Q² above zero for all endogenous constructs.
- Primary healthcare services were associated with social support (β=0.541, p
- Community nutrition had the strongest direct association with public compliance (β=0.509, p
- Public compliance was associated with trust (β=0.297, p
Why this matters globally
The model suggests that preparedness in culturally distinct communities should connect primary care with food security, transparent communication, religious leadership, health volunteers, and trust rather than focus on access alone. It can generate hypotheses elsewhere, but coefficients from one setting are not universal benchmarks and require longitudinal and program-evaluation evidence.
Thai researcher contribution
Sujira Wichaidit, Rachadaporn Jantasuwan, and Urai Jaraeprapal are affiliated with Walailak University's School of Nursing and Excellent Center of Community Health Promotion; Usa Nuampet is affiliated with the Faculty of Nursing, Prince of Songkla University. With Desmawati of Indonesia, the team generated direct evidence from eight Muslim communities in Nakhon Si Thammarat.
Limitations to consider
The cross-sectional design measures constructs at one point and cannot establish direction or show that changing services, nutrition, or support will improve compliance and trust. Self-report creates common-method and social-desirability risks. A sample of 233 people from eight communities in one province limits generalization to other Muslim or cultural settings. Q² and SRMR describe model performance, not external prediction or reduced infection, and the COVID-19 context may not transfer unchanged to future emergencies.