Predictors of post-pandemic COVID-19 preventive behaviors among Hmong Communities in Rural Phayao, Thailand: A PRECEDE–PROCEED study

Vollständiger Abstract

Worum geht es in dieser Arbeit?

This study examined predictors of COVID-19 preventive behaviors among Hmong communities in rural Phayao, Thailand, using the PRECEDE–PROCEED framework. A cross-sectional survey was conducted between November 2022 and March 2023 among adults residing in three Hmong villages. Stratified simple random sampling was applied to household lists, yielding a final sample of 402 respondents. Data were collected using a structured questionnaire comprising sociodemographic variables, five theory-based determinants, and a multi-item preventive behavior scale. The instrument demonstrated strong content validity (item–objective congruence index, IOC = 0.97) and acceptable internal consistency (KR-20 = 0.82 for knowledge; Cronbach’s α = 0.77–0.80 for attitudinal and behavioral scales). A pre-specified hierarchical multiple regression model was fitted. Block 1 (covariates) included gender, age, two education dummies (secondary and post-secondary; reference = ≤primary/no schooling), and monthly household expenditure (per 1,000 THB, approximately 29 USD). After an initial screening, age and secondary education were excluded, and gender, post-secondary education, and expenditure were retained in Block 1. Block 2 contained five determinants (knowledge, attitudes, perceived policy, perceived access to health resources, and exposure to health advice/support) and was estimated using Stepwise selection with Block 1 fixed. The final model explained 30.4% of the variance in COVID-19 preventive behaviors (R² = 0.304, adjusted R² = 0.293, ΔR² = 0.156; p < 0.001). Attitudes, perceived access, exposure to health advice/support, and post-secondary education emerged as significant positive predictors, whereas female gender and higher household expenditure were negatively associated with preventive behaviors; age and secondary education remained statistically non-significant after adjustment.