Knowledge, attitude and practice as predictors of healthcare providers’ social prescribing readiness for psychosocial health and wellbeing in older adults: A SEM-ANN modeling

Social prescribing refers to non-medical referral approaches intended to support psychosocial health. Interest in social prescribing has increased among healthcare providers, including general practitioners, nurse practitioners, psychologists, exercise practitioners, and rehabilitation therapists. This cross-sectional study examined whether Chinese healthcare providers’ knowledge, attitudes, and perceived effectiveness of social prescribing were associated with reported social prescribing practices related to older adults’ psychosocial health. This study was conducted in China using a stratified convenience sampling method. A self-administered questionnaire was used to collect primary data.A total of 5,600 healthcare providers were included in the final analyses. The collected data were analyzed using a two-stage data analysis technique, artificial intelligence neural networks, and a structural equation modeling (SEM-ANN) approach. All five prespecified structural paths were positive and statistically significant (standardized β = 0.13–0.50, all p < .001). For reported social prescribing practice, knowledge showed the largest direct association (β = 0.34), followed by perceived effectiveness (β = 0.16), and attitude (β = 0.13). In the artificial neural network (ANN) analysis, PESP had the highest normalized importance (100%), followed by HCPA (68.3%) and HCPK (68.1%).Among Chinese healthcare providers, perceived effectiveness, knowledge, and attitudes were positively associated with reported social prescribing practices. These findings are limited to provider-level associations based on cross-sectional self-report data and should not be interpreted as evidence of direct patient-level effects on thepsychosocial health of older adults.