| Shubham Choudhary, Amit Navare |
This study analyses the CDC's Behavioral Risk Factor Surveillance System (BRFSS) data through a generational lens
with Random Forest classification, and provides very accurate predictions of mental health outcomes. Our new 0-100 risk
score approach predicted that the older generations reported more days of poor mental health, while younger generations
used telemedicine more and use drugs more. There were also differences among generations on some of the key
predictors including, but not limited to, substance use, overall health, financial hardship, and social support. The model
helps healthcare organizations design appropriate, data driven strategies for multi factor approaches, and generate
targeted interventions that combine risk factors, social determinants, and health behaviors. This is fundamentally
important since there is a void in the systematic generational analysis of mental health using population surveillance data
to inform the implementation of mental health enhancement initiatives, and this has real applications for generation
specific service delivery optimization, risk based resource allocations, and technology funding prioritization to better
manage mental health outcomes across age cohorts. |