Integrating public health data to understand mental health service patterns
This project examined trends in psychiatric emergency department visits across Texas, requiring integration of data from multiple sources including CDC datasets, American Community Survey data, and hospital administrative records.
I led the data extraction, cleaning, and merging process, then conducted longitudinal regression analyses to identify temporal patterns and demographic disparities in mental health emergency services.
Statistical tables and visualizations made the complex trends accessible to policymakers and stakeholders.
Analytical Methods & Policy Impact
The analysis employed multiple regression techniques including negative binomial regression and interrupted time series analysis to account for overdispersed count data and policy intervention effects.
I scraped and cleaned external datasets using Python, merged them with hospital data in Stata, and built comprehensive statistical models to control for demographic, geographic, and temporal factors.
The resulting visualizations and statistical tables clearly showed annual patterns in psychiatric ED utilization, identifying periods of increased demand and underserved populations. These findings informed mental health policy recommendations and resource allocation decisions at the state level.
