Improving Healthcare Utilization Models and Uncovering Health Behavior Patterns in Asian Americans Through Data Disaggregation

Although Asian Americans are made up of 20+ distinct ethnicities with unique health outcomes, previous healthcare research has treated them as a monolithic group, covering up many important disparities in health outcomes and healthcare utilization patterns. This study examines how disaggregating Asian American health data by ethnicity improves models analyzing healthcare utilization patterns, focusing on routine physicals, dental checkups, alternative medicine consultations, and preference for ethnic concordance. Using data from the 2015 Asian American Quality of Life survey (2,609 adults in Austin, Texas), binomial generalized linear models (GLMs) analyzed the impact of income, age, gender, insurance status, and ethnicity on healthcare utilization. Model comparison via Bayesian Information Criterion and analysis of deviance revealed that disaggregating ethnic subgroup data significantly improved predictions for routine physicals (ΔBIC = -3.38), dental checkups (ΔBIC = -28.53), and alternative medicine consultations (ΔBIC = -24.26). Results showed Vietnamese individuals were 57% more likely to get a physical check-up than Chinese counterparts (OR = 1.57), while Koreans were less likely to undergo physicals (OR = 0.60) or dental checkups (OR = 0.66). Filipinos also had lower odds of using alternative medicine (OR = 0.45). These results highlight key healthcare utilization differences between Asian American ethnicities and the importance of disaggregating Asian American health data to address disparities in healthcare access. Improved performance of statistical models can better inform targeted interventions and promote health equity for all Asian Americans. Future research could look at other regions with substantial Asian American populations and examine factors like acculturation.