Naiya Patel

Biography

I'm a well trained public health data scientist. My thesis (https://ir.library.louisville.edu/etd/4129/) focused on clinical and environmental factors affecting early stage lung cancer treatment receipt and survival outcomes using U.S. national cancer registry data, EPA, AHRF, and global weather data. My expertise could be categorized into quantitative: data mining, algorithm designing, data engineering, big data analytics, predictive modeling, clinical trial protocol development, and qualitative: systematic reviews, qualitative analysis. Data mining experience utilizing United States Medicaid claims data, U.S. SEER 18 cancer registry data, U.S. environmental and global weather data, and Area Health resource files. Algorithm design experience utilizing SQL, Python, STATA, and KNIME. Experienced in utilizing these software to design study sampling procedures, and create ICD 9 to ICD 10 crosswalk for assigning chronic disease conditions comorbidity scores index to develop broader disease categories for claims data analysis. Developed algorithm codes in SQL to retrieve specific patient records from big real world Medicaid claims data for project data analysis. Experience utilizing STATA and Microsoft Excel to build an algorithm to geocode air pollution and weather exposure and assign exposure values to cancer registry patients. Data engineering experience includes converting raw daily air pollution and weather exposure values into monthly and yearly average files by nearest distance monitoring method through algorithm designing. Predictive modeling experience includes utilizing associational regression models adjusting for time-invariant unobservable in data, and clustered standard errors to reduce estimation bias. Predictive modeling utilizing causal inference methods such instrumental variables regression, survival analysis, regression discontinuity, propensity score matching, fixed effects and random effects panel data methods. Experienced in making meaningful conclusions from the predictive modeling results and inform policy decision-makings. Non-quantitative experience include developing directed acyclic graphs (DAG) to illustrate comprehensive causal relationship in a study context to inform statistical modeling variables. Systematic literature review, meta-analysis, microeconomics, health policy, and health services research also fall in my non-quantitative skills.