Liang Ni

Overview: Led statistical programming and analysis using extensive EHR/claims data across 10+ studies on cardiovascular diseases, diabetes, cancers, vaccine safety, and health utilization costs, sponsored by leading biotech firms like AstraZeneca, and internal studies monitoring hospital performance, delivering innovative insights.

• Applied analytic and programming expertise in drafting statistical analysis plans (SAPs), collaborating with cross-functional teams, including managers, investigators, and physicians located globally.
• Created SQL scripts for extracting and manipulating large-scale data from varied sources, including Kaiser Permanente's EHR (Oracle database), registries, and medical claims, providing analysis-ready datasets for projects.
• Defined key study variables, such as outcomes and medical history, using ICD/CPT codes from the utilization/encounter table, measurements from the lab and vital sign table, and treatment group using GPI/NDC codes from pharmacy and claim data.
• Developed SAS macros and programs on the Unix SAS server to create study datasets and validate analysis output, including tables, listings, and figures (TLF), according to the SAP.
• Visualized study findings by generating customized figures using R and SAS GRAPH, improving the quality and efficiency of result illustration, thus increasing the publication acceptance chance.
• Conducted survival analysis using propensity score analyses and Cox proportional hazard models to compare cardiovascular disease incidence and mortality rates among anti-hypertensive medication users, generating insights for improving medication safety.
• Collaborated with study teams from multiple sites, such as Harvard Medical Center, to assess the association of residential proximity to, and emissions from, dioxin-emitting facilities with the risk of lymphoma cancer endpoints, and the pattern across different subtypes.
• Performed multivariate logistic regression modeling to predict the BP outcomes among diabetes patients, assessing the effectiveness of GLP-1 as a potential hypertension treatment.
• Calculated Kaiser members' annual healthcare costs by merging healthcare data and national cost estimates, revealing the economic impact of different hypertension therapies.

• Automated data cleaning, data processing, statistical analysis, and result presentation workflows were designed and developed using 10+ self-defined R functions, reducing the time to report for 10+ clinical studies.
• Applied expertise in R programming and shell language to streamline bioinformatics analysis on Linux, with data from fMRI and patient visits, improving efficiency and providing insightful recommendations on the automated pipelines.
• I wrote the SAP and statistical analysis sections for more than five published clinical trials, focusing on motor and cognitive interventions for Parkinson's and Alzheimer's, ensuring data integrity and quality control.
• Managed 6 REDCap databases, integrating clinical data from multiple sources, and ensuring high standards of data quality.