Samuel Nwosu

ACTIV6 - Accelerating COVID-19 Therapeutic Intervention

• Meet with trial leadership to discuss methodology and analysis plan.
• Meet with trial statisticians from the Duke Clinical Research Institute (DCRI) to discuss, SAP, trial protocols, and variable creation and definitions.
• Manage data through the REDCap database platform using the API tool.
• Define logic to be used for trial metrics including baseline and outcome variables.
• Clean datasets by formatting and creating variables in preparation for study reports and statistical modeling.
• Create bi-weekly operational reports using R Markdown.
• Perform data validation department statisticians and IT team.
• Assisted with interim analyses for trial leadership and DSMB committee.

DSMBC - Data Safety Monitoring Board Coordination

• Attend weekly meetings with study investigators to discuss protocols and results.
• Met regularly with the collaborating study statisticians to discuss their protocols and variable definitions in congruence with th VUMC studies.
• Managed data across studies for univariate comparisons.
• Cleaned and formatted variables.
• Created R Markdown reports for the DSMBC committee.
• Fit multivariable regression models to assess COVID-19 symptom severity.

VICTAS - Vitamin C, Thiamine and Steroids in Sepsis

• Meet with investigators to discuss study variables and proposed analyses.
• Managed data through REDCap using the API tool.
• Create regular data reports of illogical data values for data team to review and query.
• Clean datasets by formatting and creating variables in preparation for study reports and statistical modeling.
• Created DSMB reports for the scheduled DSMB meetings.
• Create univariate summary tables and tested for unadjusted associations.
• Fit multivariable regression models to investigate covariate effects with primary outcome.
• Construct cumulative incidence plots, and summary model plots displaying the covariate and spline term effects.
• Presented overall trial results to trial leadership and team.

CONTACT - Standard Isolation vs. Targeted Isolation Group

• Attend meetings with investigators to discuss analyses and study design.
• Managed electronic medical record (EMR) data (i.e. STARPANEL and EPIC).
• Cleaned and merged admissions, demographic, ICU and infections data to create patient and visit level longitudinal analysis datasets.
• Create descriptive tables for demographic and main outcome variables.
• Fit generalize estimating equations (GEE).
• Generate statistical reports for study team.

ISCHEMIA - National Institute of Health

• Attend weekly conference calls to discuss trial protocol.
• Format and summarize variables according to trial definitions.
• Create summary tables and plots examining baseline measurements and follow up visit rates.
• Graphically monitor participant enrollment/randomization rates.
• Monitor protocol violations according to trial guidelines.
• Generate cumulative incidence plots comparing endpoint events by treatment groups.
• Construct DSMB reports for yearly DSMB meetings

VICS - Vanderbilt Center for Health Services Research

• Clean datasets by formatting and creating variables for statistical modeling.
• Create univariate summary tables testing for unadjusted associations.
• Perform multivariable regression modeling based on pre-specified hypotheses.
• Prepare statistical reports for collaborators explaining analysis results.

Workload - VA Hospital/GRECC

• Create and maintain multi-relational database for research assistants data input and collection.
• Cleaned datasets by formatting and creating variables in preparation for statistical modeling.
• Create summary statistical tables for group comparison.
• Constructed study plots assessing severity scores.
• Fit multivariable and mixed-effect models to investigate group differences.
• Create statistical reports to present to the data team at weekly meetings.

Chair Burnout Study - Vanderbilt School of Medicine

• Reformatted multiple datasets by creating variables, parsing out variable content, and modifying existing variables for stratification.
• Created descriptive statistics for variables to be analyzed in statistical models.
• Constructed summary plots investigating the effect of contributory factors leading to chair burnout.
• Fit univariate and multivariable regression models.