The BBSR biostatisticians collaborate on developing protocols for clinical trials and research proposals for laboratory-based investigations and epidemiologic studies; perform statistical analyses; interpret study results; and co-author presentations and publications. In addition to developing investigator-initiated clinical trials and new research studies, BBSR biostatisticians serve as standing voting members on the Medical Protocol Review and Monitoring Committee (MPRMC), Population Sciences Protocol Review and Monitoring Committee (PSPRMC), and the Data and Safety Monitoring Committee (DSMC). BBSR biostatisticians routinely participate in various Cancer Center forums, such as meetings of the research programs and the Site Disease Groups, and often assist Clinical Research Services (CRS) and the Informatics Applications Management group in the implementation of various projects.
- Write-up of statistical and bioinformatics considerations sections for grants and protocols
- Data analysis of clinical trials, community-based participatory research studies, population studies, and laboratory experiments
- Write-up of statistical/bioinformatics methods and collaboration on results for manuscripts, abstracts, presentations, and reports to oversight committees and funding agencies
- Support for data management and data sharing
- Participation in SCCC educational activities through seminars, lectures, or journal clubs on study design, data collection, and data science methods for clinical trial, community-based participatory research studies, population-based studies, and laboratory studies.
- Statistical design of clinical trials, community-based participatory research studies, population studies, and laboratory experiments
- Sample size determination / justification
- Statistical analysis plan
- Consultation for database design and data management
- Early stopping guidelines for data and safety monitoring of clinical trials
- Statistical analysis plan for interim and final results for clinical trials
- Assistance with clinical trial protocol amendments
- Data analysis of population-based cancer registry data from existing databases such as SEER, NCDB, FCDS
- Data analysis of population-based data for cancer risk assessment and health disparities
- Data analysis and interpretation of findings throughout the project
- Development of prediction models: feature selection and model validation using machine learning methods; random forest; penalized LASSO regression model; support vector machine; bootstrap validation; receiver operating characteristics (ROC) plot; time-dependent ROC plot; calibration plot; decision curve plot
- Kaplan Meier analysis, Cox proportional hazards regression model, and competing risks regression model
- Regression models such as linear, survival, logistic, Poisson for fixed, random, and mixed effects models for cross-sectional, longitudinal, and repeated measures studies
- Tumor growth modelling for pre-clinical studies (mouse, pig, etc.): IC50/ED50 analysis; tumor growth curve comparison based on adjusted area under the tumor growth curves; dose-response analysis; combination index plot for drug-drug interaction
- Meta-analysis, meta-regression, multivariate meta-analysis, longitudinal meta-analysis, and network meta-analysis: PRISMA diagram; funnel plot; forest plot; bubble plot; network diagram; surface under the cumulative ranking (SUCRA) curve
- Provide randomization schema for cancer clinical trials and cancer center sponsored studies
- Statistical report including publication quality tables and figures summarizing results of data analysis
- Serving on institutional scientific committees as Biostatistician reviewers for the Protocol Review & Monitoring Committee (PRMC) and for the Data Safety Monitoring Committee (DSMC).
- Statistical software (commercial): SAS, SPSS, PASS (Power Analysis and Sample Size for Windows), NCSS (Number Cruncher Statistical System), GESS (Gene Expression Statistical System), MS Excel developed modules: BAYES-R, BAYES- S, BERT and SET.
- Database software: MS Access and REDCap (http://project-redcap.org/).
- Public domain software: R, SEER*Stat, EWOC (Bayesian dose escalation / de-escalation), PH1ATD (analysis of Phase I trials with accelerated titration designs), OTSD (optimal two-stage designs for Phase II clinical trials), STPLAN (study planning calculations), R code for Bayesian phase II adaptive randomization.
To request support from the BBSR, complete the Investigator Request for Statistical and Bioinformatics Request Form and submit by clicking the “Submit” button on the form or email the form to firstname.lastname@example.org.