Services

Mayo Clinic's Biostatistics Core collaborates with investigators across all stages of a research project, from concept and design, through data collection, to analysis and dissemination. Because early involvement improves a study's rigor and competitiveness the most, we encourage engaging the core during concept development — before data are collected.

Concept and protocol design

The best time to start collaborating with a biostatistician to discuss the concept and study design is before data are collected. Biostatisticians and data science analysts help investigators make decisions about concept and protocol design.

For example, the team may:

  • Formulate testable hypotheses or refine aims that can be addressed through data.
  • Identify study designs best suited to the aims and hypothesis and propose a study design.
  • Determine the number of subjects or samples needed to appropriately power the study to detect relevant effects or associations.
  • Develop and write a statistical analysis plan, identifying appropriate methodology.
  • Write statistical sections for protocols and collaborate on protocol development before submission to the Institutional Review Board or for committee review.
  • Write statistical sections of grant proposals and provide feedback throughout the writing process.

Database and data collection

Biostatisticians, data science analysts and statistical programmers are well suited to collaborate on database builds, data collection and monitoring of study data accrual. The team may:

  • Design and build electronic data capture for clinical trials or observational studies using REDCap, Medidata Rave, Mayo Clinic's scientific data management system or other systems.
  • Extract data from electronic health records while providing full compliance with privacy regulations, state law and Mayo Clinic policy. Staff have extensive knowledge of existing data sources, the data elements that the sources contain, and the methods used to extract or retrieve data.
  • Extract or obtain data from institutional databases, including the Rochester Epidemiology Project and other sources.
  • Integrate data from diverse sources into a combined dataset or database.
  • Maintain and monitor ongoing databases.
  • Develop data dictionaries.
  • Create statistical programs for data monitoring, such as issuing queries for missing data or flagging potentially erroneous data entry, or summarizing data for data safety and monitoring board meetings.
  • Clean and prepare data for analysis.

Analysis and write-up

The core team has the expertise to develop statistical analyses, write statistical code to analyze data and collaborate in dissemination of results. The team may:

  • Develop a statistical analysis plan and select appropriate statistical and advanced analytical methods based on the study aims or hypotheses.
  • Write and document code using SAS, R, Python or other statistical software.
  • Generate tables, figures and summaries of results for publication or presentation.
  • Develop predictive models and apply advanced analytics or machine learning techniques to address research or operational questions, including supporting deployment of models into practice.
  • Co-author manuscripts, including drafting a statistical methods section and critically revising the entire manuscript with special consideration for the interpretation of statistical results.
  • Respond to reviewers' comments on co-authored manuscripts.