Services

MGH Biostatistics provides support to investigators in the design and analysis of research studies, as well as long-term scientific collaboration and data coordination for large, multi-center research programs.

Use the button below to initiate a request for a consultation with MGH Biostatistics. The information will be reviewed and a member of our team will contact you. Please note: These consultation services are typically available to investigators at MGH only. 

  • Study Development and Design
  • Data Collection and Coordination
  • Analysis

We collaborate closely with physician scientists and investigators on rigorous design of clinical studies, including multi-site clinical trials and large observational cohort studies.  This work includes:

  • Grant application support
  • Statistical analysis plans (SAPs)
  • Sample size and power calculations
  • Preliminary data development including feasibility studies
  • Randomization plans
  • Study protocol development
  • Data safety and monitoring board (DSMB) plans
  • Manual of operations/procedures
  • Website and software development

We serve as the data coordinating center for several large multi-site clinical studies.

  • Electronic data capture
  • Integrated database construction and management
  • Case report form design
  • Implementation/validation of web-based randomization plans
  • Data queries, feature extraction, derivation of variables and data checks
  • Regular expression syntax for natural language processing
  • Data management/integration and pre-processing
  • Data security/de-identification
  • Screening, enrollment, adverse event (AE), severe adverse event (SAE) reports

We support analysis of data generated in clinical trials, observational studies and Electronic Health Record (EHR) and other real-world data-based studies.

  • Generalized linear and non-linear multivariable regression modeling
  • Correlated and clustered data methods
  • Longitudinal and repeated-measures data analysis
  • Survival analysis
  • Missing data methods
  • Machine learning methods
  • Supervised and unsupervised learning
  • Causal inference
  • Bayesian methods
  • Genetic epidemiology and -omics analysis
  • Statistical learning for mHealth
  • Methods for electronic health record (EHR) data
  • Novel statistical methodology development

If you are looking for our earlier technical reports or software or our decision tree on which statistical tests to use, you can visit our historical website: hedwig.mgh.harvard.edu/biostatistics