Statistics Collaborative - Design and analysis for biomedical research


Statistical analysis plans (SAPs)

Statistics Collaborative, Inc. (SCI) has extensive experience reviewing and writing statistical analysis plans (SAPs) for a variety of trial types, analysis settings, and therapeutic areas. Our strength lies in our ability to synthesize information into a series of thoughtful displays that convey the relevant information to the intended audience. Rather than presenting a large volume of output, the reports resulting from our analysis plans are concise. Our carefully selected presentations and accompanying interpretation tell the full story of a trial's experience and results.

We tailor our SAPs to convey our thoughtful, appropriate, and comprehensive analysis of the data. Our plans include justification for trial design and outcomes, methods for analysis, and sensitivity analyses to account for different conventions, as well as other anticipated data issues. Analysis plans contain text explaining the pre-planned methods; some plans include table and figure mock-ups with defined conventions, populations, and exceptions. A SAP then serves as a complete set of instructions for those who eventually program the analyses to produce a report. The breadth of analysis plans that we prepare reflects the range of statistics and design with which we are familiar.

Examples of SCI's statistical analysis plans:

  • Final analyses of Phase 1, 2, and 3 trials of vaccines: SCI has developed SAPs for numerous clinical and field trials assessing candidate malaria vaccines. SCI has worked with sponsors and Principal Investigators to select appropriate statistical analyses for safety, immunogenicity, and efficacy outcomes, which may differ considerably depending on whether investigators conduct the trial in a clinical setting or in an endemic area with people who have prior exposure to the disease.
  • Integrated Summary of Efficacy (ISE): SCI has developed a SAP as well as performed the analyses for an ISE of trials assessing the efficacy of Kuvan in children with phenylketonuria (PKU). SCI worked with the sponsor to select appropriate statistical analysis for efficacy outcomes and contributed to the report.
  • Post-hoc analyses: For a trial of a putative cardioprotective agent for which SCI served as the independent unblinded statistician, SCI's analyses revealed an overwhelming benefit on the trial's composite endpoint of death or MI. This benefit was tempered by a clear excess in the incidence of a major adverse event in the active arms. To aid the DMC in its deliberations, SCI performed a series of post-hoc analyses in short succession in an attempt to identify a subset of subjects who were at the greatest risk of the adverse event, while still preserving the apparent benefit of the agent for others.