Statistics Collaborative - Design and analysis for biomedical research

About UsAbout Us

Collaborating with SCI
Common questions and answers

If my company wants to work with SCI, how do I get started?

If you wish to speak to a specific statistician at SCI, just send an email or call 202-247-9700 and ask for that person. If you don't know any of us, please email or call 202-247-9700.

How are project teams organized at SCI?

Each project at SCI has a Project Manager-usually a senior-level statistician-who oversees the day-to-day activities of the project. SCI believes stable project teams are important for maintaining consistently high quality work throughout the lifetime of the project. Therefore, we make every effort to keep the same Project Manager involved for the life of the project. To the extent feasible, we try to keep the teams constant as well.

Project teams have at least one statistician and a project assistant responsible for the administrative needs of the project. Large teams may also have additional statisticians, programmers, and administrative staff. For some very large projects, we designate a second Project Manager as a back-up to the main Project Manager. Regardless of the size of the project team, SCI's communication procedures ensure every project team member has the information necessary to do work efficiently.

Who would be my point of contact at SCI?

If your project is small, your point of contact will be the primary statistical consultant. For projects that involve a team from SCI, the Project Manager will generally be your point of contact. Our senior-level statisticians travel frequently, so if you have an urgent need for statistical advice, other staff can usually step in.

Is it really worth auditing my trial's randomization process?

Absolutely! An accurate randomization process is essential to maintaining the integrity of a trial's results. Validation should begin before the trial commences and continue until patients are no longer being randomized into the trial or receiving study drug. We have uncovered many different randomization and drug resupply errors during our reviews.

Some examples of what we have seen follow:

  • Permuted-block lists that were not generated at random.
  • Systems that inadvertently assigned all patients both active and control medication kits.
  • Dynamic algorithms that were inconsistent with the trial's randomization specification document.
  • Patients who were randomized by an investigator more than once in an effort to correct a misstratification. This negatively impacted subsequent randomization assignments in the trial.

Reviewing randomization processes on an ongoing basis will address and correct problems quickly.

How early in my study should I contact SCI for statistical support?

SCI works with our clients throughout the life cycle of study development. Starting from the initial stages of defining endpoints, writing protocols, calculating sample sizes, designing case report forms, and continuing on to the later phases of analysis, interpretation, and regulatory submission, our statisticians provide an experienced perspective on the entire process. We are most effective collaborators when we begin working with the study team early in the process, but we also have considerable experience joining project teams during an ongoing study or after an unsuccessful regulatory submission.

If I am setting up a DMC, when should the committee first meet?

We recommend forming the committee and holding an organizational meeting to establish operating procedures prior to enrolling the first participant.

Has SCI ever provided a second opinion on the analyses someone else has produced?

SCI often reviews analyses performed by others. We may simply review the methodology to see if it is appropriate for the question asked and the data collected. We may suggest other methods of analysis that take into consideration the disease under study, the conduct of the trial, or issues with the data. In many cases, clients provide us with the data and we perform our own analyses and write a summary report.

Will SCI write the statistical analysis plan for my study that another group would implement?

We believe it is important for the statisticians involved in the final statistical analysis of a trial to be the primary statistical authors of the analysis plan. To ensure the accuracy and interpretability of the results, those involved in the analyses should be involved in preparing the plan. If, however, the company or contracted organization doing the final analysis would like SCI to provide an additional statistical review of the analysis plan, SCI is able to provide that support.

Do statisticians from SCI participate in organizational and investigator meetings?

We encourage our clients to include SCI at their trials' organizational and investigator meetings. SCI prefers to join projects as early as possible so that we can provide input into the protocol, analysis plan, and charter. DMCs typically express their opinions about their review process at organizational meetings; therefore we recommend that the statisticians from SCI who will be presenting to the DMC participate in these discussions.