Biostatistics Consulting

Faculty and staff participating in consultation have a wide range of expertise, which is matched to the specific needs of the investigator. The Biostatistics Consulting Service is directed by Karen Steger-May.

Please communicate any deadlines as far in advance as possible. We strive to accommodate deadlines, but it generally takes weeks to months to conduct most analyses or to provide thorough grant preparation assistance. Contact the consulting service as early as possible with an awareness of the following guidelines:

• Please schedule a meeting with a consultant as early as possible in the research process. Most requests require at least two months’ notice prior to the requested final deadline. General recommendations include:

• Turnaround times for data analyses are contingent upon the data quality, format, amount, complexity, as well as the number and maturity of the study questions addressed.
• Ideally, please contact us several months prior to a grant deadline. If it is a first submission and we do not have adequate time, our contribution will demonstrate statistical involvement but could be incomplete. If it is a re-submission, we need adequate time to improve the score.
• As the deadline approaches, our ability to provide substantial support diminishes.
• Investigators are expected to work with us to help meet their deadlines. Timelines set for providing data, protocol drafts, other information needed by the consultant should be adhered to, or the timeline will be revised.
• Projects run most smoothly when there are frequent, clear and focused communications with a strong commitment from all persons involved.

The Biostatistics Consulting Service is a fee-for-service facility that assists WashU investigators and external, non-WashU investigators. The Biostatistics Consulting Service follows the guidelines set forth within Washington University Recharge Center Policy and Federal Uniform Guidance 2 CFR 200. The cost of the service can be charged to grants under the ‘Other Expense’ category.

The rates for WashU investigators and external, non-WashU investigators are indicated below.

*The rate assumes that the investigator’s institution and the paying entity are the same; rate is determined by paying entity.

^†The initial consultation meeting is intended to enable the consultant to assess the project’s needs and to provide a basic level of consultation to facilitate ongoing support. Initial consultation meetings do not include computations or analyses. Initial consultation meeting times in excess of one hour will be charged at the above rates.

^‡Up to four hours of free consultation is available when related to the preparation of a new grant that will be submitted for funding, provided that: (a) Center for Biostatistics and Data Science personnel will be included in the execution of the protocol, and (b) the submitted grant budget includes a percent effort for one or more named statisticians from the Center for Biostatistics and Data Science.

To schedule a consultation with the Biostatistics Consulting Service, please submit a request using the service request intake form.

The Biostatistics Consulting Service may be reached at biostatistics-consulting@wustl.edu.

The Biostatistics Consulting Service works with investigators throughout the life cycle of study development. Consultants can provide expertise starting from the initial stages of defining hypotheses and endpoints, writing protocols and grants, calculating sample sizes, designing data management systems, and continuing to the phases of analysis, interpretation and publication. We are most effective when we begin working with the study team early in the process, but also have experience joining project teams during an ongoing study or after data collection is complete.

The Consulting Service believes that stable project teams are important for maintaining high-quality work throughout the lifetime of a project. To the extent possible, we try to keep the same consultant(s) involved for the life of a project. With the goal to enhance science through biostatistics and education, our hope is that successful consultations will turn into long-term collaborative relationships where Center for Biostatistics and Data Science members are an integral part of your research team.

The initial consultation is intended to enable the consultant to assess the project’s needs and to provide a basic level of consultation to facilitate extensive ongoing support, if necessary. As consultant expertise varies, if a project requires particular skills, we will try to find an appropriate person to fulfill that need. Initial consultation meetings do not typically include computations or analyses.

At the initial consultation meeting, the investigator and consultant(s) should work together to:

• Provide a description of the project.
• Define particular skills required.
• Identify the materials needed by the consultant to perform the work (see guidelines below for materials needed for each type of assistance).
• Identify the hourly billing rate for the project.
• Estimated the investigator’s budget for the project, even if a best guess. This estimate will be used to determine a communication plan about the budget and effort used.
• Estimate timelines, deadline(s), and a communication plan for work needed and workflow (e.g., agendas, flow charts, checklists, shared meeting notes).
• Establish expectations of the work to be performed and deliverables (e.g., analysis section write-up, publication ready tables and figures).
• Discuss expectations of authorship.

Mentor(s) are strongly encouraged to attend the initial consultation meeting for a project that involves their trainee.

An important function of the Consulting Service is to assist with grant preparation for research studies. The following guidelines for new grant preparation will apply:

• Please provide the consultant(s) with the most current draft and highlight areas where help is needed. If there are comments from other advisors, please provide those comments.
• The consultant(s) will be given at least two months to work on the submission prior to its due date.
• Statistical aspects of a grant are intertwined with other elements of the study such as hypotheses, study outcomes, sample size calculations, randomization procedures and the timing and method of data collection. Our consultants will work closely with you to ensure that the statistical guidance we provide is appropriate to your specific study and objectives.
• Consultants will rarely write the statistical analysis plan for a study where another group is expected to perform the analyses. To ensure accuracy and interpretability of results, we believe it is important for the statistician(s) involved in the final analysis of a study to be the primary statistical author(s) of the statistical analysis plan. Thus, those involved in the analyses should be involved in preparing the analysis plan.
• For future statistical service, submitted grant budgets should include a percent effort for one or more named statisticians from the Center for Biostatistics and Data Science.
• The consultant should review a final grant application prior to submission to ensure consistency in the design, analysis and sample size sections as well as ensuring that reasonable funds for statistical support are allocated in the event of funding.

Please provide the consultant(s) with the most current draft and highlight areas where help is needed. If there are comments from other advisors, please provide those comments. While working with the consultant, please keep the following in mind:

• Statistical aspects of a protocol are intertwined with other elements of the study such as hypotheses, study outcomes, sample size calculations, randomization procedures and the timing and method of data collection. Our consultants will work closely with you to ensure that the statistical guidance we provide is appropriate to your specific study and objectives.
• Consultants will rarely write the statistical analysis plan for a study where another group is expected to perform the analyses. To ensure accuracy and interpretability of results, we believe it is important for the statistician(s) involved in the final analysis of a study to be the primary statistical author(s) of the statistical analysis plan. Thus, those involved in the analyses should be involved in preparing the analysis plan.

For assistance with sample size estimates, please consider the following prior to the initial consultation:

• Identify a few plausible statistical endpoint(s).
• Identify a plausible, clinically relevant therapeutic effect that you want to be able to detect.
• Locate previous studies that report the same endpoint(s) in a similar clinical setting.
• Identify estimates of the variability of the endpoint(s) in the population to be studied.
• Estimate the number of patients that may be available to consent and accrue and the maximum sample size you could manage.
• Sample size calculations are intertwined with other elements of the study including the hypotheses, study outcomes, randomization procedures, timing and method of data collection, missing data and the statistical analysis plan. Our consultants will work closely with you to ensure that the sample size estimates are appropriate to your specific study and objectives.

The Consulting Service will assist with data analyses and preparation of abstracts, papers, and presentations. The following recommendations will apply:

• Be prepared to provide the study proposal or manuscript’s introduction and methods sections. If those are not available, please bring the research questions, hypotheses or objectives and brief background information about your research and methodology.
• Be prepared to provide specifications of the data collection schedule, groups, subgroups, randomization, stratification, matching, crossover, primary and secondary outcomes and potential covariates.
• Provide the data to be analyzed, following the ‘Data Policy’ guidelines described below.
• The investigator is expected to define the format of deliverables (e.g., analysis section write-up, publication ready tables).
• The consultant will be given at least one week to review final drafts of an abstract or paper prior to submission or resubmission. Review includes checking tables and other numbers in the text for errors, as well as assessing the methods and results sections for accuracy and appropriateness.
• The consultant who performed the analysis will be offered authorship in accordance with our ‘Policy for Authorship’ below.

For help with a reply to a manuscript or grant review, please provide the consultant(s) with the original manuscript/grant, the original reviewer comments and the revised manuscript/grant with reply to review (if available). Manuscripts/grants developed with the assistance of the Consulting Service may require revisions after submission. The time needed by the Consulting Service to complete the necessary revisions and/or perform any additional analyses will be charged to the investigator.

Data provided to the consultant(s) should adhere to the following guidelines:

• Data provided should be complete and in a format that requires minimal pre-processing and quality control (i.e., ‘cleaning’). All time needed to process the data for analysis will be charged to the investigator. We recommend that the investigator and consultant work together to define the level of data quality control needed.
• To facilitate interpretation of the data, consider providing the consultant(s) with metadata and/or a data dictionary that may include the research instruments, a description of how and when each variable was collected, the data format, history of changes to variables, algorithms for derived data elements and the structure of the database and relationship between elements.
• Personally Identifiable Information (PII)
  • The investigator and the consultant(s) should work together to define the level of PII that is needed for the analyses.
  • Consultant(s) must be included as research team members on the Institutional Review Board (IRB) project for studies where they are provided data that includes PII.
  • Unless necessary for analysis, data should not contain PII such as names, addresses, medical record numbers, and social security numbers. It is incumbent upon the researcher to remove PII that is not needed before it is provided to the consultant. Failure to do so will result in delays and extra time charged to the investigator. Note that ‘hiding’ columns in Excel or other databases does not constitute removal.

Consultants who contribute substantially to a manuscript are expected to be offered co-authorship with the Washington University policy on authorship in mind:

An author is generally considered to be an individual who has made substantial intellectual contributions to a scientific investigation. All authors should meet the following three criteria, and all those who meet the criteria should be authors:

1. Scholarship: Contribute significantly to the conception, design, execution, and/or analysis and interpretation of data.
2. Authorship: Participate in drafting, reviewing, and/or revising the manuscript for intellectual content.
3. Approval: Approve the manuscript to be published.

Authorship will not be considered in lieu of compensation. Authorship should not be dictated strictly by the time spent on a project; the quality of time and nature of the contribution are relevant. Given the above justifications, it is expected that consultants be considered for authorship. Appropriate credit for the contributions of other individuals to the work described in the publication should be made as an acknowledgment. To avoid misunderstandings, please discuss authorship early in the consulting relationship. The consultant reserves the right to decline authorship when offered.