Chapter 15 Project Communication
15.1 Communication as a Core Biostatistics Competency
In clinical trials, strong statistical work is necessary but not sufficient. If statistical insights cannot be understood by non-statisticians, they cannot meaningfully influence decisions. Project communication is therefore not a “soft skill” add-on—it is a core professional responsibility for the Project Biostatistician.
Many project failures are not caused by incorrect analyses, but by: - Statistical risks identified too late - Risks communicated too weakly or too strongly - Misalignment between statistics, medicine, and operations - Late-stage surprises that could have been prevented earlier
Effective communication enables the statistician to act as a translator, risk signaler, and stabilizer—protecting scientific credibility while supporting timely, pragmatic decisions.
15.2 Participation in Project Team Meetings
15.2.1 Objectives in Regular Project Meetings
In routine project meetings, the biostatistician should avoid turning status updates into statistical lectures. The goal is to provide a decision-oriented summary that is understandable, actionable, and consistent with the study plan.
A practical three-question framework for each meeting update is:
- What does the data currently suggest?
- How confident are we—and why?
- What could still change the conclusion?
This structure keeps the discussion focused on progress, uncertainty, and decision risk.
15.2.2 What to Bring to Every Meeting (Practical Checklist)
Bring clear, simple updates on:
- Status
- What analyses are complete, in progress, or pending
- Dependencies (data cleaning, database updates, programming timelines)
- Risks
- Emerging issues such as imbalance, missingness, dropout patterns, protocol deviations
- Items that could affect power, interpretability, or the primary endpoint
- Decisions Needed
- What the team must decide (e.g., clarify rules, adjust timelines, request additional outputs)
- What information is needed to make the decision
15.2.3 Common Communication Pitfalls
Common pitfalls that reduce statistical influence include:
- Using technical jargon without translation
- Providing excessive detail instead of a clear summary
- Raising concerns without ranking priority or explaining impact
- Staying silent until issues become critical (silence is often interpreted as “no risk”)
15.3 Participation in Medical Discussions
15.3.1 Bridging Statistical Evidence and Clinical Meaning
Medical discussions often focus on clinical relevance, benefit–risk balance, and real-world interpretability. The biostatistician’s role is to clarify what the data can and cannot support while aligning statistical interpretation with clinical decision needs.
Effective approaches include: - Emphasizing the difference between “numerical trend” and “statistical evidence” - Translating uncertainty into clinically meaningful terms - Avoiding rigid language that sounds like rejection when the true message is uncertainty
Examples:
- Instead of “not statistically significant,” say:
“The data are compatible with both a small benefit and no benefit.”
- Instead of “wide confidence interval,” say:
“The estimate is still unstable due to limited information.”
15.3.2 Handling Disagreement Respectfully
When statistical conclusions conflict with clinical enthusiasm: - Acknowledge the clinical perspective and the rationale behind it - Restate the statistical constraints calmly and factually - Emphasize uncertainty and evidentiary boundaries rather than negation
A useful mindset is:
You are not saying “no.” You are saying “not yet proven.”
15.4 Participation in Regulatory Communication Meetings
15.4.1 What Regulators Expect From Statisticians
In regulatory meetings, statisticians are expected to: - Support scientific consistency across protocol, SAP, TFLs, and CSR narratives - Answer questions clearly and directly - Avoid speculation beyond submitted materials - Maintain precise language and consistent terminology
Regulators focus on statistical reasoning and defensibility—not programming details.
15.4.2 Practical Communication Principles in Regulatory Settings
- Answer the question asked—no more, no less
- Be precise rather than verbose
- Use consistent terms aligned with the SAP and CSR
- Avoid introducing new analyses unless explicitly requested
- Separate confirmatory and exploratory evidence clearly
Over-explaining can unintentionally create new questions.
15.5 Explaining Statistical Concepts in Non-Statistical Language
15.5.1 The Translation Responsibility
A Project Biostatistician must translate statistical concepts into decision language. If stakeholders cannot restate the message correctly, communication has failed—regardless of statistical correctness.
Practical translations:
| Statistical Term | Decision-Oriented Translation |
|---|---|
| p-value > 0.05 | The data do not rule out chance as an explanation. |
| Confidence interval | The plausible range of the true effect based on the data we have. |
| Sensitivity analysis | We tested whether conclusions change under reasonable assumptions. |
| Missing not at random | Dropouts may differ in a way that could bias the result. |
| Model assumption | This conclusion depends on assumptions about how outcomes behave over time. |
| Multiplicity | More comparisons increase the chance of false positives unless controlled. |
15.5.2 Practical Tips for Clear Statistical Explanations
- Start with the decision implication, then provide the statistical reason
- Use analogies carefully (only when they clarify, not oversimplify)
- Use ranges and scenarios rather than single-point certainty
- Repeat key messages consistently across meetings and documents
15.6 Communicating Risk to Sponsors Without Creating Panic
15.6.1 Balanced Risk Communication
Sponsors need early warning signals—not alarm bells. Effective risk communication is calm, structured, and solution-oriented.
A recommended structure is:
- What we are seeing
- Why it may matter
- What could happen if nothing changes
- What we can do now
This format provides clarity without creating unnecessary fear.
15.6.2 Avoiding Common Sponsor Communication Mistakes
Common mistakes include: - Overstating uncertainty and creating panic - Downplaying real risks to avoid difficult conversations - Using overly technical language that reduces comprehension - Communicating risks too late, causing surprise and mistrust
Trust is built when sponsors feel informed, not surprised.
15.7 Building Trust: The Statistician as a Project Advisor
The most effective project statisticians: - Speak early rather than late - Speak clearly rather than technically - Speak honestly rather than defensively
They become trusted advisors who: - Interpret risk - Support decisions - Protect scientific credibility - Keep the team aligned and calm under uncertainty
15.8 Key Takeaways
- Communication is a core statistical competency, not a soft skill.
- The goal is shared understanding, not technical completeness.
- Early, calm risk communication prevents crisis.
- Non-statistical language increases statistical impact.
- A trusted statistician shapes decisions without creating unnecessary alarm.