Part I of this series asked one question: what decision could this board materially change? Without a live uncertainty and real consequence attached to the advisors’ input, a board can run flawlessly and still produce nothing but polished slides and a report citing “broad alignment.”  In part 2 of this series, we look at how two key documents can determine the outcome, before anyone sits down in the room.

By the time the meeting begins, the outcome is largely already set. It was decided by two documents most teams treat as logistics.

Four Questions Before a Single Slide Is Drafted

The pre-read and discussion guide are usually assembled as soon as the topic is decided. But rushing to create these two documents without first defining the decision at stake and the uncertainty around it can result in agendas built on assumption rather than a genuine gap in knowledge, and a meeting that generates discussion without generating evidence. Before either document is created, the internal team should have answers to the following questions:

1. What decision is under consideration?
2. What uncertainty is preventing that decision?
3. What assumptions support the current direction, and what evidence could change them?
4. Whose experience can reveal what the company cannot see from inside its own data?

These answers become the yardstick for judging the value of the board afterward. Did the discussion surface something genuinely external to the organization, expose a weakness in the current interpretation, or distinguish a shared view from one that depends on geography, setting or patient population? Importantly, can the resulting advice be traced to an actual decision, even a decision to proceed unchanged?

The Pre-Read Frames the Discussion

A pre-read is far more than simply homework sent out so participants arrive informed. It determines what advisors treat as worth discussing and what they never think to raise. What’s included, what’s omitted, and what’s buried in an appendix are all editorial choices that frame the outcomes.

Two failure modes in pre-read creation are equally damaging:

• Too thin. Advisors arrive with different assumptions about the topic, and the first hour is lost reconciling them instead of debating substance.
• Too bloated. Dumping the whole data room signals that the sponsor hasn’t done the work of deciding what matters. It wastes a senior expert’s time, which is the one thing they won’t forgive.

A strong pre-read presents evidence honestly, including whatever complicates the sponsor’s preferred story, and keeps genuinely open questions open. Concision is the fastest signal of respect a senior clinician will notice. A pre-read curated to “lead the witness” produces a board that behaves exactly as led.

Practical test: before sending out the pre-read, ask yourself whether your document gives a skeptic enough ammunition to disagree with your working hypothesis, or only enough to agree with it. If there’s no evidence in the document that could support disagreement, revise it.

This is a quality-control check for pre-read design. If the document only contains evidence that supports the sponsor’s preferred interpretation, for example, it omits a subgroup analysis or a competing trial that complicates the story, an advisor who privately disagrees has nothing to argue from. They either stay quiet, or raise an objection that sounds unsupported. Either way, the sponsor loses the very input the board was convened to get.

The Discussion Guide Engineers the Conversation

If the pre-read frames the topic, the discussion guide determines whether participants will feel safe disagreeing, or only safe when agreeing. That safety isn’t a courtesy, it’s what ensures the discussion is truly open and productive. Crafting a discussion guide that makes disagreement feel safe involves two key elements: question construction and sequencing.

Question construction. A question can be built two ways:

1. Phrased as a proposition to endorse or worded so dissent feels like friction. This produces agreement, not insight.
2. Framed as a request for information rather than validation, asked before the group anchors on one interpretation. This produces reasoning.

Sequencing. Order carries as much weight as wording:

1. Open with a broad, low-stakes question to build shared vocabulary.
2. Move to contested material only once the room is warmed up.
3. Place the most consequential question in the middle third of the agenda, never first (guardedness) and never last (rushed).
4. Build in deliberate silence after a pointed question. The quietest expert with the contrary data point often needs three extra seconds before speaking.

A well-built guide should feel like a natural conversation, not a script.

Building In Routes for Dissent

Most advisory boards are not consensus exercises. Their job is to map the field, not to make it converge, and a genuinely divided field that gets reported as aligned has lost real information. Four safeguards protect against premature convergence:

1. Individual reflection before open debate. A brief written response or private poll before discussion opens prevents the first confident voice from setting the frame for everyone else.
2. Questions that don’t disclose the preferred answer. Neutral wording is a discipline, not a courtesy.
3. Polling before senior voices speak. Capturing initial positions before a respected KOL weighs-in preserves the independence of everyone else’s answer.
4. Documentation that separates majority opinion from substantive minority reasoning, including why the minority disagreed, not just that they did.

Balanced participation is a data-quality control, not etiquette.

If four of five advisors defer to the fifth, the sponsor didn’t get five opinions. It got one, expensively confirmed.

Done well, none of this engineering is visible in the room. It’s a mechanism for finding out what the room actually thinks, before the meeting’s momentum decides for them.

Facilitation Is Structured Elicitation, Not Hosting

Even a well-designed guide can fail in the room if facilitation is weak. Two failure patterns occur most often:

• Anchoring by a single confident participant, who sets a position the rest of the room quietly declines to contradict, without any ill intent on anyone’s part.
• Drift into anecdote. Clinical stories are illuminating and often the most memorable part of a meeting, but a board that runs on war stories produces impressions, not findings. Good facilitation keeps the discussion on track and tied to what the evidence supports without flattening the clinical texture that makes expert input valuable in the first place. This is the same discipline guideline panels use to keep evidence and opinion visibly distinct.

Format Is Downstream of Design

Format choices, whether in-person, virtual, hybrid or asynchronous, get outsized attention relative to their actual influence on output quality. Preferences have shifted: roughly three-quarters of clinicians surveyed in one study of advisory board and small-group meetings preferred virtual or hybrid formats over in-person-only options [1]. This pattern is echoed in other meeting-attendance surveys [2-4]. However, some clinician groups still prefer to meet in-person [5]. Format affects who can attend and how candid people are willing to be, so it’s worth polling participants in advance to find the format best aligned to your board.

But format is a channel, not a substitute for architecture. A virtual board with a thin pre-read and a leading discussion guide underperforms exactly as reliably as an in-person one. A well-designed asynchronous exchange can outperform a beautifully catered meeting whose questions were built to be agreed with.

The Discussion Ends. The Work Doesn’t.

A board that is well framed, well sequenced and well facilitated will generate something genuinely valuable in the room. But discussion is ephemeral, and a transcript is not insight. What happens after the meeting wraps is where a surprising amount of a board’s return on investment quietly disappears.  Part 3 of this series addresses how that value gets captured, synthesized and turned into something a sponsor can actually act on.

1. Petrus C, Lam H. Considerations for Planning Effective and Appealing Advisory Boards and Other Small-Group Meetings with Health Care Providers: Importance of Participant Preferences. Pharmaceut Med. 2024 Jul;38(4):311-320. doi: 10.1007/s40290-024-00531-0.
2. Laohawetwanit T, Gonzalez RS, Bychkov A. Learning at a distance: results of an international survey on the adoption of virtual conferences and whole slide imaging by pathologists. J Clin Pathol. 2024 Aug 16;77(9):632-638. doi: 10.1136/jcp-2023-208912.
3. Hameed BZ, Tanidir Y, Naik N, Teoh JY, Shah M, Wroclawski ML, Kunjibettu AB, Castellani D, Ibrahim S, da Silva RD, Rai B, de la Rosette JJMCH, Tp R, Gauhar V, Somani B. Will “Hybrid” Meetings Replace Face-To-Face Meetings Post COVID-19 Era? Perceptions and Views From The Urological Community. Urology. 2021 Oct;156:52-57. doi: 10.1016/j.urology.2021.02.001. Epub 2021 Feb 6. PMID: 33561472; PMCID: PMC8556060.
4. Kim KJ, Kim SR, Lee J, Moon JY, Lee SH, Shin SJ. Virtual conference participant’s perceptions of its effectiveness and future projections. BMC Med Educ. 2022 Jan 3;22(1):10. doi: 10.1186/s12909-021-03040-9. PMID: 34980077; PMCID: PMC8721183.
5. El Zouhbi A, Assaf L, Honein-AbouHaidar G, Khabsa J, Akl EA. Virtual versus in-person meetings for practice guideline panels: A qualitative study. J Clin Epidemiol. 2025 Dec;188:111974. doi: 10.1016/j.jclinepi.2025.111974. Epub 2025 Sep 12. PMID: 40946877.