August 12, 2026
Six advisors just spent four hours dissecting your data. The follow-up report calls it a success with broad alignment, valuable insights, and strong engagement. Six months later, no one on the sponsor side can say whether that alignment was the result of real scrutiny or simply that the agenda went unchallenged.
That gap between a board that felt productive and one that actually was, is what this series examines.
Groups asked to solve a problem using information scattered across their members are eight times less likely to find the right answer than groups who have all the information up front [1].
If that statistic makes you uneasy, it should. A board can run flawlessly with polished slides and a report citing “broad alignment” and still fail at the one thing that justifies convening it: putting a real decision under real scrutiny. That failure is avoidable, but only if the sponsor team asks the right question before the agenda is drafted: what decision could this board materially change?
That doesn’t mean abandoning a strategy. A board can refine a development plan, expose an implementation barrier, shift the interpretation of emerging data, or confirm that a direction would survive external scrutiny. But to do this, it needs a live uncertainty and real consequences attached to what the advisors say. Without that, an advisory board can be exceptionally well executed and yet scientifically inert, leaving important questions unanswered:
That ambiguity matters. Pharmaceutical advisory boards require substantial investments of expert time, internal preparation, scientific content and company resources. Their stated purpose is to obtain external advice and knowledge that are not otherwise available within the organization. For this investment to pay off there must be a clearly defined purpose, appropriately selected expertise, transparent documentation and reporting of outcomes [2].
Meeting those requirements establishes board legitimacy. However, they do not, on their own, ensure that the discussion generated useful knowledge.
The premise of an advisory board is straightforward: bring together people with different expertise and the collective discussion should reveal more than any participant could have contributed alone. But research on group decision-making has demonstrated that outcome is not always guaranteed.
“Hidden-profile” experiments closely map to how advisory boards are convened. Information is distributed across a group. Some facts are known to everyone. Other facts are held by individual participants. The best answer becomes visible only when the group surfaces and integrates those separate pieces of knowledge.
A meta-analysis of 65 studies involving 3,189 groups found a striking and consistent imbalance. Groups discussed substantially more commonly held information than uniquely held information. When the correct answer depended on combining the information distributed across members, those groups were eight times less likely to identify it than groups in which everyone had access to the full information [1].
How does this relate to advisory boards?
Every advisor has seen the pivotal trial, the major congress presentation and the treatment guideline summarized in the pre-read. Those materials create a common and necessary evidentiary foundation, and this information is the most easily repeated once discussion begins.
However, the knowledge that justified inviting external experts is usually less uniform. One physician may have encountered an unusual tolerability problem. Another may understand why a theoretically attractive treatment sequence fails in community practice. A third may see referral delays, diagnostic gaps or access constraints that are largely invisible from the sponsor’s vantage point.
Those observations are dispersed across the room. They may be incomplete, locally specific or difficult to reconcile with the emerging group narrative. They are also where much of the board’s potential value resides, and ordinary discussions do not reliably draw them out.
The same shared-information bias impacts clinical reasoning. In one study, 24 teams composed of a medical resident, an intern and a third-year student reviewed patient cases in which some information was shared, and other information was available to only one team member. Shared facts entered the discussion significantly more often. Importantly, when the correct diagnosis depended on the uniquely held information, teams were significantly more likely to reach the wrong diagnosis [3].
Specialized expertise in the room doesn’t guarantee better outcomes. The group must be structured to effectively use it.
Armed with this knowledge of group dynamics, one might begin to think that agreement could be evidence of a poorly designed board, but this is not necessarily true: genuine convergence may be a highly useful finding. But its meaning depends on how the agreement was produced.
Consider a discussion that follows a detailed sponsor presentation supporting a preferred interpretation of the evidence. The first question asks advisors to identify barriers to implementing that interpretation. Several participants respond within the frame provided. Later speakers hear the direction in which the conversation is moving and add examples consistent with it. By the end, the room appears aligned. That alignment may reflect the evidence. But it may also reflect anchoring, selective discussion and the ordinary social momentum of a group.
Experimental research has shown that groups remain susceptible to anchors, even when participants are working cooperatively, but there is a way to counteract this.
A group becomes less susceptible to anchors when they have to document the process used to reach their judgment, rather than merely being asked to provide an answer [4].
What does this mean in practice? For an advisory board, the sponsor’s initial framing can become the anchor. It appears in the objectives, the evidence selected for the pre-read, the order of the slides and the language of the questions. None of those choices is neutral, even when every statement is scientifically accurate.
This is unavoidable to some degree. The sponsor convenes the board because it has a question to resolve. The scientific discipline lies in recognizing the directional force of that framing and deliberately creating room for information that challenges it.
One of the most useful questions to ask is counterintuitive: Was the board designed so that a credible expert could have changed the sponsor’s interpretation?
Could an advisor reject the premise of the question without derailing the agenda? Could the group identify an option that the internal team had not anticipated? Was there enough time to examine an inconvenient observation before moving to the next slide? Would a minority view survive the final report, or would it disappear beneath a summary statement describing overall agreement?
When the answer is ‘no’ to any of these, reassurance may still be produced, but its value as external advice is limited.
Advisory boards are often described in terms of alignment, consensus and agreement. Those words sound productive, particularly in a final report. Yet the field may genuinely be divided. Practices tend to vary by institution, patient population or geography. Evidence may support more than one defensible interpretation.
Compressing that variation into a unified recommendation removes important information that is needed to improve patient outcomes across a heterogenous population.
In an experiment involving 135 three-person groups, differences in participants’ initial preferences improved the consideration of uniquely held information and increased the likelihood of a better decision [5]. What researchers observed was that dissent interrupted the easy repetition of shared knowledge and forced the group to examine information that would otherwise have received less attention.
A good advisory board does not manufacture conflict. It makes disagreement safe, visible and analytically useful.
That requires more than asking at the end whether anyone has a different opinion. By then, the dominant interpretation has usually acquired momentum. Alternative views need an intentional route into the discussion, built before the meeting rather than requested at the end of it.
In Part 2 of this series, we examine where the groundwork is laid for a productive advisory board outcome: the pre-read and discussion guide, two documents that often appear administrative but carry enormous influence over the outcome. If not framed to correctly, they can quietly shut out the very insights the board was convened to find.
References