The Board's Role in Trial Go/No-Go Decisions
There is no decision in clinical-stage biopharma that carries more weight than the trial go/no-go decision. When the board approves advancement from Phase II to Phase III, it is committing the company to spending tens of millions of dollars — sometimes hundreds of millions — on a bet that the science will hold up in a larger, more rigorous trial. If the bet is right, the company may be worth billions. If it is wrong, the company may not survive.
Most boards handle this decision poorly. Not because they lack intelligence or diligence, but because they have not established a governance framework that distinguishes their role from management's role, that accounts for the information asymmetry inherent in clinical data, and that creates space for genuine deliberation rather than rubber-stamping a recommendation the CEO has already made.
The Governance Boundary
The first thing a board must establish — long before any go/no-go decision arrives — is where the governance boundary sits. The board's role is not to make the scientific judgment. It is not the board's job to analyze the pharmacokinetic data, interpret the dose-response curves, or evaluate the statistical analysis plan. That is management's job, supported by the company's scientific advisors and clinical operations team.
The board's role is to ensure that the process by which management reaches its recommendation is sound, that the information presented to the board is complete and honestly characterized, that the risks are identified and weighed against the company's capacity to absorb them, and that the capital implications of the decision are understood.
This distinction matters because boards that blur the governance boundary create two problems. First, they waste time on technical discussions that most directors cannot meaningfully contribute to, crowding out the strategic and fiduciary questions that are genuinely within the board's competence. Second, they give themselves a false sense of confidence — the feeling that they have evaluated the science, when in reality they have listened to a presentation about the science and nodded along.
A well-governed go/no-go decision is one where the board can answer these questions with confidence: Did management follow a rigorous process? Was the data presented honestly, including the uncertainties? Were dissenting views within the organization surfaced? Does the company have the capital to execute? And are we prepared for the consequence if the trial fails?
The Information Problem
The central challenge of the go/no-go decision is that the board is being asked to approve a major capital commitment based on information it cannot independently verify. The clinical data comes from the company's own trials, is interpreted by the company's own scientists, and is presented by a management team that has strong incentives — career, financial, psychological — to recommend proceeding.
This is not to suggest that management teams are dishonest. Most are not. But they are human, and the same optimism bias that drives scientists to pursue difficult problems also drives them to interpret ambiguous data in the most favorable light. A board that does not account for this bias is not governing. It is hoping.
Effective boards address the information problem in several ways.
Independent data review. Before a major go/no-go decision, the board should have the clinical data reviewed by an independent expert — not someone on the company's scientific advisory board, not someone who has consulted for the company, but someone with genuine independence and relevant expertise. This does not need to be a formal process. A single phone call between an independent expert and the board chair, or a brief written assessment, can surface issues that the management presentation glosses over.
Pre-read materials distributed early. The board should receive detailed materials at least a week before the meeting where the decision will be discussed. Materials distributed the night before — a practice that is depressingly common — do not allow directors the time to read carefully, formulate questions, or consult their own networks.
Management's obligation to present the bear case. The board should require management to present not only the data supporting advancement, but the strongest case against advancing. What are the risks? What could the data mean in a less favorable interpretation? What would a skeptical FDA reviewer focus on? If management cannot articulate a credible bear case, the board should be concerned — either the risks have not been thoroughly analyzed, or they are being withheld.
Separation of the recommendation from the data. The presentation should clearly separate the factual data from management's interpretation and recommendation. Directors should have the opportunity to look at the data before hearing what management thinks it means. The framing of the presentation shapes the discussion, and a board that receives the data pre-framed has already ceded much of its governance role.
The Questions That Matter
When the go/no-go discussion reaches the boardroom, certain questions separate genuine governance from performance.
"What would change your recommendation?" This is perhaps the most important question a director can ask. It forces management to articulate the boundaries of their confidence. If the answer is "nothing" — if there is no data point, no risk scenario, no competitive development that would change management's view — then the recommendation is not analytically grounded. It is conviction masquerading as analysis.
"What is the Phase III trial designed to show, and what is the probability it shows it?" Management should be able to articulate, in plain language, what the primary endpoint of the Phase III trial is, what effect size the trial is powered to detect, and what the realistic probability of success is given the Phase II data. If they cannot do this clearly, the board does not have enough information to make the decision.
"What happens to the company if the trial fails?" This is the capital and strategic question that is entirely within the board's competence. How much will the Phase III trial cost? How much runway will the company have if the trial fails? Are there other programs that could sustain the company? What is the exit strategy if Phase III is negative? A board that approves a Phase III trial without understanding the failure scenario is not exercising its fiduciary duty.
"Have we sought external input?" Has the company discussed the Phase II data with the FDA in a pre-Phase III meeting? What was the FDA's feedback on trial design, endpoints, and statistical approach? Has the company discussed the data with potential partners? External validation — or the absence of it — is a meaningful signal.
"Who disagrees, and why?" In any organization, there are people who have reservations about the leading recommendation. The board should ask whether dissenting views exist within the company and what those views are. If the Chief Medical Officer is enthusiastic but the head of biostatistics has concerns about the statistical robustness of the Phase II data, the board needs to hear both perspectives.
The Pressure to Proceed
Boards should be aware of the forces that create pressure to advance programs even when the data is ambiguous.
Management's careers and compensation are tied to advancing programs. Investors who funded the company expect to see clinical milestones achieved. The company's employees have dedicated years to the program and want to see it succeed. The patient community may be aware of the program and hoping for a new therapy.
None of these pressures are illegitimate. But they are not governance considerations. The board's obligation is to the company and its shareholders, which means making the decision that the data and the company's strategic position warrant — not the decision that everyone wants to hear.
The most difficult go/no-go decisions are not the ones where the data clearly supports or clearly contradicts advancement. Those are easy. The difficult ones are where the data is ambiguous — where Phase II showed a trend toward efficacy but did not reach statistical significance, where the safety profile raised questions that might or might not matter in a larger trial, where the competitive landscape shifted since the trial was designed.
In these ambiguous cases, the board's governance role is most critical. It is the board's job to ensure that ambiguity is acknowledged rather than explained away, that the decision to proceed (if that is the decision) is made with clear eyes about the risks, and that the company has a plan if the bet does not pay off.
Building the Framework Before You Need It
The worst time to figure out how your board handles go/no-go decisions is when one arrives. The governance framework — the process, the information requirements, the independent review mechanism, the questions that will be asked — should be established well before the first major program decision.
This means the board should discuss and agree on its approach to clinical decision governance during a period when no decision is imminent. It should establish expectations with management about what information will be provided, in what format, and on what timeline. It should identify independent experts who can be called on when needed.
Boards that build this framework in advance make better decisions when the moment arrives. They spend less time figuring out process and more time engaging with substance. And they send a clear signal to management that the board takes its governance role seriously — which, in turn, tends to produce better recommendations from management.
The go/no-go decision is where governance matters most. It deserves a process equal to its importance.