Reading a Clinical Trial Protocol as a Board Member
By the time a clinical trial reads out, most of what determined the result was decided long before the first patient was enrolled. It was decided in the protocol — the document that specifies who will be studied, what they will receive, what will be measured, how many patients are needed, and what will count as success. A protocol is not paperwork. It is the place where a program's odds are largely set, and where a company can win or lose a trial before it has spent a dollar dosing anyone.
Almost no board member reads the protocol. It is long, technical, and easy to regard as management's domain — the kind of specialist document a director is not expected to engage with. And it is true that a director should not read a protocol the way a clinical scientist does, second-guessing the medical detail. But the instinct to skip it entirely gives up something valuable, because a protocol's most consequential choices are not, in fact, deeply technical. They are strategic decisions dressed in technical clothing, and a director who knows where to look can see them clearly and ask about them intelligently.
This article is a guide to that kind of reading — not how to evaluate a protocol as a scientist, but where a board member should look, what a handful of choices commit the company to, and which questions surface the decisions that matter most. It closes this pillar because it underlies everything else in it: nearly every clinical-stage governance question this series has examined traces back to a choice made in a protocol.
Why the Protocol Deserves a Director's Attention
The case for a board member engaging with the protocol rests on a single fact: the protocol encodes the strategic bets, and once the trial is running, those bets are expensive or impossible to change. A board that engages only at the readout is governing the consequences of decisions it never saw made. By then the endpoint is fixed, the population is fixed, the comparator is fixed, the statistical bar is fixed — and if any of those was set wrong, the trial can succeed on its own terms and still fail the company, or fail despite a drug that works.
The protocol is also where several of the failure modes examined elsewhere in this pillar originate. The gap between statistical and clinical significance is set by the endpoint and the power calculation in the protocol. The regulatory acceptability of the eventual result depends on whether the protocol's design matches what the agency agreed to. The strength of a subgroup or secondary finding depends on whether the protocol pre-specified it. To govern those downstream questions well, a director benefits enormously from having seen where they were decided.
None of this requires the director to master the protocol. It requires knowing which of its many sections carry the strategic weight, and reading those with attention while leaving the rest to the specialists.
Where to Look: The Sections That Carry the Weight
A protocol has many parts, and most of them a director can pass over. A small number deserve real attention.
The objectives and endpoints. This is the first place to look and the most important. The primary endpoint is the single measure on which the trial's success formally depends, and everything else in the protocol is built to serve it. A director should read the primary endpoint and ask a plain question: if the trial succeeds on exactly this endpoint, is that a result that matters — clinically, commercially, and to a regulator? An endpoint can be easy to hit and mean little, or meaningful but harder to achieve than the program can support. The secondary endpoints matter too, because they are often where the commercially important claims live, and because a program that is quietly counting on a secondary endpoint to carry the story has a fragility the board should understand.
The eligibility criteria. Who is allowed into the trial determines what the result will mean and how it will generalize. Narrow, restrictive criteria — a carefully selected, relatively healthy, tightly defined population — make a trial easier to win but can produce a result that does not transfer to the broader, sicker, messier population the drug will actually be used in. Broad criteria make the trial harder and the result more generalizable. Neither is wrong, but the choice is strategic, and a director should understand which way it was made and why. A useful instinct: when the inclusion and exclusion criteria are unusually restrictive, ask what the trial is being protected from, and whether winning in that protected population will translate into winning where it counts.
The comparator and control. What the drug is being tested against shapes what a positive result is worth. A trial against placebo answers a different question than a trial against the current standard of care, and a trial against an outdated standard answers a different question than a trial against what physicians will actually be prescribing at launch. This choice is often buried in the design section and passed over, but it is one of the most consequential in the document. The director's question is the one from the significance discussion applied at the design stage: is the comparator the thing we will really be competing against?
The sample size and statistical plan. A director does not need to evaluate the statistics, but should understand what the power calculation assumes. Buried in the sample size justification is an assumed effect size — the magnitude of benefit the trial is built to detect. If that assumed effect is larger than the drug is likely to deliver, the trial is underpowered and may miss a real but smaller benefit. If the trial is powered to detect an effect smaller than the market requires, the company can win a trial that proves something commercially irrelevant. The assumed effect size is one of the most revealing numbers in the whole protocol, and asking "what effect size is this trial built to detect, and how does that compare to what we expect and what the market needs?" is among the highest-yield questions a director can pose.
The pre-specified analyses. The protocol, and its associated statistical analysis plan, specify in advance which analyses will be treated as confirmatory — including which subgroups and which secondary endpoints. This pre-specification is what separates a credible result from a hopeful one. A director who knows what the protocol committed to in advance is equipped, at readout, to tell the difference between a finding that was planned and one that was discovered in the data after the fact. Reading this section is how a director earns the standing to ask the pre-specification question later and mean it.
The stopping rules and interim analyses. Many protocols specify points at which the trial may be stopped early — for success, for futility, or for safety. Understanding these tells a director how the trial can end before its planned conclusion and what each early ending would mean. It also reveals how safety is being monitored, and by whom, which connects to the board's own oversight of safety.
What a Director Is Actually Looking For
Reading those sections, a director is not checking the science. They are looking for a small number of things that are genuinely within a generalist's competence to assess.
Is the trial designed to answer the question that matters? The deepest failure a protocol can encode is to be beautifully designed to answer a question that, once answered, does not help the company. A trial can be rigorous, well-powered, and clean, and still be pointed at an endpoint, a population, or a comparator that makes its result commercially or strategically irrelevant. A director's most valuable contribution is to hold the whole picture — does success here actually advance the company? — while the specialists, appropriately, are focused on making the trial methodologically sound.
Is the trial designed to win honestly, or designed to win? There is a difference between a trial designed to give the drug a fair test and one engineered to maximize the probability of a positive result regardless of what it proves. A very narrow population, an easy comparator, a soft endpoint, a low bar — each can be individually defensible, but in combination they can produce a trial the company is likely to win and unlikely to learn anything durable from. A director should be alert to a design that seems built for the press release rather than for the truth, because a trial optimized to succeed can leave the company with a result that does not survive contact with regulators, payers, or the real-world population.
Does the design match what the agency agreed to? If the company has had regulatory interactions about the program, the protocol should reflect what the agency indicated it would accept. A protocol that quietly diverges from agency feedback — a different endpoint, a different population, a different analysis than what was discussed — is carrying a regulatory risk that may not surface until it is very expensive. A director does not need to reconcile the two documents personally, but should ask whether the design is consistent with the agency's input, and be uneasy if the answer is vague.
What has to go right, and what happens if it does not? Every protocol embeds assumptions — about enrollment rate, about dropout, about effect size, about the behavior of the control arm. A director reading well is asking which assumptions the trial depends on most heavily and how fragile they are. This is the same downside-branch discipline that good governance applies everywhere: not pessimism, but knowing the shape of the risk before committing to it.
How to Use the Reading
The point of a director reading the protocol is not to approve or redesign it — that is management's responsibility, informed by expertise the board does not have and should not pretend to. The point is to engage the strategic choices while they are still choices, and to do so in a way that strengthens rather than second-guesses the team.
The most productive posture is to treat the reading as the basis for a conversation, not an audit. A director who comes to management with "walk me through why the primary endpoint is the right one, and what we give up by choosing it" is doing something useful; a director who comes with "I've decided the endpoint is wrong" is overreaching into territory they are not equipped for. The questions in this article are framed as questions deliberately. Their value is in surfacing the reasoning behind the design so the board can satisfy itself that the strategic bets were made consciously and well — not in substituting the board's judgment for the team's on the medical particulars.
It also helps to do this reading at the right time: early enough that the answers can still influence the design, before the protocol is finalized and the trial is a committed, running thing. A board that engages the protocol only after enrollment has begun has, once again, arrived after the decisions were made. The window in which this reading has leverage is before the trial locks.
For directors who feel they lack the background to read even the strategic sections confidently, the answer is not to skip the protocol but to secure an independent perspective — an advisor who can help the board understand what the design commits the company to, separate from the team that wrote it. The goal is never for the board to out-expert its own scientists. It is for the board to be able to see, and ask about, the strategic choices that a protocol makes on the company's behalf.
The Underlying Principle, and the Close of the Pillar
A clinical trial protocol is where a program's future is quietly decided, in choices that look technical but are strategic: what to measure, whom to study, what to compare against, what bar to clear, what to promise the analysis in advance. A board that never engages with these choices is governing a clinical-stage company with its eyes closed during the moments that matter most, and opening them only at the readout, when everything consequential has already been fixed.
This closes the pillar on clinical-stage decision making, and it is a fitting place to end, because nearly every question the pillar has examined lives, in the end, inside a protocol. Whether the board can tell statistical significance from clinical meaning depends on the endpoint and the power calculation set here. Whether a pivot rests on honest evidence, whether a biomarker strategy is sound, whether the regulatory basis for approval is real rather than assumed — all of it traces back to decisions encoded in a document most directors never open. The clinical-stage board's task, across all of it, is the same: to ensure that the choices which determine the company's fate are made consciously, tested honestly, and understood for what they commit the company to. Learning to read the protocol is where that task begins, because it is where those choices are written down.
Lawrence Fine is CEO of AGCP Farmacêuticos and has advised on licensing, regulatory, and partnership strategy across the pharmaceutical and advanced materials sectors.