Serving as an Expert Witness: Duties, Pitfalls, Getting Started

Expert work is one of the few genuinely new skills a surgeon acquires after becoming a consultant. It uses your clinical knowledge but almost none of your clinical habits: you are not managing a patient, you are not making a decision, and — the part that catches people out — you are not on anyone's side.

Done well it is intellectually satisfying, well remunerated, and a real service to a process that cannot function without it. Done badly it damages your reputation in a way that is public, permanent and searchable, because judgments naming criticised experts are published.

This post covers what the role actually requires and where surgeons come unstuck. It sits within our medico-legal series.

The duty that defines everything

In England and Wales this is codified: CPR 35.3 provides that an expert has a duty to help the court on matters within their expertise, and that this duty overrides any obligation to the person instructing or paying them.

The underlying principles were set out earlier, in National Justice Compania Naviera SA v Prudential Assurance Co Ltd — universally known as The Ikarian Reefer [1993] 2 Lloyd's Rep 68. Expert evidence must be:

• The independent product of the expert, uninfluenced by the pressures of litigation
• Objective, unbiased opinion within the expert's field
• Clear about the facts and assumptions on which the opinion rests
• Explicit when a question falls outside the expert's competence
• Explicit when the opinion is provisional because further information is needed
• Promptly corrected if the expert changes their view

Every expert report in England and Wales must carry two declarations under CPR 35.10: that the expert understands and has complied with their duty to the court, and a statement of truth that the report is their true and complete professional opinion.

Those are not formalities. They are the hook on which an expert who has drifted into advocacy gets caught.

The single most common failure

Becoming an advocate for the instructing party.

It happens gradually and often without the expert noticing. The solicitor is friendly. The case has a narrative. Your report is one of the few documents that helps their client. It becomes tempting to phrase things helpfully, to omit the qualification that weakens the point, to not mention the paper that cuts the other way.

The tell is straightforward, and you can apply it to your own draft: would you have written this same report had the other side instructed you?

If the answer is no — if your opinion would shift with the letterhead — you have stopped being an expert.

A related discipline: you are required to consider material facts that detract from your own opinion, and to say where a contrary view is tenable. A report that acknowledges the strongest point against it is more persuasive, not less, because it demonstrates that the rest was reasoned rather than assembled.

Staying inside your expertise

The second great trap, and one orthopaedics is peculiarly exposed to because our subspecialties have diverged so far.

If you are an arthroplasty surgeon who has not managed a paediatric supracondylar fracture in fifteen years, you are not an expert on paediatric supracondylar fractures — however well you could answer a viva question on them. The standard is not "could I pass an exam on this" but "do I practise this, currently, such that I can speak to what a reasonable practitioner would do".

Courts are unforgiving here, and rightly. An expert whose own practice does not cover the matter in issue can have their evidence rejected entirely, taking the instructing party's case with it.

Say no. The willingness to decline instructions outside your field is the single clearest marker of a good expert, and it does not reduce the work you are offered — it increases it, because solicitors who get a clean, well-scoped refusal come back with the right case.

Hindsight bias is the occupational hazard

You will be asked to comment on a decision made at 3am by someone who did not know how it turned out, while you sit with the full record, the outcome, and unlimited time.

That asymmetry is the central intellectual difficulty of expert work. The question is never "knowing what we know now, was this right?" It is "was this a reasonable decision, at the time, on the information then available, in those circumstances?"

Practical guards:

• Read the records chronologically, and form a view of each decision before reading what came next.
• Ask what was actually available — which imaging had been reported, which results were back, who was on site.
• Apply the substitution test: would a reasonable body of practitioners of similar seniority, in those circumstances, plausibly have done the same?
• Distinguish sharply between "I would have done it differently" and "no reasonable surgeon would have done this." Only the second is a breach. Conflating them is the commonest analytical error in medical expert reports, and it is transparent to any experienced judge.

For the underlying legal standard, see The Medico-Legal Duties of a Surgeon.

The kinds of report

Keep breach and causation genuinely separate in your reasoning. A great many reports establish substandard care and then assume the harm followed from it — but causation is a distinct question with its own answer, and it is frequently where a claim actually fails.

The condition and prognosis report is the usual entry point. It is less adversarial, more clinical, and closer to work you already do.

How jurisdictions differ

The US is the genuine outlier in culture rather than principle. The retained-expert model, deposition practice and admissibility challenges make it a different experience from the court-facing model elsewhere — though the underlying obligation to give honest, independent opinion is the same everywhere.

Concurrent evidence, where opposing experts give evidence together and are questioned side by side, is worth specific preparation if you work in a jurisdiction that uses it. It is markedly harder to maintain an overstated position when the other expert is sitting next to you.

Accurate as of July 2026. Procedural rules differ by jurisdiction and change; confirm the requirements in the court you are reporting to.

Getting started

When. Generally once you are an established consultant with a current practice in the relevant area — typically several years in post. You need enough ongoing practice to speak to the current standard, and enough experience that your view carries weight.

How.

1. Do a recognised expert witness course. The legal framework, report structure and courtroom procedure are genuinely taught, and doing this untrained is how people come unstuck.
2. Start with condition and prognosis reports before taking breach of duty work.
3. Tell your indemnity provider. Expert work may need separate cover, and this is not optional.
4. Check your employment position — many contracts require declaration of external professional work, and it interacts with job planning.
5. Take a mentor. Have an experienced expert read your early reports before they go out.
6. Keep your clinical practice current. The moment you stop practising, your value as an expert begins a countdown.

Practical craft

Read everything. Every page of the records, not the solicitor's summary. Cases turn on a nursing note or an observation chart far more often than on the operation note.

Set out your instructions and materials. State what you were asked and exactly what you were given. If something material is missing, say so and ask for it.

Separate fact, assumption and opinion. Signpost each explicitly. Where a fact is disputed, give your opinion on each version rather than picking one.

Write for a non-clinician. The judge is not an orthopaedic surgeon. Explain the anatomy and the reasoning in plain English — this is a skill and it is noticed.

Give a range where a range exists. Real clinical practice is a distribution, not a point.

Quantify your uncertainty honestly. "I cannot say on the balance of probabilities" is a legitimate and sometimes vital answer.

Update if you change your mind. Promptly, in writing, and without embarrassment. Doing so protects you; concealing a changed view does not.

In the witness box

• Answer the question asked, then stop. Volunteering material is how experts get led somewhere they did not intend.
• Address the judge, not the barrister asking.
• Concede what should be conceded. Fighting an unwinnable point destroys your credibility on the points you should win.
• "I don't know" and "that is outside my field" are complete, respectable answers.
• Do not become irritated. Cross-examination is a technique, not a personal attack, and losing composure is precisely the intended effect.

The reputational point

Judgments are published, and a judgment criticising an expert names them. That criticism is permanent, searchable, and will be put to you by opposing counsel in every subsequent case.

The experts who attract it are almost never the ones who got a clinical detail wrong. They are the ones who argued a position they could not sustain, strayed outside their field, or wrote for the party rather than the court.

The summary

Your duty is to the court and it overrides the person paying you. Decline anything outside your current practice. Read the record chronologically and judge decisions on what was known at the time, not on the outcome. Keep breach and causation separate. Concede what should be conceded, and say plainly when you do not know.

The best expert reports read as though the author has no interest in the result. That is not a stylistic choice — it is the entire job.

---

This article describes procedural frameworks for educational purposes and is not legal advice. Requirements vary by jurisdiction — take advice from your indemnity provider before accepting instructions.