How Scientific Evidence in Medical Negligence Claims in the UK Helps Establish Medical Causation
Showing that care fell short is only half of a clinical negligence claim. The other half is showing that the shortfall caused the harm, and that is usually where the science comes in.
NHS Resolution’s annual report for 2025-26, published in July 2026, shows how often claims do not get through. It resolved 11,841 clinical claims that year, and 56% ended with damages being paid. Some of the rest will have failed on breach, but many will have turned on whether the harm was down to the negligence or would have happened anyway. Answering that takes more than a clinician’s view of what usually happens.
What causation actually asks
The test is the ‘but for’ question: on the balance of probabilities, would the claimant have avoided the harm if the care had been reasonable? If the answer is probably no, the claim fails even when the care was poor.
Take a hypothetical delayed cancer diagnosis. Say earlier treatment would have raised a patient’s chance of surviving five years from 30% to 45%. That's a better chance, but it is not better than even, so the claim on that basis fails. If earlier treatment would have lifted it from 30% to 70%, the position is different. The House of Lords took the same line in Gregg v Scott (2005), where a delayed cancer diagnosis reduced the claimant’s chance of a cure, and the court refused to compensate a lost chance of a better outcome when that outcome was less likely than not.
Those percentages do not come from the medical records. They come from studies, and someone has to read them properly.
Where scientific evidence comes in
A clinical expert can say what they have seen in practice, but experience is a narrow sample and it doesn't always match the published evidence. Scientific evidence tests the clinical opinion against what research shows about how a condition progresses, how well a treatment works and how big a risk really is.
That raises practical questions. Was the study large enough? Did it include patients like this claimant? Does a headline figure describe the relative risk, which can sound alarming, or the absolute risk, which is often much smaller? Courts have at times treated a doubling of risk as a rough marker for ‘more likely than not’, but as a guide and not a rule, and applying population figures to one person is always contested.
What a scientific evidence analyst does
Searches the literature systematically, so the evidence is not picked to suit one side
Grades the studies by design, size and risk of bias
Checks whether the papers a clinical expert cites actually support the point being made
Separates relative risk from absolute risk, and population figures from individual ones
Flags where the evidence is thin or missing, and says so plainly
Writes it up in language a judge can follow
How the role differs from a clinical expert
The clinical expert addresses the standard of care and what happened to this patient. The analyst addresses the evidence behind those opinions, and the two work best together. Practice Direction expert to refer to the literature they rely on and to summarise any range of opinion, and an analyst can make sure that part of the report stands up to scrutiny. This is often called expert analysis of medical evidence, though titles vary.
Where it tends to matter and when to instruct
Delayed diagnosis is the obvious example. So are arguments about whether a drug or infection caused a particular injury and birth injury claims, where the question is when the damage occurred and whether different management would have changed the outcome. A scientific evidence expert for solicitors is also useful for testing the other side’s report, when the causation opinion rests on a study that does not quite say what it is cited for. Where most claims settle before trial, an early view on causation helps a firm decide whether to continue, and what the claim is realistically worth.
Choosing the expert
Look for training in epidemiology, biostatistics or clinical research methods, and experience of systematic reviews. A clinician who has searched PubMed is not the same as someone who assesses evidence for a living. Check the expert has written reports under CPR Part 35, be clear about the exact question, and ask for a sample report. A good report explains the method and the limits of the evidence as well as the conclusion.
If you're a scientist considering expert work
Your duty is to the court, not the party instructing you. Keep records of your searches, so you can show what you looked at and what you left out. Be careful with wording around uncertainty, because ‘no evidence of an effect’ and ‘evidence of no effect’ mean different things, and a barrister will test the difference.
Expert Witness Gateway lists scientific evidence analysts by discipline and experience, so solicitors working on clinical negligence claims can find someone whose background fits the causation question.