
AI and the Widening Liability Gap in Clinical Practice
9th July 2026Introduction
In Dickinson v Newcastle upon Tyne Hospitals NHS Foundation Trust, Charles Bagot KC, sitting as a Deputy High Court Judge, gave judgment in a clinical negligence claim arising from a failure to administer prophylactic Thiamine. The defendant admitted breach of duty. Causation was fiercely contested. Alongside his findings on the medicine, the judge made pointed observations about the conduct of two of the defendant’s experts. These criticisms carry lessons for anyone instructing or acting as an expert witness.
Late Disclosure of an Audit
Dr Wills, the defendant’s neurologist, revealed for the first time in his evidence in chief that he had carried out his own audit of Thiamine administration at his hospital. This had not been mentioned in his report or in the joint expert statement. The claimant objected that the point had never been put to their expert and could not fairly be tested. The judge allowed the questioning but reserved his position on what weight, if any, it should carry.
Dr Wills later explained that he had not thought the case would reach trial and so had not included the audit in his written evidence. The judge did not accept this as sufficient justification.
A Calculation Introduced Too Late
A further difficulty arose when Dr Wills introduced a new calculation of how long the claimant’s Thiamine stores would have lasted. This was raised only in re-examination, after the claimant’s neurologist had already given his evidence and could no longer respond to it.
The judge applied the principle set out in his own recent decision in Mew v The General Dental Council. Experts must not hold back relevant material for oral evidence. Fairness requires that opposing experts and counsel have a proper opportunity to test new material. The judge declined to place any weight on either the audit or the calculation.
Errors in the Neuropsychiatric Report
Professor Carson, the defendant’s neuropsychiatrist, faced separate criticism concerning the accuracy of his report. He wrongly referred to the case as involving an accident. He recorded the claimant’s age incorrectly. He also appeared to attribute to the claimant certain facts, including studying and physical build, that in fact belonged to the claimant’s wife.
Under cross examination Professor Carson accepted these errors but maintained there was no reason for him to have invented the details. The judge nonetheless treated the mistakes as relevant to the overall reliability of the report.
Lessons for Practice
This judgment reinforces two familiar but frequently overlooked duties. First, an expert’s report and any joint statement should contain the whole of the expert’s relevant material. Matters cannot be reserved for trial simply because a settlement was anticipated. Second, factual accuracy matters. A report containing basic errors about age, event type, or which party said what invites scrutiny of every other conclusion in it.
For those instructing experts, the case is a reminder to review reports carefully before exchange and to ensure experts understand that everything relevant to their opinion belongs in writing, not held in reserve for oral evidence.




