Sepsis claims are rarely a disagreement about whether sepsis is serious. They are an argument about what a clinician saw, what they wrote down, and what happened in the hours that followed. The clinical medicine is usually common ground. The chronology is not.
In short: a sepsis negligence claim alleges that a deteriorating patient was not recognised or escalated in line with published national guidance, and that earlier treatment would have changed the outcome. Breach of duty is often the more straightforward half, because NICE sets out what should happen and by when. Causation is where these claims are decided, because the claimant has to show that the delay itself made a difference rather than that the delay was simply wrong.
What a Sepsis Negligence Claim Actually Alleges
Almost all of them allege a failure of recognition or escalation rather than a technically poor treatment decision. The allegation has a standard shape. Observations were taken and scored. The score should have triggered a response within a set time. The response came later than that, or did not come at all, and the patient deteriorated in the interval.
That shape is why these cases are documentary. A claim about surgical technique needs an expert to say what a competent operator would have done. A claim about sepsis recognition needs someone to lay the recorded observations, the scores, the referral times and the drug administration times side by side and say where the gap opened.
Why the 2024 Change to the NICE Guidance Matters
Because the standard a clinician is judged against depends on the date of the alleged breach, and that standard changed. NICE guideline NG51, published in 2016, covered sepsis across all ages and populations. In 2024 NICE replaced it with three separate guidelines, splitting the advice by population: adults and young people aged 16 or over, under 16s, and people who are pregnant or recently pregnant.
This matters more than a reference change. A pleading that cites NG51 against a 2025 admission is citing withdrawn guidance, and a defence that relies on NG51 wording for an incident after the split is doing the same thing in reverse. Anyone preparing a chronology needs to establish which guideline was in force on the date in question before saying anything about what should have happened.
What NICE Expects Once Sepsis Is Suspected
For adults in an acute hospital setting, the adult sepsis guideline stratifies risk using the National Early Warning Score 2, and attaches a different expected response to each band. The bands, and what they require, are the backbone of most breach arguments.
| NEWS2 score | Expected response | Antibiotics |
|---|---|---|
| 7 or more | Urgent assessment, and referral to a senior clinical decision maker as soon as possible. Score recalculated every 30 minutes | Within 1 hour, with an intravenous fluid bolus in the same hour |
| 5 to 6 | Clinical review within 1 hour of assessment, including venous lactate. Score recalculated hourly | Within 3 hours, deferral within that window permitted |
| 1 to 4 | Assessment within 1 hour by a registered health practitioner. Score recalculated every 4 to 6 hours | Within 6 hours, deferral within that window permitted |
| 0 | Review arranged, with recalculation under local protocol | Not applicable at this level |
Two features of that table decide cases. The first is that the clock starts at the point of assessment, so the time the observations were taken is as important as the score itself. The second is that the guideline builds in time-critical escalation if signs of poor perfusion or further deterioration appear at any level, which means a low score early on does not license a long period of no further review.
Where Breach of Duty Usually Sits
In the gap between a score and a response, not in the treatment given. The recurring patterns are a set of observations taken and scored correctly but never acted on, a referral made to a junior rather than to a senior decision maker, a repeat score deferred past its interval, and antibiotics prescribed within the window but not administered until well outside it.
The standard applied is the ordinary one. A clinician is not negligent for reaching a defensible view that another competent clinician would have reached, which is the Bolam and Montgomery framework doing its usual work. What is harder to defend is a documented trigger with no documented response, because there is no clinical judgement on the record to defend.
Why Causation Is the Harder Half
Because the claimant has to show that earlier treatment would probably have produced a better outcome, not that the delay was a breach. Sepsis makes that difficult in a specific way. The patients who are most obviously deteriorating are often the ones already furthest into organ dysfunction, and a defendant will argue that the outcome was fixed before the negligent delay started.
The question is decided on the ordinary civil standard, which is the balance of probabilities, and it needs a counterfactual rather than a criticism. That is the same structure as delay claims in cauda equina syndrome, where breach is often conceded and the case still turns on whether the window for a better result was still open. Claims of this kind fail on causation regularly, including where the breach is admitted, and an expert who treats the breach as the end of the analysis is of little use to either side.
What Martha’s Rule Changed About Escalation
It created a documented route by which a patient, a family member or a member of staff can demand a rapid review, which means the absence of that review is now itself a recorded event. Martha’s Rule followed the death of Martha Mills in 2021, after she developed sepsis in hospital having been admitted with a pancreatic injury from a cycling accident. The coroner concluded she would probably have survived had she been moved to intensive care earlier.
NHS England began implementation at 143 pilot sites in May 2024, extended it to the remaining acute trusts in April 2025, and has set 31 March 2027 as the date by which all three components should be in place. For anyone building a chronology, the practical effect is a new class of evidence: whether the escalation route existed on that ward at that date, whether it was advertised, and whether a family request for review was logged and answered.
Which Records Decide the Case
The timed ones. Observation charts with the time each set was taken, the calculated score and the initials of whoever took them. The electronic prescribing record, which shows the prescription time and the administration time as two separate entries. Referral and bleep logs. Nursing entries recording a family member raising a concern. Ambulance records, where the first set of observations often predates arrival by a considerable margin.
Narrative entries written after the event are the weakest material in the bundle, on both sides. A retrospective note explaining why a score was not acted on carries far less weight than a contemporaneous one, and a claimant’s recollection of how long a wait felt is not a substitute for a timed entry.
Why a Single Normal Observation Set Proves Little
Because sepsis is a trajectory and a single score is a snapshot. A normal set of observations at one point is consistent with a patient who was well at that moment and consistent with a patient about to deteriorate. What carries weight is the sequence: how many sets were taken, at what intervals, and whether those intervals matched what the score at the time required.
This cuts both ways. A defendant relying on one reassuring score is overstating it. So is a claimant relying on one alarming score without addressing what the next set showed.
What Makes a Sepsis Chronology Hard to Challenge
Anchoring every step to a timed document and separating the three questions that tend to get merged. The first is what the records show happened and when. The second is what the guideline in force on that date required. The third is what would probably have followed had that requirement been met. Merging them produces an opinion that reads as advocacy and is taken apart in cross examination.
The same discipline is what makes causation in medico-legal cases survivable under scrutiny. An expert who states the counterfactual, the range of outcomes that were still available, and the point at which that range closed has given the court something it can use. An expert who states only that the delay was unacceptable has not.
Instructing the Right Discipline
Sepsis recognition and escalation sit with emergency medicine, acute medicine, intensive care and microbiology, and a report on the sepsis pathway itself should come from one of those specialties. Instructing outside the relevant field is one of the more common reasons a report is given little weight, and it is avoidable at the point of instruction rather than later.
Where a plastic surgical opinion does belong is in the consequences. Necrotising soft tissue infection, the debridement and reconstruction that follow, tissue loss, amputation and the resulting scarring and functional deficit are questions of plastic and reconstructive surgery, and they are frequently the largest component of the damages claim once the pathway argument is settled.
Frequently Asked Questions
Is a missed sepsis diagnosis automatically negligent?
No. Early sepsis can present without a clear source and with observations that do not yet trigger a high score, and a clinician who assessed the patient properly and documented a defensible view is not negligent because the diagnosis was made later. The question is whether the response matched what the recorded score required at the time.
Do antibiotics have to be given within the hour in every case?
No. The one hour target applies to the highest risk band. The adult guideline allows deferral within a three hour window at moderate risk and a six hour window at low risk, so a two hour delay can be compliant in one case and a clear breach in another. The band at the time of assessment decides it.
Does a claim still work if the patient survived?
Yes, where the delay caused harm that earlier treatment would have avoided. Amputation, organ damage requiring long term treatment and extensive soft tissue loss are all outcomes that can be attributed to a period of delay. The analysis is the same, and the counterfactual still has to be established.
How long is there to bring a claim?
The general rule under the Limitation Act 1980 is three years from the date of knowledge, with different rules for children, for people who lack capacity and for claims arising from a death. Whether a particular claimant is in time is a question for their solicitor, not for a medical expert.
Related reading
- What Happens When an Expert Witness Is Criticised by the Court
- How to Write a High Quality Expert Witness Report
- Why Independent Clinical Opinion Matters in Medico-Legal Cases
Prof. Hindocha prepares CPR Part 35 compliant reports for solicitors and insurers across England and Wales in his own areas of practice, and applies the same approach to a disputed chronology wherever it arises. Details of his medical negligence expert witness services and how to instruct him are on the instruction page.
This article is general information about how sepsis recognition and escalation are assessed in clinical negligence claims in England and Wales. It is not medical advice, and it is not a substitute for legal advice on a specific claim. Anyone who is acutely unwell with a suspected infection should be assessed as an emergency.