NHS in England Closed After Initial Enquiries Search on PHSO website

Hull University Teaching Hospitals NHS Trust

P-002443 · Statement · Decision date: 22 February 2024 · View Hull University Teaching Hospitals NHS Trust scorecard
Complaint (AI summary)
Ms F complained her daughter was admitted to ICU too late, the source of internal bleeding wasn't found or treated, and the death certificate listed an incorrect cause.
Outcome (AI summary)
Complaint closed. No wrongdoing found with ICU transfer or bleeding investigation. A medication error occurred but had no significant impact on the sad outcome.

Full decision details

The Complaint

5. Ms F complains that in April 2022: • M was not taken into intensive care (ICU) until she had been in hospital for over 30 hours • the Trust failed to find the source of M’s internal bleeding or give any treatment to try and stop it • the death certificate says M died of an upper gastrointestinal (GI) bleed and decompensated alcoholic liver, when a bleed was never found and haemochromatosis (an inherited condition that can cause iron overload in the body) could have been as much of the cause of her liver damage as alcohol.

6. Very sadly, M died in hospital. Ms F says if M had been taken into intensive care sooner and had the source of her bleeding found, she would have got the necessary treatment much sooner and she may not have died. Ms F thinks M’s death was avoidable. She says she cannot accept the recorded cause of M’s death and this has only worsened her distress and grief, at an already difficult time.

7. To resolve her complaint, Ms F would like the Trust to accept its failings. She also wants improvements at the Trust, for lessons to be learned and action taken to make sure these events do not happen to anyone else in future.

Background

8. Ms F told us about her daughter’s addiction to alcohol and other health problems starting in 2011. She also told us about M’s hospital appointments and the treatment she had received since January 2020.

9. The admission in question started in April 2022. Records note M was admitted with three health complaints: haematemesis (the medical term for vomiting blood), decompensated alcohol-related liver disease (a term to describe the complications of an advanced disease when the liver cannot function normally) and metabolic acidosis (a serious imbalance in the body’s acid-base).

10. M was first assessed at 7pm on the day of admission. Her past medical history was noted to include hypertension (high blood pressure), anxiety, liver cirrhosis (scarring and failure of the liver), asthma and alcohol excess. The clinician noted M’s most recent oesophagogastro duodenoscopy (OGD) had been done in March. An OGD, sometimes known as a gastroscopy, is a procedure where a flexible tube with a camera is inserted through the mouth to look at the throat, oesophagus (food pipe) and stomach.

11. The OGD in March had shown small oesophageal varices (enlarged veins in the oesophagus that most often happen in people with serious liver disease), mild duodenitis (inflammation in the lining of the small intestine) and gastritis (inflammation in the lining of the stomach). The clinician thought M had an upper GI bleed and requested another OGD.

12. The OGD was done and did not find clear evidence of bleeding. A CT scan and an ultrasound were requested to investigate further and these also did not find clear evidence of bleeding.

13. Records show M continued to receive care and treatment and stayed stable until 1pm the next day when she deteriorated. Specialists in critical care (who work in ICU), gastroenterology (specialists in diseases of the digestive system) and hepatology (specialists in liver diseases) reviewed M and consulted with each other throughout that afternoon, to consider whether she was at the end stage of her disease. In the early evening, gastroenterology and hepatology clinicians confirmed her care as critical.

14. Arrangements were made for transfer and M was admitted to ICU by 9.30pm that evening. M’s clinical picture overnight and into the next day suggested she had a slow internal bleed and liver and kidney ischaemia, which means there was restricted blood supply to those organs. Another scan, a CT angiogram, was arranged to continue to investigate the source of M’s bleeding.

15. Very sadly, before the CT angiogram could go ahead, M sadly died. The consultant who completed the relevant paperwork documented the cause of M’s death as decompensated liver disease and upper GI bleed.

Findings

Wait for ICU

19. Our adviser is unaware of any guidance that says when patients should be admitted to ICU. Instead, this is a clinical decision taken by the team involved in treating the patient. It is important that full consideration of the patient’s individual clinical circumstances takes place, as admission to ICU is reserved for patients who not only need more intensive treatment but who are also felt to have potentially recoverable conditions. This is set out in the BMJ clinical review, which says:

‘Intensive care has been defined as “a service for patients with potentially recoverable conditions who can benefit from more detailed observation and invasive treatment than can safely be provided in general wards or high dependency areas.” It is usually reserved for patients with potential or established organ failure.’

‘As with any other treatment, the decision to admit a patient to an intensive care unit should be based on the concept of potential benefit. Patients who are too well to benefit or those with no hope of recovering to an acceptable quality of life should not be admitted.’

20. While Ms F complains of a 30-hour wait, records show the decision to admit M to the ICU took place 24 hours after her admission. She was first medically assessed at 7.01pm and clinicians documented agreeing the decision for ICU transfer at 6.20pm the next day. A discussion with family about this decision is noted at 6.45pm. The ICU care plan notes M’s admission as 9.30pm, 26 hours after she was first seen.

21. While M was unwell to the point that she needed hospital care, records show that from her admission up until early afternoon the next day, her condition was otherwise stable. This is supported by records of her national early warning score (NEWS). This is a nationally used scoring system, where six physiological observations are checked regularly and scores given to each, with a higher score indicating a patient’s deterioration or greater clinical need.

22. There is no evidence that M’s NEWS score was high at any point from her admission to early afternoon on the next day. This supports our adviser’s view that M stayed clinically stable. We have not seen any clinical concerns with M’s care and treatment outside of the ICU during this time.

23. A note of a clinical review at 12.45pm on day two found M to be clinically stable. This changed at 1pm, when an arrest call was made. We hope to assure Ms F that records show quick action was taken by the Trust at that time. The critical care outreach team attended and the critical care consultant requested a review from the gastroenterology team because there was a query about whether M had end stage disease.

24. It was appropriate that full consideration of M’s circumstances was completed by the relevant teams, critical care and gastroenterology, to decide whether her condition at this point of deterioration was recoverable. As the BMJ clinical review says, it is reasonable to not admit a patient to the ICU if their condition may not be recoverable.

25. Gastroenterology input is documented at 1.15pm, with records showing active treatment was continuing to be given to M. The gastroenterology clinician reviewed M at 3.44pm and further gastroenterology review with input from hepatology took place at 5.20pm. They decided M should be transferred to the ICU. Once this decision was made, the ICU consultant attended and documented agreement with that decision and the move was made at 6.20pm.

26. In summary, we find evidence to show M was clinically stable until 1pm on day two of the admission. She deteriorated at 1pm, was given ongoing active treatment while being seen quickly that afternoon by the relevant teams. This was appropriate, as we have explained the need to consider whether ICU transfer was appropriate, or whether she was at the end stage of her disease.

27. Those discussions and reviews took place over an appropriate timeframe, and ICU was agreed reasonably quickly. Our adviser explains that once that decision is made, some time is then always needed to make the necessary arrangements for the patient to be transferred safely into an available bed space. We find no concern with the time taken here.

28. We know Ms F is concerned that M had a long wait for ICU care. We hope to assure her that we find nothing of concern to suggest undue delay in M’s transfer or that this compromised the care she was given. We do not think anything went wrong.

Bleeding

29. Ms F complains that the Trust failed to find the source of M’s internal bleeding. In response to her complaint, the Trust explained it did an OGD to look for M’s bleeding. It said while the OGD could not find an obvious cause, M showed signs of bleeding. It said M suddenly became clinically too unwell to have more investigations.

30. We can assure Ms F that the Trust’s actions to investigate M’s bleeding were in line with guidance and the diagnostic conclusions reached were reasonable. Our adviser confirms that M’s presentation suggested she had upper GI bleeding and this was a reasonable conclusion for the Trust to have reached. The Trust acted in line with the NICE guidance that says:

‘1.3.2 Offer endoscopy within 24 hours of admission to all other patients with upper gastrointestinal bleeding.’

31. The Trust did the OGD within an hour of M’s admission. This was the right investigation to do first, to help with the diagnosis. The OGD did not find any active bleeding or varices. On these findings, the Trust then arranged for a CT scan, which our adviser says was the right investigation to do next. The CT scan report recommended an ultrasound and records show the Trust did this next. The report notes limited findings as M did not wish to continue with the scan.

32. Records show the Trust appropriately continued to investigate by planning a CT angiogram. Unfortunately, events overtook clinical plans and M deteriorated before this could go ahead.

33. We have not seen anything to suggest the Trust failed here. Evidence shows the Trust took appropriate action in continuing to do the right investigations, in an attempt to find the source of M’s bleeding. Unfortunately, nothing more could reasonably be done to investigate before M’s deterioration.

34. Ms F also complains that the Trust failed to provide treatment to try and stop M’s bleeding. We assure her the recorded evidence shows M was given appropriate treatment in the form of aggressive fluid resuscitation (treatment to give large amounts of fluid to a patient) including dextrose and blood transfusions. She also had inotropes (medication to help the heart pump more blood with fewer heartbeats) and antibiotics.

35. The Trust also started M on terlipressin, a drug used to help manage low blood pressure. Records note awareness of M’s OGD in March which had found small oesophageal varices, so terlipressin was given to M because it was suspected she had variceal bleeding. This was appropriate treatment because NICE guidance says:

‘1.5.1 Offer terlipressin to patients with suspected variceal bleeding at presentation. Stop treatment after definitive haemostasis has been achieved, or after 5 days, unless there is another indication for its use.’

36. While the OGD from the day of admission did not find any active bleeding or varices, it was known that M had small oesophageal varices a month earlier. Our adviser explains that when oesophageal varices become fully engorged (filled with blood) they then bleed. Once bled, the varices can collapse and can sometimes not be seen on OGD. Although there was no evidence of variceal bleeding at first, considering earlier findings and M’s symptoms in the absence of clear bleeding elsewhere, it was appropriate for the Trust to give terlipressin as treatment.

37. But, the recommendation made after the OGD was to stop terlipressin. This seems to have been clinical judgment. As the OGD did not show varices, the clinical team likely felt there was no suspicion of variceal bleeding, meaning NICE guidance no longer applied. As M’s bleeding had not necessarily been controlled and the Trust did not find another cause for it, we think terlipressin could have reasonably been continued in line with the guidance. We find this suggests a service failure. We have carefully considered what impact, if any, this may have had.

38. We start by clarifying that the Trust did give M appropriate treatment in the absence of evidence of the source of any bleed. It is not the case that we find she was without any treatment. As we have explained, the Trust could have continued terlipressin, as even in the absence of varices on the OGD, small varices were known to have been reported a month earlier.

39. Our adviser explains that continuing M on terlipressin may have helped to control any bleeding, but we must make it clear that it may not have done. Our adviser says it would not necessarily have stopped it.

40. Importantly, if M stayed on terlipressin, we think it is unlikely that this would have prevented her death. This is because M was getting appropriate treatment with intravenous fluids (fluids inserted into the veins) and blood transfusions and despite this, she was still very unwell. Her blood pressure stayed very low, her lactate (substance produced when the organs are not functioning well) was high (indicating her condition was very serious and not resolving) and the earlier CT scan showed poor function of the liver. This means her liver was not getting enough oxygen or blood.

41. Our adviser says this, on top of an already damaged liver, is significant and meant the blood supply within the liver was greatly compromised. Her liver enzymes were high meaning the cells of the liver were damaged, her blood sugar dropped, and her prognosis was poor. Our adviser explains that even if terlipressin had continued, it is unlikely this would have reversed the seriousness of her condition.

42. We know it will cause Ms F distress to know there was one medication that M could have continued to have to try to help her condition. But, we have not seen evidence to suggest this had any negative impact or that it would have changed the very sad outcome.

Cause of death

43. We know how concerned Ms F is about the documented cause of M’s death. In response to her complaint, the Trust said it was reasonable to issue the death certificate based on a probable cause of upper GI bleed and decompensated alcoholic liver disease and that this is supported by the clinical evidence available.

44. Our adviser confirms that M’s presentation and the results of her hospital investigations all pointed towards her having an upper GI bleed and decompensated alcoholic liver disease. It was reasonable that these were documented as the likely cause of her sad death.

45. While M’s blood results did show levels that can be a sign of haemochromatosis, those same levels can also be a sign of decompensated liver disease. Our adviser explains that a formal diagnosis of haemochromatosis can only be made with a genetic test and these results were not known at the time of M’s death. The Trust would not have been able to say whether M did or did not have genetic haemochromatosis when she died.

Our Decision

1. Ms F complains about the care given to her daughter, M, by Hull University Teaching Hospitals NHS Trust (the Trust) and the documented cause of her death. We extend our sincere condolences to Ms F for her considerable loss.

2. From the evidence we have carefully considered, we do not see that anything was wrong with the timing of M’s transfer to intensive care, or the actions taken by the Trust when trying to find the source of her internal bleeding. We have not found anything wrong with the documented cause of M’s death because it is supported by clinical evidence.

3. We see the Trust provided appropriate treatment at first, but we think one medication should have been continued. While this suggests a mistake, we have not found evidence to show it would have had any big impact or changed the very sad outcome.

4. For these reasons, we are not investigating further. We know how important this complaint is to Ms F. We are very sorry to hear about what happened and about her understandable distress. We appreciate the details Ms F kindly shared with us.

Other Decisions About Hull University Teaching Hospitals NHS Trust

P-005052 · 18 Mar 2026
Mrs Z complains Hull University Teaching Hospitals NHS Trust (the Trust): • did not arrange a referral • did not …
Closed After Initial Enquiries
P-004969 · 3 Mar 2026
Mr C complains about the care and treatment his mother received in November and December 2023 in relation to gallstones.
Not Upheld
P-004378 · 28 Nov 2025
Mrs L complains about the delays in her aunt's care at the Trust and complains that the Trust misdiagnosed her …
Closed After Initial Enquiries
P-004252 · 12 Nov 2025
Mrs O complained about Hull University Teaching Hospitals NHS Trust (the Trust) regarding the care provided to her mother Mrs …
Partly Upheld
P-003782 · 28 Aug 2025
Mr D complains about two hospital admissions his late mother had in April 2024. He said the way she was …
Closed After Initial Enquiries
View all decisions for this organisation →