Torbay and South Devon NHS Foundation Trust

5. Mrs L complains about aspects of care and treatment provided to her mother, Mrs O, at Torbay and South Devon NHS Foundation Trust between 27 and 30 November 2022. Specifically she says: • there was a delay in admission to the emergency department and Mrs O waited outside in an ambulance for 13 hours • once admitted Mrs O did not have an endoscopy to find the cause of bleeding from her stomach • this led to a delay in diagnosis and meant Mrs O was too unwell to have surgery to treat the cause of her symptoms, and • Mrs O died shortly after receiving an injection on 30 November 2022.

6. Mrs L says her mother’s condition deteriorated while waiting in the ambulance, and the delay in admission meant her mother was too unwell to have an endoscopy and subsequent surgery to treat the cause of her vomiting blood. She says if her mother had been admitted earlier, she may have survived. Mrs L is also concerned the injection caused or hastened her mother’s death. Mrs L believes no action was taken to diagnose and treat the cause of bleeding as her mother was terminally ill with a different health issue. She says her mother did not need to die at that time and in that manner.

7. Mrs L seeks an acknowledgement of failings and apology, and service improvements to prevent this happening to other patients. She also seeks a financial remedy for distress caused by witnessing events and her mother’s death.

8. On 27 November 2022 Mrs O started vomiting blood. She was taken to Torbay hospital by ambulance, arriving at 11.21pm. At the time, she was terminally ill with stage four cirrhosis, a degenerative disease of the liver which leads to liver failure. Mrs O had had similar symptoms seven months earlier, when she was diagnosed with a pancreato-duodenal artery aneurysm (a bulge in the pancreato-duodenal artery, the blood vessel which supplies the pancreas) which was treated with embolization (a procedure to prevent aneurysms from bleeding).

9. The ambulance arrived at hospital at 11.21pm. Mrs O waited in the ambulance outside the emergency department for nearly 13 hours before she was transferred to the emergency department at 12.55pm on 28 November. During this time doctors visited Mrs O in the ambulance, and plans were made to book Mrs O in for an endoscopy the following day, pending a consultant review in the morning.

10. Unfortunately, by 29 November Mrs O’s condition had deteriorated, and she developed hepatic encephalopathy (a serious complication of cirrhosis that leads to changes in the brain). Mrs O sadly died on 30 November 2022.

Delay in admission to the Emergency Department

14. Mrs O waited outside the emergency department in an ambulance for nearly 13 hours before she was admitted into the hospital. Understandably, Mrs L has concerns this contributed to her mother’s deterioration and death.

15. We can see from the healthcare records Mrs O arrived outside the emergency department by ambulance at 11.21pm on 27 November 2022. She was seen by the triage nurse at 11.24pm. On 28 November Mrs O was taken into the emergency department at 01.50am to see a doctor and returned to the ambulance at 02.35am. From what we can see, doctors from both the emergency and acute medical teams then visited Mrs O in the ambulance on six occasions before Mrs O was transferred to a bed in the emergency department at 12.55pm on 28 November 2022. This meant Mrs O waited in the ambulance outside the emergency department for 12 hours and 34 minutes before she was moved inside the hospital.

16. The Trust told us it had acknowledged and apologised for the delay in Mrs O being brought into the Emergency Department. It said this was beyond the emergency department’s control at the time as the hospital was at capacity, meaning a hold on the back of the ambulance was unavoidable. It said on 27 November 2022, all appropriate escalation measures had been taken, but demand for its services still exceeded its capacity. It says its services were under extreme pressure at the time and this was also experienced at other hospitals in the area.

17. The Trust said it thought Mrs O was appropriately supported in the ambulance while waiting to be moved into the hospital. She was seen by a senior emergency doctor, where her history was taken along with bloods, she had an X-ray, and was started on fluids. Mrs O was then referred to the medical team who also went out to the ambulance to assess her several times. It says it does not think the delay in admission contributed to Mrs O’s death.

18. As a background to this case, there has been a national crisis in emergency departments for the last five years. Nationally, emergency departments are often overwhelmed with patients and unable to meet the targets set out by the Department of Health. Mrs O’s 13 hour wait in the ambulance outside the Emergency Department is unfortunately not unique to this Trust nor an isolated incident at NHS Trusts across the country. While we recognise a patient waiting for 13 hours in an ambulance once it arrives at hospital is never appropriate, and the Trust recognises this, there was nothing the emergency department could do about patients waiting in ambulances at the time as it had more patients than it had capacity to treat.

19. The Royal College of Emergency Medicine guidelines state triage of patients should take place within 15 minutes of arrival. Emergency departments use a triage system to assess patients and assign a priority to them. Those patients who are assigned a priority of 1 or 2 have imminently life-threatening problems and should be seen immediately or within 10 minutes respectively. Given the current crisis in emergency departments nationally, patients with a priority of 3 (potentially life threatening – see a clinician within one hour), 4 (potentially serious/not serious -see a clinician within two hours ) or 5 (not serious – see a clinician within 4 hours) may have to wait significantly longer than the target time to see a doctor when the emergency department is overwhelmed.

20. The term ‘handover’ in emergency department records refers to the transfer of clinical responsibility from ambulance staff to the emergency department service. In Mrs O’s case, her clinical care was handed over to the emergency department promptly on her arrival with triage within three minutes of arrival.

21. The triage nurse assigned Mrs O to triage level 3 (target to see a doctor within one hour). Our emergency adviser says Mrs O was triaged appropriately for her condition on arrival. Taking into account the information available in the clinical records regarding Mrs O’s presentation, symptoms and National Early Warning System (NEWS) scores, we think she was given the appropriate triage category in accordance with the Trust’s own criteria (ROSE score) and the Manchester Triage System it is based on.

22. Mrs O should have seen a doctor within one hour of arrival. She was seen by an emergency doctor at 01:50am (two and a half hours after arrival). Our emergency doctor says this time seems excessive and she should have seen a doctor sooner with her presentation on arrival at hospital. This is not in line with the Royal College of Emergency Medicine’s guidelines.

23. Good Medical Practice states doctors must adequately assess the patient’s condition, taking account of their history, and where necessary examine the patient. Doctors must promptly provide or arrange suitable advice, investigations or treatment where necessary and refer a patient to another practitioner when this serves the patient’s needs.

24. The emergency doctor ordered blood tests, a chest X-ray, and intravenous anti-emetics (drugs to prevent vomiting) and fluids. The emergency doctor recorded the working diagnosis as haematemesis (vomiting blood) with unknown cause and referred Mrs O to the on-call medical team at 02:46am. Our emergency adviser says the emergency doctor requested the appropriate tests and we think their actions were appropriate and in line with Good Medical Practice.

25. NICE CG141 (upper gastrointestinal bleeding) states the Blatchford score, a risk assessment used to calculate risks with upper gastrointestinal bleeding, should be done at the first assessment. The Trust’s guidelines also state the Blatchford score should be done. The purpose of this is to identify patients who are at higher risk and may need admission, and those at lower risk who can be discharged home to wait for an outpatient appointment for endoscopy. There is no indication the Blatchford score was assessed while Mrs O was under the care of the emergency department doctors. However, our emergency adviser explains blood test results are needed to calculate a patient’s Blatchford score. While the emergency doctors arranged blood tests, those results were not available until after Mrs O was transferred to the care of the on-call medical team and are first recorded in the records at 05:43am by the medical doctor.

26. At the point of referral to the medical team, Mrs O’s care became the acute medicine team’s responsibility. From that point, the emergency doctors would only have continued to be involved with her care if her condition had deteriorated whilst she was waiting to see the medical team.

27. The Royal College of Physician’s NEWS2 score is a national method of monitoring a patient’s deterioration, their clinical risks and the need for urgent interventions. A score of 0-4 is low risk, a score of 5-6 is medium risk and needs an urgent response by an acute clinician (doctor), and a score of 7 or higher is high risk which requires an urgent or emergency response from an acute clinician.

28. Mrs O’s NEWS score fell during this time from 6 (medium risk) at 11.34pm, shortly after triage to 5 (medium) at 01.23am, shortly before she first saw the emergency doctor. Following this her NEWS score continued to fall, measuring 4 (low risk) at 02.37am, and measuring 2-3 between then and 06.14am when it had increased to 4 again. By this time the medical team were actively involved in Mrs O’s care. The fall in NEWS score, from medium to low risk, following triage indicated Mrs O did not need intervention from the emergency doctors before she was seen by the medical team at 05.43am.

29. We appreciate Mrs L’s concerns about her mother’s long wait in the ambulance before she was admitted to the hospital. We recognise the Trust acknowledges this wait was unacceptable but outside of its control. We think Mrs O was appropriately scored at triage and provided appropriate treatment by the emergency doctors which we think was in line with Good Medical Practice.

30. In conclusion, we have not seen any failings in the decisions made in the emergency department at triage and by the emergency medical team. While we do find there was a delay in admission caused by the department being overwhelmed, the Trust recognises this and we accept the situation was outside of the Trust’s control. We do not think the delay in being moved to a hospital bed affected the outcome, as we can see Mrs O continued to be assessed and monitored by the ambulance crew and doctors until her transfer to a hospital bed. In coming to our view, we do not underestimate the concern these events caused to Mrs L and her family, and we hope our work provides some reassurance.

Failure to do an endoscopy and delay in diagnosis

31. Mrs L says the failure to do an endoscopy meant that by the time her mother had deteriorated, it was too late to do anything. She says not enough was done to find the cause of her mother’s bleeding.

32. The Trust told us it explained to Ms L in a local resolution meeting, people with upper gastrointestinal bleeding can present in two ways. They may have a lot of blood very suddenly, when they need to be rushed into surgery. Or they may bleed more slowly, in which case they need to do everything to ‘optimise’ the patient before taking them for an endoscopy, which is an invasive procedure. The Trust said Mrs O was very frail and the patient’s stability and appropriateness for the procedure must be carefully considered. The Trust felt it was the right decision not to do an endoscopy at the time as Mrs O was very poorly and had significant underlying conditions which were not recoverable.

33. Mrs O’s working diagnosis once she was seen by the medical team was that she had suspected oesophageal variceal bleeding (enlarged veins lining the oesophagus, which can leak or burst, causing bleeding. These can be caused by complications from cirrhosis or other liver problems).

34. NICE CG141, BSG guidelines and the Trust’s guidelines all say a patient with suspected variceal bleed should be given terlipressin (a drug which causes veins to constrict and so can help to slow bleeding caused by varices) and intravenous antibiotics. They add the patient should be discussed with the on-call gastroenterologist for urgent endoscopy. Emergency endoscopy is advised for a patient with suspected variceal bleeding as soon as they are stable enough for the procedure. These suggest that a discussion should be had with an on-call gastroenterologist urgently, and that an emergency endoscopy should have been considered.

35. We can see from the records the acute medical team had a telephone discussion with the on-call gastroenterologist at 11.08am on 28 November. The decision to contact gastroenterology was made during the consultant ward round at 8.51am. The medical team did not contact the gastroenterology team until 11.08am, over two hours later. We think the medical team should have contacted the gastroenterology team sooner for an urgent discussion.  We also note that neither the medical team nor the gastroenterology team calculated the Blatchford score or considered the use of terlipressin or antibiotics. Our gastroenterology adviser says they could not see any clear reasons why terlipressin or antibiotics were not given or any contraindication to doing so.

36. The Trust shared with us a letter from Mrs O’s recent outpatient gastroenterology appointment on 14 November 2022, which set out Mrs O was becoming increasingly frail and receiving support from the hospice team. The Trust added Mrs O was struggling to eat, had experienced dizziness for the three weeks prior to her admission, had had three falls in the previous 24 hours and had multiple bruises. It said these showed a level of frailty which was significant to the decisions made about her care and treatment.

37. The Trust explained Mrs O’s Hb (haemoglobin) level was 93 on her admission. This was slightly lower than her previously measured Hb level of 95. It says this suggested she had low volume bleeding. Given Mrs O’s frailty, likely dehydration and no evidence of a large volume of bleeding, terlipressin was not given as it has potential to cause harm. The Trust accepts the decision whether or not terlipressin was used was not recorded in the paper based notes, and this meant the decision-making process was not clearly documented.

38. The British National Formulary now advises avoiding the use of terlipressin in patients with severe liver disease, unless the benefit outweighs the risk. Our gastroenterology adviser says the Trust’s explanations why terlipressin was not given are appropriate, but noted there was nothing in the healthcare records to explain why this was not given. We can now see that the Trust initially considered the possibility of variceal bleed, but then excluded this, and so terlipressin was not indicated. The standards we refer to in paragraph 34 do allow for deviation where clinically appropriate to do so. Taking this into account, we do not find it a failing that terlipressin was not given.

39. NICE CG141, BSG guidelines and NICE cirrhosis guidelines all say unstable patients with severe acute upper gastrointestinal bleeding should be offered endoscopy immediately after resuscitation. All other patients who are stable should be offered endoscopy within 24 hours of admission. Our gastroenterology adviser says suspected variceal bleeds should be investigated with endoscopy immediately after resuscitation (that is, as soon as a patient is clinically stable for the procedure), with all other suspected causes of bleeding investigated by endoscopy within 24 hours.

40. The Trust told us the healthcare records show the gastroenterology team was contacted at 11.08am on 28 November, and a decision was made to keep Mrs O ‘nil by mouth’. It says while it is not expressly documented, this infers the plan for an endoscopy was made at that time, although it concedes this is not clear from the notes. Also documented is a call from endoscopy at 5.17pm the same day, confirming Mrs O had been booked onto the endoscopy list for the afternoon of 29 November, if she was felt to be appropriate and stable for the procedure after clinical review in the morning. The Trust acknowledged the decision whether to prescribe terlipressin and the decision making process around the appropriateness and timing of endoscopy is not recorded clearly.

41. The Trust also told us due to frailty and other complications of liver disease, including acute kidney injury and slurred speech, suggesting possible encephalopathy and coagulopathy, a decision was made not to do the endoscopy immediately. It said Mrs O would be reassessed the following day to see if she had improved enough to safely perform an endoscopy, or if she had deteriorated – in which case, an endoscopy would likely cause distress or possible harm and would not help her.

42. When Mrs O was reassessed by the consultant gastroenterologist at 10.50am on 29 November, she was felt to be too unwell for an endoscopy. Sadly, Mrs O’s condition continued to deteriorate, and she developed significant worsening of hepatic encephalopathy (a common complication of cirrhosis, especially where there is acute illness such as happened here) and was frail. Our gastroenterology adviser says not proceeding with the endoscopy on this basis was appropriate.

43. Our gastroenterology adviser considered the further information provided by the Trust and the explanations why an endoscopy was not done were clinically appropriate. We see no failings in the decision not to proceed with an endoscopy or in the timings of when this procedure was first considered.

44. After considering further information, sadly we do not think there is anything more the Trust could have done to prevent Mrs O’s deterioration and death. We do not find there were failings in the decisions made about Mrs O’s clinical care and treatment after her admission.

45. However, the healthcare records do not include any decisions made in relation to providing terlipressin, or include information about Mrs O’s frailty and recent medical history. The timings of decisions about endoscopy and the decision making process around whether endoscopy should be done are not clearly recorded in the healthcare records. The Trust acknowledges this, and we find there was a failing in record keeping.

46. This meant we initially did not have all of the relevant information when we first shared our provisional views. As a consequence, this has added an additional three months to the time it has taken for us to complete our investigation, and this means Mrs L has had to wait longer than necessary to receive our final report and understand more clearly what happened to her mother.

Injection on 30 November 2022

47. Mrs L says her mother died shortly after receiving an injection, and she did not know what this injection was or why it was given. Mrs L’s concerns that this injection caused or hastened her mother’s death are understandable and we can see why this situation would have been so upsetting for her at the time.

48. The healthcare records show that oxycodone (a strong opioid similar to morphine) and hyoscine butylbromide (an anti-spasmodic drug which is used to relieve breathing problems at the end of life) were given to Mrs O at 1.30pm on 30 November. Mrs O sadly died at 3.05pm the same day.

49. The Trust told us they were sorry not to have the opportunity to discuss this with Mrs L at the time she made her complaint. It said on 30 November the ward doctor contacted Mrs O’s family to explain she had deteriorated and was approaching the end of her life. The family attended the hospital and this was discussed with them further at Mrs O’s bedside. It said as Mrs O was dying, they were moving to comfort-based care to keep her comfortable and allow her to die with dignity. Mrs O was moved into a side room and palliative care medications prescribed. At 1.30pm she was given oxycodone and hyoscine butylbromide subcutaneously (injected into the fatty layer between the skin and muscle) at the same time. It said these medications were given to provide comfort and maintain Mrs O’s dignity at the end of her life.

50. NICE NG31 (Care of dying adults) advises drug interventions to manage breathlessness towards the end of life, including opioids and hyoscine. Our physician adviser says these drugs were given in line with these standards and at a low dose – 2.5mg oxycodone was given in the injection. Our physician adviser explains this dose is very low, and as it was given subcutaneously, this would have controlled the delivery of the drug so that it was slowly absorbed. They say it is unlikely such a small dose of opioid given subcutaneously hastened Mrs O’s death.

51. From what we have seen, we think the injection was given appropriately and did not cause or hasten Mrs O’s death. We understand why Mrs L, who witnessed her mother’s death shortly after she was given the injection, had concerns about this. We hope the explanations provided by us and the Trust provide reassurance and closure to Mrs L on this issue.

52. We thank Mrs L for bringing her concerns to us. We recognise Mrs L and her family experienced significant distress at witnessing the events and continue to be deeply affected by Mrs O’s sad death. We hope the explanations provided in this report go some way to providing reassurance to Mrs L on aspects of care and treatment provided to her mother.

1. We find there was a delay in admission to the Trust, and although we think a delay in nearly 13 hours of being moved from the ambulance to the emergency department was not appropriate, we have taken into account the Trust recognises this and that this was out of its control. We also think Mrs O should have seen a doctor sooner and been referred to the gastroenterology team sooner. Sadly, we do not think these delays affected the progress of her condition or that her death could have been prevented.

2. We have not seen failings in the decisions made about Mrs O’s clinical care and treatment, or that the injection given on 30 November caused or contributed to Mrs O’s death. However, we have seen failings in record keeping which mean it has taken longer for us to complete our investigation, leading to upset and uncertainty to Mrs L and her family about what happened. However, we do think these failings understandably caused Mrs O’s family additional distress at a very difficult time for them. We therefore partly uphold Mrs L’s complaint.

3. We partly uphold Mrs L’s complaint. At the end of this report we make recommendations for the Trust to put things right. We recommend the Trust writes to Mrs L to acknowledge the failings we identified and apologise for the impact they had. The Trust should also create an action plan to improve its services to prevent those failings happening again.

4. We hope the explanations we provide in this report provide reassurance to Mrs L and her family.

53. Mrs L seeks an acknowledgement of failings and apology, and service improvements to prevent this happening to other patients. She also seeks a financial remedy for distress caused by witnessing events and her mother’s death.

54. We think there were failings in record keeping which meant it took longer than necessary for us to complete our investigation with a full picture of what had happened and the decisions made in relation to Mrs O’s care and treatment.

55. In considering our recommendations, we have referred to the ‘NHS complaint standards’. The Complaint Standards support organisations to provide a quicker, simpler and more streamlined complaint handling service. They have a strong focus on: • early resolution by empowered and well-trained people • all staff, particularly senior staff, regularly reviewing what learning can be taken from complaints • how all staff, particularly senior staff, should use this learning to improve services.

56. The NHS complaint standards say organisations should be fair and accountable, and apologise when things go wrong. We recommend the Trust writes to Mrs L within one month of the date of our final report, to acknowledge the failings identified in this report and apologise for the impact of those failings on Mrs L and her family. We ask the Trust to send a copy of its apology letter to us.

57. The NHS complaint standards say organisations should use learning from complaints to improve their services. We recommend the Trust creates an action plan to state what it has done, or what it will do, to learn from this complaint and prevent the failings we have identified from happening again. We ask the Trust to create this action plan within one month of the date of the final report, and to send a copy of this action plan to Mrs L, NHS Resolutions, the CQC and us.

58. Mrs L also seeks a financial remedy as one of the outcomes of bringing her complaint to us. Our final decision is to partly uphold Mrs L’s complaint, on the basis that there was poor record-keeping which meant we relied on additional information from the Trust to reach our final decision, and our investigation took longer than it needed to. Taking account of our severity of injustice scale, we consider the injustice arising from this sits at Level 1, and feel an apology and service improvements around record keeping are sufficient to put right the injustice caused. For this reason, we do not make a recommendation for financial remedy.

59. We thank Mrs L for bringing her concerns to us. We recognise Mrs L and her family experienced significant distress at witnessing the events and continue to be deeply affected by Mrs O’s sad death. We hope the explanations provided in this report go some way to providing reassurance to Mrs L on aspects of care and treatment provided to her mother.