17. On 29 December in addition to her stomach pain, tests showed Mrs P’s INR was raised. An INR (international normalised ratio) is a test of how quickly blood clots. A higher INR score shows it is taking the blood sample a longer time to clot.
18. The Trust diagnosed Mrs P with hyponatraemia thought to be secondary to her raised INR and it stopped her prescription of warfarin (a blood thinner used to prevent excess clotting).
19. Mrs P returned on 31 December as her symptoms had not resolved. She complained of feeling unwell and of ongoing stomach pain. She also had reduced urinary output and facial swelling.
20. The A&E doctor took bloods, completed an abdominal exam and continued to withhold her regular prescription of warfarin as her INR remained raised. She was assessed as being dehydrated and again diagnosed with hyponatraemia caused by vomiting and diarrhoea.
21. The A&E doctor noted a plan to consider an OGD. This is a procedure where a camera is used to enter and explore the stomach through the mouth.
22. Mr P told us at the inquest clinicians discussed the likelihood Mrs P’s duodenal ulcer was present on admission. He told us the Trust did not do enough to explore her symptoms, particularly given the A&E doctor’s plan to consider an OGD. Mr P says had such a procedure happened the duodenal ulcer would have been located and treated.
23. We looked at whether the Trust’s initial assessment should have identified a duodenal ulcer.
24. NHS guidance says symptoms of a stomach ulcer (including duodenal ulcer) include abdominal pain and feeling or being sick. These are also recognised symptoms of hyponatraemia.
25. GMC guidance says, ‘In providing clinical care medical practitioners must adequately assess a patient’s condition(s), taking account of their history, including symptoms.’
26. As part of our investigation our adviser explained it is very common to experience a stomach ulcer and most people will have one at some point in their lives, and they may not realise it. They often go unnoticed and self-heal.
27. Our adviser explained in their experience Mrs P’s symptoms on presentation at A&E did not suggest she had a duodenal ulcer at that time. It was reasonable to initially treat her symptoms as hyponatraemia on account of her blood results which showed low sodium.
28. NICE QS38 recommends completing an OGD within 24 hours if a patient is vomiting blood. There is no specific recommendation otherwise. Our adviser explained the decision to complete an OGD would be a clinical judgment based on the doctor’s assessment of the stomach symptoms.
29. Although Mrs P had a history of vomiting. Mrs P was not vomiting blood on admission. Our adviser told us in their view the doctor completed an assessment of Mrs P in line with what we expect to see.
30. They told us whilst the A&E doctor was thinking about an OGD there was an opportunity for her stomach issues to resolve without the need for an OGD procedure. For this reason we are not critical of the Trust’s clinical investigations at that time.
31. We recognise a discussion took place during the inquest which suggested the duodenal ulcer may have been present for some weeks. We also recognise this discussion took place with the benefit of hindsight.
32. We have not seen any evidence the Trust should have identified an ulcer based on Mrs P’s presentation at that time.
33. Taking account of the relevant guidance and evidence provided by our adviser, we are satisfied the Trust completed an initial assessment in line with GMC guidance.
Initial management 34. Before attending hospital, Mrs P had been taking pantoprazole (a gastroprotection agent). The Trust’s investigation found it did not correctly record this medication and it was not continued.
35. Mrs P was also taking regular aspirin and this continued during her admission.
36. On 2 January, the Trust prescribed Mrs P prednisolone (a corticosteroid) as part of its treatment plan for her facial swelling.
37. CKS guidance on hyponatraemia explains gastroprotective agents can be a cause of hyponatraemia.
38. However, CKS guidance on oral corticosteroids recommends gastroprotective protection in patients at risk of gastrointestinal bleeding or dyspepsia (upper abdominal pain/ nausea).
39. It also says adverse reactions are more likely with higher doses or prolonged use, of steroids, but they can occur with lower doses or shorter durations. The guidance says the lowest dose should be given for the shortest time to reduce the likelihood of side effects.
40. NICE QS38 says aspirin can cause ulcers to form and pre-existing ulcers to bleed.
41. The treating clinicians told the inquest they disagreed gastroprotective agents should have been prescribed. They told the inquest these drugs can contribute towards hyponatraemia and it would not have been appropriate to prescribe them on admission.
42. They also said in their clinical experience despite the recommendations in some guidance, they found there is strong evidence oral corticosteroids (steroids) do not increase the risk of stomach ulcers forming. It was their view gastroprotective agents would not have prevented an ulcer.
43. Our adviser explained given the presentation of hyponatraemia, they agree omitting a gastroprotective agent initially may have helped manage Mrs P’s hyponatraemia.
44. We have considered Mrs P’s presenting hyponatraemia and taken account of the treating clinicians’ thinking, CKS guidance, and our adviser’s view.
45. It is our view the Trust would likely have decided to initially withhold gastroprotective agents, had these been correctly documented on admission. We have not seen any evidence to date this was an active decision made by the doctors at the time. But we cannot see this makes a difference to the treatment Mrs P should have initially received.
46. We are satisfied this prescribing was in keeping with guidance until 6 January and it was acceptable to withhold Mrs P’s regular gastroprotective agent.
47. Between 1 and 6 January, Mrs P took aspirin and steroids without gastroprotection agents. Studies looking at aspirin and the risk of gastrointestinal complaints found patients aged 60 and over with dyspepsia are at high risk of developing gastrointestinal issues.
48. Gastroenterology journal research shows gastroprotection agent use was associated with a lower risk of upper gastrointestinal bleeding in the general population.
49. The Oxford study found gastroprotective agents reduce the chance of an upper gastro bleed in patients taking low dose aspirin by up to 90%.
50. A BMJ study found patients who are prescribed steroids in hospital have a 40% higher chance of a gastrointestinal bleeding or perforation than patients in the community.
51. We have considered this research alongside the fact Mrs P was over 60 and taking low dose aspirin and regular gastroprotection agents in the community. We can see there was a significantly increased risk of developing a gastrointestinal issue on these medications and whilst in hospital.
52. Whilst we are not critical of the prescribing in this period, it is our view it is most likely the ulcer unavoidably formed in this period between 1 and 6 January.
Further management 53. Mr P says his wife was reporting continuing stomach pain, diarrhoea and constipation that the Trust did not adequately explore and investigate.
54. Records show Mrs P frequently reported stomach pain and dyspeptic symptoms. On 5 January the Trust diagnosed urinary retention and constipation, and it unsuccessfully attempted catheterisation. This is when a small tube is used to help drain the bladder.
55. The following day a catheter was successfully inserted. The Trust says the intervention resolved the issue. However, the records show Mrs P continued to report stomach pain. This matches Mr P’s recollection.
56. By 6 January, Mrs P’s diarrhoea and vomiting had resolved, although she was still experiencing abdominal discomfort. It had become clear hypovolemia (a decreased volume of blood in the body) was causing her hyponatraemia, not her vomiting and diarrhoea as first suspected.
57. Our adviser explained at this stage Mrs P’s clinical picture had changed and the Trust should have restarted gastroprotective agents on account of her discomfort, as well as risk of gastrointestinal bleed.
58. To date we have seen no evidence of the Trust’s consideration of gastroprotective agents and a rationale to withhold them from 6 January. This was not in line with CKS guidance on corticosteroids and NICE QS38 guidance on aspirin. It is our view this was a failing.
59. As set out above, the expectation under GMC guidance is for doctors to adequately assess a patient’s condition, taking account of their symptoms.
60. Our adviser explained following catheterisation, persistent abdominal pain should have prompted the medical team to do a comprehensive abdominal assessment. They explained if this level of investigation had taken place the doctor would have likely considered if a stomach ulcer could be the cause for the abdominal pain.
61. We note the doctors also told the inquest they would have completed an OGD had Mrs P’s stomach pain not resolved, however they said it had.
62. Additionally on 6 January the Trust says a doctor gave a verbal direction to stop the steroid prescription but this did not happen. The Trust’s investigation identified this failing in its management of Mrs P’s steroid prescription so we have not commented on this further.
63. Records show Mrs P received at least a further dose of steroids on 7 and 8 January. The Trust’s investigation found record keeping was poor and the drug remained on Mrs P’s active medication list until 11 January. It is possible she received doses on 9 and 10 January but these are not documented.
64. The BMJ study highlights that steroid use without gastroprotection can mask the symptoms of an ulcer due to steroids’ anti-inflammatory properties.
65. Considering Mrs P’s ongoing pain and the lack of a comprehensive abdominal assessment, we think the Trust did not act in line with GMC guidance and carry out appropriate clinical reviews of Mrs P’s abdominal pain. This was also a failing.
66. We looked at what difference these failings made. Our adviser explained the Trust would have completed further investigations such as a OGD or CT scan (a test that takes internal images of the body). These tests would have shown if there was an ulcer present. The Trust would have also thought about her medication in relation to her stomach pain.
67. CKS guidance on managing stomach ulcers says this primarily involves applying the STOPP/START processes and monitoring or adjusting medication. This is evidence-based criteria used to review medication regimens in elderly people.
68. STOPP (screening tool of older persons' potentially inappropriate prescriptions) aims to reduce medicines-related adverse events from potentially inappropriate prescribing and polypharmacy (prescribing of multiple medications to the same person).
69. START (screening tool to alert to right treatment) can be used to prevent missing the opportunity to prescribe appropriate medicines in older patients with specific conditions.
70. Had the Trust completed a further consideration of Mrs P’s stomach pain, it is our view, in line with this guidance, the Trust would have immediately begun prescribing gastroprotective agents. It also should have stopped the prescription of steroids on 6 January. This would have significantly reduced the risk of a perforated ulcer occurring.
71. We recognise further investigations (such as a scan or OGD) may have taken some time. There was no guidance in place at the time that tells us how quickly the scan should have taken place. Given Mrs P ongoing stomach pain, and the likelihood further examination would have considered an ulcer we would expect this to have been completed promptly. This would have given the Trust a full clinical picture.
72. Studies show, prior to perforation, ulcers are relatively easily managed and not life threatening. The BMJ study highlights once an ulcer perforates it is a much more serious and associated with considerable illness and death.
73. On 8 January, the Trust restarted Mrs P’s warfarin with a view to preparing her for discharge. NICE guidance on anticoagulation says warfarin should be given with caution in people at increased risk of gastrointestinal bleeding.
74. Gastroprotective agents are not routinely recommended for patients taking warfarin as they can increase the efficacy of warfarin meaning patients’ blood becomes too thin.
75. Mrs P was regularly taking warfarin and a gastroprotective agent before admission. We have not seen any evidence to date the Trust considered this in its decision- making process around restarting warfarin without gastroprotective agents.
76. For this reason, we are not satisfied the Trust has acted in line with GMC guidance when providing patient care. This was a failing.
77. When the Trust restarted a prescription of warfarin, Mrs P’s risk of serious complication from gastrointestinal bleeding increased further.
78. The inquest found Mrs P had a gastrointestinal bleed on 11 January which remained stable until 13 January. At that point it became uncontained and caused her to have a heart attack. It is our view the prescription of warfarin added to the seriousness of this bleed.
79. We recognise treatment options were significantly limited following perforation and Mrs P’s other health conditions, although well managed, meant she was a high risk for surgery.
80. Overall, we have seen a failing to provide gastroprotective agents and sufficiently explore Mrs P stomach pain from 6 January. Had this happened it is more likely than not the perforation could have been avoided. This means the significant bleed from the perforation on 13 January would not have happened.
81. For these reasons we have found Mrs P’s death was avoidable. We have made a recommendation for the Trust regarding this point at the end of our report.
Pain 82. Our review of the records found Mrs P was given very little medication for pain during her admission, despite reporting her pain to multiple members of staff.
83. The Trust’s investigation identified the failings in its response and management of her pain throughout her admission. We have not seen anything which would cause us to contradict the Trust’s response so we have not commented on this further.
84. We think there is more the Trust should do to recognise the impact of this failing. We have made a recommendation for the Trust regarding this below.