Hospital discharge 18. Mrs W complains her husband was not fit enough for discharge on 19 December 2021. She tells us she had to call the doctor within days of him returning home as he was so unwell and was struggling to breathe.
19. The Trust has not made any specific comment about Mr W’s discharge on 19 December 2021. It has only provided an overview of the treatment he received during his admission between 17 and 19 December.
20. Mr W was diagnosed with a fast arterial fibrillation, a worsening of his COPD causing extra fluids in his body, and a chest infection. On the day of discharge, the ward round documents show that Mr W was feeling better, his observations were normal, he was no longer requiring oxygen, and he was not on intravenous fluids or medication.
21. The relevant guidance for when a patient can be discharged from hospital is the NHS Hospital Discharge Service: Policy and Operating Model, March 2020 (the hospital discharge policy).
22. Annex D of the hospital discharge policy lists the criteria indicating patients need to remain in hospital. This includes such factors as needing care in the intensive treatment unit (ITU) or high dependency unit (HDU), needing oxygen therapy and intravenous (IV) fluids or medication, and others.
23. We understand from our physician adviser that if a patient does not meet any of criteria listed in Annex D of the hospital discharge policy, then they can be discharged. The clinical evidence shows Mr W did not meet these criteria, and so we have not seen any indications of mistakes by the Trust in discharging him at that time.
24. We acknowledge that Mrs W does not think her husband should have been discharged, given how unwell he became shortly after his discharge, and we appreciate this has been a cause of upset for her.
25. Having looked at all the evidence and comments from our physician adviser, we found that Mr W did not meet the NHSE criteria to remain in hospital and the Trust’s decision to discharge him on 19 December was in line with the relevant standards.
Diagnosis of leukaemia 26. Mrs W complains staff did not recognise her husband had leukaemia earlier than 7 January 2022. Mr W had blood tests taken during his admission on 17 and 19 December 2021 and he had blood tests after he was discharged as he became unwell.
27. At the time of Mr W’s diagnosis, the internationally accepted diagnostic criteria for almost all haematological cancers, including various types of leukaemia, was the WHO guidance.
28. The WHO guidance describes that anyone who has 20% or more of blasts in the white blood cells within their blood has an acute leukaemia.
29. The Trust says that when Mr W was discharged on 19 December 2021, he was not showing any signs of leukaemia. It explains that when a patient has leukaemia, abnormal immature white blood cells, called blasts, multiply uncontrollably and prevent production of other cells important for survival. Leukaemia is diagnosed when 20 out of every 100 white blood cells in a blood sample is a blast cell. The Trust’s explanation in regarding blast cells is supported by the WHO guidance.
30. On 7 January 2022, Mr W’s blood results showed that 33 out of every 100 of Mr W’s white blood cells was a blast cell, which fulfils the criteria for a diagnosis of AML.
31. Our haematology adviser explains that, in addition to having blast cells at a greater rate of 20% within the blood, AML prevents the production of important cells by the bone marrow. This includes platelets, meaning a low platelet count is a common feature of AML.
32. Mr W had blood taken on 17 December during his admission, and again on 24 December because he was feeling unwell. Both results were normal, in that both the total white blood cell count and the counts of the various types of white blood cells that make up that total count were within normal range. As such, there was nothing to indicate on those dates that the Trust had a basis to suspect AML.
33. Mr W’s platelet counts were 213 and 119 on 17 and 24 December. We understand from our haematology adviser that a normal platelet count is between 140 and 450. Although the reading on 24 December was low, our haematology adviser confirmed that the samples did not show any abnormalities which would raise suspicions of AML. The sudden drop in platelet cells from the samples taken on 17 December and 24 December indicated that closer monitoring of Mr W’s condition was required but, sadly, given his swift deterioration after 24 December there was no opportunity for the Trust to put this in place.
34. There are no national guidelines that outline when a falling platelet count should prompt consideration of AML, but the NHSE guidance on shared care with hydroxycarbamide outlines safe thresholds for the platelet count for patients taking that medication. Hydroxycarbamide is a drug sometimes prescribed for patients with PRV.
35. We know Mr W was prescribed hydroxycarbamide as part of the treatment for his PRV, so we have thought about whether this guidance indicates the Trust acted in line with this standard in considering his platelet count. Our haematology adviser confirmed that, although this guideline was published in 2022, which is after the care Mr W received, it reflects practice from the years prior to its publication and is the result of a national patient safety alert about oral chemotherapy drugs, including hydroxycarbamide, published in 2008.
36. The NHSE guidance on shared care with hydroxycarbamide says that a platelet count of less than 100 is a contraindication to starting hydroxycarbamide, which means clinicians should consider whether it is safe to prescribe it. While Mr W’s platelet count had not reached 100 on 24 December, it was rapidly approaching it.
37. Our haematology adviser explained that PRV does not cause a rapidly falling platelet count. For a patient with this condition the platelet count is typically normal or increased and, although a low platelet count can occur in a small number of patients, it is a slow process when this happens. However, low platelet levels can be a common side effect of hydroxycarbamide. Our adviser explained it is unlikely the hydroxycarbamide was a cause in Mr W’s rapid fall in platelets, as he had been on the same dose for some time, but it would be reason to monitor his condition more closely. As explained above, the Trust did not have the opportunity to put that monitoring in place, given Mr W’s deterioration very soon after 24 December.
38. The NHSE guidance on shared care with hydroxycarbamide outlines that a patient with well controlled PRV should be reviewed every three months. This was the frequency of Mr W’s reviews until his platelet results on 24 December, when he was then reviewed two weeks later, on 7 January 2022.
39. On Friday 7 January, Mr W’s platelet count was at 71, and this is when the blast cells showed at 33 and AML was first suspected by Mr W’s doctors. Doctors arranged for further investigations after the weekend, to take place on Monday 10 January, including a bone marrow biopsy and a review in the Haematology Day Unit. Our haematology adviser explained that completing a bone marrow biopsy is the only way to determine a diagnosis of AML.
40. Taking into consideration the guidance and our haematology adviser’s comments, Mr W’s white blood cell count and the fall in his platelet count on 24 December was not enough to indicate he had AML, but it was enough to start close monitoring of the potential development of AML or for the need to adjust the dose of hydroxycarbamide, in line with the guidance mentioned above.
41. Unfortunately, Mr W’s health significantly deteriorated over the weekend and, as we will go on to discuss, he was taken via ambulance on 10 January to the ED before he could attend for the haematology review and the bone marrow biopsy. Although the Trust was not able to perform the tests to definitively confirm if Mr W had AML, there was no earlier opportunity for the Trust to suspect this as a diagnosis and the Trust acted in line with applicable guidance in considering Mr W’s symptoms and test results during this period.
42. We understand that, although Mr W was unwell in December 2021, this was due to a chest infection and not due to AML. It must have been extremely difficult for Mrs W to see how quickly her husband’s condition deteriorated, and we can see how this would have led her to fear there was an earlier opportunity to identify AML.
43. We hope we have been able to clearly explain how we have reached our view based on the evidence, and that it provides Mrs W with some reassurance here.
Treatment pathways in ED on 10 January 44. Mrs W complains that the Trust did not follow the appropriate treatment pathway for a patient with leukaemia and sepsis. She believes that if treatment had started earlier when he attended the ED on 10 January, her husband’s death on 11 January 2022 could have been avoided.
45. The Trust says that when Mr W arrived at the ED he was taken to the resuscitation area, due to how unwell he was. In its response of 25 May 2022, the Trust noted that ‘there may have been some leukaemia, but [Mr W] was not [receiving] any treatment currently’. We know that Mr W had not received a definitive diagnosis, as at that time he was going through the diagnostic process for AML and was due to have further tests completed the day he attended the ED, but given the development of his symptoms, records show that staff recognised leukaemia was suspected.
46. Overall Mr W had three conditions which posed an imminent and serious threat to life by the time he attended the ED on 10 January 2022; hyperleukocytosis (an extremely elevated blast cell count due to the AML), clinical tumour lysis syndrome (TLS), which is a very serious condition caused by the rapid breakdown of malignant cells, and sepsis.
47. When Mr W attended the ED at 11am via ambulance, he was clearly unwell and had a National Early Warning Score (NEWS) score of 7. NEWS is a clinical assessment tool used to determine the degree of illness in a patient, based on their vital signs. The NICE Innovation Briefing 205 places a patient with a sore of 7 or above in the high-risk category, which means an emergency assessment is required by the critical care team, usually leading to patient transfer to a higher-dependency care area.
48. The Trust made a diagnosis of sepsis with an unknown source, followed by an irregular and fast heart rate, known as supraventricular tachycardia. The Trust says that, whilst in the resuscitation area, Mr W had bloods taken, was given IV fluids and antibiotics (directly into his vein) and had regular observations, along with reviews from the anaesthetic team, the medical team and the haematology team.
49. The Trust’s Sepsis Management Policy also categorised Mr W as a high-risk patient. The policy states that he should receive intravenous antibiotics and includes the warning ‘This is time critical, immediate action is required now’.
50. Blood tests that became available at 12.30pm showed that he was not neutropenic and that he did not have neutropenic sepsis, which is a complication of a very low white blood cell count.
51. Mr W had arrived at the ED at 11 am. Although the wording of the Trust’s sepsis management policy does not provide a time frame for when antibiotics should be given, it does say that immediate action is required, suggesting that antibiotics should be given at the first suspicion of sepsis. The records show that the earliest recorded administration of antibiotics is 4.06pm, 5 hours after Mr W attended the ED. We consider this delay to be a failing.
52. In addition to the consideration of earlier antibiotics for sepsis, our haematology adviser says that, because Mr W was suspected to have AML, staff should have also considered developing complications, hyperleukocytosis, and clinical TLS, as these posed an imminent and serious threat at the time he attended the ED on 10 January 2022. However, our adviser also explains that Mr W could not have received treatment for clinical TLS due to how unwell he was when arriving at the ED.
53. We have also not seen any evidence that staff considered treatment for hyperleukocytosis. The haematology publication says that ‘if hyperleukocytosis is present, immediate cytoreduction is advised’. Cytoreduction is treatment to reduce the white blood cell count. The most commonly used treatment in this scenario is hydroxycarbamide.
54. As the Trust did not consider treatment for hyperleukocytosis in line with the haematology publication detailed above, we considered the likely impact. Our haematology adviser explains that even the prompt use of hydroxycarbamide will not produce immediate results. While the reduction in the white blood cell count is seen after even one dose in many patients, it takes hours for this to occur when it does. As such, we have not seen anything to suggest not considering hyperleukocytosis led to any clinical impact to Mr W.
55. Mrs W believes that if treatment for his infection and AML complications had started earlier, her husband’s death on 11 January 2022 could have been avoided.
56. When Mr W attended the ED on 10 January, he was severely unwell and had symptoms developing over the weekend. Our geriatrician adviser explains that each of the AML complications Mr W developed has a high chance of mortality, especially the clinical TLS, which has a mortality rate of up to 80%. This means that the combination of the hyperleukocytosis, clinical TLS, and sepsis would have meant his chance of surviving this event was very small.
57. The NIH paper reports that the absolute mortality of people being treated with sepsis increased between 0.3 and 1.8% every hour treatment was delayed. This means that Mr W did not get the medication he needed as soon as he should have received it due to the delay in giving antibiotics, but even if that delay had not occurred, due to the additional complications Mr W had developed and the fact he was so unwell, his chance of survival was still small.
58. This means we cannot give any view that this delay led to or caused Mr W’s sad death at that time. We do realise that the delay caused a degree of uncertainty to his wife around not knowing whether it might have made a difference, even though Mr W was so unwell at that time, and this will be a source of distress to Mrs W. In our view, this is an injustice to her which has not yet been put right.
59. We have made recommendations for remedy which are set out below.