31 December to 2 January
17. Mr M was admitted on 31 December 2020 with a diagnosis of AML which had been made following a bone marrow biopsy. He was complaining of worsening shortness of breath for a few days prior to admission with associated dry cough and chest pain.
18. Mr M was assessed and this revealed borderline low oxygen saturations at rest of 95-96% on air, a respiratory rate of 20 breaths per minute, which is at the upper end of the normal range, and an elevated temperature of 38.9 degrees. An ECG was done on 31 December which was reported as normal with no definite features of a heart attack. Severe anaemia (deficiency of red blood cells) can be a cause of shortness of breath and therefore a full blood count was also done. When Mr M was admitted he was not anaemic, but there was a marked drop in his haemoglobin by 1 January. A chest X-ray completed on 1 January was reported as ‘clear’. Also on 1 January, Mr M was swabbed for Covid-19 and various cultures were performed to look for evidence of a chest infection. These were all negative. The antibiotic tazocin was started to treat possible neutropenic sepsis.
19. On 2 January Mr M said that his breathlessness was worse when he was lying down. NICE CKS guidance on the management of breathlessness explains that breathlessness, and in particular worsening breathlessness when laying down, is a potential symptom of heart failure. This guidance also says that acute pulmonary oedema (fluid build up in the lungs) can be a cause of breathlessness.
20. NICE CKS guidance on the management of breathlessness and the BMJ best practice assessment of dyspnoea, lists numerous investigations which might be appropriate in the assessment of breathlessness. Our haematology adviser has noted that amongst these investigations a CT scan is listed as a possible investigation and has explained that a CT scan would have helped determine whether Mr M’s shortness of breath was due to an infection, possible pulmonary embolus (a blockage of an artery) or fluid overload.
21. Our haematology adviser has commented that it seems in view of Mr M’s fever and shortness of breath these were assumed to be caused by a chest infection. Our haematology adviser went on to explain that this is very common in patients with AML as their white blood cells are not working properly and this increases the risk of infections.
22. A number of relevant investigations took place to try and understand the cause of Mr M’s breathlessness, but the reports of worsening symptoms on lying down should have prompted the Trust to consider heart failure or the possibility of fluid overload as potential causes.
23. GMC Good medical practice guidance states that:
‘If you assess, diagnose or treat patients you must: promptly provide or arrange suitable advice, investigations or treatment where necessary.’
24. The NICE CKS guidance clearly sets out the potential causes of breathlessness and we consider that to not look into potential heart failure and fluid overload was not in line with the GMC guidance above and indicates failings in the actions of the Trust.
25. Whilst it seems a CT scan had the potential to reveal more about the cause of Mr M’s breathlessness, this is not set out in any guidance as a specific requirement for diagnosis and therefore we are not persuaded this was a failing.
3 January
26. On 3 January, the decision was recorded to ‘stop IV fluids’, but a further two litres of fluids were given after this time which had already been prescribed.
27. GMC good medical practice guidance says:
‘In providing clinical care you must prescribe drugs or treatment, including repeat prescriptions, only when you have adequate knowledge of the patient’s health and are satisfied that the drugs or treatment serve the patients needs.’
28. To continue to provide fluids after the decision had been made to stop intravenous fluids, is not in line with this guidance as the decision had already been made that these did not serve Mr M’s needs. This indicates failings in the actions of the Trust.
29. Mr M’s oxygen saturations were monitored throughout his time in hospital, and at times he was given supplementary oxygen. On 3 January there were repeated times when his oxygen levels were recorded at 94%, but he was not given supplemental oxygen at these times.
30. NICE CKS guidance on the management of breathlessness says that oxygen should be given when a patients oxygen saturations drop to 94% or less. The Trust’s actions here were not in line with this guidance and indicate failings.
4 January
31. In the evening of 4 January Mr M had a PICC line inserted at which point his oxygen saturations dropped to 88%. Supplementary oxygen was appropriately provided which continued when he returned to the ward.
5 January
32. Mr M’s chemotherapy with daunorubicin and cytarabine commenced in the early afternoon on 5 January, and the antibody treatment gemtuzamab was given at around 7pm on the same day.
33. There is no specific guidance on the timing of intensive chemotherapy in patients with newly diagnosed AML unless the white blood cell count is dangerously high (in which case urgent administration is necessary), but this was not the case for Mr M. Our haematology adviser has explained that in the presence of an active uncontrolled infection it would be standard practice to treat the patient with antibiotics to try and get the infection under control before starting chemotherapy, and if necessary to control the white cell count with oral hydroxycarbamide chemotherapy whilst waiting for the infection to improve.
34. Mr M had high temperatures over the first few days of his admission suggesting that an infection was ongoing. His CRP blood test (checks for inflammation) was rising from 30 mg/l on admission, to 45, 63 and then 224 mg/l on 7 January, which is a sign of possible ongoing infection. However, no organisms were grown on any of the blood cultures to indicate a confirmed infection. The antibiotic tazocin had been started on 1 January.
35. Mr M’s NEWS2 score was stable at level two up until 5 January. Scores of one to four are considered low risk on the RCP NEWS2 scoring system. Mr M was scoring one point for his tachycardia (fast heart rate) and one point either for an intermittent raised temperature or intermittent low oxygen saturations.
36. There had been times on 3 January, as outlined above, when supplementary oxygen should have been given, potentially increasing Mr M’s NEWS2 score to four. This would still have been considered to be low risk.
37. Bearing in mind there is no specific guidance for starting chemotherapy in these cases, the low risk NEWS2 scores and lack of a confirmed infection, indicate it was appropriate to start Mr M’s chemotherapy on 5 January. Our haematology adviser noted that it was recorded by the clinicians at the MDT meeting on 6 January to ‘stop tazocin’, indicating that any infection was resolving. Our haematology adviser went on to explain that it is sometimes impossible to clear pre-existing infections in patients with active leukaemia, and therefore is sometimes necessary to give chemotherapy to AML patients with ongoing infections as the infection may only resolve when the patient’s leukaemia cells are cleared, and healthier cells return.
38. There is evidence in Mr M’s medical records that his deterioration therefore began in the evening of 4 January, before the chemotherapy started.
6 and 7 January
39. Having reviewed Mr M’s records our haematology adviser commented that Mr M showed a significant deterioration at 10.45pm on 6 January. At this point his oxygen level dropped to 91% and his respiratory rate increased to 24 breaths per minute, with his oxygen requirements increasing to 15 litres. These results gave Mr M a NEWS2 score of ten.
40. The RCP define a NEWS2 score of seven or over as high risk requiring ‘emergency assessment by critical care team, usually leading to patient transfer to higher dependency area’.
41. NICE CG50 guidance is also relevant here. This guidance says a graded response should be delivered for patients at risk of deterioration and for the high-risk scoring group says:
‘an emergency call to a team with critical care competencies and diagnostic skills. The team should include a medical practitioner skilled in assessment of the critically ill patient, who possesses advanced airway management and resuscitation skills. There should be an immediate response.’
42. A NEWS2 score of ten clearly put Mr M in a high-risk category.
43. Mr M’s medical records show that at 11.30pm on 6 January the nursing team referred him to the critical care outreach team who said they would see him. However, there is no evidence they did. No doctor of any grade appears to have reviewed Mr M at this time indicating failings in the actions of the Trust’s response to Mr M’s deterioration.
44. NICE NG51 sepsis guidance is also relevant. This guidance explains that patients who are having treatment for cancer are at a higher risk of developing sepsis. The first part of this guidance explains how people with suspected sepsis should be identified and recommends using a risk stratification tool for this purpose. In Mr M’s case this tool says that for people of any age with a possible infection ‘think, could this be sepsis?.’
45. The risk stratification tool then explains the risk factors for sepsis and lists ‘impaired immunity due to illness’ as one of these risk factors. This applied to Mr M because of his chemotherapy and the stratification tool therefore says that sepsis should be suspected and lists specific high, moderate to high and low risk criteria that should be considered.
46. Of the moderate to high-risk criteria Mr M met two of these. He had an impaired immune system and tachypnoea (a respiratory rate of 21-24).
47. The risk stratification tool flow chart indicates that if a patient has two moderate to high-risk criteria, then venous blood and blood gases tests should be completed with a clinician/results review within one hour. Blood cultures were done for Mr M at 6.55am on 7 January and blood gas/lactate levels were completed at 3.21pm on 7 January. The timing of these sepsis tests was not in line with the relevant guidance, indicating failings in the actions of the Trust.
48. At 6.55am on 7 January Mr M had a NEWS2 score of eight, again high risk. Blood cultures were taken but no escalation to a doctor or critical care team took place and no medical review occurred.
49. At 8.10am on 7 January the NEWS2 score was still high at seven, at which point the critical care outreach team were called and the on-call doctor was bleeped. The critical care outreach team reviewed Mr M at 9.30am and requested a senior review which occurred at 9.55am when Mr M was reviewed by the consultant. At this time, a plan was made to give further antibiotics, get microbiology advice, give diuretics, and get a chest X-ray. No referral to the critical care team was made at this time.
50. It was not until 3.15pm on 7 January when a junior doctor reviewed Mr M again and did a venous blood gas and spoke with the critical care outreach team who advised that a senior doctor liaise with intensive care. This prompted a review by another haematology consultant at about 3.30pm when it was noted that the chest X-Ray requested in the morning had not been done. Further changes were made to the antibiotics and a plan to refer Mr M to the ITU was made.
51. The decision to escalate to critical care to deal with Mr M’s deterioration, occurred approximately 16 hours after Mr M first started having high scores on the NEWS2 scale. NICE NG50 guidance requires an emergency call to the critical care team and an immediate response. The actions of the Trust in dealing with Mr M’s deterioration were not in line with the relevant guidance and indicates failings in its actions.
52. As we have explained above, we have seen evidence that Mr M’s high NEWS2 scores and deterioration was not appropriately dealt with as emergency calls to the critical care team were not made.
53. Had these calls been made then the critical care team would have been involved with potential decisions to admit Mr M to the ITU and when would have been the best time to do this. As the emergency calls to the critical care team did not take place, as set out in NICE guidance NG50, it is difficult for us to say at what point ICU referral and/or admission would have been required, as there was no critical care involvement to inform these decisions.
54. Our haematology adviser has explained that although Mr M had a high NEWS2 score of ten at 11.30pm on 6 January, this figure did quite quickly reduce. Mr M should have been assessed more thoroughly at this stage as we have explained above, but our haematology adviser commented that it was not unreasonable for him to remain on the haematology ward at this point because his NEWS2 scores reduced.
55. The following morning at 6.55am Mr M’s NEWS2 score was high again at eight, and at this point, it remained so. Again, Mr M should have been assessed more thoroughly. Mr M’s NEWS2 scores remained high with scores between seven and nine until the point at which he was ultimately referred to the ITU which was around 6pm.
56. Based on the evidence we have seen, we consider that the referral to the ITU should have therefore taken place around 7am on 7 January. Taking into account the fact that the decision to admit Mr M to the ITU was made when the referral was finally made at around 6pm, we consider that, on the balance of probabilities, it was more likely than not that the same decision would have been made in the morning of 7 January if the referral had been made as it should have been. A further opportunity was missed when Mr M was reviewed by the haematology consultant at 9.55am, when he remained very unwell, but a referral to critical care was still not made.
57. Relevant guidance regarding the timing for admission to critical care is the NHS England, Adult critical care service specification which says:
‘Admission to Critical Care must be timely and meet the needs of the patient. Admission must be within 4 hours from the decision to admit wherever possible and should adhere to the appropriate standards.’
58. Therefore, had the referral been made at around 7am Mr M should have been admitted to the ITU before 11am. This was around ten hours earlier than the time he was admitted at 9.11pm and indicates failings in the actions of the Trust.
59. Mrs M is concerned about a delay on 7 January in Mr M receiving an X-ray. As noted above, a chest X-ray was requested in the morning of 7 January, but this did not take place until around 6pm after it had been noticed it had not taken place. There is no specific guidance which specifies how quickly a chest X-ray should take place after it has been requested for acutely ill patients in hospital. However, it was clear when the request was made at around 10am on 7 January that Mr M was deteriorating and therefore the results of the X-ray were to be used to help find the cause of the deterioration.
60. GMC Good medical practice guidance states that:
‘If you assess, diagnose or treat patients you must: promptly provide or arrange suitable advice, investigations or treatment where necessary.’
61. When the X-ray had not taken place, a portable X-ray was requested around 3.30pm on 7 January with the X-ray finally being completed at 6pm. A delay of eight hours for a deteriorating patient was not in line with the GMC guidance specifying prompt investigations, and indicates failings in the Trust’s actions.
Impact of the failings identified
62. Mrs M has explained that she believes with more timely treatment, Mr M would have survived and has said she was given no indication that he was likely to die. This must be so upsetting and difficult for Mrs M. Taking into account the failings we have identified, and the treatment Mr M received and could have received, we will now consider what impact different actions by the Trust may have had for Mr M.
63. As explained above, from the evidence we have seen we consider there were failings in the management of Mr M’s breathlessness as the possibility of heart failure or fluid overload as the cause of the breathlessness should have been considered based on Mr M’s presenting symptoms.
64. Heart failure itself was not indicated on Mr M’s coroners report but the possibility of heart failure and fluid overload does not seem to have been considered until a medical review on 6 January when a raised JVP (jugular venous pressure - an indirect indicator of central venous pressure and fluid status) was noted and Mr M was given frusemide (medication to treat fluid retention). Up until this time point Mr M was being given IV fluids and blood products regularly. His fluid balance charts continuously showed that he was in a positive balance and his weight was recorded as having increased from 101.4 kg on admission to 106.5 kg on 7 January.
65. Had different tests and/or assessments taken place earlier it is possible that his breathlessness could have improved and medication for fluid overload could have been given earlier. Had fluids been stopped when indicated on 3 January this also had the potential to improve Mr M’s breathlessness.
66. As explained, there were also times when supplementary oxygen was indicated by the relevant guidance, but not provided. Whilst Mr M’s oxygen levels fluctuated over the days he was in hospital, he should have been given oxygen when his saturations dipped to 94% on 3 January, again potentially improving his breathlessness.
67. When Mr M deteriorated, we have identified failings related to an X-ray delay, delays contacting the critical care team and in completing sepsis investigations.
68. When the X-way was completed at around 6pm on 7 January, it was reported as being abnormal with airspace opacities suggestive of either infection, pulmonary oedema, or ARDS.
69. It is difficult to be clear about the impact of the delayed X-ray, but had these results been available sooner it may have prompted an earlier referral to the critical care team, which in turn could have led to an earlier referral and/or admission to the ITU.
70. Mr M’s high NEWS2 scores should have led to escalation to the critical care team. Had the relevant guidance been followed and calls been made to the critical care team, earlier referral and/or admission to the ITU may have happened.
71. Had Mr M been admitted to the ITU earlier our critical care adviser has explained that the focus of his therapy would have been to support any organ failures and confirm a diagnosis to ensure appropriate treatment was being administered.
72. In Mr M’s case, there was a strong suspicion of infection. Our critical care adviser has explained that a normal chest X-ray does not preclude a diagnosis of respiratory infection in patients such as Mr M, and wherever possible a CT scan would be done to help with making the diagnosis. It may not be possible to do a CT scan if the patient has a high oxygen requirement and is unable to lie flat. On those occasions a scan would be performed at the next available opportunity. For example, if the patient deteriorated and required mechanical ventilation, then the CT scan would be considered after that point if the patient was stable. It was not possible to do a CT scan after Mr M’s ICU admission, but had he been admitted earlier, it is possible the CT scan could have been performed as he may have been more stable, potentially leading to a diagnosis for his deteriorating condition.
73. The delay in referral and admission to the ITU also meant there was no time to discuss treatment escalation plans because of Mr M’s deterioration. Our critical care adviser has explained that Mr M’s AML, neutropenia (low white blood cells) and chemotherapy meant he was at high risk for an infective complication. Our critical care adviser went on to say that treatment escalation plans should be considered before a patient deteriorates to allow consideration of the person’s wishes and desired outcomes.
74. In this case this would normally be done between the haematology team and Mr M (and his family if desired) and involve the critical care team for its opinion. There is evidence the haematology team considered the possibility of deterioration and exploration of Mr M’s values at the point of admission as a RESPECT (recommended summary plan for emergency care and treatment) form was completed on 31 December, setting out Mr M’s preferences.
75. Relevant guidance here is the Faculty of Intensive Care Medicine’s guidance, Care at the end of life, which says:
‘All staff should be open to such discussions and be able to respond to initial questions. Questions beyond their knowledge should trigger the involvement of experienced colleagues.’
76. As a result of the delays in referral and admission to ICU there was no involvement from the critical care team about Mr M’s clinical deterioration and treatment escalation. We consider this was a missed opportunity to explore treatment wishes and values with Mr M and his family in the context of an uncertain diagnosis and the fact that intensive care treatments are more burdensome and associated with increased morbidity and mortality than ward treatments.
77. Had the earlier ITU referral been made there was potential for a discussion during the day of 6 January between a consultant haematologist and a consultant in intensive care medicine. This may have enabled a clear plan regarding admission to intensive care and what treatments would be offered. This in turn removed the opportunity for Mr M’s family to be clearer about his condition and what might happen next.
78. On admission to the ITU, CPAP was attempted but Mr M could not tolerate the treatment after a period of ten minutes. High flow nasal oxygen was also tried as an alternative method of improving Mr M’s oxygen levels, although our critical care adviser has explained, this does not provide the same level of patient benefit as CPAP. At this point Mr M was intubated.
79. Our critical care adviser has explained that CPAP is a rapidly beneficial therapy, particularly if there had been any element of fluid overload or heart failure contributing to shortness of breath. By providing a positive intrathoracic pressure, the work of breathing is reduced, cardiac function can be improved and fluid overload on the lung can be transferred back into the circulation and subsequently removed. This can also improve blood oxygen levels. The benefits of CPAP can be achieved in some people within a few hours and by 24 hours of therapy, it is possible to see if a person is responding to CPAP.
80. Potentially one of the most significant benefits of earlier ICU admission would have been the ability to start CPAP therapy for Mr M earlier. Although he was intolerant of the CPAP within a short period after admission, it is possible that had he been admitted earlier he may have been able to tolerate the CPAP at that point and been able to obtain benefit from it.
81. Had there been earlier sepsis tests completed, as indicated by the relevant guidance, Mr M could have potentially received different treatment.
Our balance of probabilities view
82. It is not possible for us to say with any certainty what may have happened if Mr M had received different and/or more timely treatment. This is where we must take a balance of probabilities view as to what was more likely than not to have happened.
83. As outlined in the section above, there are many unknowns in this case as we cannot know what impact different or more timely treatment may have had or whether Mr M would have reacted positively to this treatment.
84. Mr M was undoubtably very unwell and his diagnosis of AML meant that his chemotherapy was started as soon as possible. Because of Mr M’s condition and his chemotherapy, he was at high risk of a number of complications.
85. Our haematology adviser has explained that Mr M had a sub-type of AML that is generally associated with a low risk of relapse and a good outcome once remission has been achieved. Recent evidence (Kantarjian, 2021) suggests that the five-year survival rates is 80% in patients with this type of AML and treated with chemotherapy and gemtuzamab (although this paper post-dates the treatment here, it is relevant here to demonstrate potential survival rates).
86. However, Mr M had only just started his chemotherapy when he became critically unwell and was three to four weeks away from achieving remission, and our haematology adviser therefore commented that his longer-term prospects for survival were significantly reduced.
87. Even with timely management a recent Frontiers in Haematology paper shows that in a study around one third of patients with similar presenting conditions as Mr M died when admitted to the ICU (although this paper post-dates the treatment, it is relevant here to demonstrate potential survival rates).
88. A study in the Journal of Clinical Oncology quotes the requirement for invasive ventilation, as Mr M required, as a poor prognostic factor with a mortality rate of 60%. A study in the Blood Science Journal found that induction chemotherapy, as Mr M was receiving, was predictive of non-survival if patients were admitted to the ICU.
89. Our critical care adviser has commented that in view of the early stage of Mr M’s AML diagnosis and his chemotherapy treatment, combined with the speed of his deterioration, offering invasive ventilation earlier would not have had any impact on Mr M’s overall chances of survival.
90. The ability to start CPAP earlier was a potential missed opportunity to prevent deterioration, and possibly the need for invasive ventilation, and offer Mr M a better chance of survival. Even accepting that if he continued to deteriorate on CPAP then his chances of survival were very low.
91. Our haematology adviser commented that even if Mr M had received optimal treatment for his problems there is a significant possibility that he would have died from these issues at a similar time or soon afterwards.
92. Despite Mr M being very unwell, there were a number of missed opportunities to do things differently. The identified failings have shown a number of possibilities for improvements in Mr M’s condition if different or more timely actions had been taken.
93. We have identified areas where treatment was not optimal, but we have not seen any firm evidence or guidance which suggests that optimal treatment would have been more likely than not to have changed the sad outcome. It is possible that different or earlier treatment may have had an impact, but it is also possible that Mr M would not have been able to tolerate the treatment, or it may not have made any positive difference. With so many variables and the overarching seriousness of Mr M’s condition, we are not able to conclude that different treatment would, on the balance of probabilities, have meant that Mr M would have survived his hospital admission in early 2021.
94. This is a finely balanced decision and one we have not taken lightly as we understand the distress and upset these issues have caused Mrs M.
95. With regard to the earlier involvement of the critical care team, we can be clear that this should have happened earlier. The outcomes of possible treatment as a result of earlier involvement are included in our balance of probabilities view above, but earlier involvement would have led to better information and involvement for Mr M’s family.
96. Mrs M has explained how she was not suitably informed of Mr M’s deterioration or the potential that he may die. Had Mr M’s deterioration been suitably acted upon, this would have presented the opportunity to involve Mrs M and provide her with the opportunity to be with Mr M. This opportunity was taken away from Mr M’s family and it is clear from Mrs M’s correspondence this has caused her great distress.