23. The evidence we have considered leads us to the view that there were aspects of Mr I’s care which were not managed appropriately and in line with the standards set out above. In addition to this, there is a lack of thorough documentation in the medical records which makes it challenging to reach a view on the balance of probabilities as to what happened at certain points in the care and treatment given.
24. The National Clinical Guidelines for Stroke (2016, 3.1.1 A, B) tell us that patients seen by ambulance clinicians outside hospital with a sudden onset of focal neurological symptoms should be screened for stroke or Transient Ischaemic Attack (TIA) using a validated tool, and those people with persisting neurological symptoms who screen positive should be transferred to a hyperacute stroke unit as soon as possible. Patients who are negative when screened but in whom stroke is still suspected should be treated as if they have had a stroke until the diagnosis has been excluded by a specialist stroke clinician.
25. NG128 advises that for patients who are admitted to the ED with a suspected stroke or TIA, clinicians should establish the diagnosis rapidly using a validated tool, such as ROSIER (Recognition of Stroke in the Emergency Room) (1.1.3).
26. As we can see from the background of events, Mr I was identified as being FAST positive by the ambulance clinicians. This was again noted during the initial ED triage. The advice we have received confirms that there was sufficient evidence upon Mr I’s arrival to the ED to regard his presentation primarily as an acute stroke. As such, the stroke team should have been alerted at the time of his arrival in the ED and Mr I should have been transferred to the stroke unit as soon as possible. This did not take place and so we consider this is a failing. This is because although Mr I arrived at hospital and was triaged just after 6pm, a referral was not made to the stroke team until 9:10pm. We cannot see they were alerted to his presence at the time of arrival. Additionally, despite the referral taking place at 9:10pm, Mr I was not transferred until the next day, and this was to the EAU instead of the stroke unit.
27. The National Clinical Guidelines for Stroke (2016, 3.4.1 A, B) advise that patients with suspected acute stroke should be admitted directly to a hyperacute stroke unit and be assessed for emergency stroke treatments by a specialist physician without delay. They should receive brain imaging urgently and at most within one hour of arrival at hospital.
28. NG128 advises that brain imaging should be performed immediately with a non-enhanced CT scan for those with suspected acute stroke where there are indications for thrombolysis or thrombectomy, indications of a depressed level of consciousness, for example where the Glasgow Coma Score (GCS) is below 13, and where there are unexplained progressive or fluctuating symptoms (1.3.2). Thrombolysis is a treatment to dissolve blood clots in blood vessels, and thrombectomy is the surgical removal of a blood clot in an artery. This is particularly important, as if provided swiftly and within a certain window of time, thrombolysis or thrombectomy can limit the long-term physical harm caused by stroke.
29. The National Clinical Guidelines for Stroke (2016, 3.4.1 D) tell us that patients with ischaemic stroke who are eligible for endovascular therapy, i.e. accessing arteries and veins with catheters, should have a CT angiogram from aortic arch (part of the body’s main artery) to skull vertex (the top of the head) immediately. A CT angiogram is where contrast materials are inserted into the blood vessels so that clinicians can visually identify any problems. However, this should not delay the administration of intravenous thrombolysis.
30. We have considered the above guidelines and reviewed the ED triage records. However, there is no documentation of what discussions or actions were held with the stroke nurse or team to address the suspicion of stroke urgently.
31. Mr I’s ED records show us that his consciousness was fluctuating on the GCS between 13 and 14. In line with this guidance, it was appropriate for the ED team to request a CT head scan and CT spine scan at 6:03pm. We can see that this was performed promptly and within one hour of Mr I’s arrival at hospital. The CT scan was reported on at 6:57pm. The guidelines suggest that scanning should be performed immediately, and as they refer to immediate scanning the results should also be reviewed without delay. The results were not reviewed until 8:35pm, which is a considerable delay and so significant as to be a failing in our view. At that time, the results did not show any evidence of a stroke or haemorrhage.
32. Despite this, Mr I should have still been treated as having had a stroke as per the National Clinical Guidelines for Stroke quoted above (2016, 3.1.1 B), as the diagnosis had not been excluded by a specialist stroke clinician.
33. We understand that Mr I should have been reassessed following the results of the CT scan and, upon review, a focused clinical history and examination should have been performed and recorded. This would have assisted our understanding of the events and would have allowed us to more fully determine what would likely have happened next if the stroke team had been called. In our view, lack of reassessment and examination following the CT scan is a failing.
34. The National Clinical Guidelines for Stroke (2016, 3.5.1 A, B) tell us that patients with acute ischaemic stroke, in whom treatment can be started within three to four and a half hours of known onset, should be considered for treatment with alteplase (a drug administered as part of the thrombolysis process which dissolves the blood clot.. The NIHSS (National Institutes of Health Stroke Scale) score should be calculated to determine if patients can be considered for combination intravenous thrombolysis and intra-arterial clot extraction. Further to this, patients with acute ischaemic stroke should be given 300mg aspirin as soon as possible but certainly within 24 hours, unless contraindicated (NG128 1.4.9). We have not seen that any NIHSS score was calculated or that aspirin was prescribed at this time.
35. It is possible that at this time, Mr I may have been a candidate for intravenous thrombolysis with alteplase. This is a time critical treatment with a maximum time of 4½ hours, at which point it will no longer be effective (NG128 1.4.1). We do not know if such treatment options were considered, as there is a lack of documentation of Mr I’s examinations and neurological condition. However, the fact that there was a delay in reviewing the CT scan results also meant that time was lost to consider critical treatment options. The Trust has explained that there were significant pressures and high demand within the ED on this day, and we understand this likely contributed to some delay in diagnosis. However, we are not satisfied that this accounts for all of the delay due to the lack of reassessment and examination following the CT scan.
36. We understand that if Mr I’s NIHSS was 4 or more but less than 25, he should have been offered thrombolysis. From the records we have, we have been unable to determine if this was indicated. If it was, thrombolysis could and should have been offered by 8:50pm at the latest. This would have been 4 ½ hours after the onset, which we understand was between 3:20 and 4:20pm. We are aware that the efficacy of IV thrombolysis is time critical but is also dependent on age and NIHSS score. The NIHSS score was not measured and because the records are poorly completed it cannot now be calculated retrospectively.
37. If his NIHSS had been calculated and found to be less than 4 or improving, thrombolysis would not have been offered due to the risk of internal bleeding. In this situation NG128 recommends immediately starting aspirin 300mg daily. We understand that in a review at 8:52am, the doctor did not find a focal neurological deficit (though it is noted Mr I was drowsy and the reason for this was not explored further). A focal neurological deficit is a problem with nerve, spinal cord, or brain function. Based on this review, it is more likely that the NIHSS could have been less than 4 or improving and if this was the case, 300mg of aspirin should have been given in the evening of 4 February 2019 to aid in the prevention of stroke recurrence.
38. A study by Rothwell et al suggests it takes 24 hours for aspirin to effectively reduce the risk of stroke by more than 50%. As such, it is likely that if aspirin had been given between 7-8pm on 4 February 2019 it would not have prevented what appears to be a second onset of symptoms on 5 February 2019. Due to the poor record keeping, we cannot determine if Mr I suffered from a second stroke in the evening of 5 February 2019 as we do not have a clear timeline of his symptoms between the onset on 4 February 2019 and the review in the evening of 5 February 2019, the time of which has not been recorded in the records.
39. Although there was a review at 10:45pm on 4 February 2019, the neurological findings were not recorded. It is noted that Mr I was drowsy, and this is attributed to the administration of morphine. The dose of morphine was prescribed at 8:35pm. However, the records suggest to us that it was not administered until 11:10pm. Therefore, the conclusion drawn in this review does not appear to be supported by the medication charts.
40. In addition to this, based on the advice we have received, we consider the review at this time was inadequate. The doctor has not recorded any neurological examination and given the fact that Mr I was recorded as FAST positive in the triage records, there should have been a follow up of these symptoms. The doctor noted that the diagnosis was possibly concussion in reaction to the fall, and despite Mr I’s symptoms of stroke no alternative diagnosis was considered. The National Clinical Guidelines for Stroke (2016, 3.10.1) tell us that patients with acute stroke should have their clinical status monitored closely, including their level of consciousness, blood glucose, blood pressure, oxygen saturation, hydration and nutrition, temperature, and cardiac rhythm and rate.
41. In the assessment at 8:52am on 5 February 2019, there is no documented clinical examination other than comments on Mr I’s right shoulder pain and drowsiness. The diagnosis given at this time was musculoskeletal (MSK) injury following a fall. There are no indications that stroke was being considered as a potential diagnosis. It is not recorded that Mr I had any weakness on one side of his body. Based on the advice we have received, it would have been appropriate and in keeping with the guidelines mentioned previously to repeat the CT head scan at this time, as there was no alternative explanation proposed for Mr I’s drowsiness and left sided weakness on 4 February 2019.
42. This review appears to compound the failings of the review at 10:45pm on 4 February 2019. There appears to be a lack of consideration of the ED records and a failure to recognise Mr I’s previous presentation as FAST positive. No neurological examination was carried out, and it is unlikely that the morphine administered on 4 February 2019 contributed to the apparent drowsiness in the morning of 5 February 2019. Based on the advice we have received, it may be the case that a stroke had already taken place by this time. This is supported by the notes of the Therapy Assistant Practitioner who reviewed Mr I at 11:30am on 5 February 2019.
43. Mr I should have been admitted to the stroke unit for observation on 4 February 2019 as this would have ensured continuity of observation by nursing and medical staff who are familiar with stroke and the relevance of a change in the neurological observations, and it would likely have been much clearer when this drowsiness began.
44. Following the assessment at 8:52am on 5 February 2019, we understand Mr I should have been subject to hourly neurological observations in line with the National Clinical Guidelines for Stroke (2016, 3.10.1). We know he was drowsy, but his GCS score has not been recorded. We understand that the doctor should have been monitoring for other signs of drowsiness and carrying out further investigations into the alternative causes of drowsiness. In doing so, a further CT scan or MRI should have been arranged during the day and before the deterioration at around 7pm.
45. The medical records suggest a new onset of symptoms at 6:50pm during a nursing review, at which time Mr I was noted as being very drowsy and unable to move his left arm or leg. A later review, of which we do not know the time because it has not been recorded, did identify a focal neurological deficit and considered stroke as a possible diagnosis. It is unclear why this diagnosis could not have been made sooner on the same basis, and because of this we consider the Trust failed to suspect and diagnose the stroke at the time of admission and failed to monitor Mr I’s presentation between his admission and the evening of 5 February 2019.
46. At this time, if Mr I had no signs of focal neurological deficit, he could have been re-considered for thrombolysis. A CT scan should have been performed immediately but was not performed and reported on until 9:25pm. The records do not suggest to us that intravenous thrombolysis was considered at this time and there is no explanation of any discussion or consideration with the stroke team in the records.
47. The doctor’s notes from this review state that if no haemorrhage is present, 300mg of aspirin should be given. The patient transfer list from 9pm tells us that Mr I had been prescribed PR aspirin, but it had not been given because the ward he was on did not stock it. We can see that Mr I was reviewed again at 2:18am, at which time a definite stroke diagnosis was made. Aspirin was prescribed again at this time but was not administered until 7am, which is a considerable delay given Mr I’s diagnosis.
48. As the aspirin was administered at 7am when Mr I was on the stroke ward (C8), we can see that PR aspirin was available on this ward, despite being unavailable on AMU1. He had been transferred to this ward at 9pm on the previous evening and the transfer checklist indicated PR aspirin was required. Because aspirin is a known blood thinner, it would not have been appropriate to provide it until the results of the CT scan had been reviewed to rule out a haemorrhage (NG128 1.4.9). Had the CT scan and medical review occurred immediately after the scan was reported on, we consider aspirin could have been given much sooner. Both the National Clinical Guideline for Stoke and NG128 advise that aspirin should be given as soon as possible and we therefore consider there is a failing to provide aspirin as soon as it was clinically indicated.
49. Overall, we consider the Trust did miss opportunities to diagnose Mr I’s stroke and to put him on the most appropriate treatment pathway. It appears that this led to a delay in his diagnosis and a delay in providing appropriate care and medication. We will now consider the impact this had on Mr I and his family.
Impact
50. Following the admission Mr I required around the clock care. He resided in a care home and spent most of his day in bed. Mrs M says he could no longer attend to his own personal care and that as a consequence he lost his mobility and his dignity. Mrs M explains that this was very distressing and had a huge impact on his three children, who witnessed his deterioration.
51. We have considered if an urgent referral had been made to the stroke team at the time of admission, what difference this likely would have made to Mr I’s care and treatment pathway. Because the referral was not made, it is difficult to give any view on the balance of probabilities if this would have made a difference to the clinical outcome, as this assumes the diagnosis would have been made following the assessment. However, we can say the pathway would have been different.
52. If Mr I had been referred to and seen by the stroke team, it is likely a stroke would have been diagnosed and plans would have been made to transfer Mr I to the stroke unit directly and without delay in line with the National Clinical Guideline for Stroke 2016 (3.4.1). It is our understanding that this would have been the main difference to Mr I’s care. Having looked very carefully at the clinical evidence available to us and the independent advice we have received, we do not consider we can go so far as to say on the balance of probabilities that the delay of several hours in eventual admission to the stroke unit made a difference to the clinical outcome.
53. However, this did mean that Mr I lost the opportunity to be considered for intravenous thrombolysis on two occasions, although due to poor record keeping it is not possible to say if he would have been suitable for this treatment. In addition to this, aspirin could have been prescribed and given much sooner than it was although, again, it has not been possible to determine on the balance of probabilities what impact this had on the eventual clinical outcome, based on the evidence available to us.
54. We also explored if there were any other alternative treatments which should have been considered or provided, had a diagnosis been made in the ED. As set out earlier, for patients with ischaemic strokes who are unable to have thrombolysis, a small number can be treated by a thrombectomy. The procedure is only effective at treating ischaemic strokes caused by a blood clot in a large artery in the brain and is most effective when started as soon as possible, similar to the considerations for thrombolysis.
55. The decision whether to take forward thrombectomy depends in large part on the Trust either having the capabilities to perform it or having the means of referring cases to a centre capable of doing so. We understand this would have depended on the speed of stroke diagnosis and referral, as the procedure would have needed to start within five hours of stroke symptoms starting.
56. If the stroke diagnosis had been made earlier and a large clot in the brain identified in the CT scan, a discussion should have occurred with the relevant specialists or specialist centre. However, there is no guarantee this treatment would have gone ahead or been successful and we have been unable to reach a view on the balance of probabilities in part due to the lack of observations recorded in the medical records. We can only say that taking forward this discussion would have been in line with the National Clinical Guidelines for Stroke (2016 3.5.1), whatever the outcome of that discussion might have been. Given that the diagnosis was delayed, Mr I also lost the opportunity of being considered for the thrombectomy treatment.
57. We appreciate that this was a very difficult and challenging time for Mr I’s family as they saw the deterioration in his condition from the moment he suffered the stroke and through his admission and eventual discharge to a care home. Although we have been unable to give any view on the balance of probabilities as to how the delay impacted on Mr I’s clinical deterioration, it is important to consider the emotional impact of these events and of the mistakes we have identified. Due to the delay, the family will now never know if the outcome would likely have been different, had the diagnosis been made and treatment given sooner than it was. The uncertainty and distress it caused and continues to cause to the family is an injustice.
58. We also understand that on the 4 and 5 February 2019, Mrs M tells us she was given the impression that Mr I had not suffered a stroke and she that she was told that his condition was due to a bang to the head. Although we cannot reach any robust view as to what exactly Mrs M was told, from the documentation we have reviewed it does appear that a stroke was not being seriously considered as a diagnosis. Therefore, the communication of this message to the family is likely to have caused them serious concern and frustration. Had the team identified the symptoms and followed the relevant pathway, this would have likely given reassurance to the family that the staff were concerned about Mr I’s symptoms and were considering the possibility of a stroke with appropriate urgency.
59. In its complaint response, the Trust said that staff were concerned about Mr I’s symptoms and were still considering the possibility of stroke. However, it then went on to say that there was a delay in the two CT head scans because of a change in symptoms. It explained that upon his initial presentation to AMU, Mr I did not show signs of a stroke, which was in keeping with the CT scan carried out by staff in the ED. The Trust said that there is a lack of documentation relating to the history of stroke symptoms as the doctor did not feel that Mr I had suffered a stroke.
60. We consider that this is contradictory to the first statement made, and not supported by the clinical evidence available to us, as we can see from the documentation that Mr I presented as FAST positive and the stroke team had been alerted to his condition upon admission. As such, this response does not appear to be in line with the Local Authority Social Services and NHS Complaints Regulations 2009 (the Regulations) or our Principles of Good Complaint Handling. The Regulations say that complaints should be investigated in an appropriate manner, and that the response should include an explanation of how the complaint has been considered and the conclusions reached in relation to the complaint. Our Principles explain that public bodies should be open and honest when accounting for their decisions and actions and should give clear, evidence-based explanations and reasons for their decisions. Due to the contradictory nature of the letter, we do not consider the Trust’s response to be clear and evidence-based. Although Mr I’s symptoms were being monitored initially, it appears the lack of action after his admission caused a delay in providing treatment. We consider it is likely that this contributed to the distress faced by the family.
61. We recognise that the Trust has provided an apology for the speed with which the morphine was injected and advised this would be discussed with the nurse who administered the morphine so they could reflect on this in future practice. Although we have not identified that the morphine had an impact on the examination at 10:45pm on 4 February 2019, we consider it was in line with our Principles that the Trust acknowledged and apologised for this concern.
62. The Trust has also apologised that Mr I was not referred to the stroke team more urgently after receiving the CT head scan report and that the potential diagnosis and concerns about Mr I’s symptoms were not properly communicated to the family at the time. It has acknowledged that had this been explained to the family, it may have offered some assurance that staff were concerned about Mr I’s symptoms and were considering the possibility of a stroke.
63. Whilst the Trust has acknowledged some of its errors and the impact they caused, we do not consider it has recognised the full extent of the failings we have found or fully remedied the impact we have identified.