Call to 999 and ambulance response
20. Miss A complains that the first 999 call to the Trust was incorrectly categorised, resulting in a delayed ambulance response time.
21. According to the Principles of Emergency Medical Dispatch, which is the guide used by the Trust, EMDs must assess the clinical urgency of a 999 call depending on the presenting issue. This is done in the first instance during the ‘case entry’ phase of the call. This phase of the call is the point from which it is answered to when a protocol is chosen by the EMD. The information gathered by the EMD in this phase informs subsequent protocols they follow (and questions they might ask as indicated by the chosen protocol). It can also inform how the call should be categorised.
22. Ambulances are a finite resource that need to be allocated according to clinical need. The most urgent cases are categorised as Category 1, under the Ambulance Response Programme, and should be responded to within a maximum time of 15 minutes. There are four categories within the Ambulance Response Programme, with Category 1 being the most clinically urgent category and Category 4 being less urgent calls that can usually be dealt with over the telephone. Which category the EMD places a patient in is based on their assessment and so the EMD’s assessment informs how quickly an ambulance is likely to reach the patient.
23. The Principles of Emergency Medical Dispatch provide guidance when the problem reported in a call is about someone having breathing difficulties. The guidance indicates that if an ‘ineffective breathing’ descriptor is used during the ‘case entry’ phase of the call, the EMD should categorise the call as Category 1.
24. The Trust has given further information on what these ‘ineffective breathing’ descriptors are. These phrases (about breathing) include:
• barely breathing
• can't breathe
• fighting for air
• making funny noises
• just a little
• turning blue or purple.
25. If any of these phrases are used during the ‘case entry’ phase of the call the patient’s breathing should be recorded as ‘ineffective’. Where this is recorded, the Principles of Emergency Medical Dispatch say that a call should be categorised as Category 1, the highest clinical priority. The Trust’s operating system classifies a call as Category 1 when such information is inputted. Therefore, if Mr R’s family used any of these descriptors during the ‘case entry’ phase of the call, the EMD should have inputted ‘ineffective breathing’, and this should have resulted in Mr R being categorised as a Category 1 patient.
26. In the ‘case entry’ phase of the call that Mr R’s family made, the EMD first asked if the patient (Mr R) was breathing. Miss A responded and said ‘yes, but he is finding it very difficult’.
27. Thereafter, Miss A confirmed Mr R’s address to the EMD, and that she was with him. The EMD then asked Miss A to tell them what had happened. She explained how Mr R’s condition had got worse during the day. This included telling the EMD he was lightheaded and very hot. During this description, Miss A said to the EMD that he ‘can’t breathe’.
28. After this, the EMD confirmed Mr R’s age. They then started to ask questions indicated by protocol six of the Principles of Emergency Medical Dispatch. These are questions an EMD should ask to assess a patient’s breathing when the EMD records that they can breathe, rather than their breathing being ‘ineffective’.
29. This indicates that just after confirming Mr R’s age, the EMD selected the protocol to follow as they started asking the questions protocol six indicated. At this point, the ‘case entry’ phase of the call was over.
30. Before this, Miss A had used the descriptor ‘can’t breathe’. This was during the ‘case entry’ phase of the call. Therefore, the Principles of Emergency Medical Dispatch indicate that Mr R’s breathing should have been recorded as ‘ineffective’, and the call should have been categorised as Category 1. Unfortunately, this did not happen. The call was categorised as Category 2.
31. We note that three different statements were used to describe Mr R’s breathing during the ‘case entry’ phase. Right at the start of the call, when Miss A was asked if Mr R was breathing, she said ‘yes, but he is finding it very difficult’. 40 seconds into the call she then went on to say, ‘he is finding it really difficult to breathe’. It was 48 seconds into the call when she said, ‘he just can’t breathe’.
32. In a time-pressured situation, where different statements were given to the EMD, we can understand why they might have considered Mr R was breathing rather than selecting the ‘ineffective breathing’ descriptor. Despite the conflicting information, Miss A used an ‘ineffective breathing’ descriptor during the ‘case entry’ phase of the call. This should have been the overriding factor in recording whether Mr R was breathing, or if this was ‘ineffective’. The Trust has confirmed this in its comments on our provisional views.
33. We therefore consider that the information given to the EMD during the ‘case entry’ phase of the call should have resulted in Mr R’s breathing being recorded as ‘ineffective’. This meant he was not categorised as a Category 1 patient when he should have been. As this is not what the Principles of Emergency Medical Dispatch indicate should have happened, we consider this is a failing in service that the Trust provided.
34. Having identified this failing, we next considered what, if any, impact this might have had on Mr R. Miss A says this error led to an unnecessary delay in her father receiving the treatment he needed. She says the delay in receiving this treatment caused her father’s death.
35. We have found that, had the EMD correctly categorised the call, it is more likely than not that the call would have been coded as Category 1. The family contacted 999 a further three times, two of which provided a possible opportunity to rectify this error. However, as identified by the Trust, there were serious departures from policy and clinical standards, particularly in call three. This means the Trust appears to have missed further opportunities to review and recategorise the call before it eventually sent an ambulance.
36. As outlined above, the target response time for a Category 1 call is a maximum of 15 minutes. Evidence from the Trust shows that at the time of the first 999 call there were resources available to respond to the call about Mr R if it had been correctly categorised. The Trust has identified this in a retrospective review of the incident. There were two emergency ambulances that could have responded within one to seven minutes. There was also a rapid response vehicle that could have responded within four minutes. A rapid response vehicle is a smaller vehicle that can move quickly through traffic, staffed by a qualified paramedic.
37. The ambulances available at the time of the first call were showing as ‘restocking’ but ‘available’ and, according to the Trust, could have ‘reasonably been considered for allocation had the incident been correctly categorised’. Even if they had not been able to allocate these ambulances, there was a rapid response vehicle nearby that could have attended and the paramedic could have assessed the care Mr R needed. Therefore, we find that the failing we have identified led, on the balance of probabilities, to a delay in an ambulance attending to Mr R. It is more likely than not that, had the call been correctly categorised during the first call, allowing for the 15 minute maximum target, the ambulance would have arrived by 5.38pm at the latest, some 2 hours and 51 minutes sooner than it actually did.
38. We considered what would likely have happened next, if the ambulance had arrived by 5.38pm. It is difficult to know exactly how long the crew would have been on scene had this happened. Guidance from the Resuscitation Council (UK) advises a maximum on scene time of 20 minutes. The crew that attended the scene that evening were there longer than this due to Mr R having a cardiac arrest when they arrived. However, this was not a factor at the time of the first 999 call and so we can see no reason why the ambulance would have been at the scene longer than 20 minutes had it arrived sooner.
39. We consider that it is more likely than not that Mr R would have been conveyed to hospital within 20 minutes of the crew arriving, and therefore departed for hospital by 5.58pm at the latest. The journey to the hospital would likely have taken approximately 10 minutes as the address is less than two miles away. Mr R would, therefore, more likely than not have arrived at hospital by 6.08pm had the call been categorised correctly.
40. We obtained advice from our Cardiology Adviser to inform our thinking as to whether it is more likely than not Mr R could have lived if he had arrived at hospital and received treatment when he should have.
41. Mr R died from a cardiac arrest. This means there was a loss of blood flow around his body due to the heart being unable to function effectively. The pathologist could not formally establish the reason why Mr R’s heart was not functioning properly during the post-mortem. However, they did note there may have been a heart rhythm disturbance caused by scarring to the heart. Our Cardiology Adviser supports this view and explained that a cardiac rhythm disturbance was most likely responsible for the breathlessness Mr R experienced. We explain this in more detail below.
42. Our Cardiology Adviser explained that when the ambulance crew arrived, they found Mr R was in a state of pulseless electrical activity (PEA). This means there is electrical stimulation still present in the heart, but no effective output (no blood being circulated). There are several different reasons this can happen but most of these were excluded in the ED’s tests and during the post-mortem. The only cause not ruled out was profound acidosis.
43. The venous blood gas analysis taken in the ED when Mr R arrived reveals low pH of the blood. A normal reading for the blood’s pH level is quite narrow, being 7.35 - 7.45. Mr R’s was 7.1, which means he was in acidosis at the time he arrived at the ED. It also showed a significantly high lactate level of 11.9, with a normal level being up to 1.6. This is normally caused by poor oxygen supply to the body. According to our Cardiology Adviser, this shows Mr R’s cardiac function had been severely impaired for approximately six hours prior to his cardiac arrest.
44. There was no evidence of any coronary atheroma (furring of the arteries) or thrombosis (new clot formation). Further, there was no evidence his heart was enlarged, which suggests he had no chronic or longstanding impairment of cardiac function. Therefore, it is more likely than not his presenting issue was acute when his family called 999. (This means there had been no known longer-term issue with his heart.)
45. The only abnormal finding at post-mortem was a small piece of scar tissue on the left ventricle of the heart. Our Cardiology Adviser’s view is that this most likely caused Mr R’s acute heart rhythm disturbance. On balance, based on the medical evidence available, the most likely disturbance was ventricular tachycardia, which is a fast, abnormal heart rate that begins in the heart’s ventricles (lower chambers of the heart). This can become a life-threatening condition and commonly presents with breathlessness and dizziness, which were the symptoms Mr R was experiencing.
46. Mr R’s breathlessness was present for several hours on 2 January and, from what our Cardiology Adviser said, this was likely due to his declining cardiac output.
47. Our Cardiology Adviser explained that, on balance, if Mr R had arrived at the ED before 7.00pm it is more likely than not his cardiac arrest could have been avoided. Had the ambulance or rapid response vehicle arrived before then, the paramedic would, most likely, have performed an ECG. This would have established that Mr R had an abnormal heart rhythm at that time. In line with the estimated timescales outlined in paragraphs 38 and 39, he would then have been transported to hospital where the diagnosis would have been confirmed and either pharmacological or electrical interventions could have been used to restore his normal cardiac rhythm. Our Cardiology Adviser’s view is that had he reached hospital by 7pm, there would have been enough time to administer treatment to prevent the cardiac arrest that happened at 8.35pm. We shall go on to explain why.
48. When Mr R arrived at the ED, if his heart was pumping blood with a stable force, known as being haemodynamically stable, then his ventricular tachycardia could have been effectively treated with an intravenous injection, or infusion, of a drug called amiodarone which can restore normal heart rhythm.
49. If he had arrived at the ED with haemodynamic instability, meaning his blood was being pumped with an unstable force, this could have been treated with synchronised cardioversion. This is where targeted, low-level, sensor-guided electrical shocks are used instead of standard defibrillation.
50. Our Cardiology Adviser explained that as Mr R did not lose consciousness until after 8pm and because there is no evidence of any underlying heart condition, if his normal heart rhythm had been restored then, on balance, he would have survived if he had arrived at the ED prior to 7pm.
51. Instead, Mr R’s abnormal heart rhythm continued unabated for at least four hours. This led to a reduced cardiac function over this time, reducing blood flow to vital organs and causing acidosis. This eventually resulted in his cardiac arrest. Based on the evidence available, we consider that this could have been prevented had the ambulance arrived with him when it should have.
52. Therefore, Mr R’s death was most likely avoidable. This is the most serious injustice a person can suffer. His family were present during his clinical decline and had to witness him suffer and deteriorate whilst waiting for the delayed ambulance. This was compounded by making repeated calls to 999, in which the Trust’s staff made further errors and escalated the family’s distress. The family will also have to live with the terrible knowledge that Mr R’s death could have been avoided. This is a significant injustice to them.
53. Having seen evidence of failings resulting in injustice, we next considered what, if anything, the Trust has done to put things right and learn from this incident, and whether that is enough. We can see the Trust has taken this complaint seriously, having done a very thorough investigation, and it has put an action plan in place to learn from this issue. However, it has not accepted that Mr R’s death was avoidable.
54. Miss A wants service improvements at the Trust so that events like the ones she complains about do not happen again. She also wants the Trust to acknowledge the impact these events had. As the Trust has made changes to its service as a result of its investigation and action plan, we have considered if these changes are appropriate and whether there is anything further it needs to do to learn from what happened and improve.
55. Our Principles for Remedy say that part of a remedy may be to ensure that changes are made to policies, procedures, systems, staff training, or all of these, to ensure that the poor service is not repeated. It is important to ensure that lessons learnt are put into practice. We have considered if the Trust’s changes are in line with this.
56. As we explained earlier, despite being given information that indicated, according to the Principles of Emergency Dispatch, Mr R’s breathing should have been recorded as ‘ineffective’, the EMD did not do this. As the EMD inputted that he was breathing, this meant the case was categorised incorrectly. If the EMD had inputted ‘ineffective breathing’, the Trust’s system would have coded the call correctly at Category 1.
57. Therefore, what happened is attributed to an individual EMD making an error and inputting information different to what the guidance indicated. The Trust’s investigation identified that other EMDs had made the same mistake on 15 occasions between December 2017 and May 2018. The Trust therefore made changes to its operating system in response to this. These changes went live in June 2018.
58. On its operating system, the Trust has attached alerts for when an EMD asks the breathing related questions indicated by the Principles of Emergency Medical Dispatch. This includes the responses that indicate when ‘ineffective breathing’ should be recorded. This information is then processed by the operating system to appropriately categorise the call. The Trust has provided training to its EMD staff on this change.
59. We consider, as the inputting of ‘ineffective breathing’ would have resulted in the correct call categorisation, this is an improvement that minimises the chances of Mr R’s circumstances occurring again. Had the alert that the Trust has now put in its system been in place at the time his family contacted the Trust, we consider there would have been information available that would have prompted the EMD to input the appropriate information. The Trust’s system would then have categorised the call appropriately.
60. In our consideration of the Trust’s service improvements, we also looked at other similar complaints we have received about it. We did this to see if there are cases since June 2018 where similar problems have occurred and, therefore, if these changes have been effective and whether there is further action it should take.
61. We have not found cases where the same mistakes have happened. We saw that there were instances since June 2018 where EMDs asked the breathing related questions the Principles of Emergency Medical Dispatch indicate, and that the call was correctly categorised. This indicates that when presented with similar circumstances, the changes the Trust has made have resulted in the appropriate categorisation of similar calls.
62. Given that we can see evidence the changes have been effective, and the mistakes that happened in Mr R’s case appear not to have been repeated in other cases we have looked at, we do not consider there are further meaningful service improvements the Trust can make. While we recognise this cannot change what happened in Mr R’s case, we hope this provides some assurance to his family that the Trust has made changes so that others do not have the same experience. This is in line with our Principles for Remedy.
63. Although the Trust appears to have made appropriate service improvements, it has not yet taken action to acknowledge the impact to Mr R’s family that we consider the events caused.
64. Our Principles for Remedy say that where poor service has led to injustice or hardship, the organisation responsible should take steps to provide an appropriate and proportionate remedy. An acknowledgment, apology, explanation, or a combination of these things may be a sufficient and appropriate response.
65. Miss A seeks an acknowledgment of and an apology about the impact of the events she complains about. As the Trust has not yet acknowledged Mr R’s death was avoidable and apologised for this, we consider it should take action, in line with our Principles for Remedy, to do this. We have set out what recommendations we are making at the end of our report.
66. Miss A complains that the ambulance did not arrive with adrenaline, that the treatment with adrenaline was delayed, and that it was not administered within the appropriate timeframe.
67. There are no national guidelines for what items, such as medicines, should be kept within an ambulance vehicle. However, as adrenaline is an important component of advanced life support in adults, it seems reasonable to expect that this would be available to the paramedics who first attended the scene. We note the paramedics would likely not have been aware Mr R was about to suffer a cardiac arrest, however, this is something clinical staff should have access to in case cardiopulmonary resuscitation (CPR) is required.
68. The Trust interviewed its staff regarding this incident and both accounts are consistent that the basic response bag was carried into the house, alongside an oxygen tank. There is no evidence to inform our thinking as to whether there was adrenaline in this bag. The staff’s accounts are consistent that the bag was brought into the house and that intravascular (IV) access was attempted in order to give the adrenaline before the other ambulances arrived. The post-mortem report supports the assertion that IV access was attempted several times. Our Paramedic Adviser explained that it is not unusual for IV access to be difficult to achieve until support staff arrive.
69. We cannot know for sure whether there was adrenaline stored in the ambulance, or whether the staff brought adrenaline with them into the house. However, it is more likely than not that this was available to the paramedics in one way or another. The response bag was brought into the house and the paramedics were attempting to obtain IV access in order to administer adrenaline. This tells us it is more likely than not the crew had adrenaline and were attempting IV access to administer this prior to further resources arriving.
70. With regards to whether there was a delay in administering the adrenaline doses, guidance from the Resuscitation Council (UK) states that for a patient with a ‘non-shockable’ rhythm, which is a heart rhythm that cannot be corrected by the use of a defibrillator, the priority is the commencement of chest compressions. We know from the records that Mr R was in a state of PEA, which is not a shockable heart rhythm.
71. We also know that the paramedic had difficulty obtaining IV access to administer the drug. The priority was to continue chest compressions with minimal interruption as this can help to restore blood flow whilst the heart is not functioning properly. The staff appear to have correctly prioritised chest compressions over IV access to administer adrenaline, in line with the resuscitation guidelines.
72. When further support arrived, paramedics supporting the first crew were able to gain intraosseous access. This is where a drug is injected directly into the bone marrow. This is used when IV access is not obtainable, as in this case. We can understand why the family thought the first crew had not brought adrenaline with them. However, the evidence indicates this was not the case. The support crew was able to give adrenaline because they did not need to prioritise chest compressions and could administer the drug into the bone whilst other staff continued the chest compressions. Therefore, on balance, whilst there may have been an initial delay in administering adrenaline, this was more likely than not due to poor IV access. We consider the ambulance staff managed the initial administration of adrenaline in line with the applicable national guidance.
73. The family are also concerned there were delays in the time taken to administer the repeated doses of adrenaline to Mr R after the support crew arrived.
74. Guidelines from the Joint Royal Colleges Ambulance Liaison Committee (JRCALC) state that adrenaline should be administered every three to five minutes during resuscitation. We can see this happened, with the exception of the fourth, seventh and eighth dose. Paramedics gave these doses six, two and eight minutes later than the guidelines, respectively.
75. The evidence shows that paramedics gave the seventh dose around the time the crew began to move Mr R to transport him to hospital. Therefore, the delay with this dose was likely unavoidable.
76. Whilst there was a delay in giving the other two adrenaline doses, most of the doses were given within the timeframe set out in the guidelines. Therefore, we consider that the paramedics’ actions did not fall so far short of the guidance that they amount to a failing in how they gave the adrenaline. The staff were clearly trying to manage Mr R’s resuscitation as best they could, and we acknowledge this was a clinically complex situation.
77. This experience was naturally very distressing for Mr R’s family but also for the ambulance staff who responded. This was a complex medical emergency and, from the evidence we have seen, we are satisfied that the staff’s actions were, on the whole, in line with what should have happened regarding the adrenaline.
Defibrillator and resuscitation
78. Miss A complains that the crew did not come in with a defibrillator or begin resuscitation soon enough. She adds that her father was not transferred to hospital in a timely manner by the paramedics.
79. The Trust’s policy says the paramedics should have carried the defibrillator to the house in the first instance, as the paramedics were aware Mr R was having breathing difficulties, even though the call was a Category 2 call. This is reflected in the EMT1’s witness account of the incident, which states ‘I recall the incident was passed to us as a difficulty breathing’. Staff did not bring the defibrillator with them into the house in the first instance. We consider this to fall so far short of what should have happened that it amounted to a failing.
80. Again, we next consider what, if any, impact this might have had on Mr R. The fact the ambulance crew did not bring the defibrillator/heart monitor in the first instance led to a delay in staff identifying the need for CPR to commence. Consequently, there was a delay in staff starting resuscitation attempts.
81. Our Cardiology Adviser explained that based on the medical evidence from Mr R’s hospital records and post-mortem report, by the time the ambulance arrived at 8.29pm, regrettably, Mr R’s condition was already too poor for the staff to correct.
82. The survival rate for an out of hospital cardiac arrest, which is what Mr R had, is approximately 8.6%. The prognosis is better when a patient has a shockable heart rhythm, which he did not have. Our Cardiology Adviser explained that the outcome for cardiac arrest following PEA is poor, irrespective of the treatment the ambulance staff could have given at this point. Information from the Resuscitation Council (UK) states that ‘survival following cardiac arrest with asystole or PEA is unlikely unless a reversible cause can be found and treated effectively’.
83. As the outlook was, sadly, very poor for Mr R by the time the crew arrived, it is highly unlikely that, had the crew brought the defibrillator into the house in the first instance, this would have prevented his death. However, the fact this did not happen caused worry to the family, at an already incredibly distressing time, and we can understand why they felt this may have had an impact in his survival. However, based on the medical evidence we have seen, this error had no impact on the outcome for Mr R.
Transfer to hospital
84. The Trust explained that it advises paramedics to continue resuscitation for 20 minutes at the scene before transporting the patient to hospital. This is in line with guidance from the Resuscitation Council (UK).
85. The ambulance was on scene for longer than 20 minutes. However, there is nothing about when to stop treating a cardiac arrest where PEA is present in the JRCALC’s Ambulance Service Clinical Practice Guidelines. Resuscitation Council (UK)’s guidelines also state that deciding when to stop resuscitation attempts for PEA can be very difficult to determine.
86. Transferring Mr R would have interrupted resuscitation efforts, such as chest compressions, and we consider that it was reasonable for staff to have prioritised the resuscitation attempts. Therefore, the fact the ambulance was on scene for longer than 20 minutes does not fall so far short of what should have happened that it amounts to a failing in the care provided to Mr R. However, we do acknowledge, understandably, that the prolonged time at the scene was very distressing for Mr R’s family.