DNACPR decision
21. Mrs A and Mrs O say doctors implemented a DNACPR decision on 28 September, against the wishes of the family and without consultation with them. They say this caused them immense distress.
22. The Trust say clinicians had a discussion with Mr I about resuscitation on 28 September and they assessed he was competent to make his own decisions. It says Mr I was supported by a doctor and a nurse during this conversation, and he agreed with the DNACPR decision. The Trust apologise for the lack of communication with Mr I’s family about this.
23. BMA and RCN joint guidance on decisions relating to cardiopulmonary resuscitation acknowledge that clinical decisions on whether to attempt CPR can raise sensitive and potentially distressing issues for patients and their loved ones. CPR is most successful for otherwise healthy patients, following an acute coronary event (reduced blood flow to the heart).
24. In other cases, including chronically unwell patients, the probability of success is much lower. Successful CPR comes with significant risks. If unsuccessful, the clinician also risks subjecting a patient to traumatic treatment in their final hours and depriving them of a dignified death.
25. GMC guidance on cardiopulmonary resuscitation (CPR) says:
‘As with other treatments, decisions about whether CPR should be attempted must be based on the circumstances and wishes of the individual patient. This may involve discussions with the patient or with those close to them, or both, as well as members of the healthcare team…
If CPR may be successful in restarting a patient’s heart and breathing and restoring circulation, the benefits of prolonging life must be weighed against the potential burdens and risks. But this is not solely a clinical decision. You should offer the patient opportunities to discuss (with support if they need it) whether CPR should be attempted in the circumstances that may surround a future cardiac or respiratory arrest. You must approach this sensitively and should not force a discussion or information onto the patient if they do not want it. However, if they are prepared to talk about it, you must provide them with accurate information about the burdens and risks of CPR interventions, including the likely clinical and other outcomes if CPR is successful’.
26. The involvement of family members in discussions about DNAR is not mandatory. Our adviser says it is good practice to tell the next of kin that that a DNACPR decision has been made. Mrs A and Mrs O say a doctor informed them about the DNACPR decision after it had been made.
27. We consider the Trust’s doctors acted in line with GMC guidance. On 28 September, the completing doctor noted CPR would be inappropriate due to Mr I’s advanced COPD. Our adviser confirmed that CPR would not have been successful in Mr I’s case due to his underlying health conditions.
28. The records show the doctor explained the reasons for this decision to Mr I. We also note that Trust doctors had completed DNACPR forms for Mr I during previous admissions (on 10 June and 1 July) for the same reason, and the records show they had explained these decisions to him. As such, we consider it likely Mr I would have understood the reasons for the 28 September decision.
29. We recognise it was very distressing for Mrs A and Mrs O not to be involved in the DNACPR decision making process and we are sorry to hear about this. We can see the Trust acted in line with the GMC and the joint BMA and RCN guidance by discussing the DNACPR with Mr I. While we acknowledge Mr I’s family disagree with how the Trust approached this, we cannot see any evidence it got something wrong. On this basis, we do not uphold this part of the complaint.
Treatment for infection
30. Mrs A and Mrs O say the Trust did not properly treat their father’s infection. They say this contributed to his death.
31. NICE guidelines on sepsis make recommendations for adults who have suspected sepsis, according to the risk classification criteria. The risk tool looks at the patients’ history, respiratory rate, blood pressure, circulation, hydration, temperature, and skin.
32. Patients who are identified as at high risk of sepsis (including those with a raised breathing rate) should have an immediate review by a senior clinician. The guidance also says the patient should have venous blood tests and a broad-spectrum antimicrobial (such as antibiotics) within one hour.
33. Mr I’s chest X-ray on admission did not show pneumonia and the C-reactive protein, a blood marker of inflammation, was only mildly elevated. The admitting doctor decided that antibiotics were not indicated at this point.
34. A doctor reviewed Mr I at 3:30pm on 28 September. They noted Mr I looked unwell and was sweaty with a raised respiratory rate. Examination of his lungs revealed crackles at the left lung base. The doctor noted their clinical impression was sepsis. They took blood, urine, and sputum samples, requested a chest X-ray, and started Mr I on intravenous co-amoxiclav and gentamicin (antibiotics).
35. During the ward round the following morning, doctors prescribed Mr I with a dose of an intravenous furosemide (a diuretic to reduce fluid on the lung) and morphine, as required. They noted Mr I’s care did not need escalating but it’s not clear from the notes on what basis this conclusion was reached.
36. At 12pm, a member of the critical care outreach team (COOT) reviewed Mr I. They noted the furosemide had a good response but did not explain what this meant and whether he was clinically better. They recommended reducing the oxygen treatment to aim for oxygen saturations of 88-92%.
37. A doctor identified Mr I’s gentamicin level was raised during a review on 29 September. They decided to withhold this and to recheck the gentamicin level the following day.
38. Mr I had arterial blood gas samples taken in the evening on 28 September. The first sample demonstrated low oxygen and normal carbon dioxide. The second sample demonstrated normal oxygen and carbon dioxide levels.
39. The BTS guidelines on oxygen use in emergency care state that the target oxygen saturation for patients with normal blood CO2 is 94-98%. Our adviser says Mr I’s oxygen treatment was titrated (continuously measured and the balance adjusted) appropriately to maintain oxygen saturations within the target range of 94-98%.
40. From the records, we can see doctors’ appropriately identified Mr I as high risk for sepsis on 28 September. They arranged tests and prescribed treatment in the form of broad-spectrum antibiotics. This was in line with the NICE guidance on the management of sepsis, as set out above. We have not seen any evidence to suggest the Trust failed to treat Mr I’s infection appropriately.
Call buzzer
41. Mrs A and Mrs O say they are concerned nursing staff ignored the buzzer when their father was asking for help in the hours before his death. They say when they arrived at the hospital, after their father had died, a nurse told them Mr I had pressed the buzzer twice that evening. They said the first time was because he had spilled a drink, but they could not remember the reason for the second buzzer.
42. The Trust say the nurse who was looking after Mr I on the night of 29 September recalls checking on him every 10 to 15 minutes to make sure he was settled. It says there was a medical emergency on an adjacent ward which staff attended around midnight. It acknowledges there may have been a slight delay in call buzzers being answered during this time. When the nurses returned, they found Mr I’s breathing had become more laboured, and he was becoming unresponsive.
43. The Trust acknowledge the nursing team may not have responded to Mr I’s buzzer in a timely manner. It says between 25 September and 1 October, there were 983 nurse calls and 93% of these were answered in less than five minutes, meaning 7% were not.
44. The Trust acknowledge the overall percentage of calls not answered in five minutes is not acceptable. In its second response, dated 22 May 2019, the Trust say it is unable to retrieve any further data about Mr I’s call buzzer. We therefore do not know when and how many times Mr I pressed the buzzer during the evening of 29 September and if/ when the nurses responded. There is no mention in the nursing notes if Mr I used his call buzzer to request help on the evening he died. We do not consider witness testimony from the staff would be useful at this point, given the low likelihood of staff remembering the required level of detail in routine tasks after a period of time.
45. We appreciate not knowing this information is very distressing for Mrs A and Mrs O. Due to the lack of available data, we are not able to say whether the Trust did respond to Mr I’s call buzzer in a timely manner on 29 September. We cannot therefore say there is an indication something has gone wrong here.
Observations
46. Mrs A and Mrs O say staff did not undertake sufficient observations of Mr I on 29 and 30 September.
47. The Royal College of Physicians (RCP) National Early Warning Score (NEWS) is used to determine how unwell a patient is. It based on a scoring system. A score is allocated to the following physiological measurements:
1) respiration rate (number of breaths per minute)
2) oxygen saturation
3) systolic blood pressure
4) pulse rate
5) level of consciousness or new confusion
6) temperature
48. Each of the six physiological measurements are allocated a score according to the extent of disturbance to each one. The individual scores are then added up, along with a score of 2 for use of supplemental oxygen, to create the total ‘NEW’ score for the patient.
49. If a patient has a NEW score of seven or more, this puts them in the ‘emergency response threshold’. For this, the RCP guidance recommends:
· continuous monitoring of vital signs
· immediately informing the medical team
· emergency assessment by the critical care team
· consideration of transfer to a high dependency unit (HDU) or intensive care unit (ICU).
50. Mr I’s observation charts show he had NEW scores of seven or more on multiple occasions on 26 September up to the time of his death. There is no evidence in the notes that Mr I had continuous monitoring of his vital signs. Staff did not consider transferring him to a higher dependency unit until 29 September.
51. After midday on 29 September, the records show Mr I’s observations were taken only three times at 1:10pm, 5:05pm and 8:20pm. There are no other recorded observations after this. The timing and frequency of the observations was not in line with RCP guidelines which, as above, says there should be continuous monitoring of vital signs for patients with NEWS above seven and a consideration of transfer to an HDU.
52. On 29 September Mr I was diagnosed with acute pulmonary oedema (fluid on the lungs) due to heart failure. He was given an intravenous dose of furosemide (a diuretic) at 8.15am and the review at 12.00pm documents that this was given with good response. Subsequent observations showed Mr I’s oxygen requirements went back up to 80%. This is near the maximum that can be given and is sign that the air and oxygen are not able to get through lungs. This may have been due to recurrence of the pulmonary oedema, which our adviser says could have been relieved by further doses of intravenous diuretic.
53. Our clinical adviser says there are no specific clinical guidelines that can be applied to Mr I’s treatment on 29 and 30 September. On the basis that Mr I had clinical signs of sepsis and fluid on the lungs, it was appropriate to continue with antibiotics and to give the furosemide diuretic treatment.
54. Diuretics are medications that increase water excretion from the kidneys. They are indicated for acute pulmonary oedema (water on the lungs, usually due to heart failure). Giving a diuretic makes the water leave the lung tissue and go back into the blood then out in the urine. This allows air and oxygen back into the lung tissue which helps improve blood oxygen levels and relieve breathlessness.
55. The Trust did not carry out observations for Mr I in line with the guidance. Had the Trust taken Mr I’s observations more frequently and transferred him to a high dependency unit, our adviser says it is likely his condition would have been more intensively monitored and treated. Specifically, additional diuretic may have been given to treat the fluid on his lungs. Our adviser says the Trust should have reassessed Mr I during the day on 29 September as to whether further doses of the diuretic were required, which they did not do. We will consider the impact of this below.
Regular Medications
56. Mrs A and Mrs O say staff failed to give medications to Mr I in the hours before his death, despite the Trust saying they did. They say this made them lose confidence in the Trust’s complaint handling.
57. They Trust says Mr I was administered oral morphine for pain at 8:14pm and then had his regular medications from 10:24pm to 10:39pm.
58. We can see from the electronic medication records, there is evidence staff gave Mr I his regular medications during the evening of 29 September. This correlates with the information the Trust gave in its response about the timings of the medication. We cannot see evidence the Trust failed to give Mr I his regular medications or that they provided inaccurate information about this during the complaints process.
Time of death
59. Mrs A and Mrs O say when they arrived at the hospital, after receiving the phone call to say their father had died, the nurse told them they saw their father take one last breath before he died. They say the nurse’s account of Mr I’s time of death, and that they were with him when he died, is incorrect. Mrs A and Mrs O are concerned that because their father was not monitored properly, he may have died earlier than the Trust say he did.
60. Mr I’s death was verified by the matron at 2:15am. Mrs A and Mrs O say a nurse called them at around 2am and told them ‘your father has stopped breathing’. They say when they arrived at the hospital at around half an hour later, Mr I had ‘pooling’ on the left side of his face, which they say suggests he could have died hours earlier.
61. The Trust say a nurse checked on Mr I every 15 minutes during the evening, except for a period around midnight when there was a medical emergency on another ward. There is nothing in the notes to say whether or not these checks happened as the nurse says. As we have identified in the section above, there were no recorded observations in Mr I’s notes after 8:20pm. We would therefore not be able to say with any certainty whether the information recorded about when Mr I died is correct and whether the nurse was with him when he died, as they say.
62. We recognise that due to the lack of observations, Mrs A and Mrs O may never know for sure what time their father died and we appreciate this is very distressing for them. We will consider the impact of this below.
Impact
63. Mrs A and Mrs O say the Trust’s actions contributed to their father’s premature death. They say because he was not being observed, no one knew he was deteriorating, and staff could not help him. They also say this meant they were not given the opportunity to see him before his death, which caused them immense distress.
64. As identified above, we have seen the Trust failed to take Mr I’s observations in line with guidance and consider further diuretic treatment.
65. The coroner’s post-mortem report says Mr I’s cause of death was:
1a Fibrosing organising pneumonia 1b Chronic obstructive pulmonary disease with aspiration 2 Hypertensive heart disease
66. The post-mortem report describes ‘widespread pulmonary oedema’ (fluid on the lungs). It also describes ‘left ventricular failure’ which is a type of heart failure commonly caused by hypertensive heart disease (listed as a cause of death).
67. Our adviser explained that further doses of diuretics were likely to have reduced the amount of water that was in Mr I’s lungs, which then would have improved his blood oxygen levels. Oxygen is an essential requirement for the major body organs to function and very low blood oxygen is a common cause of cardiac arrest.
68. Our adviser says it is unlikely Mr I would have received additional treatment on HDU other than the diuretics. Although not explicitly stated in the notes, given that a DNACPR decision had been made (on this basis of his severe COPD), he would not have been offered life support on a ventilator either.
69. The main reason for moving him to HDU would have been for continuous monitoring and prompt intervention if required. The nursing ratios on HDU are generally one nurse to two patients, whereas on a general ward it is usually one nurse to at least eight patients. It is much more challenging to closely monitor a very sick patient on a general ward.
70. Our adviser says additional diuretics would not have had any effect on Mr I’s fibrosing organising pneumonia or COPD with aspiration, which were the primary causes of his death. It is likely he would have died of these two conditions even if he had been observed more regularly and given further doses of diuretics.
71. Based on how ill Mr I was with a severe case of pneumonia, our adviser says further diuretics may have delayed his death by a matter of hours. They would not have prevented his death.
72. We recognise it is very upsetting for Mrs A and Mrs O that their father was not monitored properly, and to feel he did not get the help he needed towards the end of his life. We also appreciate the uncertainty around the time of Mr I’s death caused them additional distress.
73. We also recognise that if Mr I had been monitored in line with guidelines, staff could possibly have been alerted to the decline in his condition and his family could have been informed of this sooner. There was a potential lost opportunity for Mrs A and Mrs O to be with their father when he died.
74. It is understandable this caused Mrs A and Mrs O distress and added to their grief over the loss of their father at an already very difficult time.
75. We do not think the Trust has done enough to put this right and we have made recommendations below.