Observations
20. Mr B complains the nursing team in the ED did not complete full observations to monitor his mother, and did not document them, as they should have. He says the incomplete and lack of timely observations meant the ED team may have missed identifying when her condition started to deteriorate.
21. The Trust has said on 17 March 2023 it was extremely busy and was in Operational Pressures Escalation Level (OPEL) 4. This means pressures on services prevented the hospital from being able to deliver comprehensive care. It said it had limited nursing staff in the ED and the high number of people in attendance affected Mrs C’s care. It said this was not the standard of care it seeks to provide.
22. The NMC’s The Code says nurses must ‘accurately identify, observe and assess signs of normal or worsening physical and mental health in the person receiving care’. They should keep ‘clear and accurate records’.
23. The ED nursing team used the RCP’s NEWS chart when taking Mrs C’s observations. The NEWS chart instructs nurses to take six key observations: • respiration rate • oxygen saturation (the level of oxygen in the blood) • blood pressure • pulse rate • level of consciousness or new confusion • temperature.
24. Each reading earns a score between zero and three. After taking all the observations, the nurse should go on to calculate the overall NEWS score. The more abnormal the observations, the higher the overall NEWS score.
25. RCP guidance says the NEWS ‘should be used to inform the frequency of clinical monitoring, which should be recorded on the NEWS chart’. The tool also assists nursing teams to know when to escalate care and helps medical staff to identify acutely ill patients.
26. The records show Mrs C arrived in the ED at 12.22pm. A nurse took her observations at 12.27pm which included a temperature of 37°C, a respiration rate of 18 breaths a minute, a pulse rate of 82 beats per minute and blood pressure of 122/64. These results were all in normal ranges.
27. The nurse did not record Mrs C’s oxygen saturation level, or overall NEWS score. Our ACP adviser has said it is likely the nursing team was under immense pressure at the time Mrs C was in the ED, however, they should be documenting observations in-line with NMC and RCP guidelines. This did not happen and we find this a failing.
28. Our ACP adviser told us it is not possible to say what Mrs C’s oxygen saturation level would have been at this time. This is because different factors can affect this, such as age and lifestyle, and we do not know what was normal for Mrs C. Due to this missing information, we do not know what the overall NEWS score would have been at this time.
29. A nurse took full observations at 1.10pm. Mrs C’s temperature was 36.2°C, her respiration rate was 18 breaths a minute, her blood pressure 130/69 and her oxygen saturation level was 93%. They calculated her overall NEWS score as one. However, on the NEWS chart an oxygen saturation level of 92-93% should score two, so the Trust made an error here.
30. Our ED adviser has commented that apart from her slightly low oxygen saturation level, Mrs C’s other observations were all in the normal ranges. This means they would have been zero scores on the NEWS chart.
31. The NEWS guidance says that for an overall score between one and four, nursing staff should monitor the patient at a minimum of every four to six hours. While the NEWS score of one was inaccurate, it indicated the same level of monitoring as the correct NEWS score of two. Following careful consideration, we consider the error in calculating the NEWS score a shortcoming, but it does not fall so far below expected standards we would call this a failing.
32. An ED doctor assessed Mrs C at 2.30pm and noted tests showed she had high lactate levels. The body produces lactic acid when there is inadequate blood flow to the tissues. The doctor suspected Mrs C may have an aortic dissection and she was waiting for a CT scan. We consider this assessment in more detail further on in our report.
33. The NEWS guidance meant the nursing team were due to next take Mrs C’s observations between 5.10pm and 7.10pm. A nurse next completed observations at 3.30pm, sooner than her NEWS score indicated.
34. Mrs C’s observations at 3.30pm showed she had a slightly improved oxygen saturation level of 94% (score of one on the NEWS), but an increased pulse rate of 91 (also a score of one). The nurse did not document her temperature. The other observations fell within normal ranges. The nurse did not calculate the overall NEWS score.
35. Mr B has specific concern nursing staff did not check his mother’s temperature between 1.10pm and 5pm and considers this was too long for someone who was very unwell. As we set out, we think the interval was consistent with guidance and the nurse completed the NEWS sooner than indicated. However, in-line with the ACP advice we have received and with reference to the NMC guidance, we consider the Trust should have documented all observations, including temperature. It should also have documented the NEWS score. We find this a failing in record keeping.
36. In terms of what Mrs C’s temperature would have been at 3.30pm, we cannot determine this. We know what her temperature was before and after this time but we do not know when it started to fall outside of the normal range. This means we do not know what the overall NEWS score would have been.
37. At around 4.50pm, the ED team moved Mrs C from a wheelchair in the corridor to the resuscitation area (‘resus’) of the ED and onto a trolley. Resus is for patients who are most critically ill or injured. The notes say there was now space available.
38. By this time, further tests showed Mrs C’s condition was worsening. The ED team also reviewed Mrs C’s CT scan and did not consider this showed an aortic dissection. They were now considering alternative diagnoses to explain her symptoms.
39. A nurse recorded full observations at 5pm. Mrs C’s oxygen saturation level, blood pressure and pulse rate readings were all outside of normal ranges and she also had a raised temperature of 38.2°C. The NHS website says normal body temperature is generally around 37°C but this can vary throughout the day. A high temperature is 38°C or above.
40. The nurse did not calculate the overall NEWS score which we consider they should have done in-line with NMC and RCP guidance. From the recorded information, we can calculate the NEWS score would have been eight.
41. The NEWS chart states that for a NEWS over five, the team should ‘think sepsis’ and escalate care. For a NEWS score over seven, the NEWS guidance says an emergency response is required and the patient should have continuous monitoring of their vital signs. The NEWS guidance also gives further instructions on the clinical response required for a high NEWS and we have considered this in detail below.
42. In the resus area, Mrs C was put on machines that continuously monitored her vital signs, and although the team had not calculated her NEWS score at 5pm, starting continuous monitoring is consistent with some of what was required under the NEWS guidance.
43. On review of the NEWS charts from 5pm, there continue to be gaps in staff fully documenting all Mrs C’s observations. At 5.34pm, her oxygen saturation level, pulse, level of consciousness and temperature are all missing. The charts following this continue to miss consistently recording her temperature and calculating her overall NEWS score.
44. Our ACP adviser has commented the continuous monitoring Mrs C was on from 5pm may explain the gaps in the NEWS charts. However, staff should be completing the records accurately and in full, in-line with the RCP guidance.
45. The nursing team continued to partly complete NEWS charts approximately every hour, up to her transfer to the intensive care unit (ICU) at around 2am on 18 March. We have not seen guidance applicable to the time of these events to say how often staff should document the NEWS when a patient is on continuous monitoring.
46. In summary, we consider the nursing team monitored Mrs C’s vital signs at a frequency that meets with the NEWS guidelines. However, we find the missing information throughout the charts does not meet with NMC or RCP guidelines. The missing information indicates the nursing team was not completing all the required checks. We therefore find a failing in this part of the complaint.
47. Mr B has told us he considers the inadequate monitoring and documentation meant the medical team did not have all the information available to them when making decisions on his mother’s care. We understand he has a key concern that staff may have missed identifying if his mother was showing earlier signs of sepsis and this delayed her treatment.
48. As noted above, we do not consider we can determine what the missing results in the observations taken at 12.27pm and 3.30pm would have been. However, Mrs C’s observations at 1pm were largely normal.
49. During this time, the ED team were also investigating Mrs C for an aortic dissection. From around 5pm, Mrs C’s condition stated to worsen which her observations reflect, and the team started to treat her for an infection.
50. Both our ACP adviser and ED adviser have commented that from the information available about Mrs C’s condition, there were no signs before 5pm suggesting an alternative diagnosis, including sepsis. We therefore do not consider we can say the missing documented observations led to a clinical impact to Mrs C. However, we recognise it has caused Mr B on-going worry and distress that the team did not have all the information available to them while making decisions for his mother’s care.
51. The team did not calculate Mrs C’s NEWS score was eight at 5pm. Had they done so, this score should prompt an emergency response and a sepsis screen. Around this time, the clinical team had already become aware Mrs C’s overall clinical picture was worsening, and they had moved her to the appropriate area of the ED. While the team took some appropriate actions here in-line with the NEWS guidance, they did not do a sepsis screen.
52. We have reviewed the care the team provided to Mrs C from 5pm and the impact of not doing a sepsis screen in detail further on in our report. However, we recognise this was a missed opportunity to prompt action from the ED team and this remains a source of considerable concern for Mr B.
53. In terms of the incomplete NEWS charts from 5pm, Mrs C was on continuous monitoring of her vital signs from this time. This means the staff caring for her could see information on her vital signs while reviewing her. However, the purpose of the NEWS charts is to show trends in a patient’s condition. The NEWS score is a clear indication of whether they are getting worse, staying stable or improving.
54. Our senior lead clinician team has commented that in someone who is very unwell, this is valuable information for the medical teams who are assessing the patient and making decisions on the management of their care. It is a source of evidence and justification for the clinical decisions made. It assists in joined-up care between teams.
55. We cannot determine if fully completed charts would have changed the decisions the clinical teams were making from 5pm onwards to manage Mrs C’s condition. We recognise they would have been reviewing different sources of information during this time to determine her condition. However, it has created on-going doubt for Mr B of whether this could have made a difference to his mother’s care.
56. The Trust’s ED was clearly under pressure on this date which affected the level of nursing care it could provide. We acknowledge there are limits to what a Trust can do in such circumstances. However, in its responses to Mr B, it has not explained what action the Trust was taking at that time to manage the pressure on its services and the risk this caused to patient safety.
57. We therefore consider that to address the impact to Mr B, the Trust should explain the actions it plans to take, or has already taken, to safely manage this level of demand in future. It should also explain how it plans to improve the standard of monitoring and record keeping in the ED. We have set out our recommendations at the end of this report.
Initial assessment
58. Mr B complains the doctor who assessed his mother at 2.30pm did so while she was sat in a wheelchair in a corridor and this meant they did not physically examine her. He also says he does not think his mother’s test results and symptoms pointed to an issue with her heart, and the ED team should have suspected an infection or sepsis.
59. The GMC’s Good Medical Practice says doctors who assess, diagnose or treat patients must, ‘adequately assess the patient’s conditions, taking account of their history’, and where necessary, ‘examine the patient’. They should also, ‘promptly provide or arrange suitable advice, investigations or treatment where necessary’.
60. The symptoms of an aortic dissection are a sudden and severe chest pain that spreads to the neck or back, and shortness of breath.
61. Mrs C’s GP had told her to go to the ED for assessment because they were concerned the symptoms she was reporting may be connected to her existing heart condition of aortic root dilatation.
62. Following her arrival to the ED and while waiting for a full medical assessment, a doctor requested a CT aortogram for her (a CT scan involving injection of a dye to clearly show the arteries and blood vessels in the heart). At 2.30pm, an ED doctor assessed Mrs C while she was sitting in a wheelchair.
63. The doctor noted Mrs C’s medical history and documented her account of a stabbing chest pain that radiated to her back, that she had been nauseous, had pain in her upper teeth, had no cough but was shorter of breath than usual. She reported no fevers, but had a decreased appetite, decreased fluid intake and diarrhoea.
64. The doctor identified Mrs C had an irregular heart rate and on review of her blood tests, saw she had a high lactate level of 5.8mmol/l. She also had abnormal kidney results and her blood was acidic (known as acidosis).
65. The doctor suspected she may have an aortic dissection, and a severe acute kidney injury (AKI). AKI is sudden kidney damage or failure. They noted she was still waiting for the CT scan. The doctor called the radiology team to chase this up. They planned initial treatment of fluids, pain killers and to admit her.
66. Our ED adviser explained aortic root dilatation is a risk factor for aortic dissection. This is because the enlarged part of the aorta can weaken making it vulnerable to rupture. Our adviser has said the presence of chest pain in someone with a known risk of aortic dissection is a warning sign that should alert doctors the patient needs emergency investigation and treatment.
67. Our ED adviser further explained an aortic dissection could also account for why Mrs C had high lactate levels because this condition can affect the blood flow to the tissues. AKI can be associated with people who have an aortic dissection. There can be several possible causes for acidosis, it indicates a patient is generally unwell.
68. Overall, our ED adviser has said there was no reason for the doctor to suspect an alternative diagnosis at this time. They said Mrs C’s medical history and symptoms justified the doctor’s working consideration she may have an aortic dissection, and doctors had requested the appropriate tests to help determine this diagnosis.
69. In consideration of the advice we have received, we find the available information supported the ED team’s initial suspicion of an aortic dissection and it was appropriate for them to want to first rule this out. We consider this approach was in-line with the GMC guidance quoted in paragraph REF _Ref208241184 \r \h 59.
70. However, the notes say the doctor was unable to physically examine Mrs C because there were no trolleys available in the hospital and she should be re-examined when one became free.
71. The GMC’s Good Medical Practice says, ‘if patients are at risk because of inadequate premises, equipment or other resources (…) you should put the matter right if possible’ and ‘make a record of the steps you have taken’.
72. Our senior lead clinician team considered these circumstances. They have commented the ED team suspected Mrs C may have an aortic dissection and this is a medical emergency. For this reason, they said Mrs C needed a full assessment including an examination, and so she needed a trolley to allow the doctor to do this. They consider the team should have considered the facilities available to enable a full examination.
73. We consider inadequate resources affected the level of care the ED team provided to Mrs C. While we recognise the significant pressures on the ED that day, there is no evidence in the records of the actions the team took to explore options that would have allowed them to fully examine Mrs C at 2.30pm. In consideration of the GMC guidance referred to in paragraph REF _Ref208241206 \r \h 71, we find this a failing.
74. Mr B says because the doctor did not fully assess his mother at 2.30pm, it meant the team was unaware she had a rash on her torso for around five hours after her arrival. He questions the difference earlier identification of the rash could have made.
75. A medical note that is undated but lies between entries made at 4.50pm and 5pm on 17 March say an examination showed Mrs C had a ‘diffuse erythematous rash’, this a red rash that is usually an allergic reaction to medicine or is due to an infection. Mrs C was unable to tell the doctors when the rash appeared but did not think it had been there before she had come to the hospital.
76. Doctors did not document consideration of the cause of the rash until an ICU review at 11.30pm when they questioned if the rash may be a drug rash caused by the penicillin. Although when doctors first prescribed Mrs C antibiotics at 5.39pm, this was after the ED team first saw the rash.
77. When staff later moved Mrs C to the ICU at around 2am on 18 March, doctors recorded the rash was a blanching rash. This is a rash that fades under pressure. The surgical team reviewed Mrs C and they questioned if the rash was necrotising fasciitis. This is an aggressive bacterial infection that affects soft tissue. Mrs C underwent a surgical procedure on 18 March but the surgical team did not consider there was convincing evidence of necrotising fasciitis.
78. By 19 March, the rash had spread to Mrs C’s thighs and upper arms. It was now a ‘non blanching haemorrhagic rash’, this is a rash that does not fade under pressure and appears reddish-purple in colour. No cause for the rash was found before Mrs C sadly died on 20 March.
79. Samples later taken from the affected tissue on Mrs C’s torso showed results after her death that she had invasive group A streptococcus. This is a common bacteria carried on the skin, however in rare cases, it can cause a fatal infection. The Trust informed the family the bacteria led to necrotizing fasciitis, and this caused sepsis.
80. Our intensive care adviser has explained group A streptococcus is a common cause of necrotising fasciitis, and this leads to sepsis which happened in Mrs C’s case. While the surgical team did not consider there was evidence of necrotising fasciitis, our intensive care adviser has said this condition is uncommon and can be difficult to diagnose. It is an aggressive condition that spreads quickly. The only way to treat it is to remove the affected tissue.
81. Necrotizing fasciitis usually presents as a rash that is very painful. Mrs C did not report pain, she was just a ‘bit uncomfortable’ over her abdominal area. She also had no temperature on arrival to hospital. Without these red flags early on, our intensive care adviser has said it would have been challenging to make this diagnosis and it only became clear from the later pathology results.
82. On review of the records, it is evident Mrs C’s rash spread and worsened during her admission. Our physician adviser has said it is not possible to say if the rash would have been present at 2.30pm on 17 March, but the records indicate it was developing over time. This suggests that even if it had been present, its appearance may not have been as bad.
83. It is possible that had the rash been visible and the team seen this earlier, it could have led doctors to more quickly be able to rule out a drug rash, which was the initial suspicion, and consider other causes.
84. A red flag for sepsis is a non-blanching rash, and the rash initially was a blanching rash. This means we cannot say it would have been an indicator of sepsis if doctors had seen it earlier on 17 March.
85. Our physician adviser has also commented the team was initially not too concerned about the rash. They noted it, but it was not causing Mrs C any pain and they did not consider it was due to an infection.
86. We consider it is possible earlier sight of the rash may have played into the overall clinical consideration of Mrs C’s condition, though we cannot know if it was present at the time of the 2.30pm assessment in a wheelchair. However, in consideration of the physician advice we have received and on review of the evidence, we do not think this would have changed how the ED managed her care or impacted the overall outcome.
87. We consider the lack of an examination at 2.30pm meant there was a missed opportunity for the team to have all the information available about Mrs C’s condition. While we cannot say what this examination would have showed or that it would have changed what happened over the following days, this uncertainty continues to cause Mr B concern and distress.
88. The Trust has not yet explained what it has done or plans to do to help prevent this failing from happening again, to improve patient safety. We have therefore set out our recommendations at the end of this report.
Sepsis diagnosis and treatment
89. Mr B complains that by around 5pm, his mother’s observations suggested a systemic infection but the clinical teams did not pick up on this. Certainly by 5.50pm, he considers his mother’s symptoms indicated she had sepsis and this was a basic consideration the team should have made. He complains the team delayed in identifying and treating her.
90. NEWS guidance says if a patient has an overall NEWS score of more than five, the clinical response should be to consider sepsis and to ‘complete a sepsis screen’.
91. The Sepsis Screening Tool lists the criteria clinicians should consider to determine if a patient may have an infection. It also lists amber and red flags for sepsis. Amber flags include a systolic blood pressure between 91-110mmHg and a heart rate between 91-130. If any of these flags are present, the patient must have blood tests and there should be a senior clinical review within one hour.
92. Red flags for sepsis include an altered mental state, a heart rate of more than 130 beats per minute, a non-blanching rash, a requirement for oxygen, and lactate levels of more than 2mmol/l. If any red flags are present, doctors should follow the ‘sepsis six’ which is a set of steps that must be completed within one hour.
93. As we have set out above, on review of the records we do not consider Mrs C was showing signs of sepsis prior to around 5pm. We can see her condition started to change around this time. A venous blood gas test at 4.38pm showed her lactate level had increased from 5.8mmol/L at 12.46pm to 6.1mmol/L, this was despite her receiving fluids to try to reduce this level. Sepsis can cause high lactate levels.
94. A doctor reviewed her at 4.50pm along with the results of her CT scan and in discussion with a senior doctor and consultant, they agreed this ruled out an aortic dissection. They questioned if she may have a pleural effusion (a build-up of fluid around the lungs), a benign intestinal tumour, pneumonia or possibly bowel ischemia (this is when the bowel is not getting enough blood flow). A tumour and bowel ischemia were later ruled out.
95. The consultant advised the CT scan suggested Mrs C had pneumonia and instructed the team treat her for a chest infection. This led to the team giving her IV antibiotics at 5.39pm.
96. Mrs C’s observations at 5pm showed her temperature was high at 38.2°C, her oxygen saturation level had dropped to 91%, her pulse rate had increased to 130 beats a minute and her blood pressure had also dropped to 91/52.
97. Another doctor reviewed Mrs C at around 5pm when they described her as being disorientated, giving a confused story, and having a rash over her torso. At 5.05pm, the team noted an ECG showed Mrs C had atrial fibrillation, this is an abnormal and fast heart rate.
98. The clinical team planned to continue to treat Mrs C’s high lactate levels with fluids and if she did not improve, they would consider a repeat scan of her abdomen, they would discuss her case with the intensive care team and ask the surgical team to review her.
99. Our physician adviser has said Mrs C’s test results and observations around 5pm should have prompted the team to screen her for sepsis. With reference to the Sepsis Screening Tool, Mrs C’s new confusion/ disorientation, low oxygen saturation level (and requirement for oxygen), high heart rate and high lactate level were all red flags for sepsis. This did not happen, and in consideration of the evidence, we find this a failing.
100. A nurse completed further observations at 5.34pm. They show Mrs C’s blood pressure had improved slightly, but she was now on oxygen.
101. Although not calculated on the form, Mrs C’s NEWS score was eight from 5pm. Despite the NEWS guidance that a score over five should result in a sepsis screen, there is no reference to this happening, or to the teams considering sepsis at this time. We consider it is possible the incomplete NEWS form may have contributed to this error.
102. The records show at 9.45pm, a doctor planned for blood culture tests if her temperature spiked which suggests they were considering if Mrs C was showing signs of sepsis. The ICU doctor reviewing her at 11.30pm documented she had pneumonia and diagnosed septic shock. They reviewed her antibiotics and instructed she should have a full septic screen.
103. Mrs C was moved to the ICU shortly after 2am where they continued to treat her for these conditions.
104. To consider the impact of the delay in diagnosing sepsis, we have looked at the steps the clinical team should have taken in response to the red flags at 5pm. The Sepsis Screening Tool says if red flags are present, doctors should complete the following actions within one hour: 1. ensure a senior clinician attends 2. give oxygen if required 3. take bloods and send for tests including cultures 4. give intravenous (IV) antibiotics (maximum dose of a broad spectrum therapy) 5. give IV fluids 6. monitor.
105. Although the Trust did not recognise and act on red flags for sepsis, we have considered the actions the team took after 5pm as we recognise some steps were taken despite this failing.
1. Senior clinician attendance
106. The records show a senior doctor reviewed Mrs C at 4.50pm and discussed her case with a consultant at 5.50pm to decide on a management plan. Our physician adviser has confirmed senior doctors in the ED were reviewing Mrs C during this time which was appropriate.
107. We therefore consider the team met the first action in the Sepsis Six guidelines, and this is what should have happened had they identified the red flags for sepsis.
2. Oxygen
108. At 5.34pm, the team put Mrs C on oxygen to help with her breathing. Had the team been reacting to suspected sepsis at this time, this action would meet with the Sepsis Tool guidelines.
3. Blood tests
109. Mrs C had a blood test at 5.35pm, however, the team did not request blood cultures. Blood cultures look for micro-organisms in the blood which can cause infection. The first reference to this being necessary appears to be at 11.30pm when a doctor noted Mrs C should have a full sepsis screen, including viral swabs. This is nearly six hours later.
110. If the team had suspected sepsis at 5pm, in-line with the Sepsis Tool guidelines, our physician adviser has confirmed the team should have requested blood cultures within the hour. This did not happen and we find failing with this. We have considered the impact of this below.
4. Antibiotics
111. A consultant advised Mrs C should have antibiotics to treat suspected pneumonia. This led to the team giving her the antibiotic co-amoxiclav intravenously at 5.39pm at a dose of 1.2g every eight hours.
112. NICE guidance for the treatment of pneumonia lists co-amoxiclav as a suitable antibiotic choice for patients with a serious infection.
113. The BNF provides information to health professionals on selecting, prescribing, dispensing and administering of medicines that are available in the NHS. The BNF describes co-amoxiclav as a broad-spectrum antibiotic. It says the dose to treat infection in an adult is 1.2g.
114. Our physician adviser has said the choice of antibiotic was correct for treating sepsis, even if this was not the purpose at the time doctors prescribed it. On review of this information, we therefore consider this prescription meets with the Sepsis Tool guidelines.
5. IV fluids
115. The ED team had already been administering Mrs C with fluids following her arrival to hospital. There are two entries on 17 March that do not have times recorded but were written before 5.24pm with the fluids due to be given over two hours. The entry at 5.24pm prescribes 1l of saline over two hours.
116. The clinical team were prescribing fluids initially to address Mrs C’s high lactate levels, and then in response to her low blood pressure. Although not given in response to suspected sepsis, the fluids administered meet with the Sepsis Tool guidelines.
6. Monitor
117. The Sepsis Tool says to monitor the patient, clinicians should use the NEWS tool and measure the person’s urinary output, and the patient may require a catheter to do this.
118. As we have set out above, while the nursing team was using the NEWS to monitor Mrs C, we have found failings in this monitoring due to the gaps in the charts.
119. In terms of urinary output, the records say Mrs C had passed only a small amount of urine on 17 March. Nursing staff started completing a fluid balance chart at 5pm, with input entries at 5pm, 6pm, 7pm, 8pm and midnight. There are no output values, meaning Mrs C had not passed any urine during that time.
120. Sepsis can affect kidney function which then reduces urinary output. Our physician adviser has said a urinary catheter is the most accurate way to closely monitor urinary output and to predict an acute kidney injury and its progress.
121. Our physician adviser has said an ED team will typically administer fluids and watch and wait to see the effect on a person before deciding to catheterise them. By 5pm, our physician adviser has highlighted Mrs C’s blood pressure had started to drop. This trend continued in the observations taken between 6.20pm and 8.35pm. This was despite her having received around four litres of fluid by this point.
122. NICE guidelines for sepsis say to escalate a patients care to a consultant if they have not improved after two litres of fluid. The guidance also says that adults being managed in an acute hospital setting with high-risk criteria and a lactate level over 4 mmol/L should be referred to ‘critical care for review of management including need for central venous access and initiation of inotropes or vasopressors’. These are drugs that help improve blood flow in the body.
123. Our physician adviser has said the deterioration in Mrs C’s blood pressure despite treatment should have prompted the ED team to call the ICU team for advice. It should also have prompted consideration of giving Mrs C a catheter due to her low urine output despite continued fluids, and due to suspected sepsis.
124. Our intensive care adviser has agreed this escalation should have happened at this earlier time, in-line with the NICE guidelines on sepsis. With earlier review, they have commented the ICU team would likely have recommended Mrs C be catheterised to measure the effectiveness of her fluid treatment.
125. We can see from the records Mrs C was still showing red flags for sepsis following the first hour of ‘sepsis six’ actions. She had a low systolic blood pressure less than 90mmHG and still required oxygen. Her lactate level at 7.13pm was still high at 4.9mmol/l. In-line with the NICE guidelines, we consider this further indicates the ED should have escalated Mrs C’s care to the ICU team earlier.
126. A senior doctor reviewed Mrs C at 9.45pm and decided she should have a catheter, and this happened at 10.30pm. The doctor also spoke with the ICU to request their input, and an ICU doctor attended at 11.30pm to review Mrs C in person. They agreed she should be admitted to the ICU and she was transferred there after 2am.
127. In consideration of the advice from our physician adviser and ICU adviser we consider the ED team should have contacted the ICU for advice between 6.20pm and 8.35pm, in-line with the NICE guidelines on sepsis. We also consider the team should have given Mrs C a catheter when her observations showed her blood pressure was continuing to drop. The fluid charts showed she had passed no urine despite all the fluid she had received since her arrival.
128. The Trust delayed taking these actions and we find this a failing. We have considered the impact of this below.
129. To monitor the patient, the Sepsis Tool also says clinicians should ‘repeat lactate at least once per hour if initial lactate is elevated or if clinical condition changes’. As noted above, NICE guidance for sepsis refers to a lactate level of over 4 mmol/L as criteria for escalation to intensive care.
130. Mrs C had an elevated lactate level on arrival to hospital. As we have set out above, the ED team were initially concerned she had an aortic dissection, and this could have accounted for her high lactate level. However, the ED team ruled this condition out at between 4.50pm and 5pm when they discussed her CT scan results.
131. Tests through the evening show Mrs C had a lactate level 4.9 mmol/L at 7.13pm, 5.3 mmol/L at 8.32pm and 3.2 mmol/L at 11.59pm.
132. Although the team did not identify the red flags for sepsis at 5pm, Mrs C met the criteria for requiring hourly lactate testing due to her elevated levels. This did not happen. However, our physician adviser has highlighted there is differing guidance on measuring lactate levels.
133. The Surviving Sepsis Campaign: International guidelines for the management for severe sepsis say when adults have sepsis or septic shock, ‘we suggest guiding resuscitation to decrease serum lactate in patients with elevated lactate level, over not using serum lactate’. The guidelines describe this as a ‘weak recommendation’ because during acute treatment for sepsis, ‘lactate level should be interpreted considering the clinical context and other causes of elevated lactate’.
134. Previous versions of the Surviving Sepsis Campaign guidelines recommended using lactate target levels, but it has since changed its view because, ‘normal serum lactate levels are not achievable in all patients with septic shock’ and in recognition not all settings have access to be able to measure lactate levels.
135. On review of this guidance, we understand it may not be possible to correct lactate levels in a person who is very unwell with sepsis. The guidance says that while there should be a goal to reduce high lactate levels, they should not be the only factor guiding treatment. Doctors should consider the full clinical picture. The guidance suggests that lactate levels alone do not necessarily tell doctors how the patient is responding to treatment as other factors can affect this.
136. Our physician adviser has commented the team could have done more frequent lactate testing, but it is not of significant concern this did not happen.
137. In consideration there is difference in guidance on measuring lactate levels, and the advice we have received, we do not consider there is a failing in the ED team not re-testing Mrs C’s levels every hour from 5pm. By this time, Mrs C was becoming more unwell as evidenced by her observations and symptoms. The team already knew she had high lactate levels, and our senior lead clinician team have commented it is unlikely additional testing would have added new information to her clinical picture or changed the approach to her treatment.
138. We are sorry to hear Mr B’s serious concerns about the delay in the ED identifying his mother had sepsis, and for the continued distress this causes him and his family. We have found failing the ED team did not take all relevant actions to treat Mrs C for sepsis from 5pm and escalate her care when indicated. We have set out our consideration of the impact of this below.
Impact
139. Mr B continues to question the difference prompt and appropriate treatment for sepsis could have made to his mother. He knows it is not possible to say if she would have lived, but he wonders how this affected her chances of survival. We are very sorry for how upsetting it was for him to see her sudden deterioration and to feel the Trust did not provide the level of care he expected.
140. We have carefully considered the impact of the failings. In terms of testing blood cultures, our physician adviser has explained this is of less value if the test happens after the person has already started taking antibiotics, as occurred in Mrs C’s case, but should still be tested because the results can help inform the management of treatment.
141. A doctor requested blood cultures at 12.50am on 18 March. The laboratory result on 23 March reported no growth after five days incubation, which means the test did not detect any microorganisms in the blood.
142. The journal article, ‘If Blood Cultures Were Not Done Before Starting Antibiotics, Is It of Any Value to Obtain Them Later?’ says in a study of patients, the percentage of blood culture samples with significant growth was 23.2% positive in patients with sepsis when the test happened within the first hour after starting antibiotics.
143. The study says these positive results ‘declined significantly in the period 1–12 hours after IV antibiotics were started’. However, the ‘residual positivity rate remains high enough that blood cultures are still clinically worthwhile’. This means it is still possible to obtain useful results, even if blood cultures tests are done after the patient has started antibiotics.
144. In Mrs C’s case, she had been taking antibiotics for over seven hours by the time the team took a sample. On review of the above clinical study, it is reasonable to conclude this impacted the results. However, we recognise we cannot say what the results would have been if the team taken a sample at 5pm. We also recognise it would still have taken time for the results to be ready.
145. While we cannot say what difference this test result would have made to how the ED team managed Mrs C’s treatment or to her outcome, we recognise the uncertainty around the difference this could have made is an injustice to Mr B.
146. In terms of a delay in the medical team contacting the intensive care team for advice, we can see the ED team referred Mrs C at 9.45pm and she had an ICU review at 11.30pm. They determined Mrs C had septic shock and should have further fluids, a different antibiotic, a full sepsis screen and transfer to the ICU for care, including administering treatment for septic shock.
147. Our intensive care adviser has commented that had the ED team referred Mrs C to the intensive care team from around 6.20pm, they would likely have reviewed her earlier and she may have been moved to the ICU by around 8pm. However, even if an ICU bed had not been available at that earlier time, the ICU team would have been able to instruct on her care while she was waiting for this move.
148. Once in the ICU, the team’s focus would be on providing organ support, which is what we can see happened. The team would have started earlier invasive blood pressure monitoring; this is when a cannula is placed into the artery to monitor blood pressure. It also allows for frequent blood gas testing, which includes lactate testing.
149. Our intensive care adviser has further explained with earlier input from the ICU, Mrs C would likely have had earlier treatment for her blood pressure with vasopressors. This medication effectively squeezes the blood vessels to drive blood flow to the organs.
150. In terms of the overall difference this earlier care could have made, our intensive care adviser has highlighted necrotizing fasciitis (which caused Mrs C to develop sepsis) has a high mortality rate.
151. The clinical study from the World Journal of Emergency Surgery: ‘recommendations for the management of skin and soft-tissue infections’ refers to a 2009 study which found a mortality rate of 40.6% in patients with necrotizing infections in the ICU. The study emphasises the importance of early detection of the infection, and early surgery to remove the affected tissue.
152. In Mrs C’s case, the ICU team did not diagnose necrotizing fasciitis while she was still alive. Our intensive care adviser has commented that despite the surgical team not making this finding, the ICU did manage her condition as necrotizing fasciitis from 18 March.
153. In summary, we consider earlier referral to the ICU would have allowed the team to have input into Mrs C’s care from around 8pm. This would likely have led to earlier organ support, and earlier admission to the ICU where she would have had one-to-one nursing care. However, we understand necrotizing fasciitis is an aggressive infection that causes quick deterioration. Our intensive care adviser has said on review of the records, it is unlikely earlier ICU treatment would have made an overall difference to the sad outcome.
154. While we cannot say earlier referral and treatment from the ICU team would have likely changed the outcome, we recognise Mr B has suffered distress in knowing that his mother did not receive the prompt care she needed. We understand this was very difficult for him to see at the time of events, and that he and his family have since been left questioning the impact of this. We consider this distress is an impact to Mr B.
155. In its response to the complaint, the Trust did not identify it delayed diagnosing Mrs C with sepsis, or that it delayed referring her to the ICU team for advice. Therefore, it has not yet taken actions to address the impact linked to these failings. We have set out our recommendations below. We hope these actions will go in some way to assuring Mr B of the difference his complaint will make to improving services at the Trust.