Antibiotic treatment
20. Miss R complains the Trust did not provide her father with the right antibiotic treatment for sepsis. She questions why the Trust only gave him one antibiotic and believes it should have also given him gentamicin (an antibiotic used to treat severe or serious bacterial infections).
21. In its first complaint response, the Trust said the admitting medical team gave Mr R co-amoxiclav (a combination antibiotic) according to hospital policy for the admission diagnosis. It said the Emergency Surgeon then reviewed Mr R while he was on the MSSW and did not feel they needed to change antibiotics as he was responding positively.
22. In its second response, the Trust said staff prescribed co-amoxiclav before moving Mr R to the MSSW. It said he had a working diagnosis of pneumonia when the Emergency Surgeon reviewed him, and he was stable. Pneumonia is a breathing condition where there is inflammation or an infection of the lungs or large airways. The Trust said the Emergency Surgeon spoke with microbiologists who also felt there was no need to change antibiotics.
23. During our investigation, the Trust told us it admitted Mr R with a diagnosis of community-acquired pneumonia. It said a patient can have sepsis of unknown origin but, if doctors think they know the source of the infection, they will treat that condition. It said doctors treated Mr R’s pneumonia with co-amoxiclav in line with the Trust’s Antibiotic Formulary, as this is what they thought the source was.
24. The Trust said doctors could have given Mr R clarithromycin (another antibiotic) as well as co-amoxiclav, but clarithromycin is only used to treat atypical pathogens (less common bacteria). It said it is unlikely clarithromycin would have had any added benefit as Mr R responded to co-amoxiclav. It said giving him gentamicin would have put him at risk of its side effects.
25. We looked at Mr R’s clinical records to see what happened.
26. ED records for 9 October say a chest X-ray showed no signs of pneumonia and Mr R’s chest was clear, but he had an ‘extremely raised’ lactate. The doctors recorded a diagnosis of ‘decompensated ALD +/- sepsis’. They started Mr R on co-amoxiclav, and suggested staff repeat blood tests to check his lactate level. The Trust then transferred him to the MSSW.
27. The Trust’s Clinical Care Outreach Team saw Mr R that night as he needed oxygen. They noted he was being treated for ‘first presentation of decompensated liver disease +/- sepsis’ and had a raised lactate, though the cause of this was unknown.
28. They recorded that his chest was clear, and a repeat chest X-ray had shown no new acute changes. They also noted he had a new oxygen requirement ‘secondary to sepsis ?source’ but his lactate was improving with antibiotics and fluids. They decided to continue treatment with co-amoxiclav and repeat blood tests, including lactate, the following morning.
29. The Trust’s Clinical Care Outreach Team saw Mr R again in the morning on 10 October. The records say doctors were on the ward at the time and saw him on their ward round instead. They noted ‘transmitted sounds’ in his chest, and listed community-acquired pneumonia as one of their findings.
30. We can see other references to the initial diagnosis of possible sepsis on admission. There is a record on 17 October which lists ‘decompensated ALD +/- sepsis’ under ‘admission information’. There is also a record on 19 October that says, ‘was treated for sepsis initially on admission – completed course of abx [antibiotics]’.
Diagnosis
31. We considered the Trust’s diagnosis by looking at Mr R’s clinical records alongside relevant guidance and clinical advice from our two advisers.
32. The clinical records show doctors in the ED did not diagnose pneumonia and instead suspected Mr R may have had sepsis. They show it was not until doctors reviewed him the following morning that they diagnosed community-acquired pneumonia.
33. Our first adviser confirmed there is no mention of suspected pneumonia on admission and Mr R’s chest X-ray showed no signs of pneumonia. They said doctors in the ED recorded a diagnosis of ‘decompensated ALD +/- sepsis’ which shows the doctors thought he may have had sepsis. They also said doctors did not record the suspected source of sepsis but planned further tests to investigate this.
34. Our second adviser said Mr R did not have the features of pneumonia as his white cell count was not elevated, his chest X-ray was normal, and he was not coughing. We note doctors in the ED also recorded his chest was clear. They said the records show the doctors were thinking he had sepsis not pneumonia.
35. Our second adviser also noted Mr R’s elevated lactate, which they said can be a marker for sepsis. Our first adviser also said lactate is a blood test highlighted in NG51 as being raised in patients with sepsis. Our second adviser said it appears doctors did not follow this up at the time but planned to test it again, which also suggests they were thinking about sepsis.
36. NG51 says doctors should assess people with any suspected infection to identify a possible source of infection:
1.1 Identifying people with suspected sepsis 1.1.5 Assess people with any suspected infection to identify: · Possible source of infection · Factors that increase risk of sepsis · Any indications of clinical concern, such as new onset abnormalities of behaviour, circulation or respiration
37. Our second adviser said there was no evidence of a possible source of infection as Mr R’s white cell count and chest X-ray were normal.
38. NICE NG51 then lists the following as risk factors for sepsis:
1.2 Risk factors for sepsis 1.2.1 Take into account that people in the groups below are at higher risk of developing sepsis: · The very young (under 1 year) and older people (over 75 years) or people who are very frail
· People who have impaired immune systems because of illness or drugs, including: o People being treated for cancer with chemotherapy o People who have impaired immune function (for example, people with diabetes, people who have had a splenectomy or people with sickle cell disease) o People taking long-terms steroids o People taking immunosuppressant drugs to treat non-malignant disorders such as rheumatoid arthritis
· People who have had surgery, or other invasive procedures, in the past 6 weeks · People with any breach of skin integrity (for example, cuts, burns, blisters or skin infections) · People who misuse drugs intravenously · People with indwelling lines or catheters
39. Our second adviser said Mr R did not have any of these risk factors. However, he was at moderate/high risk of sepsis considering the signs of sepsis set out within this guidance. Our first adviser agreed Mr R had at least two of the high-risk criteria for sepsis:
1.4 Stratifying risk of severe illness or death from sepsis
1.4.2 Recognise that adults, children and young people aged 12 years and over with suspected sepsis and any of the symptoms or signs below are at high risk of severe illness or death from sepsis: · Objective evidence of new altered mental state · Respiratory rate of 25 breaths per minute or above, or new need for 40% oxygen or more to maintain oxygen saturation more than 92% (or more than 88% in known chronic obstructive pulmonary disease) · Heart rate of more than 130 beats per minute · Systolic blood pressure of 90mmHg of less, or systolic blood pressure more than 40mmHg below normal · Not passed urine in previous 18 hours (for catheterised patients, passed less than 0.5ml/kg/hour) · Mottled or ashen appearance · Cyanosis of the skin, lips or tongue · Non-blanching rash of the skin
1.4.3 Recognise that adults, children and young people aged 12 years and over with suspected sepsis and any of the symptoms or signs below are at moderate to high risk of severe illness or death from sepsis:
· History of new-onset changed behaviour or change in mental state, as reported by the person, a friend or relative · History of acute deterioration of functional ability · Impaired immune system (illness or drugs, including oral steroids) · Trauma, surgery or invasive procedure in the past 6 weeks · Respiratory rate or 21-24 breaths per minute · Heart rate of 91-130 beats per minute or new-onset arrhythmia, or if pregnant heart rate of 1—130 beats per minute · Systolic blood pressure of 91-100 mmHg · Not passed urine in the past 12-18 hours (for catheterised patients, passed 0.5-1ml/kg/hour) · Tympanic temperature less than 36C · Signs of potential infection, including increased redness, swelling or discharge at a surgical site, or breakdown of a wound.
40. Our second adviser explained this was because Mr R was confused, and his blood pressure was less than 90. They said he also had at least one of the moderate risk criteria, as his temperature was less than 36˚C. However, our second adviser noted Mr R had an acute kidney injury and low potassium level which could have also caused his low blood pressure and confusion.
41. We recognise Mr R’s presentation changed following his move to the MSSW. We can then see references to increasing oxygen requirements on 9 October and chest crackles (bubbling, clicking or rattling sounds that occur when small air sacs in the lungs fill with fluid) on 10 October.
42. Our second adviser said the presence of these crackles indicates some irritation to the lungs and this would have been sufficient grounds on which to diagnose pneumonia. We also note references to Mr R having a wheeze on 11 and 12 October, and a cough on 12 October.
Treatment for sepsis
43. We considered the treatment provided by the Trust by looking at Mr R’s clinical records alongside relevant guidance and advice from our two advisers. We started with the treatment for sepsis.
44. NICE NG51 says adults who have suspected sepsis and one or more of the high risk criteria should be reviewed by a senior clinician, discussed with a consultant, and given a broad-spectrum antimicrobial (antibiotic) within one hour:
1.6 Managing and treating suspected sepsis in acute hospital settings
Adults, children and young people aged 12 years and over with suspected sepsis who meet 1 or more high risk criteria 1.6.1 For adults, children and young people aged 12 years and over who have suspected sepsis and 1 or more high risk criteria:
• arrange for immediate review by the senior clinical decision maker to assess the person and think about alternative diagnoses to sepsis • carry out a venous blood test for the following: o blood gas including glucose and lactate measurement o blood culture o full blood count o C-reactive protein o urea and electrolytes o creatinine o a clotting screen
• give a broad-spectrum antimicrobial at the maximum recommended dose without delay (within 1 hour of identifying that they meet any high risk criteria in an acute hospital setting) in line with recommendations in section 1.7 • discuss with a consultant
45. NG51 goes on to say that adults with a suspected infection but no confirmed diagnosis should be given an IV antimicrobial from the agreed local formulary:
1.7 Antibiotic treatment in people with suspected sepsis
1.7.7 For people aged 18 years and over who need an empirical intravenous antimicrobial for a suspected infection but who have no confirmed diagnosis, use an intravenous antimicrobial from the agreed local formulary and in line with local (where available) or national guidelines.
46. The BASL’s guidance similarly recommends treatment with antibiotics in line with the local policy for suspected sepsis in patients with decompensated cirrhosis. Cirrhosis is scarring of the liver caused by long-term liver damage.
47. Our advisers both said the Trust should have followed its guidance for sepsis of unknown origin/sepsis in a patient with ALD on admission, as Mr R had no features of pneumonia and doctors did not suspect he had pneumonia.
48. The Trust’s Antibiotic Formulary does not include guidance for sepsis in a patient with ALD but recommends co-amoxiclav and gentamicin for sepsis/infection of unknown source. This shows the Trust should have prescribed Mr R co-amoxiclav and gentamicin when doctors in the ED suspected he may have had sepsis.
49. This would have also been in line with the GMC’s Good Medical Practice which says doctors should provide treatment to serve the patient’s needs:
Apply knowledge and experience to practice
15. You must provide a good standard of practice and care. If you assess, diagnose or treat patients, you must: a. adequately assess the patient’s conditions, taking account of their history (including the symptoms and psychological, spiritual, social and cultural factors), their views and values; where necessary, examine the patient b. promptly provide or arrange suitable advice, investigations or treatment where necessary c. refer a patient to another practitioner when this serves the patient’s needs.
16 In providing clinical care you must: a. prescribe drugs or treatment, including repeat prescriptions, only when you have adequate knowledge of the patient’s health and are satisfied that the drugs or treatment serve the patient’s needs b. provide effective treatments based on the best available evidence c. take all possible steps to alleviate pain and distress whether or not a cure may be possible d. consult colleagues where appropriate
50. Our second adviser said prescribing gentamicin to Mr R would not have caused him any harm but would have treated the worst case scenario (sepsis). They said this would have meant his treatment was in line with the GMC’s Good Medical Practice. They said it also would have been in line with the standing principle of medical ethics, which is to take action that benefits the patient and causes the least harm.
51. Our second adviser said the doctors should have sought advice from a microbiologist if there was any doubt about the best antibiotics to prescribe. The Trust’s complaint response says the Emergency Surgeon sought advice from a microbiologist, who recommended continuing with co-amoxiclav as Mr R was stable.
52. We found no evidence of this input within the clinical notes and when we queried this with the Trust, it could find no record either. However, the Emergency Surgeon did not feel this changed the Trust’s response. They said Mr R was already stable when they saw him, meaning there was no reason to change the antibiotic or speak to a microbiologist at that point.
53. We think the Trust is referring to events on 12 October, as this is when the records show the Emergency Surgeon saw Mr R and said to continue with the current antibiotic treatment. However, we note this was three days after doctors in the ED suspected sepsis and started him on co-amoxiclav.
Treatment for pneumonia
54. We looked at the treatment for pneumonia.
55. The Trust’s Antibiotic Formulary recommends staff give patients with community-acquired pneumonia a CURB-65 score and then prescribe antibiotics based on this score. CURB-65 is a tool used to predict mortality in patients with community-acquired pneumonia and infection.
56. The Trust did not give Mr R a CURB-65 score. However, our first adviser said Mr R was confused, had a urea of 19.1 (a substance excreted by the kidneys in the urine), and his blood pressure was on occasion less than 90. They said this would mean a CURB-65 score of at least 3 (high severity) and our second adviser agreed.
57. The Trust’s Antibiotic Formulary recommends staff give patients with a CURB-65 score of 3 or more co-amoxiclav as well as clarithromycin, for at least the first 48 hours (continued after 48 hours if atypical pathogens are suspected). NICE NG138 similarly recommends co-amoxiclav and clarithromycin for patients with a score of 3 or more.
58. Both our advisers said the choice of co-amoxiclav was in line with NICE guidance and the Trust’s Antibiotic Formulary for community-acquired pneumonia. However, they also said doctors should have included clarithromycin based on Mr R’s CURB-65 score.
Overall
59. We have found evidence of diagnostic indecision by the doctors in the ED following Mr R’s admission. They suspected he had sepsis but prescribed a broad-spectrum antibiotic to treat any possible infection, without being certain of a diagnosis. This treatment was inadequate for both sepsis and pneumonia. It was also not in line with NICE guidance or the Trust’s own Antibiotic Formulary. We find these to be failings.
60. We also note the Trust’s complaint response is not in line with our Principles of Good Complaint Handling, which say public bodies should be open and honest when accounting for their decisions and actions. This is because the Trust did not diagnose pneumonia on admission or provide treatment in line with its own guidance as it told Miss R.
Impact
61. We have considered what impact these failings had on Mr R and, in turn, Miss R. Our second adviser did not consider they contributed to Mr R’s death. They said Mr R was a very ill man, and it is unlikely he would have survived had the Trust provided different treatment.
62. We therefore do not find these failings had any impact on Mr R. However, it is clear they had an impact on Miss R and caused her a great deal of distress at an already difficult time. We have made recommendations to put this right in paragraphs 98-101.
‘Nil by mouth’ sign
63. Miss R complains a healthcare assistant on the MSSW gave her father food and drink when he was supposed to be nil by mouth. She says this caused him to develop aspiration pneumonia.
64. In its first complaint response, the Trust accepted a Clinical Support Worker (CSW) removed a nil by mouth sign and gave Mr R food/drink. It said the CSW was undertaking their care certificate, which ensures they are trained with the most up to date knowledge about patient care. It also said the CSW would receive ongoing support and it would monitor any competency issues.
65. The Trust said it was highly unlikely this incident caused Mr R’s aspiration pneumonia alone but acknowledged the risk and the distress it had caused. The Trust said it would address this with the CSW and take appropriate action in line with its policies.
66. In the Trust’s second response, it said the CSW would be working more day shifts to enable better monitoring of their performance. It also said the CSW now had their care certificate and had completed some refresher training.
67. The Trust said the ADN had spoken with the CSW and would meet with them every month. The Trust also said it now has more senior nurses on duty at night as well as a dedicated member of staff to observe and guide staff through their development.
68. According to Mr R’s clinical records, he was made nil by mouth on 10 October with staff recommending a nasogastric (NG) tube. An NG tube is where a tube is passed into the stomach via the nose to either remove or add substances to the stomach.
69. Mr R then had an ultrasound on 11 October, and a note later that day said he could eat and drink again. However, the records suggest he may have remained nil by mouth throughout 11 October.
70. There is a note of a conversation with the family on 11 October about concerns they had raised around his nutrition and nil by mouth status. It says staff explained he had been nil by mouth prior to the ultrasound but was then able to eat and drink.
71. The note says staff apologised for any ambiguity and explained Mr R’s swallow and conscious level had fluctuated over the past 24 hours. It also says staff were going to start an NG tube for nutrition if Mr R tolerated this.
72. We can see doctors then started Mr R on an NG tube that evening but he pulled it out. There are then notes later that night, and early on 12 October, about keeping him nil by mouth again. The Speech and Language Therapy Team (SALT) then saw him that afternoon and recommended a fork-mashable diet and thickened fluids.
73. We can see nothing within the records to clearly show when the incident with the CSW happened. The family raised concerns about Mr R’s nil by mouth status on 11 October which suggests it happened sometime between the Trust making him nil by mouth on 10 October and this meeting.
74. However, it is clear a member of staff did give Mr R food and drink when he was nil by mouth as the Trust has already acknowledged this. We have therefore focused our investigation on what impact this had on Mr R and, in turn, Miss R.
Impact
75. Our first adviser said staff recorded concerns about Mr R’s oxygen levels on 12 October. However, there was no clear sudden deterioration of lung function suggestive of aspiration during the time he was nil by mouth in the lead up to the ultrasound on 11 October.
76. Our second adviser said patients with liver disease often experience reflux, meaning food can get into the lungs or go down the wrong way. This can lead to aspiration pneumonia, but secretions can also cause it.
77. Our second adviser said Mr R could have developed aspiration pneumonia without staff giving him food or drink while nil by mouth, and even with an NG tube in place. They said it is possible the incident contributed to him developing aspiration pneumonia, but they could not say it was the sole source.
78. Overall, we cannot say, even on the balance of probabilities that giving Mr R food or drink, while he was nil by mouth, either contributed, or directly led, to him developing aspiration pneumonia.
79. However, we can say this incident understandably caused Miss R distress. The Trust has already acknowledged this, apologised, and outlined action it has taken to help prevent something like this from happening again.
80. We think the Trust has been open and accountable about what happened here. We also think it has taken appropriate action to put things right for Miss R and to prevent the same thing from happening to another patient. We also note the Trust gave Miss R detailed information about how it had addressed the matter with the CSW involved.
Laxatives
81. Miss R complains the medical team prescribed laxatives for her father, but nursing staff did not always give them to him. She says the clinical records also show nursing staff did not record why they missed doses.
82. Miss R says her father was encephalopathic and needed laxatives to open his bowels two to three times a day, to flush out toxins. Encephalopathy is confusion, altered levels of consciousness, and coma, caused by the accumulation of toxic substances in the blood stream.
83. In its second complaint response, the Trust said staff prescribed lactulose and phosphate enemas (both laxatives) to encourage Mr R to open his bowels two to three times a day. It said staff administered this regime at a level to achieve the required bowel frequency and reduce the risk of hepatic encephalopathy. This is encephalopathy in patients with ALD.
84. The Trust said staff gave Mr R two to three doses of lactose plus phosphate enemas for 48 hours, between 12 and 14 October, which resulted in excessive stool frequency. It said this was more than they would normally need to give to reduce encephalopathy and was at a level that made it right for them to omit lactulose.
85. The Trust apologised if staff did not properly explain this at the time. It said Mr R’s bowel frequency then reduced by 19 October, and he became more unwell. It said staff passed an NG tube and restarted the lactulose regime using this.
86. Our second adviser explained that bacteria in the large bowel can form ammonia. They said, in a person whose liver is not functioning properly, this ammonia gets into the blood and causes encephalopathy. They said laxatives ensure they have frequent bowel movements to promote the growth of bacteria that does not form ammonia.
87. The BASL’s guidance recommends 20-30ml of lactulose four times per day, or a phosphate enema, aiming for two soft stool per day. We can see from Mr R’s clinical notes that doctors in the ED recognised he needed regular lactulose and monitoring of his bowel movements.
88. On 10 October, doctors noted nursing staff needed to ensure he was opening his bowels 2-3 times per day. On 11 October, doctors said nursing staff needed to ensure they gave him lactulose and, if he did not open his bowels, a phosphate enema. On 13 October, doctors said nursing staff should give him 15ml of lactulose three times a day with phosphate enemas as needed with the aim of 2-3 soft stools per day.
89. However, doctors asked nursing staff to stop lactulose on 14 October due to Mr R having profuse diarrhoea and they did not ask them to start this again until 19 October. Mr R’s clinical records contain a detailed record of the frequency of lactulose and phosphate enemas between the 9 and 20 October.
90. Our second adviser said the records show Mr R had diarrhoea from 13 October onwards, which means he did not need further laxatives at that time and stopping them was the right thing to do. They said this did not cause him any harm as he had a low chance of getting encephalopathy.
91. However, we are concerned about the administration of the laxative regime between 9 and 12 October. This is because the records show that nursing staff did not always give Mr R laxatives in line with the BASL’s guidance, or the Trust’s own treatment plan, despite him having few bowel movements.
92. We can also see there were times when nursing staff did not record why they had missed lactulose doses or not given a phosphate enema. By looking at other records, we can see why some doses might have been missed, because for example Mr R had profuse diarrhoea the night before. However, with other entries it is less clear.
93. This record keeping does not look to be in line with the NMC’s Code which says nurses should keep clear and accurate records:
10. Keep clear and accurate records relevant to your practice This applies to the records that are relevant to your scope of practice. It includes but is not limited to patient records.
To achieve this, you must: 10.1 complete records at the time or as soon as possible after an even, recording if the notes are written some time after the events 10.2 identify any risks or problems that have arisen and the steps taken to deal with them, so colleagues who use the records have all the information they need
94. We find this record keeping combined with not following the BASL’s guidance or Mr R’s treatment plan to be a failing. We have therefore considered what impact this had on Mr R, and in turn, Miss R.
Impact
95. Our first adviser said the clinical records show Mr R had large amounts of stool on 13 October and 14 October while on laxatives. They said staff also noted he was agitated and trying to get out of bed at this time. They said this suggests Mr R remained confused even when the laxatives were working.
96. Our first adviser said the records show Mr R was incontinent of faeces on 16 October and 19 October. However, staff noted his encephalopathy was worsening. They said this shows he remained encephalopathic even when he was opening his bowels significantly more often than two to three times per day.
97. Our first adviser thought it was very unlikely any missed doses of laxatives would have changed the course of Mr R’s illness. This means we have not seen this failing had any impact on Mr R. However, we do recognise it caused Miss R distress. We have made recommendations to put this right in paragraphs 98-101. We sincerely hope our findings provide Miss R with some reassurance.