21. For clarity, we explain our findings under separate headings.
Prophylactic antibiotics
22. The UKALL guidance recommends that children with Down’s Syndrome should be given prophylactic antibiotics (appendix 5). That is to say, in order to try to prevent the child developing an infection, they should have antibiotics regardless of whether there is any sign of such an infection. The rationale is that those children have a significantly worse survival rate compared to those who do not have Down’s Syndrome. This is because of a higher risk of death due to sepsis.
23. The Trust did not give L prophylactic antibiotics. Mrs J says this was wrong and she believes that had L received them, her chances of survival would have increased. We carefully considered the Trust’s action in this regard.
24. In its first response to Mrs J’s complaint, the Trust explained carbapenemase-producing Enterobacteriaceae (CPE) was increasing in Manchester Children’s Hospital. CPE are bacteria which are resistant to a class of antibiotics called carbapenems. The Trust had therefore decided to avoid he routine use of Ciprofloxacin, which is the recommended antibiotic in the UKALL guidance. The Trust said it had not observed adverse effects but there would be further discussion about patients with Trisomy 21.
25. In the Trust’s second response, a consultant haematologist explained there was no convincing evidence that prophylaxis works but that it did lead to problems with antimicrobial resistance (i.e. antibiotic resistance). Due to the problem with antimicrobial resistance within the Trust, they had decided not to use prophylaxis in any high-risk patients.
26. Antibiotic resistance is a recognised problem, not only in the UK, but worldwide. This is when the overuse of antibiotics over a period of time leads to antibiotics becoming less effective. Some strains of bacteria develop resistance to many different types of antibiotics. NICE guidance NG15 makes recommendations on what organisations should do to reduce such risks. It defines ‘antimicrobial stewardship as ‘an organisational or healthcare‑system‑wide approach to promoting and monitoring judicious use of antimicrobials to preserve their future effectiveness'. The Trust’s caution was in line with this.
27. Current guidance on CPE says ‘Antimicrobial stewardship with particular attention to reducing the use of broad-spectrum antibiotic use is critical in the prevention of antimicrobial resistance.’ (UK Health security Agency, Framework of actions to contain carbapenemase-producing, Enterobacterales’. This was published sometime after these events, in 2022, and we include this as an illustration of recent concerns.)
28. The Trust’s response shows that the decision not to give prophylactic antibiotics was its own policy. The Trust therefore followed its own guidance on this: it was not a lapse in care specific to L’s case.
29. We understand why Mr and Mrs J would be so concerned that national guidance was not followed. We therefore carefully considered the Trust’s decision to implement a policy that appears to have been at odds with the recommendation in the UKALL 2011 guidance.
30. Guidance is not obligatory to follow. An organisation or individual practitioner can deviate from it. This is summarised in an article from the British Medical Journal in 2005 titled ‘Can you ignore guidelines?’, which says:
‘Guidelines are just that—they provide guidance. They do not and should not imply that doctors must suspend their clinical judgment in order to follow to the letter what has been written down… While ignorance of guidelines is a poor defence, a reasoned and justifiable decision to deviate from guidelines in an individual case, backed up by good clinical records made at the time, may be acceptable.’
31. This refers to decisions of individual doctors, but the principle holds for local policy.
32. Good Medical Practice tells doctors: ‘You must be familiar with guidelines and developments that affect your work.’ This does not place an obligation to follow the guidelines. It goes on to say ‘In providing clinical care you must… prescribe drugs or treatment… only when you have adequate knowledge of the patient’s health and are satisfied that the drugs or treatment serve the patient’s needs.’
33. We consider the Trust’s policy was a considered and reasoned decision to not follow the UKALL guidance. We also recognise the UKALL guidance says that administering prophylactic antibiotics is recommended. It does not say they must be given.
34. The Trust’s second response said the Trust was monitoring the situation and had not seen rates of fatal bacterial infection to be above those is similar units or reported in clinical trials. This evidence appears to support its decision, and shows they were monitoring the situation. It also said they were keeping under review whether children with Down’s Syndrome should be given prophylactic antibiotics as they constitute such a small proportion of patients and the use of antibiotics in this case may not affect resistance. This is in keeping with Good Medical Practice, which says doctors should ‘Contribute to and comply with systems to protect patients’.
35. The Trust gave a reasonable explanation of the issue they had with antibiotic resistance and how they had not seen any evidence prophylaxis antibiotics reduced fatalities.
36. In its responses and incident report, the Trust said the decision was made between the clinical team and microbiologists. Microbiologists’ role within a hospital is to direct the prevention, diagnosis and treatment of infectious bacterial diseases. They provide a consultancy service to colleagues about the treatment of patients with infection or suspected infection. The Trust told us that it could not find a written policy for the decision or recorded evidence of the discussion. Therefore we cannot reassure Mr and Mrs J how the decision was made and whether those making the decision had included a consideration of higher-risk patients, such as children with Down’s Syndrome.
37. Therefore, we find there were failures in the way the decision was made and lack of explanation to Mr and Mrs J.
38. Mrs J said that when she consented to the UKALL trial for L, nobody mentioned that they were not going to be following one of the recommendations; that is to say, they would not be giving prophylactic antibiotics as recommended in appendix 5. If it had, she would have insisted on the prophylactic antibiotics being given. If they had still refused, she would have taken L to another hospital.
39. The sixth of the principles in GMC’s ‘Decision making and consent’ says ‘If patients are not able to make decisions for themselves, the doctor must work with those close to the patient and with other members of the healthcare team. The doctor must take into account any views or preferences expressed by the patient [or in this case, her parents]’. We found the Trust did not act in line with this. Doctors should have explained to Mr and Mrs J why they were not following the recommendation in the UKALL guidance.
40. This is not to say that this would have led to L being given antibiotics. Doctors might still have decided to omit them even if Mrs J had made the point that she wanted L to have them. Good Medical Practice says that doctors must ‘provide effective treatments based on the best available evidence’. As explained above, there was a logical reason why they were going against it (antibiotic resistance) which was specific to this Trust’s circumstances.
41. On the other hand, they may have agreed to give them as an exception because of her Down’s syndrome. This is a recommendation that arose during the Trust’s investigation. The Trust has told us that it reviewed its practice as part of its learning from this case children with Trisomy 21 are now managed as the exception to the ‘no prophylaxis’ so now receive ciprofloxacin during induction, in line with the UKALL guidance.
42. We realise Mr and Mrs J are uncertain whether a different decision would have made a difference to the outcome. We do not consider the Trust’s decision to omit prophylactic antibiotics was unreasonable. The explanation it gave for not following the guidance was logical. We consider there were two failures in this regard. First, there is no written record of how the Trust made the decision and whether high-risk patients were part of the consideration. Secondly, doctors failed to discuss the deviation from the UKALL guidance with Mr and Mrs J when it sought their consent. As a result, Mr and Mrs J did not have the opportunity to make an informed decision about their daughter’s treatment. This will continue to cause uncertainty and therefore distress for them.
Deterioration
43. Mrs J complained that when L started to deteriorate, leading to her admission to PICU in the early hours of 30 August, this was not recognised and acted on in a timely way.
44. In its investigation report, the Trust recognised they were slow in introducing sepsis management in line with the NICE guidance on neutropenic sepsis. This led to a delay in starting antibiotics. The Trust identified that their own Early Warning Score (EWS) system was not complied with. This is a system for scoring physiological measurements that are recorded routinely, including respiration rate, oxygen saturation, blood pressure, pulse rate, level of consciousness and temperature, in order to help identify patients who are deteriorating.
45. L’s EWS was recorded as green (i.e. least concern and no need to change the frequency of observations or escalation) up to 16:00 on 29 August. Soon after, at 18:06, she was amber. Her pulse had risen to 150. This meant observations should have been repeated every hour.
46. A nurse went to record observations at 7pm. L was asleep and settled after having been unsettled for the previous days. The nurse therefore decided to let her rest and not to disturb her for further observations. The Trust recognised in its response and investigation report that this was not in line with its EWS system. L’s observations were actually done again at 20:38, which was two-and-a-half hours after the amber. By this time, she was recorded as red, which is the most concern. Her oxygen levels were at 88%, blood pressure at 81/39, and pulse at 145.
47. We found there was a failure in monitoring. It was not in line with the NMC Code, which says ‘Make sure that people’s physical, social and psychological needs are assessed and responded to’.
48. We cannot say what the EWS would have been if observations had been taken at 7pm, but our adviser says they may have shown a deterioration sooner. A doctor would probably have reviewed L sooner on the evening of 29 August, and if showing signs of sepsis then, started antibiotics. There is a potential delay of around an hour and a half, if the 7pm observations had shown L’s condition has worsened. This is based on the red score at 8:38pm. We considered what impact this had. Although the sooner the better for starting antibiotics (NICE guidance CG151 says ‘immediately’) our adviser says this is a fairly short delay. We cannot make firm conclusions here, but our adviser said it likely it would not have changed the outcome.
49. In this regard, the Trust’s investigation report said ‘Whilst it is likely that the outcome would have been the same, given her disease and co-morbidities, the optimum treatment was not provided and it cannot be determined with certainty that this would not have changed the outcome.’ From our view of the evidence, including our clinical advice, we find this is a reasonable conclusion.
50. The Trust’s investigation report has recommendations relating to the EWS score and administering of antibiotics in line with the NICE guidance. It says ‘Compliance with the existing Trust Policy: Paediatric Observation Policy incorporating ManChEWS2 (Manchester Children’s Early Warning Score) Version 2 should be improved’ and ‘Work should be undertaken to improve compliance with administration of antibiotics thereby ensuring NICE neutropenic sepsis guidelines, November 2015, are routinely followed’.
51. The Trust’s Action Plan and Learning Assurance Report from February 2022 considered its learning from L’s case. We noted it said that due to the impact of COVID-19, a planned audit of EWS for 2021/22 had delayed but would take place in the first part of the 2022/23 financial year. We asked the Trust for further information about what had subsequently happened. It said that the audit had been planned but the Trust had implemented a new electronic patient record system (called HIVE) in September 2022. In conjunction with that, the National Paediatric Early Warning Score system (NPEWS) was introduced in the hospital. The focus of training was now on the new system but the Trust assured us that training for and monitoring of compliance with the existing EWS continued to be monitored in the meantime. It explained that the new patient records system allows for immediate evaluation of compliance and continuous monitoring of the NPEWS tool.
52. EWS is of more limited value to monitor deterioration in children with Down’s Syndrome. This is confirmed in the UKALL guidance, which says ‘DS patients may not present with classic signs of sepsis such as pyrexia’ [high temperature] and ‘Be alert to early signs of shock in septic patients and refer promptly for intensive care’.
53. We therefore considered evidence of what happened during the day leading to L’s admission to the PCIU to see if there were other, earlier missed opportunities to act on a deterioration.
54. We recognise that chemotherapy makes the patient feel poorly anyway, and potential symptoms of infection might reasonably be explained as the side effects of chemotherapy.
55. Our adviser said there is no specific guidance to mitigate for the reduced effectiveness of the EWS for children with Down’s Syndrome. On 29 August, nurses took L’s observations at 6:30am, 7:20am and 9:14am. Each time, the EWS score was green. As we say, this is not as reliable an indication for children with Down’s Syndrome. However, this is not to say it is completely unhelpful and does provide some relevant information. These did not cause any concern that L was deteriorating.
56. The nurse in the last of these noted L had abdominal distention and asked for a doctor to review her. Soon after that, L was seen by a Specialist Registrar. (The time is not recorded in the medical notes.) So there is evidence that an unexpected problem (distended abdomen) was acted on. The doctor recorded that L had been unsettled overnight, was nauseous, not eating and drinking much. Her abdomen was distended. L was sleeping when the doctor saw her. The doctor found that L’s lungs were clear and that her abdomen was ‘soft, not tender’. The plan was to increase her laxative and for an abdominal X-ray if she still did not open her bowels.
57. Mrs J says the abdominal examination was cursory. We carefully considered the record of this encounter and we consider it is in line with GMP requirements on record keeping. We appreciate Mrs J, with her own experience does not have confidence at what the doctor did at this time. We understand her concern and do not dimmish her experience. It is difficult to comment more in this regard. Our adviser said he saw no indication that it would have made a significant change to L’s management.
58. L had a lumbar puncture under general anaesthetic that morning. Early Warning Scores at 1.30pm and 4pm in the afternoon were green. She was well enough to leave the ward for a while.
59. The Trust’s failure to recognise L’s deterioration as soon as it should have understandably caused distress to Mr and Mrs J. We recognise their continued uncertainty around this. In its responses, including the investigation report and the Learning Assurance report, we believe the Trust has shown it recognises that it got this wrong. While not all changes are a direct result of L’s case (for example, the newly implemented patient records system), we find the Trust has acted in line with our Principles, which say that when things have gone wrong, public bodies should give an apology, explanation and acknowledgement of responsibility and ensure that all feedback and lessons learnt from complaints contribute to service improvement.’ We did not find there was more we can reasonably expect the Trust to do in this regard.
Communication and involving Mrs J
60. Mrs J complained that clinicians at the Trust did not listen to her and did not sufficiently include her as part of discussions around the investigations they were carrying out and the care and treatment they were giving to L. Mrs J is a nurse herself (in emergency medicine) so clearly has more knowledge than the layperson. More than that though, she was L’s mother and knew her daughter better than anyone. She could communicate with L more effectively that any stranger could and could recognise when things were not right or were deteriorating. She does not think this knowledge was recognised and therefore used effectively in L’s management. Mrs J feels that staff dismissed her concerns about L’s deterioration.
61. We recognise Mrs J’s frustration about what was happening, especially when L deteriorated quickly on the evening of 29 August. Her account of her frustrations at that time is compelling. It is difficult for us to conclude exactly how communication was done. We considered how issues around communication and Mrs J’s involvement affected L’s management around this time.
62. We examined the nursing and medical notes. They show evidence of communication with Mr and Mrs J. They suggest important changes in situation were communicated with them. As such, we consider records were what we would expect and in line with Good Medical Practice, which says ‘You must be considerate to those close to the patient and be sensitive and responsive in giving them information and support’. The record of the conversation about leukaemia is short, but we recognise the discussion itself was probably longer. Records of such conversation are not usually verbatim.
63. Mrs J made it clear however, that the communication issue was more than about being kept informed about what was happening. She is worried that her input was not taken on board. Having carefully considered the recorded evidence, our adviser did not see any evidence of failures which occurred as a result of staff not listening. This is not to diminish Mrs J’s frustration and distress.
64. There are instances when nurses and doctors recorded what Mrs and Mrs J had said. While we recognise it feels below what Mrs J wanted and that it caused her distress, we did not see anything that was missed in terms of L’s care (we address pain relief below).
65. There was little in the responses to say how staff had taken into account the fact L had Down’s Syndrome. L was only three and our adviser explains that for any three-year-old, communication is primarily with the parents, regardless of whether they have a disability. We do not presume to tell Mrs J that she was wrong to be concerned about how staff involved her. In terms of L’s overall care and treatment, from what we saw, we cannot say that there is evidence that L’s care was compromised by a lack of understanding about her needs.
66. We saw no evidence of failings in this regard. Overall communication was in line with the requirements of Good Medical Practice.
Pain management
67. Good Medical Practice tells doctors:
In providing clinical care you must: • 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 • provide effective treatments based on the best available evidence • take all possible steps to alleviate pain and distress whether or not a cure may be possible
68. The Royal College of Nursing says ‘Assessing and managing pain are essential components of nursing practice’. (www.rcn.org.uk/library/subject-guides/pain). The NMC Code tells nurses ‘Keep clear and accurate records relevant to your practice’.
69. The Trust’s investigation report acknowledged nursing pain assessments were not completed during the admission. Our adviser explained there is no national guidance or standards on acute, non-surgical pain. Clinicians need to look for cause of pain, assess the level of pain and measure the response of pain relief. That is what a pain management tool is for. The Trust had its own pain management tool and this was not used. As such, we find there was a failure in managing L’s pain. In the absence of robust assessment of L’s pain, we could not conclude that her pain was managed as effectively as it should have been. We can see why this would be so upsetting for Mr and Mrs J.
70. With regard to what the records show was happening, there was little in the medical notes about pain up to 27 August. A nurse gave L oramorph (a liquid form of morphine) at 1:00pm on 25 August because she was unsettled. She also had oramorph the next day.
71. At 12:20pm on 26 August, a doctor saw L and prescribed paracetamol and PRN (as required) oramorph. A doctor reviewed L at 10:05am on 27 August, who included ‘analgesia’ in the management plan. The note says she was ‘unsettled’ rather than in pain.
72. A nurse noted around an hour and a half later that Mrs J asked for regular oramorph and paracetamol as L was in pain. She was not reviewed by a doctor again that day. In the nursing records, she was later noted to be more settled due to pain relief.
73. L was reviewed by a specialist registrar on 29 August. Later that evening, another doctor reviewed her, noting she was more settled after starting Buscopan (a medicine used for treating stomach cramps) and was currently on regular paracetamol and regular oramorph. This is when L was diagnosed with sepsis. Mrs J told the doctor L had been unwell, unsettled and in pain for the last three days and she had not been investigated properly.
74. The problem here is that the Trust cannot point to a thorough assessment and plan to show it had taken appropriate action to manage L’s pain. It is evident staff were considering L’s pain and took action to relieve it by giving her oramorph, paracetamol and Buscopan. From a medical point of view, doctors acted in line with the section of Good Medical Practice as described above. The problem was the poor assessment and associated recording. As such, we find there was a failure. Record keeping was not in line with the NMC Code’s requirements as described above.
75. The Trust’s investigation report acknowledged this. This is in line with our Principles in being ‘open and accountable’. The investigation report included a recommendation to improve pain assessments and documentation. The individual actions identified were to share learning at nursing staff team and undertake local learning sessions to influence understanding and compliance. We thought it had little detail of what actions they had taken as a result. We therefore asked the Trust for further information about improvements.
76. The Trust said a number of changes had taken place. They now have a Quality Care Round (QCR) process. This is a spot check of records, which the senior nursing team carries out monthly. The checks include compliance with pain scoring, staff’s response to reports of pain and evaluation of the effect of pain relief. The outcome is reported to the hospital’s Patient Experience Committee. The Trust also monitors pain management in patient and family surveys.
77. It went on to say that pain scoring and documentation is in the process of being made a mandatory field within the Trust’s new electronic patient recording system. The Trust explained that once this has been completed, they will be able to continuously audit and monitoring pain management and ensure records are kept appropriately.
78. Our Principles of Good Complaint Handling, say that organisations should put things right by acknowledging mistakes and apologising where appropriate’. They should also seek continuous improvement ‘using all feedback and the lessons learnt from complaints to improve service design and delivery’ and ‘where appropriate, telling the complainant about the lessons learnt and changes made to services, guidance or policy’. Some of the Trust’s changes were not implemented as a result of Mrs and Mrs J’s complaint, but we considered whether the Trust has put things right as far as we can reasonably expect. We consider they have done so in line with our Principles. This is not to downplay the distress caused by the fact we and the Trust cannot assure Mr and Mrs J that L had the best possible pain relief.
Referral to the coroner
79. L died on 7 September 2018. The Trust did not to refer her death to the coroner at that time. They decided to do so in July 2019. Mrs J complins that the referral should have been done immediately and the unreasonable delay added to her and her family’s distress.
80. Mrs J first contacted us soon after that referral. Manchester Coroner’s Office had told them ‘The Trust has informed this office of the outcome of the internal investigation, which gives the coroner grounds to open an inquest. However, L's death was originally discharged as a natural event, therefore we must apply to the Ministry of Justice and the Chief Coroner to seek authority to investigate further.’
81. In early October 2019, the Manchester coroner’s office contacted us to say the permission had been granted. We decided to allow that process to be completed before considering Mrs J’s complaint. The process was delayed and the inquest was not concluded until 7 February 2022.
82. Government guidance (‘Guidance for doctors completing medical certificates of cause of death in England and Wales’) says ‘Doctors… have a legal obligation to report certain categories of deaths to the coroner before they can be registered. These include deaths where there is reason to suspect the death was unnatural, unknown, violent or where the death occurs in custody or otherwise in state detention. This can include where the death is due to the person undergoing a treatment or procedure of a medical or similar nature: for example, a death during an operation or before full recovery from an anaesthetic. In addition, there will always be cases which may on one view be ‘natural’ which have some other element (for examples concerns about neglect or standards of care) which brings them within the orbit of the coroner. These should be reported to the coroner.’
83. Trust doctors did not think L’s death needed to be reported to the coroner because at the time, they thought her death had been natural. We have considered whether there is any indication of failure in the Trust not referring L’s death immediately to the coroner.
84. We have seen no indication that the following are disputed: when she was admitted to Manchester Children’s Hospital, L was diagnosed with ALL and she contracted an infection, which led to sepsis. The coroner concluded the L died of multiorgan failure caused by sepsis.
85. We appreciate why Mr and Mrs J were distressed by the decision to refer L’s death to the coroner months later. We recognise the Trust could have referred it to the coroner immediately. If that had happened, additional distress could have been avoided. Our consideration is not whether it would have been better course of action, but whether the decision not to refer was so unreasonable as to constitute a failure. Taking into account guidance, we do not consider that it was. The cause of L’s death was known and at that time, we think it was not unreasonable for doctors to consider her death was natural.
86. The later investigation raised some concerns about whether prophylactic antibiotics should have been considered for L, and that there was a delay in administering antibiotics when L deteriorated on 29 August. The Trust’s report did not identify these had contributed to her death, but the Trust thought it should be a matter for the coroner to consider. In other words, there were now ‘concerns about neglect or standards of care, which brings them within the orbit of the coroner’ as described in the guidance’.
87. During the inquest, the coroner said he thought it would have been appropriate to have referred L’s death to him sooner; the family should have had clarity of what processes were being taken. He had informed the Trust of this so that does not happen again.
88. We partly uphold the complaint. We found that when doctors sought consent for the UKALL trial, they did not explain to Mr and Mrs J that they would not be following the recommendation on prophylactic antibiotics. Furthermore, there is no written record of the policy to show how it was made and what evidence was considered. While we cannot say there is evidence the outcome would have been different, we recognise the uncertainty and distress these failures caused Mr and Mrs J.