Dapagliflozin 26. We know how concerned Mrs G is, that the Trust gave her husband dapagliflozin. We hope to assure her we find dapagliflozin was appropriately prescribed and administered following consultant advice at the medical ward round, on 1 September.
27. As explained in NICE TA679, NICE TA902 and in BNF guidance, dapagliflozin is licenced and clinically recommended to treat chronic heart failure with reduced ejection fraction. Heart failure is a condition that occurs when the heart is unable to pump enough blood to meet the body's needs. This is called the ejection fraction. Left ventricular ejection fraction is the amount of blood pumped by the left ventricle out to the body during each heartbeat.
28. Information on the NHS website explains dapagliflozin reduces the amount of work the heart has to do to pump blood around the body, helping to improve symptoms of heart failure such as breathlessness, tiredness and swelling of the lower legs. It also reduces the chance of the heart becoming weaker.
29. Mr G had heart failure, and echocardiograms taken during his admission showed his ejection fraction was reduced. As such, dapagliflozin was appropriately prescribed and administered in line with NICE TA679 and NICE TA902.
30. When speaking with us, Mrs G voiced concern that dapagliflozin was inappropriate to give her husband as he did not have diabetes. Our cardiology adviser says this is an understandable misconception. They explain dapagliflozin was initially developed as a diabetes drug, however after being in use for several years, it was found that people who had heart failure alongside their diabetes were seeing a benefit in their heart failure symptoms.
31. It was then found even people without diabetes would see a benefit in their heart failure from dapagliflozin. The evidence-base is considerably strong and whilst it does have a basis in diabetes care, and a separate marketing authorisation as a glucose-lowering agent for type 1 and type 2 diabetes, dapagliflozin is separately licenced as appropriate heart failure treatment.
32. Mrs G is concerned the Trust failed to monitor the impact of dapagliflozin on her husband’s blood sugar. NICE TA679 says clinicians should: ‘Start treatment of symptomatic heart failure with reduced ejection fraction with dapagliflozin on the advice of a heart failure specialist. Monitoring should be done by the most appropriate healthcare professional’.
33. Our cardiology adviser says whilst NICE TA679 recommends monitoring, this is a recommendation that can reasonably be applied to any newly prescribed medication. They say there is no required monitoring of blood sugars specifically, that reasonably, the monitoring that is recommended in NICE TA679 would primarily involve consideration of Mr G’s nutritional intake. More explicitly, BNF recommendations advise testing and then discontinuing dapagliflozin only once signs and symptoms of DKA are seen, not before.
34. DKA stands for diabetic ketoacidosis. Ketoacidosis is a condition where blood sugar and ketone (acids) levels are raised, causing an acidic environment in the body. It is often termed diabetic ketoacidosis due to it being prevalent in those with diabetes. HHS is also characterised by high blood sugar levels, but without elevated ketone levels. As explained by Diabetes UK, HHS often develops more slowly than DKA, over many days.
35. DKA is a known side effect of dapagliflozin, although importantly, it is categorised by the BNF as a ‘rare or very rare’ risk. As explained, BNF recommendations state: ‘test for raised ketones in patients with signs and symptoms of DKA… and discontinue treatment if DKA is suspected or diagnosed’. Our cardiology adviser confirms the same can reasonably be applied for HHS.
36. We do not see evidence to suggest any failure in the Trust’s monitoring of the impact of dapagliflozin on Mr G’s blood sugar. In line with BNF recommendations, specific monitoring of blood sugar or ketone levels was not required until there were apparent indications of DKA/HHS. We are not critical of the Trust for not monitoring Mr G’s blood glucose levels for this reason. Our cardiology adviser says as long as there was some nutritional intake, then dapagliflozin is typically considered reasonable to continue to administer.
37. That said, we think as a minimum the Trust should have documented its consideration of the associated risk of DKA/HHS from dapagliflozin as events unfolded. We do not find any evidence in the records to show consideration of this. We can see dapagliflozin was stopped on 9 September, when signs and symptoms in line with the above were apparent and Mr G was first found to be in HHS. This was appropriate. Yet, we find evidence it should have been stopped sooner, as early as 3 September. We go on to explain our thinking here.
38. At the ward round on 2 September, the MDT’s recommendation for heart surgery is documented. The management plan included note to withhold Mr G’s enoxaparin (blood thinning medication) from the evening of 3 September in preparation for this upcoming surgery. At the ward round on 3 September, it is noted Mr G was scheduled for surgery the following day.
39. The UKCPA Handbook recommends stopping dapagliflozin 48 hours before elective surgery. MHRA advice is to withhold an SGLT2 (a type of drug such as dapagliflozin) for patients who are hospitalised for major surgical procedures.
40. On 3 September it was known Mr G was planned for elective surgery the next day. For this reason, in line with the above recommendations, dapagliflozin should have been stopped. Records show consideration on 2 September of stopping another medication in anticipation of upcoming surgery. As a minimum, consideration for dapagliflozin, whether to discontinue or entry of a note to explain any clinical rationale for it continuing, should have been documented alongside the Trust’s consideration for other medications before surgery.
41. Records on 3 September also show Mr G’s CRP and WCC levels were rising. This was evidence of a developing infection. Our cardiology adviser says for this reason, dapagliflozin should also have been stopped or at the least, we would expect the Trust to have documented its consideration of dapagliflozin. If the plan was for it to continue, reasonably the clinical rationale should have been documented.
42. We think not appreciating the significance of continuing dapagliflozin on 3 September in the absence of any documentation about it and not stopping dapagliflozin due to rising infection markers and/or due to upcoming planned surgery, are not in line with the above cited guidelines. We identify this as service failure.
43. It is our view Mr G’s HHS almost certainly developed due to dapagliflozin. We emphasise that we find dapagliflozin was prescribed to him appropriately, including in dosage. We would not say dapagliflozin should not have been given to Mr G, even with knowledge of the associated risk of DKA/HHS.
44. We also do not find any apparent failure in the Trust’s monitoring. As we have explained, specific monitoring of blood sugar or ketone levels is not required until there are apparent signs of DKA/HHS. This was not the case for Mr G by 3 September. Stopping or consideration to stop dapagliflozin for the reasons of no nutritional intake was also not an apparent concern by 3 September, as at this time he was taking appropriate oral intake.
45. However, we think 3 September is when the Trust should have stopped his dapagliflozin for two reasons, firstly because of developing evidence of infection, and secondly because of his upcoming surgery. Even though surgery did not go ahead, this was not known 48 hours before when dapagliflozin should have been stopped, in line with UKCPA Handbook recommendation and MHRA advice.
46. Whilst the matter of upcoming surgery was no longer an issue from 4 September, we find the Trust missed further opportunities in the days between 4 and 9 September, to either stop dapagliflozin because of Mr G’s rising infection markers, or to have documented the clinical rationale for it continuing. In addition, records on 7 September note his eating and drinking was compromised. Our cardiology adviser says this is a factor that required clinical consideration for the appropriateness of continuing dapagliflozin from that date forward.
47. Our cardiology adviser confirms the HHS which was first apparent on 9 September almost certainly developed due to the continued administration of dapagliflozin. We think it entirely possible HHS would never have developed, if not for the Trust’s failure to stop it - or at the very least record a clinical rationale for it continuing with additional monitoring actions – on and from 3 September. This view is supported by the fact Mr G was not diabetic previously, meaning the only risk factor present in his case for developing HHS was the dapagliflozin itself.
48. In terms of the impact, Mrs G says her husband never recovered from the comatose state caused by HHS, which left him mostly unresponsive in his final week. She says she and their sons lost precious time to talk with and be around Mr G as he always was in the time before he died. We were very sorry to have learned directly from Mrs G and her son, about their experience at this time. It was clearly incredibly distressing and upsetting for them all as a family.
49. The recorded evidence shows Mr G had a period of reduced responsiveness on 9 September, with an early Glasgow Coma Scale (GCS) measurement of 9, and later that day a GCS of 13 after HHS treatment had been started. GCS is a clinical tool used to assess a person’s level of consciousness based on their response in three areas: eye, verbal and motor responses. The highest score is 15, indicating a person is fully awake and aware, with the lowest score of 3 indicating a deep coma.
50. At the ward round the following morning, it is noted Mr G’s consciousness had improved further, and he had a GCS of 15. The entry notes Mr G felt better, with significantly improved confusion and he was alert. A clinical review in the afternoon of 10 September recorded a GCS of 15, noting Mr G was lethargic but more alert than yesterday, with only a slight reduction later that afternoon with GCS 14 noting Mr G was sleepy.
51. Records continue in this vein. We do see times where it is noted Mr G had some confusion and delirium, yet by 13 September records note he was oriented to place, recognising staff members and expressed he was ‘doing good’. From the records we have seen, we are not left with a picture of Mr G being in a clinically comatose state through to his death, as is claimed. It remains the HHS did cause a period of some days where Mr G had reduced consciousness and confusion, and we accept that the quality of time Mrs G and her family had with Mr G from this time forward was diminished as a result. We recognise how distressing this was for Mrs G and her family.
52. As another impact, Mrs G says the HHS combined with the delayed potentially life-saving surgery put strain on her husband’s heart, causing his death. Mr G was clearly scheduled for urgent treatment (surgery), and that treatment was delayed in part because of his HHS. However, evidence shows the primary reason treatment was delayed was due to evidence of a developing infection, seen from 3 September.
53. We know Mrs G has also complained that surgery was delayed despite the Trust finding no apparent infection. We can clarify that the clinical evidence does indicate an infection was present. We continue our thinking here in relation to the issue of dapagliflozin, and we explain our thinking on the matter of apparent infection in the next section of our report.
54. Mr G’s infection did not develop because of any failing. Whilst an increased risk of infection is a known ‘common or very common’ side effect of dapagliflozin as categorised by the BNF, even if this infection was related to the dapagliflozin, that initial prescription and the giving of dapagliflozin to Mr G up to that time was appropriate.
55. Mr G developed the additional complication of HHS because of the failing we identify. The HHS was another, appropriate reason to delay his treatment, alongside the persistent infection. Yet, we do not consider these entirely separate factors. We think Mr G was denied the chance to potentially receive the necessary treatment for his heart, because of the Trust’s failure to have not stopped dapagliflozin on 3 September.
56. Our cardiology adviser says that had dapagliflozin been stopped as we think it should, the outcome could have been different. In that circumstance, Mr G’s infection would have had chance to resolve without the possible contributory side effect from dapagliflozin and the additional complication of HSS, which would likely not have developed.
57. Considering his clinical condition on 3 September, our cardiology adviser thinks Mr G would have had a better chance of recovering from his infection quicker than his recovery from infection plus HHS. This may have resulted in him having chance for the surgery to be reconsidered, rescheduled, and to have potentially gone ahead prior to his fatal heart attack. IJoC research reported the type of surgery planned for Mr G reduced the risk of mortality (death) and MI, compared to optimal medical therapy.
58. The course of events, and the outcome, may therefore have been different. We know this will be incredibly distressing for Mrs G and her family to read.
59. We must however make clear, that we cannot say the course of events or the outcome would have been different with any certainty, strong likelihood, or even on the balance of probabilities. We cannot say the infection would have responded or resolved sooner, or that surgery would have been able to have been reconsidered, rescheduled or proceeded at any earlier time before Mr G’s sad death. There are too many unknown factors and there is no quantifiable data, for us to say whether one outcome was more likely than not.
60. We can say the chance of Mr G’s infection resolving to allow for treatment to be reconsidered, potentially rescheduled, and to possibly go ahead, was affected by the continued giving of dapagliflozin. We can say, if not for the failing, there was the possibility of a different outcome and that Mr G was denied the chance of this. We can also say that knowledge of this will cause Mrs G understandable, avoidable distress. We outline recommendations to the Trust in response to these impacts, as set out at the end of our report.
Surgery and infection 61. We return to the issue referenced earlier, of Mrs G’s complaint that the Trust delayed urgent bypass surgery despite finding no apparent infection. As we have explained the decision to delay surgery due to Mr G’s raised CRP and WCC levels was appropriate. We expand further on this here, by explaining the relevant guidance.
62. RCS guidance says surgeons have a duty to comply with systems and processes that aim to reduce the risk of harm to patients. It says surgeons must recognise the risk of surgical site infection and the potential for cross-infection and follow local infection control procedures. In addition, Infection Prevention in Practice guidelines set out clinical advice on patient decontamination and the requirement to use antiseptics and antibiotics to reduce the risk of an infection at the surgical site.
63. Any active infection in the body prior to surgery would increase the risk of infection at the surgical site. Not only does this carry its own risks, but it would also reduce the patient’s chance of a straightforward and uncomplicated recovery following surgery. We know Mrs G and her family query whether surgery should have gone ahead irrespective of the infection, that whilst there was risk due to the infection, there was risk to Mr G of it not proceeding. Our cardiology adviser says the decision to delay surgery, to resolve Mr G’s infection first, was appropriate and in line with RCS guidance and Infection Prevention in Practice guidelines.
64. Addressing Mrs G’s complaint that the Trust found no apparent infection, there is a necessary distinction to make here between whether there was or was not any infection and where the source of any infection may have been. Records are clear that the clinical signs indicated that Mr G had a developing infection. Records are clear in noting that despite concerted efforts, the Trust was unable to identify the source.
65. Our cardiology adviser says it is established good practice and appropriate clinical care, to start treating signs of infection even before its cause or source has been identified. This is what the Trust did. That the tests the Trust took did not identify the source of the infection, does not mean that there was no infection present. We hope this provides Mrs G and her family further assurance on this issue and of the decision-making regarding surgery.
Nutrition 66. Mrs G complains the Trust failed to provide her husband with appropriate nutrition. We hope to assure her we find evidence of appropriate nutrition management which was given in line with clinical guidance.
67. Mr G arrived at the Trust on 30 August and was first weighed on 1 September, within 24 hours of his arrival. He was next weighed on 6 September, five days later. Our dietician adviser confirms this was appropriate practice in completing assessment and nutritional screening in hospital, in line with NICE CG32. This says all patients should be screened for the risk of malnutrition on admission, with screening repeated weekly for inpatients.
68. NICE CG32 goes on to say that screening should assess the percentage of any unintentional weight loss, the time over which nutrient intake has been unintentionally reduced, and/or the likelihood of future impaired nutrient intake. We find evidence to show the Trust complied with this guidance.
69. When he was weighed on 1 September Mr G’s weight was 82.5kg. When weighed on 6 September two entries are made, the first of 69.4kg and the next, later in the day, of 76.4kg.
70. Considering his history before admission, including family reporting he regularly played golf and did not have any apparent health problems, Mr G appeared to be keeping fit and well. Our dietician adviser says nutritional risk is typically lower in this type of patient, compared to those whose history presents a picture of them being frail and deteriorating, with a chronic illness or having struggled over a period of months.
71. Given this, alongside Mr G arriving at the Trust with pulmonary oedema and receiving medication in the form of diuretics (‘water tablets’) to remove this fluid overload, the change in documented weight between 1 and 6 September is reasonably explained by this loss of fluid. Our dietician adviser explains it is not uncommon to see such a considerable weight loss in patients with fluid overload over a short period like this, as treatment comes into effect.
72. Our dietician adviser goes on to explain that achieving an appropriate fluid balance is complex. Patients may need diuretics or other similar treatments in attempts to reduce fluid overload, yet they may also need to be encouraged to drink or even given supplemental fluids to ensure they remain hydrated. Whilst the change in Mr G’s documented weight on 6 September may seem concerningly significant, our dietician adviser explains this can be reasonably ascribed to him being given fluids then re-weighed, in the Trust’s best attempts to better manage the balance of his fluid.
73. The later recorded weight on this day was 76.4kg, and this gave Mr G a healthy body mass index (BMI) of 25. This is assurance that even after his weight loss from 1 September, Mr G was not underweight to have raised concern or required any alternative or additional action on his weight loss alone, at that time. We recognise the amount of loss from 1 September to this second measurement on 6 September was rapid, yet as we have explained, this is not uncommonly seen in patients being treated for oedema for the reasons explained.
74. In terms of recording Mr G’s dietary input, we find nursing notes are full, clear and accurate, in line with these stated expectations within the NMC Code. Daily entries from admission note Mr G was eating and drinking well up to 4 September when notes state he was not taking food, feeling sick and put off by the smell. The next day, notes state Mr G was eating and drinking well again, then on 6 September it is noted his food intake was poor.
75. This further supports our view that the weight loss documented on 6 September was most likely fluid loss, considering by 6 September Mr G had only demonstrated a reduced appetite across two separate days.
76. Appropriately, in response to observations of poor food intake, a referral to the dietetic team was made on 6 September. Mr G was seen by dietetics, thereby receiving specialist input, the next day. This action was line with the NMC Code, which says nurses should identify any risks or problems that have arisen, refer matters to other practitioners and colleagues when required, and make records for colleagues to have all the information they need.
77. The timeliness of this action was also appropriate, in line with NICE CG32. This says nutrition support – which includes specialist dietetic input – should be considered in people at risk of malnutrition. One of the criteria defining a person at risk of malnutrition is when they have eaten little or nothing for more than five days and/or are likely to eat little or nothing for the next five days or longer. Considering Mr G had only demonstrated eating ‘little or nothing’ for just two days on 4 and 6 September, we think it very timely the Trust acted in making the referral promptly, on that second occasion.
78. Records show Mr G received frequent dietetic input, on 7, 11, 13 and 15 September. Documentation shows thorough dietetic assessments took place, in line with the expectations of assessment and what should be considered, set out throughout NICE CG32.
79. Our dietician adviser comments that the first assessment on 7 September was particularly proactive, as this noted consideration for a feeding tube if Mr G continued to struggle with adequate oral intake. This is in line with a specific recommendation in NICE CG32 where it says healthcare professionals should consider using oral, enteral (via the intestine) or parenteral (via anywhere in the body other than the mouth) nutrition support, alone or in combination, for people at risk of malnutrition.
80. An NG tube was inserted just two days after this early consideration for it, on 9 September. Considering Mr G’s oral intake had started reducing from 4 September, our dietician adviser confirms tube insertion for feeding was in line with the above guidance in NICE CG32. Considering his oral intake had only started reducing on 4 and then 6 September, the date of its insertion in Mr G’s case was appropriately timely.
81. We know from speaking with Mrs G she is concerned about the NG tube being sited on 9 September, as during a conversation with a doctor on 8 September family were advised this was not yet needed and potentially not appropriate. Our dietician adviser hopes to assure Mrs G that it was entirely clinically appropriate to have sited the NG tube on 9 September, to ensure Mr G could receive adequate nutrition in the context of his reduced oral intake of food. We are sorry to hear a likely unfortunate issue with communication added to the family’s distress at that time.
82. Once sited, NG tube feeding is well-documented, and our dietician adviser confirms feeds were given in line with clinical expectations to meet Mr G’s nutritional requirements. They comment that the dietician’s documentation shows a very thorough consideration of Mr G’s circumstance, evidencing good clinical practice on each occasion.
83. Records go on to show appropriate amendment to the NG tube feeding regime when Mr G’s clinical status changed. For example, when it was seen that his sodium was high, records show good communication with the clinical team about this, in dietetic requests to amend the feeding regime around those clinical issues. This was in line with GMC guidance which says clinicians must adequately assess the patient’s conditions, promptly provide or arrange suitable treatment where necessary, and record relevant clinical findings, who is making those decisions and agreeing the actions.
84. Although Mr G was able and allowed to eat and drink after each dietetic assessment, it is documented his oral intake was negligible and so his full nutritional requirements were given via NG tube. This was appropriate action for the Trust to take, to ensure his nutritional requirements were met. The Clinical Nutrition article we cite provides evidence that even in this circumstance, this does not impact on a person’s want or ability to eat.
85. Mrs G raises specific concern that the Trust failed to provide her husband appropriate nutrition considering his raised blood sugar, poor oral intake, weight loss and drowsiness. We have addressed the issue of his weight loss and oral intake in our explanations above.
86. Regarding Mrs G’s concerns of blood sugar and drowsiness, our dietetic adviser confirms the thorough dietetic assessments took account of the full extent of Mr G’s clinical status on each occasion. They include notation of Mr G’s HSS, consideration for his blood sugar levels and need to amend carbohydrate intake within the feeding regime. Our dietetic adviser provides assurance that they find no evidence to suggest any concern with the Trust’s provision of nutrition and this not taking account of or having any direct association with Mr G’s blood sugars or drowsiness.
87. In conclusion, records show the Trust provided Mr G with appropriate nutrition and good nutritional care, in line with guidance.