University Hospital Southampton NHS Foundation Trust

7. Mrs E complains that in April 2024 the Trust overdosed her husband, Mr E, with oxygen and left him unattended for two hours or more.

8. She said that the overdose meant his condition deteriorated and he needed to use a breathing machine. She said these two things caused additional stress to his lungs and led to his death later that month, and had it not been for the overdose his death could have been avoided.

9. Mrs E says that because her husband appeared frail, the Trust did not properly explore his condition and wrongly and prematurely decided that he was at end-of-life. She said that the Trust did not give her husband the time he needed to get better, and it did not provide the level of respect due to a patient in need of care, which caused her distress. She said that he was making plans for his discharge and looking forward to getting fit and spending time with his family. She said that their family misses her husband and are distressed by their ongoing concerns that he should not have died when he did.

10. Mrs E wants service improvements and a financial remedy.

11. Mr E was a 68-year-old man. We can see from Mr E’s medical records he had end stage Chronic Obstructive Pulmonary Disease (COPD), which is a progressive lung disease, and he was frail with a body mass index (BMI) of 14.5. In early March 2024, Mr E attended the emergency department (ED) at a hospital within the Trust. He had stomach pain which was affecting his breathing. Blood tests showed that his blood was acidotic on arrival to the ED, which indicated that his lungs were struggling to remove carbon dioxide.

12. The Trust admitted him to the RHDU. It treated Mr E with non-invasive ventilation (NIV), which is breathing support which uses a machine to actively push air into the lungs and supports people who have difficulty taking deep breaths in and out, nebulisers (which help administer high doses of liquid medication by turning it into a fine mist that can be inhaled), antibiotics and prednisolone (a corticosteroid to treat respiratory inflammation). Ten days later, the Trust transferred Mr E to a ward to manage his constipation and wean him off the oxygen he had been administered in the RHDU.

13. By mid-April 2024, around 4 weeks later, Mr E’s oxygen saturation level (the level of oxygen in the blood) dropped to between 50 and 60%. The Trust recognised that this was an unsafe level. It administered high flow oxygen therapy (a form of oxygen therapy used for respiratory support) to help Mr E’s oxygen saturation level increase to a target level of 88-90%.

14. Trust staff left Mr E unattended while they administered the high flow oxygen. His condition deteriorated further. When ward staff noted his deterioration, they called the emergency team who helped support the ward staff caring for Mr E.

15. Mr E’s condition improved when staff removed the high flow oxygen. The ward transferred him back to the RHDU where the medical team started treatment for hospital acquired pneumonia with intravenous antibiotics. The Trust also administered further NIV to help support Mr E.

16. Mr E also had frequent exacerbations (a flare up when breathing becomes worse than usual), secondary pulmonary hypertension (high blood pressure in the pulmonary arteries) as documented on an electrocardiogram (ECG) taken in April 2024. The Trust completed an ECG two days after it administered high-flow oxygen therapy. This showed Mr E had an enlarged heart.

17. Despite the Trust’s efforts, Mr E’s condition continued to deteriorate while he was admitted to the RHDU. The respiratory team discussed Mr E’s condition with the palliative care team as it felt Mr E was sadly coming to the end of his life.

18. The palliative care team spent time with Mr E to discuss his wishes and decisions about the treatment he wished to continue with. Eleven days later, Mr E sadly died.

21. Mrs E said that the Trust ‘overdosed’ Mr E with oxygen in April and left him unattended for around two hours or more. She said that this resulted in his condition deteriorating, and him needing a breathing machine. She said that these two things stressed his lungs and led to his death later that month. She believes that his death could have been avoided, had it not been for the ‘overdose’.

22. The Trust explained that its treatment plan after it admitted Mr E was to manage his constipation and gradually wean him off oxygen, which staff administered to support his breathing when he was first admitted to the RHDU. When his oxygen levels dropped in early April 2024, it administered a high flow of oxygen via a rebreather mask (a mask which delivers high-concentration oxygen and prevents a patient from rebreathing the carbon dioxide they breathe out). Its response letters explained that it could not confirm the flow rate of the oxygen and it was either 10l/min or 15l/min.

23. The Trust said that administering high-flow oxygen was the correct course of action. However, in its complaint response, the Trust also recognised the level of oxygen it administered was an excessive amount.

24. When the Trust became aware that it had not monitored Mr E while he received the high-flow oxygen, it escalated his care to the emergency team who supported the ward in caring for Mr E. It removed the oxygen and Mr E’s condition improved, and it then moved him back to the respiratory HDU, where it treated Mr E for hospital-acquired pneumonia with IV antibiotics.

25. We discussed with our adviser the concentration of oxygen the Trust administered and how the Trust managed this on the day in question.

26. Our adviser highlighted that the BTS guideline sets out that the target oxygen saturation levels for patients with COPD should be between 88-92%. This is also known as the scale 2 target. Oxygen therapy can be administered to help patients such as Mr E to increase their oxygen saturations, helping them meet the scale 2 target.

27. Our adviser confirmed that Mr E’s records do not document the flow rate of oxygen the Trust administered. They said that oxygen should be titrated upwards with an aim of achieving the target saturation levels of between 88-92% as per the BTS guideline.

28. The BTS guideline makes it clear that when administering oxygen in a hospital setting, the patient should be carefully monitored for hypercapnic respiratory failure with respiratory acidosis (a critical condition where carbon dioxide levels in the blood increase, increasing blood acidity). Because Mr E was diagnosed with end stage COPD, he was at a higher risk of developing this condition.

29. It is noted that Mr E had an episode of low blood oxygen levels due to the hospital acquired pneumonia with a deterioration. The level of carbon dioxide in his blood would be higher than that of a patient who did not have pneumonia. Our adviser confirmed that even though his carbon dioxide levels were higher than in a patient without pneumonia, it was appropriate for the Trust to administer a higher oxygen concentration temporarily, to help Mr E meet the Scale 2 oxygen target. We find this decision was in line with the guidance and is not a failing in the Trust’s actions.

30. The BTS guideline explains that blood gases should have been repeated between 30 and 60 minutes after the oxygen was applied to check for rising carbon dioxide in the blood, the partial pressure of CO2 through pCO2 levels (the measure of how much carbon dioxide is pushing through the blood) or falling pH levels.

31. Mr E’s records document that alongside the end stage COPD, he had chronic type 2 respiratory failure, with frequent flare ups of his COPD symptoms. The pCO2 levels on his blood gas test results were at an elevated level, which meant Mr E was at a further risk of developing hypercapnic respiratory failure. When the pCO2 levels are elevated, further NIV may be required to treat the acute type 2 respiratory failure.

32. Mr E was left on high level of oxygen for two hours with no monitoring of his blood gases. The Trust acknowledged that it failed to appropriately monitor Mr E.

33. We find that the lack of monitoring of Mr E’s blood gases is not in line with the guidance and is a failing in the Trust’s actions. We have therefore considered the likely impact to Mr E.

34. Mr E’s pCO2 level was elevated before the Trust administered the oxygen, meaning he was at a higher risk of complications. The lack of monitoring means that on the balance of probabilities, it is likely that his pCO2 levels increased over time. Our adviser explained that the clinical features of a higher pCO2 level in the early stages are a little bit more drowsiness and mild confusion. In a scenario like Mr E’s case, these signs are normally quite subtle, and a patient’s deteriorating condition would be a gradual process over time. Our adviser explained that pCO2 levels would not acutely increase unless for some reason a patient’s airway was acutely, completely blocked, which was not the case here.

35. The American study looked at the effect of high flow oxygen over 15-20 minutes. This study explained that in patients with COPD and receiving high flow oxygen, deterioration in that scenario is a gradual process.

36. On the balance of probabilities, in a case like Mr E’s, you would see a steady deterioration over time due to excess oxygen therapy. We cannot know whether monitoring in line with guidance would have noted Mr E’s deterioration, as the changes in his condition may have been subtle. We can say that this was a missed opportunity to possibly treat the small increase in the pCO2 levels sooner and that the lack of monitoring led to Mr E needing acute NIV to treat his condition. The Trust ultimately did administer NIV and admitted him back to the RHDU, once the error was identified.

37. Our adviser explained that patients do not tend to die from pCO2 retention in the setting Mr E was in over a short period of time. It would take 10 to 15 hours for an increased pCO2 level to be clinically significant and much longer than the two hours Mr E was left unmonitored.

38. We have also considered Mr E’s comorbidities. He was frail, had severe end stage COPD, pulmonary hypotension secondary to COPD, a low BMI of 14.4, and hospital-acquired pneumonia. These are all significant and increased the likelihood of needing further NIV treatment throughout his admission, even without the unmonitored high-flow oxygen therapy. Despite this uncertainty we understand the lack of timely monitoring of Mr E’s pCO2 levels meant that it is likely there was an earlier need for further NIV treatment during the admission and the lack of monitoring meant this was slightly delayed.

39. The BODE index is a multidimensional scoring system used to predict the survival and long-term outcomes in patients with COPD. It uses a patient’s BMI, airflow obstruction, dyspnea (breathlessness), and exercise to calculate a score from 0-10. A higher score indicates a greater risk of mortality and more frequent exacerbations of the COPD and hospitalisations.

40. Our adviser explained that Mr E’s BODE index, based on the data available in his medical records, was the maximum score of 10. This equates to a four-year survival rate of 18%. The NIV report explains that the patients who require ward-based NIV to treat type 2 respiratory failure may have an increased in-hospital mortality risk of at least 20%. Patients with pneumonia are also at increased risk of in-hospital mortality with ward-based NIV. Sadly, Mr E’s mortality risk would have been higher still, due to his other comorbidities, especially his frail condition.

41. Our adviser explained that on the balance of probabilities, the contribution of the nuanced, gradual rise in pCO2 from the unmonitored high flow oxygen likely was a small factor in the overall picture of Mr E’s health. The culmination of his comorbidities were likely the key factors in Mr E’s death.

42. On balance, given Mr E’s comorbidities and frailty, and the information we have seen, we cannot say that his death was unexpected or avoidable. Nor can we say that based on Mr E’s medical records and the advice we have received that the outcome would have been any different had the high flow oxygen been monitored in line with guidance.

43. We recognise that the high-flow oxygen therapy was a major worry for Mrs E. We hope our report reassures her that we have not seen anything to indicate the error led to her husband’s death, as she feared.

44. Mrs E also said that the Trust did not properly explore Mr E’s condition, nor did it give him the time he needed to get better. She understands that her husband was frail, and she believes that the Trust prejudged her husband’s condition and incorrectly determined that he was at end-of-life. She said he went to hospital for treatment to manage his constipation, not to end up on end-of-life care. She tells us she expected the Trust to treat him with the respect and dignity he deserved but instead Mr E was met with staff with an indifferent attitude. She believes that the Trust’s view that he was end-of-life led to her husband effectively giving up.

45. Domain 1 of the GMC’s Good Medical Practice explains how clinicians should provide good clinical care. Section 6 says if you assess, diagnose or treat patients, you must work in partnership with them to assess their needs and priorities. The investigation or treatment you propose, provide or arrange must be based on this assessment, and on your clinical judgement about the likely effectiveness of the treatment options.

46. Section 7 says that in providing clinical care you must:

• Adequately assess a patient’s condition(s), taking account of their history, including: • Symptoms • Relevant psychological, spiritual, economic and cultural factors • The patient’s view, needs and values • Carry out a physical examination where necessary • Promptly provide (or arrange) suitable advice, investigation or treatment where necessary • Propose, provide or 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 will meet their needs • Propose, provide of prescribe effective treatment based on best available evidence • consult colleagues or seek advice from your supervising clinician, where appropriate • refer a patient to another suitability qualified practitioner when this serves their needs.

47. We can see from Mr E’s records that his symptoms were investigated by the appropriate clinical teams, based on his symptoms related to his constipation. He underwent ultrasounds and X-rays of his abdomen which did not show any abdominal pathology which would account for his stomach pain, aside from the constipation he suffered.

48. Mr E’s medical records show he had a short-term deterioration in his condition, which stabilised while he was first admitted back to the RHDU in April 2024. His condition slowly deteriorated after that time. Mr E did not respond to antibiotics and other treatment from the Trust to help manage his frailty and end-stage COPD. Sadly, these conditions led to his death 11 days later.

49. NICE Guideline NG142 directs organisations in terms of end-of-life and preparing for end-of-life care. It says that as adults approach the end of their life, healthcare providers should complete a holistic needs assessment, repeat assessments of their needs as circumstances change and provide highly skilled health and social care multi-practitioners to help support the adults approaching the end of their life, their carers, and their families.

50. The clinical teams involved in Mr E’s care reached a consensus to place Mr E on an end-of-life care pathway when his condition stopped improving and its treatment was no longer working. The Trust reached this decision in line with guidance, and it is supported by Mr E’s clinical records.

51. Mr E’s medical records show that when the Trust referred Mr E to the palliative care team, the team spent time with Mr E to understand his thoughts and needs as he was approaching the end of his life and understand what was important for him. This was in line with the guidance.

52. We appreciate that Mrs E has ongoing concerns that Mr E should not have died when he did. Our work shows that sadly Mr E was dying in the later period of his admission and the Trust made the right decision to place him on an end-of-life pathway when it did. This ensured that Mr E was comfortable during the last days of his life.

Impact

53. We considered the Trust’s actions after it accepted that it failed to monitor Mr E’s blood gases as set out in the guidance. We understand that Mrs E seeks service improvements and a financial remedy to put right the impact of that mistake.

54. The NHS Complaint Standards say organisations should openly identify instances when things have gone wrong, or where services have had an unfair impact, and take responsibility for these. They make sure staff can offer a range of ways to put things right for the individual. Staff should also look at what action will be taken to learn from the experience to continuously improve services and help support staff.

55. The Trust explained in its response letters it investigated the incident with its governance team. The staff member involved in the clinical error did not work on the ward itself but held a substantive role within the Trust. When they became aware of what happened in Mr E’s care, the staff member reflected on what happened and was given support and education to learn from it. Feedback was shared with both the respiratory medicine matron and the staff member’s matron.

56. We can see the Trust acknowledged and apologised that it did not monitor Mr E when it administered the high-flow oxygen. It fed this back, allowed the staff member to reflect, learn and provided additional learning around what happened. These are all appropriate steps for the Trust to take to improve the service it provides, and we are glad to see the Trust has taken those actions forward. These actions are in line with the Complaint Standards.

57. While we can see that the Trust took appropriate steps in recognising and apologising for its error, we cannot see that it has fully put right the emotional impact its error had on Mrs E. Mrs E explained to us that she has lived with the ongoing concern that her husband should not have died when he did. She said that this has caused deep distress and has had a lasting impact on her and her family.

58. We understand that the Trust’s complaint responses say that in its view, Mr E should not have been given such levels of high-flow oxygen, it was strongly contra-indicated (i.e. it should not be used) and the error likely hastened Mr E’s death. We can see that the Trust’s explanations likely contributed further to Mrs E’s understanding that the error had led to a significant clinical impact to her husband and added to her distress. We have not seen that the Trust has yet taken appropriate action to put right that impact and we set out below the recommendations which will put this right.

1. Mrs E complained that the Trust overdosed her husband, Mr E, with oxygen and left him unattended for two hours or more. We have found that administering high flow oxygen when Mr E’s oxygen saturation levels deteriorated on the ward was in line with guidance. We consider that the Trust failed to properly monitor Mr E after it administered the oxygen, which led to a short deterioration in his condition and required his transfer back to the respiratory high dependency unit (RHDU) for continuing treatment.

2. We have not found that this led to the clinical impact on Mr E in the way that Mrs E described. Based on the evidence available to us, this had a short-term impact on Mr E’s condition, which the Trust treated swiftly and appropriately. We have not seen anything to indicate that the high flow oxygen had any significant or long-term impact on Mr E’s condition. Sadly, Mr E was approaching the end of this life while admitted to the Trust. We found a failing that did not have any impact on the sad outcome for Mr E.

3. However, we have found that the failing led Mrs E to believe her husband’s death at that time may have been avoided, and that in its complaint responses the Trust’s explanations added to that fear. We have not seen any evidence that the Trust considered the distress that its error caused at what was already a difficult time for Mrs E.

4. We have not seen any evidence to show that the Trust failed to investigate or treat Mr E’s symptoms. The Trust provided treatment which was appropriate to manage his symptoms. When it became clear that his condition would improve no further, the Trust involved its palliative care team, who supported Mr E into end-of-life care. We do not uphold this part of Mrs E’s complaint.

5. From our investigation, we partly uphold Mrs E’s complaint. We recommend that the Trust pays her £600 in recognition of the distress caused.

6. We were sorry to hear of Mr E’s death and the deep distress and lasting impact that this caused Mrs E and her family, particularly as she was concerned the Trust’s actions in April 2024 may have contributed to Mr E’s death. We extend our sincere condolences for Mr E’s loss. We hope our findings help resolve any outstanding concerns that she had about the Trust and its actions during the last days of Mr E’s life.

59. We make recommendations in line with our Principles for Remedy which are reflected in the NHS Complaint Standards. These say organisations should identify instances where things have gone wrong, take responsibility for these and find ways to put things right for those involved. They should learn from complaints to improve services.

60. We expect organisations to take action to compensate people appropriately if they cannot return them to the position they would have been in if the poor service had not occurred. In some cases, a financial remedy will be required. To decide on a level of financial remedy, we review similar cases where the person has experienced a similar injustice, along with our severity of injustice scale, taken from our Guidance on Financial Remedy.

61. We have identified failings in relation to the Trust’s complaint response. We consider that its response added to Mrs E’s fear that her husband’s death may have been avoided.

62. With that in mind, we recommend that within two months of the date of this report the Trust pays Mrs E £600 in recognition of the emotional distress caused.

63. The Trust should send us evidence it has completed the recommendation we have made.