Duodenal perforation 19. Mr P had the ERCP procedure on 16 October 2023. This is a common procedure used to diagnose and treat problems with the biliary system. Langerth et al. (2019) explain that whilst common, it is also a complex procedure that requires a high level of expertise. It is the complexity of this procedure that can lead to adverse events.
20. Duodenal perforation is a well-recognised, though rare, complication of this procedure. The complication rate is approximately 10%, which means one in ten people who have the ERCP procedure will experience a short-term adverse complication. This can include perforation, which occurs in approximately 1.5% of patients who undergo an ERCP.
21. The procedure itself also carries a mortality rate of up to 6%. For individuals who experience a duodenal perforation, the mortality rate can be as high as 36%.
22. When Mr P went for this procedure on 16 October, he signed a consent form which included an explanation of the potential complications. Perforation of the bowel was listed as a potential complication, as was ‘potentially life-threatening’ complications. He signed to say this had been explained to him and that he was consenting to go ahead with the procedure.
23. Our surgeon adviser explained that the fact a duodenal perforation occurred does not indicate service failure. It is well-recognised that this can happen during the ERCP procedure. Mr P was appropriately advised that this could happen, and he decided to proceed knowing this.
24. We have found no evidence the duodenal perforation occurred due to service failure. We understand why Mrs T had serious concerns about this.
Delay in repairing the perforation
25. Ansari et al. (2019) outline the standard procedure for managing duodenal perforations. There are different options for managing a perforation, which depend on the clinical circumstances and the patient’s fitness for certain approaches. Overall, the goals of managing a duodenal perforation are usually to prevent infection, provide nutritional support, and restore the continuity of the gastrointestinal tract.
26. Achieving these goals can be approached conservatively by supporting the body to heal itself. They can be approached endoscopically or surgically, which is where the clinicians intervene to assist the body to heal.
27. Although surgical management can be an option, especially in patients where there is leakage into the peritoneal cavity (the space between the organs and the abdomen wall), this is dependent on the patient being fit enough to tolerate and survive surgery.
28. In cases where surgery carries a high risk of mortality for the patient, conservative or endoscopic options are usually preferred. This involves inserting a percutaneous transhepatic cholangiogram (PTC) drain to allow drainage of bile from the liver. Endoscopic interventions such as using through-the-scope (TTSC) clips or self-expandable metal stents (SEMS) are also appropriate approaches.
29. The perforation to Mr P’s duodenum was identified immediately during the ERCP procedure, and the surgeon took appropriate action by partially closing the defect. We understand Mrs T was very concerned the hole was not completely sealed at that time and our surgeon adviser confirmed this was necessary for draining the bile that had poured into Mr P’s duodenum. Later that afternoon, an emergency operation took place to insert a PTC drain for the bile. The surgeon also placed a fully covered metal stent, and placed TTSC clips across the defect to seal the perforation.
30. Our surgeon adviser confirmed that this was an appropriate intervention to repair the duodenal perforation. Following the emergency surgery, the Trust’s hepato-pancreato-biliary (HPB) consultant documented a detailed plan for managing the perforation. This included the prevention of infection using antibiotics, close observation, and restricting food and fluid intake until they could confirm the leak had resolved.
31. There is no standard or guideline as to how soon the intervention to repair the duodenal perforation should have happened. The hole was not fully sealed during the first procedure due to the bile that had entered Mr P’s duodenum, and the second procedure to insert a drain and seal the hole took place later the same day. Our surgeon adviser explained that this was timely and likely the soonest opportunity the Trust had to undertake the procedure. The procedure required both a general anaesthetic and highly specialist surgical expertise, which can take time to organise to perform the procedure safely.
32. Unfortunately, Mr P began to feel unwell and had signs of infection from 20 October. On 24 October the Trust identified there was likely a continued leak from the duodenal perforation. The clinical picture was also significantly complicated at this time because the same scan which identified the ongoing leak also found both a pulmonary embolism (a blood clot in the artery supplying blood to the lungs) and a subcapsular haematoma (bleeding into the liver). This meant the clinical team had to balance the risks associated with treating each of these conditions and proceed carefully.
33. Further procedures took place on 27 October, 2 November, and 24 November. These aimed to seal the perforation and resolve the ongoing duodenal leak. Unfortunately, Mr P’s duodenum did not respond as hoped with these procedures and it took much longer than expected to seal the perforation and resolve the leak.
34. The team consistently sought surgical advice throughout this period of care and the surgical team consistently advised that endoscopic management was the correct approach due to the risks associated with surgical management of the perforation.
35. Our surgeon adviser confirmed that endoscopic management of the defect was appropriate in the context of Mr P’s clinical presentation.
36. We have seen no evidence of a delay in providing appropriate care and treatment to repair Mr P’s perforation. This initially took place the same day, but the procedure was not successful. Follow up procedures also failed to resolve the leak, but this was also not due to service failure.
37. We can understand why repeatedly attempting to resolve the perforation and leak could appear like the surgeons got something wrong and delayed in successfully closing the defect. We hope our work reassures Mrs T this did not happen due to service failure.
Communication 38. The GMC’s Good Medical Practice guidelines say doctors must give patients the information they want and need to know in a way they can understand. They also state that doctors must be considerate to those close to the patient and be sensitive and responsive in giving them support and information. The guideline does not specify how often these updates should be provided, and the frequency of updates and communication can be dependent on the availability of clinicians. When there is high demand for care, doctors must prioritise patient care.
39. NICE guideline CG138 (patient experience in adult NHS services) says that service providers should clarify with the patient a first point of contact and whether they would like family members involved in key decisions. If the patient agrees to sharing information with family, the team should facilitate this. This guidance also does not specify the frequency for updates, and this will depend on the clinical demands on the service.
40. Mr P was updated most days, with the exception of some weekends, via his daily consultant-led ward rounds. He had mental capacity, and it was appropriate for staff to update him and give him the information about his care as a first priority. His family were also updated regularly by different members of the team via family updates that are consistently well-documented in Mr P’s medical records.
41. From early November, Mr P’s family began to raise concerns about communication and expressed anger at changing medical plans and not receiving updates. Documentation on 8 November recorded that Mrs P was angry about conflicting information being given about surgical repair of the perforation. A nurse documented that they explained medical plans can change regularly and agreed a plan for one of the doctors to discuss this with Mrs P. This was followed up the next day with a full family update provided by a doctor.
42. The Trust also allocated a physician to be a single point of contact for the family on 22 November and this person introduced themselves to the family the same day. Although regular updates were given by the team, the family remained unhappy and contacted the Patient Experience Team.
43. On 29 November a doctor explained to the family that the expectation of daily updates from the consultant was not reasonable because the consultants had to prioritise the care of their patients. This information was correct and whilst doctors should be sensitive to families, their first priority is communication with the patient when that person has mental capacity.
44. From 30 November the family declined any further updates from the doctors and requested all information go via the Patient Experience Team. We can see that when the team attempted to update Mr P’s family after this, they reiterated that they did not want to speak with the doctors and wanted this information via the Patient Experience Team.
45. We understand why Mr P’s family were so anxious to understand what was happening with his care and why they wanted more updates from the doctors. This is not always feasible, however, due to the demands of the hospital’s clinical services. Mr P was updated daily by the ward consultant, aside from some weekends where there was no consultant-led care provided. We have seen nothing to indicate he was unable to communicate these updates to his family.
46. Doctors must manage finite resources and their priority must always be the care of their patients. The information communicated during the documented family updates was comprehensive and these updates took place regularly. The medical plans did change in response to sudden clinical changes, for example upon discovering Mr P’s duodenal leak had not sealed. This is expected for patients with complex medical needs and P was updated on these changes daily, as documented in his medical records.
47. There were also some changes to the plans as a result of needing to manage finite resources. For example, some of his routine procedures were delayed because of the need to prioritise resources for emergency patients. We understand that this likely caused Mr P’s family some anxiety during an already distressing time, and they may have been comforted by more timely updates regarding this. Unfortunately, in a service with finite resources this is often not possible.
48. In summary, the doctors’ priority was to communicate with Mr P, and this happened almost daily. The team should also have been sensitive to his family and provided updates whenever feasible, in line with the GMC’s Good Medical Practice guidelines. We have found that this happened.
Choice and preference 49. The GMC’s Good Medical Practice guidelines state that doctors must listen to patients, take account of their views, and give them the information they want/need to know in a way they can understand.
50. Both our surgeon and oncologist advisers have confirmed that, sadly, Mr P had very few options in relation to the treatment available. From an oncology perspective, he was not medically well enough to undergo any treatment for his pancreatic cancer during his admission. This meant there were no options available to him. From a surgical perspective, the options were very limited in relation to addressing the duodenal perforation.
51. Although there were very few options available, the physicians discussed Mr P’s care with him almost daily. There are multiple documented discussions throughout his care where the doctors clearly documented they had explained the current medical plan with him and his family, and that they were happy with the plan.
52. We can also see instances where Mr P was given choices and where the plan was changed as a result of his wants and needs. For example, Mr P was given a choice about whether or not to attempt a final endoscopic closure of the defect, and the doctors documented he chose to go ahead with this. Similarly, the doctors recommended changes to his fluid intake and allowed him to have ice lollies in response to Mr P communicating his distress at being nil by mouth.
53. Another instance of providing care in the context of Mr P’s preferences included providing a catheter when not medically required as he preferred this for his comfort and to help with his difficulties sleeping at night.
54. The evidence shows there were few choices available to Mr P with regards to the direction of his care. The notes reflect that the doctors communicated with him almost daily, and consistently documented his views and/or preferences during his admission. The medical team were also responsive to his preferences, and this is reflected in making changes to his care with the aim of making him more comfortable. This was in line with the GMC’s Good Medical Practice guidelines.
Oncology care
55. Our oncologist adviser explained that whilst Mr P was in hospital, there were no options for treating his recently discovered pancreatic cancer.
56. NICE guideline NG85 (pancreatic cancer in adults) outlines that a pancreatic multidisciplinary team should decide on what course of action or treatment is needed. Mr P’s medical records show that the team caring for him brought his case to the pancreatic multidisciplinary team meeting in October and November, which was in line with NG85. Sadly, the pancreatic team consistently advised that Mr P could not receive treatment until the duodenal leak had resolved.
57. Although NICE guideline NG85 outlines surgical and chemotherapy options for treating pancreatic cancer, these are only appropriate if the patient is fit and well enough to undergo these. Mr P was not fit for chemotherapy until his duodenal leak had been sealed. This is because if chemotherapy treatment is given when patients have a leak or a perforation, it prevents the perforation from healing. Healing the perforation was an urgent medical need and our oncology adviser confirmed that chemotherapy was ‘absolutely contraindicated’ prior to healing of the duodenal leak.
58. We can also see that the Trust had considered surgical options, and that the pancreatic cancer was potentially operable. Unfortunately, the surgical team advised that Mr P was not fit and well enough to undergo major surgery, and that he would have had a high risk of death if this went ahead. This was well documented as being explained to Mr P and his family at the time.
59. We can understand why not being given active treatment for pancreatic cancer caused his family concern. Sadly, pancreatic cancer has a very poor prognosis and the lowest survival rate of the 22 common cancers. Just 30% of patients survive to one year after diagnosis, and only 10% of patients are eligible for surgical intervention to treat this cancer. This means that even if the perforation had not happened and Mr P had been able to have chemotherapy or surgery, his chances of survival and the time he potentially had left was still very low.
60. Our oncologist adviser confirmed that after the perforation occurred, there were no options to treat the cancer until the perforation healed. The only thing the Trust could do was to continue to revisit his options via the pancreatic multidisciplinary team meetings. Mr P’s medical records reflect this happened and he received the care appropriate to his needs, in line with NICE guideline NG85.
Surgical drain 61. Our surgeon adviser explained that there is no guidance or standard that requires a drain to be removed via interventional radiology, and this can be done on the ward by an appropriately trained member of the medical team. However, they also advised that it is often a better option to remove the drain via interventional radiology because they can undertake tests to check drainage and because some drains have complex locking systems which require unlocking prior to removal.
62. There is no record of the procedure to remove the drain in Mr P’s medical records; however, there are a number of clinical records that indicate, on balance, that the drain was removed via the interventional radiology team.
63. A nursing entry on 20 December documents that at 2.32pm Mr P was about to return to the ward from the interventional radiology department after having his drain removed. There is also a radiological checklist, dated 20 December at 2.01pm, that shows Mr P was in the radiology department having an intervention that concluded at 2.28pm, and the intervention was completed without complication. The medical records from this point repeatedly reference the drain had been removed.
64. In summary, although there was no requirement to remove the drain via the interventional radiology team, the balance of the evidence indicates that this happened on 20 December.
Discharge 65. The Department of Health’s guidance on hospital discharge states that the following should happen when patients are discharged from NHS hospitals:
• discharge planning should begin on admission • discharge planning should include information about post-hospital care, such as advice and information about community teams that can be contacted post-discharge • family and/or unpaid carers should be involved in discharge planning, where appropriate • on discharge from hospital people who have new or additional needs should be offered choices of onward care and support that are available at the time of discharge.
66. Mr P was independently mobilising with his walking frame and did not need assistance with his daily living activities. Therefore, he was appropriately discharged under pathway zero, which is for people who do not require a formal package of health and social care in the community. Whilst Mr P needed to be followed up in the community with outpatient care, he did not have health and social care needs that required a package of care to address.
67. There is little evidence of discharge planning in Mr P’s medical records prior to 20 December. Although the discharge timeline commenced on 20 December indicates a discharge team was monitoring Mr P from late October, this was only documented from the perspective of his mobility and there is no evidence of a multidisciplinary approach to planning his discharge. This was not in line with the Department of Health’s guidelines.
68. The medical records reflect that the HPB consultant identified that Mr P was getting close to discharge on 22 December; however, there is no evidence of a multidisciplinary approach to planning his discharge from this point. The approach from the different members of the multidisciplinary team appears to have been uncoordinated and inconsistent. There were instances of the nurses encouraging Mr P to mobilise as he was nearing discharge, and the diabetes nurses showing him how to administer his insulin, but none of this was documented or provided as part of a coherent discharge plan.
69. Things appear to have moved quite quickly from 28 December, with the multidisciplinary team documenting the following interventions in relation to Mr P’s discharge:
• referral to the community diabetes team • a referral to the District Nursing team for wound care, which was declined by the District Nurses as they were only accepting ‘housebound’ patients at that time • dressings supplied for Mr P’s former drain site.
70. The notes also reflect that Mr P was ‘very keen’ to go home at the point of his discharge.
71. There is no evidence within Mr P’s medical records which reflect his family were actively engaged in discharge planning or that the options for onward care were explained to Mr P and his family. For example, although the District Nursing service had declined to visit Mr P at home to change his drain site dressing, there is no documentation to reflect what the alternative arrangement would be or that Mr P and his family were told how his wound would be managed following his discharge.
72. In summary, whilst there is some evidence of discharge planning from 20 December, the Trust did not document discharge planning from admission, nor is there evidence of a coherent, multidisciplinary discharge plan. In addition, his family were not engaged in the discharge planning and were not provided with all the information they needed to know about what Mr P’s onward care would look like. This fell short of the Department of Health’s guidelines for hospital discharge and was a failing in his care.
73. We understand this caused his family stress and anxiety in the weeks following his discharge, whilst they attempted to understand what Mr P’s care in a community setting would look like. The Trust has not yet taken action to put this right and we have outlined recommendations to address the impact of this failing below.
Nutrition 74. NICE guideline CG32 (nutrition support for adults) outlines what good nutrition support looks like for patients in hospital. In line with this guideline the Trust should have:
• weighed and screened Mr P for the risk of malnutrition on admission, and then weekly whilst he remained an inpatient. The malnutrition universal screening tool (MUST) can be used to do this • considered nutrition support if Mr P became malnourished, which is defined based on BMI and percentage of weight loss • considered Mr P to be at risk of malnutrition if he had eaten little/nothing for five days/or was likely to do so • considered using oral, enteral, or parenteral nutrition support if he became malnourished or was at risk of malnutrition • considered parenteral nutrition when his perforated gastrointestinal tract meant he could not eat or drink safely.
75. The nursing team weighed Mr P and screened him for malnutrition using the MUST tool on admission to hospital on 16 October. He was appropriately screened as being low risk for malnutrition at that point, based on his BMI and reported nutritional intake.
76. At this stage Mr P was able to have food and fluid orally and, whilst he was to be nil by mouth for 24 hours to allow for a contrast scan on 17 October, he did not meet the five-day criteria for risk of malnutrition at this point.
77. Mr P was weighed that week, in line with the NICE guideline CG32. From 25 October he was asked to stop eating and drinking due to the duodenal leak. He was referred to the Trust’s dietetics service that day for more specialised advice on maintaining his nutrition.
78. Mr P’s nutritional needs were fully assessed by a dietician on 27 October, who identified he was at risk of malnutrition as he would be unlikely to be able to eat and drink for more than five days. The dietician documented a plan to insert a nasojeunal (NJ) tube directly into his bowel to provide enteral nutrition, which was in line with NICE guideline CG32. This was inserted that day, and upon confirming it was safely sited he received nutrition via the NJ tube on 28 and 29 October.
79. He was reviewed by the dietician again on 30 October and 1 November, when a referral to the adult nutrition team was made for parenteral nutrition. The was in line with NICE guideline CG32, which recommends parenteral nutrition for those with a perforated gastrointestinal tract.
80. On 2 November the adult nutrition team assessed Mr P and recommended parenteral nutrition, in addition to the NJ tube. They recommended that if the NJ feeds failed then a PICC line for full parenteral nutrition was required. This assessment included weighing Mr P and a calculation of his BMI, as well as consideration of his intake, in line with NICE guideline CG32.
81. The nursing team weighed and screened Mr P for malnutrition again on 5 November and documented he was at high risk of malnutrition. This was in line with NICE guideline CG32.
82. Mr P continued to receive parenteral nutrition and was reviewed every two to three days by the adult nutrition team. He was weighed at least weekly during this time. This was in line with NICE guideline CG32.
83. On 14 November the team inserted a peripherally inserted central catheter (PICC) to provide full parenteral nutrition as using the NJ tube was no longer viable. He continued to be reviewed every two to three days by the adult nutrition team. The nursing team also continued to weigh him and screened his nutritional needs weekly during his admission. This was in line with NICE guideline CG32.
84. We have found Mr Spruce’s nutritional needs were managed in line with NICE guideline CG32.