Challenge of achieving truly individualised informed consent in therapeutic endoscopy

Abstract

Objective

Guidance covering informed consent in endoscopy has been refined in the UK following the obstetric case of Nadine Montgomery, and in light of updated General Medical Council guidance. All risks likely to be material to the patient must be explored, as well as alternatives to the procedure. Despite this, departments and endoscopists still struggle to meet the current standards. In this article, we explore the challenges encountered in achieving individualised consent in therapeutic endoscopy through real-life scenarios.

Methods

Five realistic therapeutic endoscopy (hepatobiliary) scenarios are described, followed by presentation of possible or ideal approaches, with references related to existing literature in this field.

Results

The vignettes allow consideration of how to approach difficult consent challenges, including anxiety and information overload, urgency during acute illness, failure to disclose the risk of death, the role of trainees and intraprocedural distress under conscious sedation.

Conclusions

The authors conclude that a high degree of transparency is required while obtaining consent for therapeutic endoscopy accompanied by full documentation, involvement of relatives in nearly all cases, and clarity around the presence of trainees who may handle the scope. A greater focus on upskilling trainees in the consent process for therapeutic endoscopy is required.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Following the case of Nadine Montgomery (heard at the Supreme Court, UK, in July 2014) the law around informed consent changed, requiring a more patient-centred and individualised approach.

• The British Society of Gastroenterologists updated its guidance in 2016, and again in 2023 based on new General Medical Council guidance.

• There remain significant challenges in meeting the legal standards around high-risk or urgent therapeutic endoscopy.

WHAT THIS STUDY ADDS

• Highlights the challenges of obtaining informed consent in complex therapeutic endoscopy especially in endoscopic retrograde cholangiopancreatography (ERCP).

• Emphasises the need to avoid overwhelming patients with information overload.

• Describes the need to provide enhanced training in the informed consent process.

• Explores best interest decisions in emergent scenarios.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Clinical scenarios based on real-life presentations, and the brief reflections that follow, will encourage clinicians to consider their approach to informed consent.

• The scenarios can be used in training modules, tutorials, simulation sessions and courses to prompt discussion.

• Units should recognise the need for ongoing audit against best practice.

Introduction

The focus on informed consent has increased since the change in the law that moved away from a paternalistic, physician centred interpretation of ‘what is best for you’ towards a genuinely shared decision-making approach. The tragic experience of Nadine Montgomery and her baby forced this change, the courts finding in 2015 that informed consent had not been obtained before a procedure unless all risks likely to be material to the patient have been considered and discussed.¹ In 2016, the British Society of Gastroenterologists (BSG) updated its earlier guidance, and again in 2023, in light of new General Medical Council (GMC) guidance.^{2 3} The GMC described seven principles, the fourth is most pertinent to endoscopists: Doctors must try to find out what matters to patients so they can share relevant information about the benefits and harms of proposed options and reasonable alternatives, including the option to take no action. This was interpreted by the BSG into the following recommendation: Individualised consent for endoscopy requires that clinical factors related to the patient’s symptoms, medical history and specific patient-related risk factors related to the proposed procedure, as well as the patient’s personal preferences and expectations, should be explored.

Following Montgomery, there has been concern and debate about whether doctors are able to meet its standard. There is evidence in other clinical disciplines that the law is still poorly understood, that consent practices have not been able to keep up, and that the time required to fully explore all potential material risks is incompatible with daily clinical demands.^4–8 The ideal model of a consent process (often provided in a ‘two-stage’ arrangement) is well established, allowing patients to take in the information that is given to them and reflect on it, including other therapeutic options (or that of ‘doing nothing’).⁹ Even with such arrangements, patients still report that they wished to have more information; in Bassi’s study, although nearly all patients read the information leaflet that was sent to them, 64 of 127 were dissatisfied about the amount of information provided.¹⁰ In a surgical study, there was evidence of adequate discussion on the day of the procedure itself in less than 50%.¹¹ Elsewhere, older patients and those with a lower educational level have been found to have less understanding of the risks that were, in theory, explained to them.¹² In an international study by Triantafyllou et al, the proportion of units discussing alternatives, specific complication rates and mortality risks were under 25%.⁹ Although there are ‘mitigating circumstances’ where it seems perfectly understandable why a consent process has not been observed (including multiple previous identical or similar procedures, direct referral from another hospital, department or multidisciplinary meeting and urgency), these factors would probably not sway a court if litigation were to arise from a complication.¹³

Although ongoing refinements in informed consent are welcome, we must be able to recognise situations in which patients become overwhelmed. As explored by ethicists Herring and Foster, ‘Too much knowledge can inhibit rather than enable thought’.¹⁴ We may meet patients for whom detailed discussion of risks will cause anxiety and emotional distress. What do we do when a patient categorically states “I will have the procedure but do not want to know the risks”. Do we then respect their right not to know or force them to provide a true informed consent?

It is in the realm of therapeutic endoscopy that these concerns become most relevant for the gastroenterologist. Percutaneous endoscopic gastrostomy and endoscopic retrograde cholangiopancreatography (ERCP) are highlighted as particular examples in BSG communications, but units where advanced polypectomy, endoscopic submucosal dissection and ablative therapy in the upper gastrointestinal (GI) tract will also be concerned. ERCP has a reputation as one of the riskiest endoscopic interventions (sitting somewhere between parachuting and colectomy, in one well-known analysis), and because it is often required in frailer patients in emergent scenarios, arrangements for its provision must be very robust.^{15 16} Indeed, a coroner wrote to the Secretary of State for Health and Social Care in 2021, expressing concern that a service was unsafe.¹⁷ ERCP will be the focus of this article, but the principles explored are readily applicable to other procedures.

Clinical scenarios based on true events are described, followed by a reflection. The names in each scenario are fabricated and clinical details are amended to maintain anonymity. Each scenario presents a dilemma in obtaining fully informed consent that is likely to be encountered by endoscopists working in the National Healt Service (NHS). Adherence to legal requirements needs to be differentiated from ‘best practice’. Best practice is ensuring consent is understood by patients and families, clarity around indications, multidisciplinary discussions, ensuring good documentation of conversations and decisions thus avoiding complaints and inquests. Most of these scenarios describe the challenge of delivering best practice during the consent process, while issues that relate to legislation are highlighted where relevant.

Scenarios and reflections

Anxiety and overload

A 23-year-old woman was referred by her general practitioner to the gastroenterology clinic because of right upper quadrant pain. In the clinic, she saw a registrar, Ishan, who suspected biliary colic, or, in light of mildly abnormal liver function tests, common bile duct stones. A magnetic resonance cholangiopancreatography was arranged which demonstrated multiple small calculi in a moderately dilated bile duct. She was invited back to clinic and the need for ERCP was explained. She had no previous medical history or comorbidities, and was clearly fit for the procedure under conscious sedation. Ishan, an experienced registrar 1 year away from becoming a consultant, discussed potential complications including post-ERCP pancreatitis. His form of words was, “Pancreatitis occurs in around one in 20 cases, and we do everything we can to avoid it, including an anti-inflammatory suppository, intravenous fluids in some cases, and sometimes a small plastic stent into the pancreas, if our wires go into it inadvertently.”

“Is it serious?”

“It can be. Pancreatitis in the setting usually passes in a few days, but you do need to stay in hospital. Sometimes, though, it becomes more serious and the pancreas gets more inflamed. We will keep you in hospital, give you fluids and probably do a scan. In about 1 in 200 cases severe pancreatitis can develop, and this can be life-threatening.”

“How often?”

“Maybe one in 500?”

The patient, who attended with her father, became visibly anxious about this risk and asked for more time to consider the procedure. Ishan asked his consultant to attend during the clinic, and although further explanations and reassurance were given, the risk of dying from ERCP was clearly dominating her thoughts. It was agreed that she would return to the clinic a week later. However, the alternative risks that the stones could lead to like jaundice, cholangitis, or indeed pancreatitis were also explained. She became extremely upset at this point and was crying. She was led to a separate room so that the clinic could continue, but the consultant needed to go back to see her later so that the next steps could be agreed. Eventually, with her father helping to calm her and interpret the risks, she agreed to attend for the ERCP. This took place a week later. Her sustained high level of anxiety was difficult for the entire team to manage before she entered the room. The procedure passed uneventfully.

Reflection

Turton, writing in Medical Law Review, clarified that, ‘The doctor’s duty is not […] fulfilled by bombarding the patient with technical information which she cannot reasonably be expected to grasp’,¹⁸ which is also emphasised in para 90 of the Montgomery ruling. BSG guidance reflects this, explaining that: It is important […] the patient is not overwhelmed with excessive information, such that they are unable to evaluate the material risks and benefits. Sensing how much information it is necessary to share is, in the authors’ view, one of the most subtle of clinical skills. If you, the specialist, truly believe that to do the procedure is in the patient’s best interest, you will not wish to say anything that might put them off; nor to say so much that their decision-making ability is paralysed by the number of competing facts and ‘pros and cons’. However, in modulating the information, the specialist shapes the facts such that their words are likely to persuade the patient to give consent. This, we presume, is the trap that Nadine Montogomery’s obstetrician fell into: they sincerely believed that natural birth was preferable, so omitted or underplayed the risk of shoulder dystocia. In the matter of mortality following ERCP, there is a binary decision: to mention death, or not. Readers will have found their own ways of dealing with this. Personally, we find telling patients in their 30s that a potential complication of ERCP for a small stone is death very uncomfortable indeed; however, we have also lost young patients to pancreatitis. One only needs to think back to the last tragic outcome to feel more justified in broaching mortality. A consequence of such an honest policy or best practice is, unfortunately, the occasional highly emotional response. It could be argued that the cumulative emotional distress caused by such honesty is unacceptable, yet we do this to cover ourselves in the worst-case scenarios and to demonstrate (to the coroner, occasionally) that we did explore the most material of risks—the end of life itself.

The family unit

Flora, a gastroenterology trainee in her first year of ERCP training, was assiduous in obtaining informed consent. She observed her trainer obtain consent several times, and having attended lists for 3 months was now trusted to speak with patients in clinic or on the ward. When inpatients were referred to the gastroenterology team, she reviewed them, assessed the indications and discussed each case with the consultant. In the case of 78-year-old woman, Flora had a full discussion which included the risks of sepsis, bleeding, perforation and pancreatitis. She explained that pancreatitis could be severe. She was not present when the patient had her procedure, but heard that it went straightforwardly, and that there were no pancreatic injections or instrumentations. She was therefore surprised to learn that the patient developed severe post-ERCP pancreatitis and was admitted to the intensive care unit on day 3. The patient died a week later, in multiple organ failure. The coroner was informed and called an inquest. Although Flora was not asked to attend, she learnt from her consultant that the focus of the inquest was the patient’s informed consent. The family wrote to the coroner before the inquest, explaining that neither they nor the patient were aware that complications could be fatal. Although family members were not involved in the consent process (the patient had full mental capacity), they were aware that the procedure was due to take place but understood it was a ‘fairly routine camera test’ to remove stones. Ultimately, the coroner found that death had occurred through an unavoidable complication of a necessary medical procedure; however, she instructed the Trust to improve its consent procedures around this particular intervention.

Reflection

Although this scenario also turns on full disclosure of mortality as a risk, the coroner took an interest not because the patient died, but because the family wrote expressing concern. In higher risk procedures it is prudent to involve family members in the discussion, even if the patient has full mental capacity, as in this case. Unless the patient expressly forbids the medical team to talk to wider family members, it is best practise to make sure that somebody else knows what is going on, and that they hear about complications. If the patient does suffer a complication and is incapacitated for a period of time, or if they die, they will not be able to corroborate the clinician’s account of what was said during the consent process. The family, who may be angry, sceptical and in an accusatory frame of mind, are unlikely to give the benefit of the doubt. Only a discussion that is clearly documented in the medical notes can serve this purpose. In the case of urgent referrals from medical or surgical teams to the ERCP service, when contact between endoscopists and patient have been fleeting, this has involved phoning relatives at the last minute from the endoscopy room, to ensure that they are aware that their (usually elderly) relative is just about to undergo a high risk but necessary procedure. Often, this will be the first that they heard of it, and a way must be found for mortality to be mentioned.

Urgency and best interests

An 89-year-old man was admitted with jaundice (bilirubin 150 µmol/L) and sepsis on a Saturday morning. He was treated for suspected cholangitis and a CT scan confirmed a large stone (1.5 cm) in the common bile duct with gross intrahepatic dilatation. The patient was admitted from a nursing home where he had resided for 3 years due to advanced Alzheimer’s disease. On Monday morning, the gastroenterology team was informed, and the consultant reviewed the patient on her ward round. She contacted the ERCP consultant who, she knew, reserved an emergency slot on her ERCP list that afternoon. The hepatobiliary (HPB) fellow then reviewed the patient to confirm the findings and assess his fitness. He agreed that the patient was fit for ERCP and that given ongoing sepsis with a rising bilirubin (180 μmol/L), it should be done that day.

The HPB fellow attempted to contact the named next of kin, but he was unsuccessful. A decision was made to complete the consent form 4 in the patient’s best interest, and it was countersigned by the ERCP consultant. At ERCP, the papilla was located within a large duodenal diverticulum. Bile duct cannulation was successful, and sphincteroplasty was preferred over sphincterotomy. Unfortunately, a perforation occurred and was identified immediately. A covered metallic stent was inserted without removal of the stone, and the procedure was terminated. A CT scan demonstrated pneumoperitoneum and free fluid. The patient deteriorated quickly and died 5 days later. The next of kin was successfully contacted after the procedure and informed of the complication. He attended the ward to be with his father during his final days, but subsequently made a formal complaint that he was not involved in the decision to proceed.

Reflection

In this scenario, clinical urgency competes with the time required for comprehensive and detailed explanation of the risks, and this is further complicated by a lack of mental capacity. The BSG is clear that urgency does not justify bypassing the informed consent process. The consultant who signed the consent form 4 was taking a personal risk in doing so if the procedure had not been discussed with the next of kin (or even more critically, an individual with lasting power-of-attorney for care decisions). The consequences of not proceeding in this case could be that the patient succumbs to sepsis without the opportunity for a biliary decompression. What is the right answer? Urgency and incapacity are not unique to HPB presentations, and in the field of thrombolysis for stroke, approaches to informed consent have been developed that are acceptable to patients’.¹⁹ The routes to ERCP are many and varied, and a single consent model is unlikely to work. In the authors’ experience, the correct approach is to document fulsomely all of the considerations that have taken place before decision to proceed. This may include corroborative input from the general practitioner, senior nurses on the ward or colleagues. The mental capacity act requires a clear assessment of capacity and best interests to be documented, and in endoscopy there usually is time to do this unless the indication is catastrophic haemorrhage. Another consideration here is that of doing nothing until adequate communication has taken place—although the clinical assessment was that antibiotics were not effectively improving the situation, they were maintaining stability, and during this stable phase further attempts to contact relatives could be made. We should be honest and acknowledge that urgency is sometimes dictated by the endoscopist’s pressured time, rather than the patient’s.

The trainee

Wasim attended Dr Clarke’s ERCP lists every Wednesday afternoon. He usually obtained consent from patients before the list, and between cases, while Dr Clarke wrote the last report up. On this occasion, Dr Clarke took the first patient’s consent himself as he had been informed that Wasim would not be coming due to another commitment. However, Wasim turned up just as the patient was being sedated. Dr Clarke welcomed him into the room and encouraged him to start the procedure. Wasim was very pleased to achieve biliary cannulation quickly. The patient was a 48-year-old man with an 8 mm stone in the common bile duct, and sphincterotomy was indicated. Wasim had done six sphincterotomies before, and Dr Clarke allowed him to proceed with this one. The consultant was standing about a metre behind Wasim as he commenced the cut. With the last press of the diathermy pedal, there was immediate bleeding. Dr Clarke allowed Wasim to keep the scope, but instructed him verbally. They waited for 30 s, but the bleeding was evidently not going to stop. A decision was made to insert a covered metallic stent, and for this, Dr Clarke took the scope as he wanted to minimise the risk of losing position and losing wire access. The stent was placed successfully and haemostasis was achieved. The stone was not removed. In recovery, the complication was explained to the patient. He asked who had performed the procedure, as recalled hearing a different voice in the room. Dr Clarke had also used the term ‘we’ when describing what happened in the room. However, the patient was insistent on knowing who had done what, and then asked Dr Clarke what the trainee’s qualifications were and how many procedures he had done before. The patient was irritated, as he would now have to return for a second procedure for the stent to be removed and for the stone to be definitively treated. Dr Clarke suggested that they continue the conversation the next day, as the patient was being advised to stay in overnight for observation. During the next conversation, Dr Clarke apologised for not making it clear that the trainee would be present during the endoscopy. The patient, who was feeling much better, accepted the apology and decided to take it no further.

Reflection

The ethics of practical training in medicine are complicated, as certain compromises in the quality of care must be accepted if trainees are to ascend the learning curve,^{20 21} This difficulty has even been examined through the lens of Kantian philosophy (‘one should always treat patients as ends in themselves, and never only as a means to other ends’—the ends here being another doctor’s training) and has been explored by Atul Gawande, with reference to the need to adopt new technologies.^{22 23} In endoscopy, reduced quality may mean increased procedure time, missed pathology or increased risk of complications, even if the trainer is present and very actively engaged in the list. If we accept that there is an increased risk, although marginal, the patient must be made aware of this in order to provide fully informed consent. In this vignette, the consultant did not explicitly explain that a trainee would be present. Once the patient was sedated, he could not give consent. Therefore, strictly speaking, the trainee should not have been allowed to take the scope. From Dr Clarke’s point of view, having a trainee present was entirely normal, and part of a standard ERCP list in his position as a trainer. So here we have a clash of understanding of what is normal and reasonable between consultant and patient. A potential solution to this is to establish a departmental policy, clearly explained to patients, that all procedures are potential training opportunities. However, clearly this is signalled in letters, leaflets or on posters in the department, we cannot escape the fact that patients expect their procedures to be done by the most experienced, most skilled personnel. It is when patient expectation is betrayed that complaints arise. In this case, an apology plus assurances that the patient was at not put at higher risk sufficed. In effect, consent was achieved after the event.

Creating rules around the presence of trainees at procedures, especially in large teaching centres, is complex. Consultants have an obligation to train the next generation of clinicians, but it is not clear how much explanation about the trainee’s role should be given during the informed consent process. In surgery, there is evidence that patients have concerns about trainees performing procedures.²⁴ In endoscopy, best patient care must be balanced against the institutional and societal prerogative to develop skills in trainees requires honest and open discussion.

‘Please stop’

A 35-year-old female patient was appropriately consented for ERCP to treat a biliary stricture with significant jaundice (bilirubin 300 µmol/L). After a difficult cannulation biliary access was achieved, a tight stricture was identified. This required dilatation, sampling for cytology and placement of a stent. At this point, 25 minutes into the procedure, she signalled that she wanted the procedure to be stopped. The endoscopist believed that it was too risky to stop the procedure, as the risk of cholangitis was high (undrained system). The decision was made to complete the procedure. To achieve this, very high doses of sedative (midazolam) and physical restraint were required, the latter being distressful for the patient and the wider team.

Reflection

This challenging and contentious situation is encountered quite often in the endoscopy department. The BSG states that ‘it is the responsibility of the endoscopist to act in the patient’s best interests’. The decision to stop is at the endoscopist’s discretion and is easier to do when the procedure is purely diagnostic or non-urgent. However, ERCPs are often urgent and are usually therapeutic, so the risks of deferral must be balanced against the burden of acute distress. These situations are distressing to the team in the room and the patients may suffer prolonged psychological morbidity.^{25 26} Agreed ‘safe’ signals and excellent communication between endoscopy nurses and endoscopists are paramount. A European Society of Gastrointestinal Endoscopy position statement from 2023 provides some guidance in this scenario, stating that special consideration needs to be given to situations where terminating the procedure is harmful to the patient.²⁷ In such circumstances, it may be necessary to restrain the patient and adjust sedation. Once the situation is controlled, it is imperative to discuss the situation with other members of the team to agree on a strategy for continuation of the procedure. Endoscopic non-technical skills (ENTS) are essential in scenarios like this. ENTS were developed in 2010 and are now incorporated by the Joint Advisory Group (JAG) on Gastrointestinal Endoscopy into training of endoscopists in the UK.²⁸ This assesses the importance of communication, team work, empathy and behavioural skills in decision making during endoscopy. Staff must feel empowered to ‘speak up’ if they feel that the patient is suffering—a well-established principle in the operating theatre.²⁹ There is a growing demand for anaesthetic support for ERCP in the UK and elsewhere, and if successful the risks of procedural distress will become much smaller.³⁰ Until that change in service has been achieved, it is inevitable that some patients will experience distress, so all teams should consider how to recognise and manage this safely.

Discussion

The ethical foundations of informed consent were established at the Nuremberg trials, when the unspeakable crimes of Nazi doctors who experimented on prisoners were investigated and the 10-point Nuremberg code was developed.³¹ This related to human experimentation and research, but the medical literature includes debate about consent for surgery prior to this, in the early 20th century.^{31 32} It is salutary for doctors to consider this history, when skimming through information for procedures that to them are quite routine, but to their patients are memorable moments of stress and anxiety. However, even before Montgomery, commentators were concerned that truly informed consent was almost impossible to achieve.³³ Montgomery has successfully clarified the importance of individualising discussions but has not fundamentally reinvented the process of informed consent. Turton, writing in Medical Law Review, felt that the decision represented a ‘tidying up of the law, rather than a dramatic legal development’, and this view may resonate with clinicians who feel that good doctors always took consent to that standard anyway.¹⁸ The importance of changing the law only becomes clear when something goes wrong, and the way in which consent was obtained is scrutinised in the courts. This perspective raises an important point: when things have gone wrong and significant morbidity has occurred, the complication will of course appear ‘material’. In the Montgomery judgement, The test of materiality is whether, in the circumstances of the particular case, a reasonable person in the patient’s position would be likely to attach significance to the risk, or the doctor is or should reasonably be aware that the particular patient would be likely to attach significance to it.

To us, this stipulation undoubtedly covers the risks of mortality or prolonged admissions with pancreatitis. As we have seen, these can occur in the youngest, fittest patients, so even if these serious sequelae appear unthinkable based on the patient’s fitness, they must be considered. The impact of other adverse outcomes (eg, retroperitoneal perforation or major haemorrhage) can be estimated based on patient’s performance status, comorbidities and clinical status during the presentation. To satisfy the law, we must include all these complications when we seek informed consent, while tailoring quantitative estimates of morbidity or death according individual patient factors. The cost of this, potentially, includes discomfort (for doctors and patients), anxiety, time (so that discussions can be done in an unhurried and non-coercive way, involving family members where appropriate) and adequate organisational support. In addition, we must role-model these behaviours to our trainees and ensure that transparent and appropriately comprehensive approaches to obtaining consent are developed in the endoscopists of the future.

Though the scenarios here describe HPB endoscopy, the challenges apply to all of therapeutic endoscopy. Individualisation of consent according to risks associated with the specific procedure, patient factors and the likely impact of complications is both humanly reasonable and legally mandated. To achieve this, discussions with the patient (and family members where appropriate) should include an explanation of risk–benefit balance, possible complications and how they are likely to respond to them. Such pre-emptive disclosure reduces the likelihood of complaints or litigation, and, in the worst-case scenario, can protect clinicians and departments at inquest. Importantly, these measures also promote trust, realistic expectations and improve overall patient experience.

JAG has now introduced formal accreditation for HPB endoscopy (ie, ERCP and endoscopic ultrasound), which will improve and streamline the training process. This is an opportunity to introduce scenarios in training and allow trainees to anticipate challenging situations. It will also allow ENTS-based training via simulation courses and introduce these into JAG basic skills courses. Therapeutic endoscopy is associated with increased risks of legal and governance issues due to its association with emergent indications and more frequent complications. Unfortunately, traditional teaching methods do not prepare trainee endoscopists for many of these issues before they become consultants with ultimate accountability. The stress associated with this can be reduced through earlier exposure and mentoring.

Data availability statement

No data are available.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

Not applicable.