Informed consent in interventional radiology – are we doing enough?

Abstract

Objectives:

Obtaining informed consent is a mandatory part of modern clinical practice. The aim of this study was to identify how often complications relating to Interventional Radiology (IR) procedures were discussed with the patient prior to the procedure.

Methods:

A retrospective analysis of 100 patients who experienced a complication related to an IR procedure was performed. The patient’s procedure consent form was examined to identify whether the complication they experienced had been discussed as a possible risk. Other parts of the consent form relating to need for blood transfusion and the need for further procedures were also examined.

Results:

39% of patients who experienced a complication did not have the complication documented as a potential risk on the consent form. 14% of patients required a blood transfusion but were not consented for this. 42% of patients required a further procedure or operation but were not warned of this.

Conclusion:

The model of gaining informed consent on the day of procedure is no longer valid. Better education and the use of clinics, patient information sheets and other resources is essential.

Advances in knowledge:

The paper highlights the inadequacies of the current model in gaining consent for IR procedures. A more comprehensive consent process making use of all available resources is essential.

Introduction

Obtaining informed consent from patients prior to medical treatments is a pillar of modern medical practice. ^1–3 In the UK, the General Medical Council explicitly lays out the duties of doctors in its document entitled “Consent: patients and doctors making decisions together”. ⁴ This document mandates a discussion of risks defined as side-effects, complications and failure of an intervention to achieve its desired aim.

Interventional radiology (IR) finds itself uniquely positioned in that the types of intervention being proposed are often highly complex and carry significant risks, nevertheless the patient will often have seen a specialist in a different department who has made the decision to refer to interventional radiology. The corollary of this is that the interventional radiologist may not see the patient themselves until the day of procedure. The referring physician may not have a detailed understanding of the proposed procedure and therefore will be unable to provide specific details to the patient. The potential therefore exists for patients to not have had sufficient information or the time to comprehend this information prior to having their procedure performed.

As a result of these concerns, there has been an increasing trend in recent years for interventional radiologists to see patients in advance of procedures within IR clinics. However, within the UK, this has not yet been adopted universally with a recent survey identifying only 50% of UK interventional radiologists as having access to IR clinics. ⁵

The aim of this study was to identify patients who had developed complications related to interventional procedures and identify whether or not the complication that they had experienced had been discussed with them as evidenced by documentation on the pre-procedure consent form.

Methods and materials

Institutional board approval for a retrospective audit was obtained. The setting for the study was a UK University teaching hospital with an established IR department consisting of six Consultant IRs over the period of the study. The unit performed the full range of vascular and non-vascular interventions and was a transplant and major trauma centre. 24 h a day on call IR services were available.

The data set consisted of consecutive complications discussed in the bimonthly Interventional Radiology mortality and morbidity meeting between January 2015 and January 2019. Neurointerventional radiology complications were not discussed in these meetings as they were performed by a separate team with its own governance arrangements. Aortic interventions, although performed jointly by IR and Vascular Surgery teams were also excluded as the consent process for these cases was led by Vascular Surgery. Other exclusion criteria were patients who had consent obtained by other clinical teams, patients who were unable to consent either due to being under anaesthesia or because of dementia, patients whose consent form was not available in the electronic medical record and cases that were discussed in the meeting for shared learning, but which were not felt to represent complications such as patients with failed treatments rather than complications. After these exclusions were applied, a final data set of the first 100 patients with complications was identified.

A standard trust approved generic consent form was used for all procedures. This consent form contains the following sections: name of procedure, the intended benefits of the procedure, the possible risks involved (including those specific to the patient), what the procedure is likely to involve and the benefits and risks of alternative treatments including no treatment. There are further sections with tick boxes relating to extra procedures that may become relevant after the proposed procedure including a separate tick box for blood transfusion. The remainder of the consent form deals with questions relating to training, consent for use of medical records and tissue in research. Finally, there are sections dealing with patients who have impaired hearing or who have required a translator to help with obtaining informed consent.

During the study period, there were no formal IR clinics in place with the consent discussion between the IR team and the patient occurring on the day of the procedure, either in the ward or immediately prior to the procedure itself.

The 100 cases identified were reviewed again using the electronic medical record (EMR) system. Complications were graded according to the CIRSE complication grading. ⁶ Unlike previous systems which are based on surgical complications for example the Clavien-Dindo scoring system, ⁷ the CIRSE system is specifically designed for IR procedures. The system assigns a score of between 1 and 6 depending on the severity of the complication (Table 1). The nature of the complication was recorded and then the patients consent form for the procedure was examined. If the complication that occurred was documented as a potential risk on the consent form, then it was assumed that the patient had had that specific complication discussed. If there was no mention of the complication that occurred on the consent form, then the patient was deemed to have not had that specific complication discussed.

Table 1.

The CIRSE classification of interventional radiology complications

Grade	Description
1	Complication during the procedure which could be solved within the same session, no additional therapy, no post procedure sequalae, no deviation from the normal post therapeutic course
2	Prolonged observation (including overnight stay as a deviation from the normal post-operative course <48 h); no additional post procedure therapy, no post procedure sequalae
3	Additional post-procedure therapy or prolonged hospital stay (>48 h) required; no post-procedure sequalae
4	Complication causing a permanent mild sequalae (resuming work and independent living)
5	Complication causing severe permanent sequalae (requiring ongoing assistance in daily life)
6	Death

Although for most of the cases it was clear whether a complication had been discussed, a pragmatic approach to deciding whether a complication had been fully discussed was adopted in the event of doubt. For example, the term “bleeding” on a consent form was felt to encompass pseudoaneurysm formation as well as haematoma, and therefore pseudoaneurysms were not included as a non-discussed complication in this situation. However, death from a bleeding complication was felt to be significant enough to have warranted specific discussion. Author consensus was used to decide on any specific cases where there was doubt about the documentation.

The grade of the person filling in the consent form (IR registrar/fellow or IR consultant) was recorded. The completion of tick boxes regarding discussion around the need for further procedures and blood products was also recorded.

Results

Over the study period, 11,620 cases were performed of which 173 (1.5%) cases had complications and were discussed in the mortality and morbidity meeting. These were screened to identify the cohort of 100 cases with complications that formed the final data set for this study.

Overall, 39% of complications that occurred were not documented on the completed consent form. A breakdown by CIRSE grade of complications overall and in the non-consented group is shown in Table 2. The eight deaths in the non-consented group related to bleeding (TIPSS/angioplasty/liver biopsy, spleen biopsy and PTC), ischaemic acute liver failure (TIPSS, hepatic arterial stenting) and non-target embolisation (transarterial embolisation of HCC and GI bleed).

Table 2.

Breakdown of complications by CIRSE grade

CIRSE grade	Overall data set (n = 100)	Non-consented group (n = 39)
1	33	12
2	6	4
3	47	14
4	3	1
5	3	0
6	8	8

13/39 (33.3%) of the non-consented group were consented by trainees as compared to 28/61 (45.9%) in the patients who were appropriately consented. 14% of the overall cohort required blood transfusion although the box for this was not ticked on the consent form; 4 of these patients were within the group who experienced a complication that had not been discussed with them. 42 patients required a further procedure as a result of their complication but did not have the appropriate box marked. Of these, 18 patients had not had the complication they experienced discussed with them.

A breakdown of the 39 complications not mentioned in the consent form is outlined in Table 3. One patient had a contrast allergy that was not detected during the WHO checklist, and a second patient developed an arm neuropraxia following hepatic radiofrequency ablation. Of note, the majority (26/39 = 67%) of the complications that occurred are described in the literature of occurring with a greater than 1% frequency.

Table 3.

Summary of complications not described in the consent form

Type of complication	Number (n = 39)
Bleeding	8
Non-vascular visceral injury	7
Organ ischaemia/arterial occlusion	7
Malposition device	6
Device failure	5
Non-target embolisation	2
Arterial dissection	2
Others	2

Discussion

NHS guidance states that the principle underlying informed consent is that a person must give permission before they receive any type of medical treatment, test or examination. This must be done based on an explanation by a clinician. For consent to be valid, it must be voluntary and informed, and the person consenting must have the capacity to make the decision. Consent can be presumed, implied, verbal or written. ⁸ In the context of interventional or surgical procedures which carry potential risks and complications, the written consent form has become standard practice.

Nevertheless, there are well-established problems with the use of written consent forms with acceptance that the signature of a patient on a form does not constitute informed consent. ⁹ Rather, the obtaining of informed consent is a process whereby details of the proposed treatment, common and serious side-effects and the probability of success should be discussed. The patient should also be made aware of alternative treatment options available including no treatment and if further procedures would be required. Referencing complication rates may be helpful, but on its own is not indicative that informed consent has been obtained. ¹⁰

Such discussions inevitably require time for the discussion and temporal spacing from the time of proposed procedure in order to be effective as patients may have a quite different pre-existing expectation of the procedure to reality. A patient perception study of 200 patients undergoing peripheral angioplasty concluded that patients tended to underestimate the risks and overestimate the benefits of the procedure. ¹¹ Previous surveys from the European IR community have also suggested that more than 50% of interventional radiologists are unhappy with their local policies for obtaining informed consent with insufficient time identified as a major factor. ¹²

In the UK, the law around informed consent changed following the Montgomery ruling in 2015 with a shift in the test of duty of care with regards to material risk in the context of informed consent from the clinicians perspective to the patients. ^13,14 In practical terms, the implication is that it is insufficient to restrict discussion about potential complications to commonly occurring complications or those which the operator feels to be important. Instead a full detailed discussion including serious but infrequent risks should occur along with the other parts of informed consent outlined above and, as identified in previous studies, meeting patient expectations of consent discussions. ^15,16

IR finds itself in the eye of the storm within radiology litigation claims with rising rates of procedural complication lawsuits identified. ^17,18 In addition, there are specific risk factors involved with IR practice. ¹⁹ For example, the use of “off label” devices to treat specific cases is a relatively common occurrence which should entail specific discussion around the rationale for treatment. ²⁰ Secondly, the operator may be early on in their learning curve for a procedure, and quoted literature rates of complications (if they exist at all) may not be relevant. Thirdly, many IR procedures involve ionising radiation and there is evidence that the majority of patients would like a discussion about any potential increased cancer risk. ²¹ Finally, many procedures or pieces of equipment which are thought to be beneficial are later shown to be of dubious benefit and may actually be harmful. ²²

This study reviewed a series of complications that were experienced by a group of patients undergoing IR procedures. The headline figure of 39% of complications experienced not recorded on the consent form is alarming. The study also identified that common sequelae such as need for blood transfusion or further procedures was not discussed in a significant proportion of patients.

The study was not designed to evaluate the reasons for why the complications were not discussed, although there are several potential causes. There is the possibility that there was a lack of awareness of the risks of the procedure by the consenter. However, the majority of the complications that occurred were not especially rare or unusual, suggesting that lack of knowledge alone would be unlikely to explain the findings. Furthermore, there was no increase in non-consented patients amongst those consented by trainees.

The likeliest explanation therefore is that a lack of time is the reason why patients were not appropriately consented. There has been an increasing global trend therefore in the development of IR clinics which also necessarily entails a change in training ²³ in order to ensure that the IR is up to date with alternatives to the proposed IR treatment. In the UK, the Royal College of Radiologists recognise and advocate time and resources for IR clinics in their job planning guidance for IRs, ²⁴ however only 50% of IRs report having access to such clinics. ⁵ The reasons for this are unclear but may reflect a lack of understanding at senior hospital managerial level with regards to the nature of modern IR practice.

Of note, no patient in this cohort had death mentioned as a potential complication despite there being eight deaths overall. The reasons for this are likely to relate again to time constraints and reluctance from the consenting physician to mention this for fear of dissuading the patient from having the procedure. Nevertheless, where data exist it should evidently be mentioned in the wake of the Montgomery ruling. A UK national liver biopsy audit ²⁵ identified 4/3486 (0.11%) deaths following biopsy with all deaths occurring in a subgroup of targeted focal lesion liver biopsies (4/1222 = 0.33%). It would seem likely that a “prudent patient” would like the small but real chance of death to be discussed with them.

Informed consent should always be done in an appropriate environment with sufficient time at hand to discuss all possible complications and their treatment options. GMC guidance about seeking informed consent is invaluable. In addition to clinics, other well-evidenced strategies should be employed to further inform patients about the risks of procedures. These include patient information sheets, ²⁶ produced locally or by IR societies, i.e. BSIR and CIRSE and the use of specialist nurses ²⁷ to answer questions that patients may have. The use of procedure specific consent forms may also be helpful ²⁸ provided that there is space to explore consent issues pertinent to that specific patient.

Limitations to the study

This was a retrospective study with the usual limitations of such an approach. Nevertheless, to identify a significant number of complications for analysis in a prospective manner would have required a very large cohort of patients and taken a number of years to perform. In the event, more than 11,000 procedures were performed, from which this data set was identified.

The local IR mortality and morbidity meeting relies on open discussion of all cases. The patient’s hospital ID and a brief summary is noted in a “Complications book”. The culture of the department is to try and record all complications regardless of importance or whether the submitter was the operator in the case. Nevertheless, there will be bias in the types of case submitted with more “interesting” or more serious cases likely to be listed for discussion. The spread of cases submitted encompassed a wide range of CIRSE grades suggesting that this was not a major issue.

Secondly, informed consent is a process not just a box filling exercise and it may be that the complication that occurred was discussed with the patient but not documented on the consent form. Most of the patients in the study cohort would have seen another clinician prior to referral to IR and it may be that complications were also discussed at that stage. Without audio-visual consultation recording however these data are impossible to obtain.

Conclusion

This real-world data set of complications shows that despite increasing awareness of IRs with issues around consent there remains a significant proportion of patients who are not appropriately counselled about the potential complications of IR procedures. Time to discuss specific risks in dedicated IR clinics and the use of ancillary supporting resources is mandatory in modern IR practice.