Introduction
Identifying and learning from mistakes is central to improving healthcare. The process enables transparent discussions to explain why something has gone wrong and what should be done to avoid recurrence. In many countries, there are professional and moral obligations to be candid with patients when harm occurs. However, in England this concept was formalized in 2014 with the introduction of Duty of Candour (DoC) legislation [1 ].
The regulation applies to National Health Service (NHS) providers in England registered with the healthcare regulator and mandates disclosure of “notifiable safety incidents” to patients as soon as possible. A notifiable safety incident is one that:
must have been unintended or unexpected
must have occurred during the provision of a regulated activity
has, or might, result in death or severe or moderate harm.
This is a legal requirement and failure to comply can lead to prosecution [2 ]. When a notifiable safety incident occurs, the following actions are required.
Tell the person (or, their advocate, relative, or carer) that something has gone wrong.
Apologize.
Offer an appropriate remedy or support to put matters right (if possible).
Explain the short- and long-term effects of what has happened to the person.
Despite the legislation being in place for 10 years, little is known about how DoC is applied across the English NHS.
Post-colonoscopy colorectal cancers (PCCRCs) are colorectal cancers diagnosed following a colonoscopy where no cancer was identified [3 ]. More than 1000 PCCRCs occur in the NHS each year [4 ]. The majority of these are avoidable and many meet the definition of a notifiable safety incident [5 ]. There is also variation in the rate of PCCRCs, suggesting that many can be prevented by higher quality colonoscopy [6 ]
[7 ]
[8 ].
A national audit of PCCRCs was undertaken to help and encourage services to identify and review PCCRCs and implement changes to prevent them. The audit provided guidance on discharging DoC and captured information about its application.
Our aim was to describe and analyze how DoC is applied to PCCRCs within the English NHS. The intention was to raise the profile and understanding of the legislation, and to enable a better understanding of the difficulties of applying the legislation in endoscopy and how these might be overcome.
Methods
Between 1 September 2021 and 10 May 2022, a national audit of PCCRCs was conducted in all acute hospitals providing colonoscopy in England. A single-center example of a similar audit has been published [6 ].
PCCRCs were identified centrally using linked population-based administrative datasets. Audit leads were informed of their PCCRCs via an online portal, which contained a template to aid root-cause analysis. It is estimated that >85% of the PCCRCs in England over the study period were identified. Audits were completed via the portal. Participation was not mandated.
The audit template prompted reflection on factors that contributed to the PCCRC and whether it was avoidable. “Avoidability” was divided into four categories: unavoidable, possibly avoidable, probably avoidable, and definitely avoidable. If a PCCRC was deemed avoidable, the audit captured information on which organizational, diagnostic, or therapeutic process had failed. The template prompted assessment of the level of patient harm. Harm was categorized into four groups: no/minor harm, moderate harm, major harm, and premature death.
Determining levels of “avoidability” and harm can be subjective. To improve consistency, the audit issued guidance. For example, the guidance recommended that judgments were made after anonymous presentation to a group of at least three experienced endoscopists to form a collective judgment.
Quantifying harm is particularly challenging and endoscopists are not necessarily best placed to make this judgment. For example, some may regard an uncomplicated endoscopic excision of a PCCRC in a polyp as no/minor harm, but a person given a diagnosis of a cancer that could have been avoided may view this harm differently. In this study, the guidance included how DoC should be applied (see Fig. 1s in the online-only Supplementary material) and examples of cases, as presented in the Supplementary material. The guidance included a detailed explanation of the law, the principles of DoC, and how to support judgments when the case was not clear-cut (Fig. 2s ).
Once “avoidability” and harm judgments had been made, the guidance recommended discharging DoC if the PCCRC was considered probably or definitely avoidable and resulted in major harm or premature death ([Fig. 1 ]). The formal audit standard was 100% adherence to this recommendation. Although incidents resulting in moderate harm are “notifiable safety incidents,” discharge of DoC was not recommended in those instances owing to the subjective nature of harm and the priority to encourage engagement. If there was uncertainty about discharging DoC, involvement of the relevant hospital clinical risk department was advised.
Fig. 1 Algorithm for decision making in Duty of Candour (DoC) discharge in cases of post-colonoscopy colorectal cancer (PCCRC).
The main outcome measures of the audit were the number of PCCRC cases, assessed harm levels, “avoidability,” and DoC discharge rates. An analysis was undertaken to determine the number and proportion of cases in which DoC was triggered when the harm threshold was set at major harm or death. The analysis was repeated with the threshold lowered to include moderate harm.
Results
A total of 2263 uploaded PCCRCs were deemed eligible for audit, of which 1724 (76%) were completed and submitted. Of the remaining cases, audits for 153 were only partially completed and 386 were not started. Of the 1724 submitted, 962 (56%) were judged to have resulted in no or minor harm, 486 (28%) in moderate harm, and 276 (16%) in major harm or death. The number of PCCRCs where DoC was discharged when the threshold was set at major harm and death is presented in [Fig. 2 ]. Of these PCCRCs, 75 (27%; or 4% of the total audited cases) were probably or definitely avoidable and met the audit standard threshold for discharging DoC. However, DoC discharge was deemed necessary in only 27 (49%) of the 55 that were probably avoidable, and in 13 of the 20 (65%) that were definitely avoidable. This reflects 53% compliance with the audit standard. There was uncertainty about DoC discharge for 2 (10%) of the definitely avoidable and 18 (33%) of the probably avoidable PCCRCs.
Fig. 2 Categorization of post-colonoscopy colorectal cancers (PCCRCs) according to whether harm was caused and how avoidable the cases were (N=1724). PCCRCs for which Duty of Candour (DoC) was discharged, and should have been discharged (yellow and red banding), are shown.
DoC discharge was deemed necessary in 128 PCCRCs that had resulted in no, minor, or moderate harm. These included 20 PCCRCs that were unlikely to be avoidable and 57 that were possibly avoidable.
If the audit threshold for DoC discharge was lowered to include moderate harm as well as major harm and death, DoC should have been triggered for an additional 116 PCCRCs, or a total of 191 (11%). DoC discharge was deemed necessary in 85 (45%) of these: 63 (33%) probably avoidable and 22 (12%) definitely avoidable (Table 1s ). This reflects 45% compliance with the audit standard.
Discussion
This study shows that, despite recommendations and guidance, there is inconsistent application of DoC following identification of a PCCRC across the English NHS. Audit standards were not met, despite being a statutory requirement with the risk of prosecution for failing to adhere to the legislation. Major factors in this inconsistency are the difficulty in determining whether a PCCRC was avoidable (unlike in radiology, images of the site of the cancer are rarely available for review) and of judging harm when the consequences of an error often appear years after the initial colonoscopy.
Multiple other factors may explain the inconsistent discharge of DoC in PCCRCs, including concerns about the time, workload, and emotional impact of initiating the process, particularly when discussing harm with patients and their advocate, relative, or carer. Additionally, there is a lack of experience in making judgments, which is compounded by the complexities of assessing the natural history of colorectal cancer and the quality of past colonoscopies. For example, proxy measures of quality such as adequacy of bowel preparation and photographs of anatomical landmark are not reliable guides of mucosal visualization. They may provide false reassurance about quality, resulting in a false sense of adequacy and subsequent conservative categorizations of “avoidability.” The imprecise natural history of polyps also complicates attributing PCCRC to inadequate technique, unless a precursor lesion was identified in the same colon segment. However, the relationship between adenoma detection rate and PCCRC indicates that most are due to missed precursor lesions and are therefore avoidable [7 ]
[8 ]. In contrast, it is easier to interpret the need for DoC when PCCRCs arise due to nonprocedural factors. Avoidable delays due to long waiting times, booking failures, and excessively long referral pathways for polypectomy are usually obvious.
Judgments about the level of harm can also be complex. While uncomplicated endoscopic treatments for cancer cause no or minor harm, categorization of harm from surgery or chemotherapy is more difficult. Excision of a rectal cancer with a permanent stoma causes more harm than uncomplicated surgery for cecal cancer. It may be obvious to discharge DoC for the former operation, but not for the latter. NHS England has recently clarified levels of harm, defining moderate harm as requiring more than one extra professional healthcare interaction, temporarily reduced independence (under 6 months), or impacting treatment effectiveness without shortening life expectancy or worsening disability [9 ]. These criteria, along with setting moderate harm as a criterion for DoC discharge, will increase the percentage of PCCRCs meeting the DoC threshold in the future. If moderate harm was applied as the threshold in this audit, the number of PCCRCs potentially meeting the requirement for DoC discharge would increase from 4% to 11%.
Even when a PCCRC meets the criteria for discharge of DoC, some may argue that it is not in the best interests of the patient to be informed. For instance, the patient might lack the capacity to grasp the significance of the incident and could become distressed when it is revealed, or it is possible that the patient might now be deceased. Such situations may explain the variation in the application of DoC for those who met the criteria. However, DoC regulations make it clear that a paternalistic approach is not appropriate. An independent review into the Republic of Ireland CervicalCheck screening program was extremely critical of nondisclosure incidents [10 ]. Further research, determining the negative consequences of discharging DoC, is necessary.
Some inconsistency in discharging DoC may be related to concerns about triggering formal complaints and/or litigation. Open disclosure is likely to have the opposite effect on litigation: health systems that adopt active medical error disclosure show a reduction in legal defense costs [11 ]. All these difficulties were raised in discussions about DoC with sites during and after the audit closed.
There is little published literature to inform judgments, and what exists is limited to a survey demonstrating variation in the interpretation of DoC between peers [12 ]. To our knowledge, this is the first study to highlight the challenges peculiar to endoscopy, when the harm is apparent so long after the procedural or system failure.
This national audit of PCCRCs has accelerated the DoC journey for endoscopy providers; a journey they would have had to make eventually. Over time, it is hoped that providers will become more familiar with the process and less wary of adverse consequences. If the audit continues, we anticipate that DoC practice will become routine. Inevitably, some variation in application of DoC will persist and ways to mitigate this will need to be identified. These should include clearer guidance and e-learning resources, similar to those released by the UK Royal College of Radiologists and Royal College of Surgeons of England [13 ]
[14 ]. Formal frameworks for the improved application of DoC in clinical practice would also help clinicians make consistent decisions, as has been proposed in other specialties [15 ].
The main aims of this paper are to raise awareness of DoC, to understand the challenges of applying it consistently in endoscopy, and to help other specialties grapple with the challenge [12 ]. The British government is planning a review of DoC regulations in England, which will explore the extent to which:
the policy and its design are appropriate for the NHS
the policy is honored, monitored, and enforced
the policy has met its objectives.
The current study will provide important evidence for this review. Greater awareness of DoC will prompt governments and health ministries beyond England to consider being more open with patients about notifiable safety incidents.
