Philip Berry: Teething problems with duty of candour

To doctors what began as a transparent human response has evolved into a series of deadlines

In my roles as departmental governance lead and, formerly, trust lead clinician for patient safety, I have helped to develop duty of candour (DoC) processes in two hospitals. Deciding when and how a “candour discussion” should be initiated has revealed that this responsibility contains grey areas and unforeseen challenges.

When should they take place?

The answer is simple: if significant harm has occurred to a patient during a healthcare episode. Yet the problem is we are not agreed on how to define significant harm. “Significant” harm encompasses “moderate” and “severe” harm. The Care Quality Commission and the National Reporting and Learning System give some example scenarios. For example, a patient whose surgery was postponed because of the failure to manage their warfarin prescription is proposed to have suffered moderate harm. To me, that sounds like an inconvenience—one which may cause psychological distress—but which causes minor or zero harm (assuming the operation was not for cancer). At the other end of the spectrum, a bowel perforation during surgery that requires a defunctioning stoma and subsequent surgeries is cited as an example of serious harm. That seems appropriate, but even this example can be quibbled over.

Harm vs known complication

Was that bowel perforation the result of an error in technique, or was it a known, statistically inevitable complication of intestinal surgery? This assessment may be subjective, relying on the attitude of the surgeon or the efficiency of the local morbidity and mortality review system. Who should make that judgment? The surgeon, or a peer? This is controversial.

To make life easy, one could propose that all complications are rated for harm: those which are agreed to have caused moderate or severe harm would then lead to a DoC conversation. Should this include wound infections? Post colonic polypectomy bleeds? Dehiscence of an anastomosis? Such a universal approach would not sit easily with clinicians who consent their patients for complications with as much care as they perform the operation. Not all complications result from error. However, the non-clinician may well retort, what’s so hard about saying sorry when the outcome of a procedure is bad?

A conversation or a process?

DoC is about more than saying sorry. After a harm event has been identified, the patient will be spoken with. Regulation 20 requires that the conversation is reiterated in a letter so that the incident, its impact, remedial action, apology, and a commitment to investigate (if appropriate) are formally recorded. It is at this stage that I suspect many departments fall down. It requires organisation. In the busy world of healthcare, that means administrative support—for instance, a governance manager. This person must also keep an eye on the state of the investigation, ensure that it is completed within 60 days, and arrange for its conclusions to be communicated back to the patient (if they want to receive them).

DoC has thus become a process, overlapping with the Serious Incident Framework. To the clinician, what began as a transparent human response has evolved into a series of deadlines. However, this overlap is clearly necessary if lessons are to be unearthed, shared, and retained by organisations, and if patients are to understand how the harm they experienced came to pass.

My experience with developing and delivering DoC has shown that there are areas of uncertainty when it comes to defining harm and differentiating harm from known complications. Furthermore, the administrative burden of DoC must be recognised, and its potential impact on the willingness of clinicians to report incidents should be considered. The DoC regulation was developed quickly, and is based on unimpeachable principles of Good Medical Practice. Yet from what I’ve seen, how it is done remains variable and is yet to be established in all clinical areas.

Philip Berry is a consultant hepatologist at Guy’s & St Thomas’ NHS Foundation Trust, and is departmental governance lead. He has interests in medical ethics relating to treatment escalation, end of life care, communication around DNACPR decisions, and response to medical error. He writes on medical ethics at, and is on Twitter @philaberry

Competing interests: None declared.

  • Palladium

    Robert Francis and many families have spoken about this. They place value on the conversation taking place and the quality of communication. This is hard enough to achieve. Adding grey areas of definition of thresholds of harm do not help matters. The other challenge is what to do about additional information beyond the scope of the agreed investigation. Deciding how, when or if to share this unforeseen information with patients or families is another challenge.