Reach Us +441474556909
All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Developing an audit instrument for breaking bad news


Macmillan Consultant Senior Lecturer in Palliative Medicine and Communication Skills

Joanne Fisher BSc PhD

Research Fellow

Heather Cooke RGN

Research Associate


Professor of Primary Care and Head of Department

University of Warwick, Coventry, UK

Corresponding Author:
Dr M Barnett
Centre for Primary Healthcare Studies
University of Warwick, Coventry CV4 7AL, UK
Tel: +44 (0)24 7657 2950
Fax: +44 (0)24 7652 8375
Email: [email protected]

Accepted date: September 2003

Visit for more related articles at Quality in Primary Care


This short discussion paper considers the nature and setting of the breaking bad news consultation,and the issues involved in conducting audit in this area. It o¡ers a brief review of some of the limited published work available on this relatively overlooked topic, and sets out the approach undertaken locally to develop an audit instrument focusing on one aspect of the process: documentation of the initial consultation.

Breaking bad news is an important and often di¤cult task faced by nearly all health professionals, although most often responsibility for the initial consultation lies with the doctor responsible for the patient’s care at that time, and this is a role expected by patients.[1]

There has been increasing research into the impact of bad news breaking on patients and its e¡ect on their psychological adjustment and their perceptions of the doctors involved.[2,3] Studies have also considered the stress this task causes doctors.[4,5] In addition, there are numerous reports of educational interventions for healthcare professionals, although only a minority of these have been fully evaluated.[69]

However, much of the research has focused on clinicians working in secondary care. Within primary care, the nature and range of breaking bad news consultations is going to be far more variable for the individual clinician. For example, while confirmation of a cancer diagnosis is likely to be given by a hospital specialist,[3] initial suspicions are likely to be discussed by the general practitioner (GP) prior to referral for further investigations. In the context of a condition such as diabetes mellitus, both diagnosis and follow up are most likely to take place in the primary care setting. A key di¡erence between primary and sec-ondary care is that the breaking bad news consultation is likely to be part of an evolving process that builds on a pre-existing relationship. This has both up and down sides: communication is likely to be easier where the patient and health professional know each other. However, the open access to a surgery appoint-ment means that the clinician has little time to prepare, and may uncover a ‘bad news’ diagnosis during a routine consultation, e.g. the six-week check on a new-born that reveals a previously unsuspected cardiac anomaly.

While there are a wide variety of texts that include pointers on ‘how to do it’, there are very little pub-lished data on policies or auditing the process of bad news breaking. Thus for hospitals or primary care trusts there is a tendency to reinvent the wheel whenever the question of guidelines arises in this area.

One of the issues is determining what to audit, as the breaking bad news process can be viewed in a number of stages:

•    system factors, e.g. the arrangements of appoint-ments to make it feasible for individuals to be scheduled specifically for a breaking bad news consultation, and the availability of privacy

•    the organisation of team roles, e.g. the availability of a clinical nurse specialist either for a first con-sultation or to provide follow-up information

•    the performance of the individual clinician

•    documentation and ongoing communication be-tween members of the secondary care team and the primary care team.

One of the very few attempts to set out process guidelines that might be auditable was by a group reporting to the King’s Fund.[10] They set up a multi-disciplinary joint working group: doctors (surgeons, oncologists and a GP), specialist nurse, counsellor and four patient representatives (including one with experience as a carer). Facilitated by an independent professional, they met with a structured remit and fixed time frame to come up with a series of guide-lines, focusing on the outpatient setting. These covered both system factors (availability of notes and test results, diversion of telephone calls) and individual performance. They also touched on follow up processes (see Box 1). The joint working group also suggested complementary interventions to sup-plement the process:


Box 1: Summary of King©s Fund guidelines10

•    referral form for GPs to communicate relevant information about and to the patient

•    form to be sent to the patient’s GP giving details of the bad news interview.

Following the recommendations of the original group, the guidelines were piloted and audited.

As the authors point out, these guidelines were intended for use when the first diagnosis of cancer was given, and applied largely to the hospital outpatient environment.

In our local acute hospital trust and cancer centre, the lead cancer clinician produced a consensus guide-line document representing cancer clinicians (see Box 2).[11] This applied to a wider setting, although still focused on secondary care. It also did not include an explicit user viewpoint.


Box 2: Summary of Walsgrave (now UHCW) Trust guidelines11

In comparing the two sets of guidelines, there appeared to be agreement on all major points, al-though there were di¡erent points of emphasis. The trust guidelines considered the process for breaking bad news throughout the hospital setting, rather than only in the clinic. Both considered that the most appropriate person to break bad news is the senior clinician involved in the diagnostic process, consist-ent with patient survey results.[1] However, trust guide-lines also acknowledged issues such as junior sta¡ involvement, and the need sometimes to stage the breaking bad news process as more information emerges or as the patient’s wish for information dictates.

As a member of the palliative care team working both in the hospital and community setting, I ob-served that within the hospital we were often ‘picking up the pieces’ emotionally after unsatisfactory

consultations. Meanwhile, within the community there was a sense of isolation, with primary care teams unaware of what had taken place. Thus, we decided to combine elements from the King’s Fund and our own trust guidelines, to audit both documentation and a subsequent examination of practice and patient experience.

First we audited case notes (n = 95), for which we used a 17-point proforma based on a combination of the two sets of process guidelines. This explicitly included points that might be recorded as opposed to observed (e.g. the name and professional identi-fication of the person breaking bad news; the presence of specified team members). While it was anticipated that some of the points might be too detailed for inclusion in every interview, all were felt to have validity. The subsequent analysis showed that docu-mentation was highly inconsistent, and less than half of the 17 points on the proforma were documented regularly (see Table 1). Key points that were infre-quently documented included terminology (i.e. what the patient had actually been told), which was only recorded in 31.6% (30/95), while the patient’s level of understanding was documented in 16.8% (16/95). Who broke bad news was recorded in 67.4% of cases, but in only 42% of those was identification complete (i.e. legible recording of name, signature, professional role). Most disturbing in terms of interprofessional communication was the lack of information passed on to the GP: only 16.4% (12/73 patients who were discharged or seen as outpatients) received written details of what the patient had been told.[12]


Table 1: Audit points used in proforma and frequency of documentation

We then conducted semi-structured interviews with 105 newly referred oncology patients. Full details of this exercise are still being analysed and will be presented in a further publication, but initial findings proved somewhat more reassuring: good practice guidelines appear to have been followed in most cases. Although documentation of areas such as patient understanding was poor in the preceding audit, the patient reports indicated that the majority did feel they understood the information given, and were able to ask questions. The only significant gaps were in pre-warning the patient to bring a companion if required, and providing contact numbers.[13]

However, while patient surveys provide a valid approach to examining the patient experience, they may not always re• ect the objective behaviour of the health professional, which may require a separate observational exercise.[14] This can involve placing an actual observer in the consultation, which is likely to prove practically di¤cult to achieve unless the obser-ver is part of the regular team (as utilised in Walker’s audit in 1996).[10] This then risks introducing subjec-tive bias, unless the rating criteria are very clearly specified, which may reduce the discriminatory power. Alternatively the consultation may be audio-or video-recorded, which can subsequently be rated independently. This is not unfamiliar to GPs, as recording actual consultations has formed part of the assessment process for the Member of the Royal College of General Practitioners (MRCGP) examina-tion since 1995.

There may be some practical constraints in record-ing breaking bad news consultations, although this has been carried out successfully in the oncology outpatient setting.[9] From the viewpoint of auditing clinician performance among those groups where breaking bad news is a less predictable element of their working day, a more standardised approach would be to use simulated consultations. While this produces a best behaviour on the part of the clinician, it is a fair way to evaluate di¡erences in performance in a standardised context, either for examination or educational purposes.[15] A simple rating schedule that we found useful was the BAS (Breaking bad news Assessment Schedule), developed in Oxford for undergraduate educational purposes, but which we have successfully applied to postgraduate assessment.[16]

It is unlikely that any one instrument could be designed to measure all these facets in all settings, but we have begun to identify key points that are amen-able to measurement, and to develop or identify tools that may be used at each step. While it will always remain a challenging task, e¡ective methods for assessing the quality of the process of breaking bad news will help health professionals to ensure that the patient experience is as satisfactory and causes as little trauma as possible. 


Select your language of interest to view the total content in your interested language

Viewing options

Post your comment

Share This Article

Flyer image
journal indexing image

Post your comment

captcha   Reload  Can't read the image? click here to refresh