The commentary by Tyrer (Psychiatric Bulletin, December 2007, 31, 447–449) summarises the reasons why the Royal College of Psychiatrists has decided to abandon the long case as a summative assessment in the MRCPsych examination. The Royal Australian and New Zealand College of Psychiatrists, however, continues to have a long case in their exams for Fellowship as well as OSCEs. The form of the long case is a 50-minute interview by the candidate who is observed by two examiners. After the interview the candidate has 20 minutes to produce a formulation and management plan, which they then discuss with the examiners.

We have persisted with the long case because it is a valid test of important skills. The most important skill it tests is the ability to prioritise information and ‘make sense of a case’ – the time limits force the candidates to work out what are the key issues for the patient. The long case gives trainees and supervisors an important message that interviewing and formulation are skills fundamental to the practice of psychiatry and it also provides an incentive for supervisors to observe their trainees’ interview.

We ensure the reliability of the long case through a number of measures. Each candidate is examined by a senior and experienced examination committee member and an invited co-examiner. Prior to the exam all examiners have a 3-hour training workshop to standardise their marking. During the viva part, examiners may only ask candidates questions from a limited menu of clarification probes. Examiners initially mark the candidate independently and then agree on a consensus mark on five domains using a 5-point scale – half of the marks awarded are identical, a further 40% are discrepant by only one grade, and less than 10% are discrepant by more than one grade. The discrepant marks are resolved by consensus between the examining pair and if this is not possible, each discrepant mark can be discussed at an examiners’ meeting at the end of the examination. While the patients may be different, what the examiners look to mark in candidate performance is generalised and standardised.

There are also important negative reasons why we have decided to keep the long case as a summative assessment in the Fellow of the Royal Australian and New Zealand College of Psychiatrists (FRANZCP) examination. We agree with Tyrer that the main question is not whether the skills tested in a long case are important and need to be assessed, but whether they need to be assessed using a summative examination. A major value of a summative assessment is that the examiners have no possible conflict of interest or even awareness of the prior training and examination history of the candidate. Making the long case part of training as a formative assessment does not get around any problems of reliability and may make the reliability worse as assessors do not have the same degree of examination training. There may also be a significant conflict of interest with local supervisors keen to get their trainees through training.

Finally, there is the wider issue of the change in culture in medicine. Increasingly there are moves to reduce medicine to a set of procedures which are laid out by guidelines, encouraged by incentive payments and evaluated by audit or other performance measures. Relying solely on OSCEs encourages this tick-box procedural approach to healthcare. We believe that what patients need when they visit a specialist is someone who can make sense of complexity, knows what procedures to use and what to do when they do not work. Dropping the long case in the examination is not good for consumers and risks reducing psychiatry to a set of simplistic procedures.

Declaration of interest

L.P. was awarded Laughlin prize for outstanding performance in old format MRCPsych exam, Autumn 2007. He is also involved in writing a multiple choice questions' book for the new format MRCPsych.