Selecting GP speciality trainees: squaring the circle?

Irish and Patterson's discussion paper reviewed the current pathway for selecting GP speciality trainees. They ask how it can be improved.1

Despite Mencken's caution that ‘for every complex problem there is an answer that is clear, simple, and wrong’,2 maybe there is an easier way.

The current structure (The National Recruitment Office for GP Training), the process (a single standardised recruitment system), and the specific outcome (successful completion of GP speciality training), coupled with continuing attempts to improve the process could make the journey to becoming a GP less of a maze and more a motorway to success.

The ideal selection outcome is that the recruited GP subsequently provides high quality care throughout a 30-year career. An optimal outcome may be that the GP provides high quality care for 5 years and then passes revalidation. The minimum outcome is that the GP passes the MRCGP licensure examination by the end of 3 years of training. However, low pass rates in the Clinical Skills Assessment (CSA) module of the MRCGP examination suggest that either the selection process is flawed, or that the CSA is not a reliable, fair, and valid test. The flaw in selection may be that it is impossible to reconcile the need for the recruitment process to fill all the training posts every year, and ensure that all these recruits pass the MRCGP examination within their 3 years of training. Any worries that the MRCGP examination is not a good test of competence needs to be addressed by the Royal College of General Practitioners (RCGP).

Although the RCGP has always wanted a longer training scheme, the recruitment process needs to select candidates who are likely to pass the MRCGP exam within their 3 years of training. The Mencken-defying improvement may be to link their selection to the results of taking real or mock modules of the MRCGP examination. These modules could be taken online at a convenient Applied Knowledge Test examination centre, remotely by reviewing the candidates' existing ePortfolios including work-place based assessments, and by a CSA at nearby GP-training practices. The candidates could pay a fee to take these assessment modules. Only those candidates with qualifying scores would be eligible to apply for GP training.

The results of these tests could inform the training of recruits to improve their chances of success in the MRCGP exam. Weaker candidates may need a few attempts over a couple of years to qualify for GP training. Perhaps the best of these weaker candidates could be offered any unfilled training posts on an ad hoc basis, but they would not be in the 3-year GP training scheme.

Educators and learners may be reluctant to concentrate their efforts on passing a test, but the RCGP – GP curriculum statements already point in this direction. The RCGP regularly updates a curriculum that describes what a GP needs to be able to do to work in general practice for the first 5 years and conducts a membership examination that should be a reliable, fair, and valid test of that ability. After the first 5 years, the RCGP's (and General Medical Council's) regular revalidation should ensure that GPs have the competencies to practice for the rest of their careers.

Mencken also advised ‘a judge is a law student who marks his own examination papers’.2 This may describe the continuing difference between our confidence and competence in our assessment of the ability of others.

• © British Journal of General Practice, 2011