Members of the meeting were:
Sandra Goldbeck-Wood
Teifion Davies
Ian Forgacs
Stephen Evans

Decision: Offer to publish in the form of a short report
Nature of Decision: Majority

Who retires early from the NHS due to ill health and what does it cost?

Thank you for giving us the opportunity to consider this paper. We feel that the point it makes would be of interest to a BMJ readership if presented in a brief way, in the form of one of our short reports. If you would be willing to revise it in this way, we would be delighted to publish it. The following few points were also made, and we would like you to attempt as far as possible to address these while staying within the constraints of the short report format. We realise that a compromise will have to be reached in order to achieve this:

1. We would want you to remove the section relating to the SF36.
2. We did not understand your description of pension enhancements.
3. As it is, your paper comes across as being rather parochial, when viewed from the perspective of the international readership of the BMJ. Since the point you make is one which might potentially have relevance to other industries, and since the NHS is Europe’s largest employer, it seems to us that this need not be the case. We would like you to rewrite the paper with our international readership in mind, trying to make clear in what way the results you present might be of wider interest. For example, if you are able to provide any references which would allow comparison with other industries or countries, this would be helpful.
4. We feel that your conclusion "a third of cases may be preventable" is over ambitious. Since this is bases solely on self-reporting, in a circumstance where bias is very likely, we feel you can draw conclusions reliably neither about cause, nor about prevention. We would want you to omit this.
5. We had some concerns over the conclusions you drew relating to "losing highly experienced employees", in view of the huge range of times spent in NHS employment (2 – 40 years). We were concerned that there were some over interpretation happening here.
6. Could you reassure us that you had prior consent from the individuals studied to use their data in future research.
7. We were concerned that April to August may not be representative of the whole year. Ideally, we would have preferred to see a random sample taken over a period of years, rather than 100% sample over months. In particular, we worried that any bias by access to financial advice might lead to particular categories of worker being more or less likely to retire at particular times of year.