Citation frustration

I was intrigued by your league table of biomedicine (March 4) - in particular the research standing of UK universities as reflected in the "citations per paper" index.

Evidence Ltd publishes an annual UK Higher Education Research Yearbook that shows each university's citation impact (citation counts rebased to the world average) in different subjects. Although biomedicine as such is not included as a category, clinical medicine, biological sciences and subjects allied to medicine are recorded, and a weighted average of the impacts in these three categories can be calculated to reflect biomedical science as a whole. In all three listed subject categories and in the calculated biomedical science category, the highest impact (two to three times the world average) is recorded for the Institute of Cancer Research; the University of Dundee also figures prominently. Neither of these is listed in your league tables.

The Institute of Cancer Research may have been excluded from the universities table because it has only postgraduate students, in which case it should have figured in the research institutes table. Some other highly cited European research institutes are also missing - NKI, Amsterdam, and DKFZ, Heidelberg.

It is not reassuring to see internationally excellent centres of research missing from listings based on citation analysis. If this is down to the varying methodologies that are used in compiling different bibliometric indices of impact, it raises issues for the wider use of such metrics for research performance evaluation and, with the next research assessment exercise looming, suggests caution in relying on the value of citations as objective research indicators.

Keith Snell
Scientific secretary
The Institute of Cancer Research
University of London