How well does journal 'impact' work in the assessment of papers on clinical physiology and nuclear medicine?

This study reports a citation analysis of 217 scientific papers on clinical physiology and nuclear medicine published in 69 different journals during the years 1985-92. The actual citation frequency was compared with the journal 'impact factor' (i.e. the average number of times a paper is cited in a particular journal in the year of publication and the subsequent year). The average impact factor per paper was 1.96 and per journal 1.92 (range 0.0-22.8). A direct relation was found between the journal impact factor and the citation factor (i.e. the actual citations in the year of publication and the subsequent year) (journals with five or more papers R=0.69, P<0.01; all journals R=0.40, P<0.001). However, the citation factor was significantly below the journal impact factor (P<0.0001). An almost linear relation was found between cumulated citations and time (R=0.99, P<0.02-0.0001), but an initial phase of no citation was identified, and the number of citations per year reached a maximum or plateau 3-7 years after publication, which was later than the general maximum of 2-3 years for all medical specialities. In a sample of 200 known autocitations, 75% of the papers had full recovery, 3% mixed recovery and 22% no recovery, giving an overall recovery of 70%. In conclusion, although a phase of no citation was identified, which gave a lower citation factor than the journal impact factor, the citation frequency increased over time, and altogether there seems to be a good agreement between journal impact factor and overall citation frequency of papers on clinical physiology and nuclear medicine.