For each of the years 2003, 2004, and 2005 the number of citations for individual papers published in Physics in Medicine and Biology was compared to the mean quality-score assigned to the manuscript by two independent experts as part of the normal peer review
process. A low but statistically significant correlation was found between citations and quality score (1 best to 5 worst)
for every year: 2003: −0.227 (p < 0.001); 2004: −0.238 (p < 0.001); 2005: −0.154 (p < 0.01). Papers in the highest quality
category (approximately 10 per cent of those published) were cited about twice as often as the average for all papers. Data
were also examined retrospectively by dividing the papers published in each year into five citation quintiles. A paper of
the highest quality is about ten times more likely to be found in the most cited quintile than in the least cited quintile.
By making the assumption that the mean number of citations per paper is a reasonable surrogate for the impact factor, it was
also shown that the impact factor for Physics in Medicine and Biology could be increased substantially by rejecting more papers
based on the reviewers’ scores. To accomplish this, however, would require a reduction in the acceptance rate of manuscripts
from about 50 per cent to near 10 per cent.