A citation study of the NIH peer review process
The average priority score given by NIH peer review groups to research grant renewal applications improved over time compared to new applications. These trends are due to the effect of increasingly tighter budgets during the early part of the period causing funding higher quality grants and therefore renewal applications increased in quality during subsequent years. The disapproval rate of renewal applications has been declining more than can be explained by increased quality. The probability that a grant will publish a frequently cited article is related to the priority score received on the application, but is more strongly related to the score received on an application to renew the grant following publication. The correlation coefficient between the score received on a renewal and the score received the previous time the grant was reviewed is 0.4. This low value can be explained in part by the output produced during the previous grant period. These observations are consistent with the existence of a good deal of uncertainty in judgments of proposed research and adaptation by the second study section to the merits of the research and away from the earlier appraisal of its potential. New projects receive a slightly better score than renewal applications with the same citation rate.