Profile of a meeting: how abstracts are written and reviewed.

We analyzed submissions to a recent scientific program to determine (1) how abstracts were reviewed and (2) what constituted a successful abstract. We found that (1) reviewers' gradings varied from 2-29%, in some instances differing significantly; (2) many (<74%) abstracts had inadequacies in form, title, introduction, aims, methods, results, and conclusions(collectively termed "content") or lacked numerical or statistical data; (3) accepted abstracts had fewer inadequacies and better "content"; and (4) abstract grades correlated closely with "content". The quality of preparation and of individual features of abstracts led to favorable review. This information is of potential value to scientists preparing and reviewing abstracts and planning programs.