We reviewed the abstracts of presentations given at meetings whose audiences were thought likely to overlap with the Association of Surgeons meeting (table 1). Papers from non-surgical departments, departments outside the United Kingdom and Ireland, and those updating multicentre trials were excluded. For each abstract the title, department, and first two authors were recorded. When the title or authors from the same department coincided the abstracts were read to detect duplication, defined as the methods being identical and the data identical or similar, including the addition of a few patients to a previous series.