Reporting quality of conference abstracts on randomised controlled trials in gerontology and geriatrics: a cross-sectional investigation.

Without transparent reporting of how a randomised controlled trial was designed and conducted and of the methods used, its internal validity cannot be assessed by the reader. A congress abstract is often the only source providing information about a trial. In January 2008, an extended CONSORT statement on abstract reporting was published. Its impact has yet to be evaluated. Using a slightly modified CONSORT checklist comprising 17 items, we thus investigated the reporting quality of randomised controlled trials published in the book of abstracts presented at the World Congress of Geriatrics and Gerontology in Paris in July 2009. A total of n=4,416 abstracts was screened for inclusion; n=129 met the inclusion criteria. The overall quality of the abstracts was remarkably poor. The primary outcome was mentioned in 34/129 abstracts (26%), none of the abstracts reported on the procedure of random allocation of participants or clusters, 21/129 abstracts (16%) reported some kind of blinding, and the attrition rate was mentioned in only 12/129 abstracts (9%). The majority of abstracts fulfilled two items: description of intended intervention for each group (102/129; 79%) and general interpretation of results (107/129; 83%). Trial status was reported in all abstracts. Both journal editors and committees organising congresses are requested to define the use of the CONSORT statement as a prerequisite in their guidelines for authors and to instruct reviewers to conduct compliance checks. Medical associations should finally endorse the indispensability of the CONSORT statement and publish it in their journals. Otherwise the intended benefits cannot be fully generated.