BACKGROUND: There is a general believe that risks associated to drugs are less frequently reported in industry sponsored studies, potentially impacting the acceptability of results. The acceptability of results obtained from epidemiological studies has been subject of discussion in journals and conferences, including past ICPEs.
OBJECTIVE: To describe the abstracts presented at 19th ICPE in terms of type of study presented, sources of data, and authors affiliation; and to ascertain the reporting of association between drug exposure and outcome in studies according to the authors’ affiliation and funding of study.
METHODS: All original research abstracts presented, oral or poster, at the 19th ICPE, Philadelphia, August 21–24, 2003 were identified based on PDS 2003, 12 (Suppl 1), and excluding those withdrawn as per the final meeting program. The three authors reviewed abstracts. Each reviewer was randomly assigned two thirds of the abstracts. Two reviewers reviewed each abstract. The non-involved reviewer resolved any disagreement on a case-by-case basis. Results presented in this abstract are previous to disagreement arbitration.
RESULTS: We reviewed 350 abstracts, 61% were posters. Drug utilization studies (28.3%), cohort (18%), case-control (12%), and analysis of spontaneous reports (12.6%) were the most frequent type of studies. One third (34%) of the abstracts presented results from data collected exclusively for the study, another third (34%), reported on studies conducted in automated databases. Adverse events databases were the source for 9.2% of the abstracts. An association between exposure and outcome was studied in 95 abstracts, 77% included authors from academia, 20% from regulatory bodies, 17% from industry and, 16% from contract research organizations or vendors. A positive association was shown in 66% of the abstracts. We did not find significant differences according to the authors’ affiliation: 67%, 63%, and 63% of those with authors from academia, industry, and regulatory bodies, respectively. Data on sponsorship, and other authorship combinations are available.
CONCLUSION: The type of work presented at last year’s ICPE was diverse. Less than one third of the studies were of analytical nature. Most of them were authored only, or in collaboration by researchers in academia. Among those abstracts reporting positive associations between drugs and outcomes, we did not find considerable differences among those with industry, academia or regulatory bodies authorship.
