Article type
Year
Abstract
Background: Papers selected for systematic reviews are generally screened by title and abstract. Since harms are reported poorly, screening using title and abstract may miss relevant papers when compared to screening full text.
Objectives: To identify if systematic reviews of harms using the traditional method of screening papers based on title and abstract will omit relevant studies, in comparison to screening based on full text.
Methods: We are proceeding with two systematic reviews of adverse events associated with propofol infusion in pediatric patients. Only clinical trials will be selected (randomised or quasi-randomised). In the first review, papers are screened by title and abstract, looking for reports of presence or absence of adverse events. In the second review, the papers are screened using full text for reports of adverse events.
Results: Both reviews are underway and full results will be shared at the Cochrane Colloquium. A total of 109 papers fulfilled the inclusion criteria and were selected, of those, 59 publications reported adverse events in title or abstract and 14 only reported at the results section. Thirteen publications had not reported the presence or absence of adverse events. A total of 23 non-English publication are in the process of being translated.
Conclusions: Although at a preliminary state, the partial data suggest that clinical trials do not consistently report the presence or absence of adverse events. Systematic reviews of harms may need to screen studies based on full text in order to capture all applicable studies and reduce type II error.
Objectives: To identify if systematic reviews of harms using the traditional method of screening papers based on title and abstract will omit relevant studies, in comparison to screening based on full text.
Methods: We are proceeding with two systematic reviews of adverse events associated with propofol infusion in pediatric patients. Only clinical trials will be selected (randomised or quasi-randomised). In the first review, papers are screened by title and abstract, looking for reports of presence or absence of adverse events. In the second review, the papers are screened using full text for reports of adverse events.
Results: Both reviews are underway and full results will be shared at the Cochrane Colloquium. A total of 109 papers fulfilled the inclusion criteria and were selected, of those, 59 publications reported adverse events in title or abstract and 14 only reported at the results section. Thirteen publications had not reported the presence or absence of adverse events. A total of 23 non-English publication are in the process of being translated.
Conclusions: Although at a preliminary state, the partial data suggest that clinical trials do not consistently report the presence or absence of adverse events. Systematic reviews of harms may need to screen studies based on full text in order to capture all applicable studies and reduce type II error.