0045 Reporting bias in observational epidemiologic research on phthalates

    loading  Checking for direct PDF access through Ovid


BackgroundObservational epidemiology studies on phthalates are diverse, have yielded contradictory results and consequently are difficult to assess. In order to better evaluate these studies we conducted a systematic review including a request for the underlying study protocols.MethodsA literature search and additional searches yielded 158 eligible journal articles. Corresponding authors were asked to participate in a short telephone interview and to provide a copy of their study protocol. Study characteristics were scored and their association with protocol provision examined.Results47 (29.7%) Corresponding authors agreed to be interviewed and ultimately 22 (14%) provided a copy of the study protocol. Of the 43 publications for which we received information about whether a protocol existed, 16 confirmed there was none and three protocols had been lost. Corresponding authors reporting their study as being positive were three times less likely to provide a copy of their protocol and to participate in the interview (OR=0.31 95% CI: 0.11–0.86). Concordance between the protocol and the publication could not be assessed because of lack of detail in nearly all protocols.InterpretationEpidemiology studies on phthalates often lack a protocol and transparency. Given these results we are uncertain whether a formal systematic review of this body of literature will provide reliable risk estimates and we refrain from conducting one. We recommend researchers conducting systematic reviews on observational epidemiology studies to obtain a copy of the underlying protocols prior to conducting the systematic review.FundingThe European Council of Plasticizers and Intermediates (ECPI)Declaration of potential conflict of interest:The study was funded by The European Council of Plasticizers and Intermediates. Gerard Swaen is a consultant to ECPI

    loading  Loading Related Articles