Quality and content of abstracts in papers reporting about drug exposures during pregnancy.

摘要: BACKGROUND: Most clinicians read only the abstract of papers in scientific journals. Therefore, it is very important that abstracts contain as much information possible, to summarize data succinctly. Our objectives were evaluate quality reporting human fetal outcomes following drug exposure during pregnancy. METHODS: We developed criteria based on previous work, modifying them for use with pregnancy outcomes. Quality scores calculated present/absent all equally weighted criteria, then expressed percentages (present/[present + absent]). We examined a random sample 100 obtained through searches MEDLINE, EMBASE, and Web Science databases from 1990 2005. Average compared across designs (cohort, case-control, meta-analysis, mixed design) Using Kruskal-Wallis ANOVA structured/unstructured formats using Student's t test. RESULTS: the overall average was 59.2% ± 14% (median, 61.5%; range, 15.4–83.3%). not significantly different (P = .16) or between structured unstructured .44). increased over time (Rho 0.23, P .02). Most frequently absent baseline risk (94%), dose (91%), nonsignificant values (72%), confounders (69%), significant (57%), difference (48%). CONCLUSIONS: Abstracts provide insufficient information, particularly values, readers make evidence-based decisions regarding pregnancy. Efforts need be made improve include critical such risk. Birth Defects Research (Part A), 2006. © 2006 Wiley-Liss, Inc.