Objective: To assess the quality of studies of nifedipine used to treat spontaneous preterm labor.
Design: A systematic review of study quality using a novel validity assessment tool, examining method-specific and topic-specific items in the domains of selection, performance and measurement biases.
Data sources: Medline (1996-2003), EMBASE (1996-2003), BIOSIS (1993-2003), Current Contents (1995-2003), DERWENT DRUGFILE (1983-2003), Cochrane Database of Systematic Reviews. Bibliographies of existing meta-analyses and systematic reviews of nifedipine as a tocolytic.
Methods of study selection: Forty-five studies evaluating the effectiveness of nifedipine were identified.
Data extraction: Each study was assessed for 40 method-specific and topic-specific items of quality in duplicate using piloted data extraction forms. Disagreements between assessors were settled by consensus/arbitration.
Data synthesis: Very few of the studies complied with adequacy criteria of quality for either method-specific or topic-specific items. There was no improvement in quality over time. The quality of method-specific items was significantly poorer when compared with topic-specific items of quality overall (P<0.0001) and in the domains of selection bias (P<0.0001) and performance bias (P<0.0001).
Conclusion: Studies of the effectiveness of nifedipine as a tocolytic are of poorer quality with respect to method-specific items than topic-specific items. These deficiencies should be highlighted in meta-analyses or systematic reviews which measure efficacy and should influence the generation of guideline statements or recommendations for the use of nifedipine as a tocolytic. A large randomized trial fulfilling the quality items is necessary to assess the real efficacy of nifedipine in preterm labor.