Study characteristics impacted the pragmatism of randomized controlled trial published in nursing: a meta-epidemiological study.

ABSTRACT

OBJECTIVES: The objective of this study was to examine the impact of study characteristics on the score of the pragmatism/explanatory continuum of randomized controlled trials (RCTs) published in nursing journals using the PRagmatic Explanatory Continuum Indicator Summary (PRECIS)-2 tool. STUDY DESIGN AND SETTING: RCTs concerning five themes of nursing care indexed in the PubMed and CINAHL databases published from 2002 to 2005 and 2012 to 2015 were selected by title/abstract. A sample of 400 was randomly selected and evaluated with the PRECIS-2 tool and reading grid. RESULTS: The median PRECIS score was 32 of a possible 45 [28; 36] corresponding to a medium level of pragmatism. Studies with "medication" as an intervention had a more explanatory PRECIS score than studies with other intervention types (P = 0.015). Studies with "placebo" and "no usual care" as comparators had a more explanatory PRECIS score (P = 0.0027). The pragmatism/explanatory level was unaffected by impact factor (P = 0.42), h-index of the first and last author (P = 0.27 and P = 0.25, respectively), funding (P = 0.32), blinding (P = 0.41), sample size (P = 0.22), and time (P = 0.11). CONCLUSION: This study highlights the pragmatism/explanatory level of nursing RCTs, the impact of the field of the article, and the comparator type on the pragmatism of these studies. Further studies are needed to confirm the astonishing result that blinding resulted in no significant difference in the PRECIS score.