Background: The success of evidence based practice and participation with healthcare decisions depends inherently on the clear and effective communication of health messages. Objectives: To evaluate the effect of negative versus positive framing of the same health information (e.g. 20% chance of dying vs. 80% chance of surviving) on understanding, perception, persuasiveness, and behavior of health professionals and consumers. Methods: We followed Cochrane methodology. We electronically searched Medline, Embase, PsycLIT and CENTRAL. We included parallel and cross-over studies with health professionals and consumers evaluating one of two types of framing: attribute framing (positive or negative encoding of a specific attribute of a single item) and goal framing (positive or negative framing of the consequences of performing or not performing an act). We standardized the outcome effects using standardized mean difference (SMD). We conducted pre-planned subgroup analyses based on the type of message (screening, prevention, and treatment).
Results: We identified 35 eligible studies reporting 51 comparisons. With attribute framing, participants understood the message better when it was framed negatively (SMD −0.58 [95% CI −0.94 to −0.22]; moderate effect size). Although positively framed messages may have been better perceived (SMD 0.36 [95% CI −0.13 to 0.85]; small effect size), there was little or no difference in persuasiveness (SMD 0.07 [95% CI −0.23 to 0.37]) and behavior (SMD 0.09 (95% CI −0.14 to 0.31)). With goal framing, loss messages were perceived as more effective for screening topics (SMD −0.30 (95% CI −0.49 to −0.10); small effect size) and may have been more persuasive for treatment topics (SMD −0.50 (95% CI −1.04 to 0.04); moderate effect size). There was little or no difference in behavior (SMD −0.06 [95% CI −0.15 0.03]). Conclusions: Contrary to commonly held beliefs, the available evidence suggests that framing may have little if any effect on behavior.