When 'unlikely' is not the same as 'likely not': Effects of likelihood framing and numerical format on patients' responses to normal diagnostic test results.

Abstract

Communicating the results of diagnostic testing is a central task in clinical practice. Even when results are normal, patients may remain concerned, particularly depending on how risk information is presented. The present study investigated how likelihood framing and numerical format influence patients' understanding, emotional responses, and intended health care use following normal diagnostic test results. In an experimental analogue study (N = 564), participants were presented with a symptom vignette suggestive of a possible brain tumour. They then viewed a videotaped consultation in which a physician communicated a normal CT result. Likelihood presentation was manipulated in a 2 × 2 factorial design, varying framing (low likelihood of disease vs. high likelihood of no disease) and numerical format (frequency vs percentage), with an additional control condition not mentioning any likelihoods. Positive framing (emphasising the likelihood of not having a disease) led to significantly lower perceived disease likelihood compared to negative framing (p = .007, η²_p =.016, 90% CI [.002,.042]). Positive framing also reduced emotional concern and decreased intentions to seek additional consultations or diagnostic tests. A significant framing by format interaction indicated that frequencies enhanced the effects of positive framing but attenuated those of negative framing. For some outcomes, the control condition was most effective, though. Subtle differences in likelihood framing and numerical format significantly influence how patients interpret and respond to normal diagnostic test results. Thoughtful communication of low-risk information may represent a simple, cost-free, and scalable approach to support patient reassurance and reduce unnecessary follow-up.