Search

Introduction: This study examined the discordance between hemodialysis patients' reports and their physicians' estimates of dietary restriction adherence and related factors in Japan. Methods: In a cross‐sectional survey of 6644 outpatients, physicians who estimated higher and lower adherence than their patients' self‐reported were categorized as overestimation and underestimation in terms of discordance, respectively. Possible factors included clinical indicators, patient characteristics related to negative stereotypes, and health beliefs related to statistical discrimination. Results: The concordance rate was 0.069 based on the weighted kappa coefficient. The coefficients of acceptable serum potassium, prevalence of diabetes, and self‐efficacy on overestimates were 0.663, −0.126, and −0.132, respectively. The coefficients of these factors on underestimates were −0.589, 0.338, and 0.145, respectively. All these coefficients were significant. Conclusions: The discordance may be high and is related to physicians' clinical data reliance, negative stereotypes about patient characteristics, and a lack of understanding of patients' health beliefs. [ABSTRACT FROM AUTHOR]

Nearly all framing studies to date presuppose unbiased estimates. If an expert says that “programme A will save 200 people”, it is tacitly assumed that this prediction is correct. In real life this is rarely the case. In the present study people were asked to evaluate such claims that eventually turned out to be incorrect. Participants in five experiments were asked to rate how correct, and how true, are predictions and reports that either overstate or understate the facts. Overall, understatements were considered more accurate than overstatements, and pessimistic statements were better than overly optimistic ones. Thus predictions of an outcome that turns out better than expected should preferably have been presented in a positive frame (e.g., money saved), whereas outcomes that turn out worse than expected will appear less “wrong” if predictions were presented in a negative frame (e.g., money lost). Parallel studies were performed with wrong predictions and incorrect factual claims (lies). In all studies, correctness ratings were affected by the way the statements were framed. [ABSTRACT FROM AUTHOR]