Abstract
OBJECTIVE: When the volume or complexity of health information exceeds the capacity to process it, patients may misinterpret or ignore critical information. Numerical information is especially challenging to process for many patients, yet no empirical data shows whether numerical information influences how well they could process and recall it. METHOD: Using natural language processing tools, we estimated the amount of numerical and probability-related (quantitative) information that was provided in 112 paired urology-radiology clinical consultations with patients who had been recently diagnosed with prostate cancer. The primary outcome measured was patient knowledge about their prostate cancer outcomes assessed before and after the consultations. RESULTS: Patients with prostate cancer, Gleason score 6 or 7, and stage Time 1 or Time 2 participated in the study. Paired consultations included on average 11,086 words spoken. The relationship between quantitative information provided in consultations and patient knowledge about their cancer outcomes followed an inverted U-shape. There was a positive association between quantitative information and patient knowledge about their cancer outcomes. However, after the amount of quantitative information exceeded 4% (422 quantitative words) in paired consultations, the relationships between knowledge and the number of quantitative words became negative. Individual differences in education were not associated with observed relationships. CONCLUSION: Despite concerns about patients' capacity to process quantitative information, we found that patients' knowledge about cancer risks is positively associated with a certain amount of quantitative information. In the consultations, patients need to receive quantitative information that is well balanced with qualitative explanations. (PsycInfo Database Record (c) 2022 APA, all rights reserved).