Discordance of patient and provider perceptions of the meaning of verbal estimates of perioperative risk

Abstract

PurposeTo determine whether the use of semiquantitative, verbal (non-numerical) language to express levels of perioperative risk is a possible cause of poor communication between practitioners and patients. MethodsSurvey forms were distributed to patients who had just completed an outpatient preoperative clinic visit at a large academic medical center. Surveys were distributed from October 2015 through March 2016. Data collected included demographic characteristics, literacy and numeracy characteristics, patient recollection of level of risk discussion, and the effect of the discussion on patient desire to pursue surgery. Patients were asked to define, using percentages, commonly used verbal semiquantitative terms for levels of risk (low, medium/intermediate, and high). Results212 patients completed surveys. Overall, 84% of patients noted that the actual level of risk was discussed in quantitative or semiquantitative terms; of these patients, 66% reported risk was described using words only. When asked to numerically define low, medium/intermediate, and high risk, patient estimates were markedly discordant, being typically higher, from the usual definitions for these terms as published in the medical literature and as used by medical practitioners. More than 55% of patients surveyed indicated that high risk would correspond to a >50% chance of adverse event, whereas clinicians generally define high risk as >5% chance of adverse event. ConclusionSemiquantitative terms used to define levels of perioperative risk may have very different meanings to patients and practitioners. To facilitate true informed consent, such terms should be avoided and more precise, numerical, quantitative estimates of risk should be used instead.