Clinician Factors Rather Than Patient Factors Affect Discussion of Treatment Options.

Abstract

BACKGROUND\nShared decision-making aims to combine what matters most to a patient with clinician expertise to develop a personalized health strategy. It is a dialogue between patient and clinician in which preferences are expressed, misconceptions reoriented, and available options are considered. To improve patient involvement, it would help to know more about specific barriers and facilitators of patient-clinician communication. Health literacy, the ability to obtain, process, and understand health information, may affect patient participation in decision-making. If the patient is quiet, deferential, and asks few questions, the clinician may assume a more paternalistic style. A patient with greater agency and engagement could be the catalyst for shared decisions.\n\n\nQUESTIONS/PURPOSES\nWe assessed (1) whether effective clinician communication and effort is related to patient health literacy, and (2) if there are other factors associated with effective clinician communication and effort.\n\n\nMETHODS\nWe combined a prospective, cross-sectional cohort of 86 audio-recorded visits of adult patients seeking specialist hand care for a new problem at an urban community hospital in the Netherlands with a cohort of 72 audio-recorded hand surgery visits from a tertiary hospital in the United States collected for a prior study. The American cohort represents a secondary use of data from a set of patients from a separate study using audio-recorded visits and administering similar questionnaires that assessed different endpoints. In both cohorts, adult patients seeking specialist hand care for a new problem were screened. In total, 165 patients were initially screened, of which 96% (158) participated. Eight percent (13) of visits were excluded since the final diagnosis remained unclear, 8% (12) since it was not the first consultation for the current problem, 5% (8) in which only one treatment option was available, and < 1% (1) since there was a language barrier. A total of 123 patients were analyzed, 68 from the Netherlands and 55 from the United States. The Newest Vital Sign (NVS) health literacy test, validated in both English and Dutch, measures the ability to use health information and is based on a nutrition label from an ice cream container. It was used to assess patient health literacy on a scale ranging from 0 (low) to 6 (high). The 5-item Observing Patient Involvement (OPTION5) instrument is commonly used to assess the quality of patient-clinician discussion of options. Scores may be influenced by clinician effort to involve patients in decision making-as well as patient engagement and agency. Each item is scored from 0 (no effort) to 4 (maximum effort), with a total maximum score of 20. Two independent raters reached agreement (kappa value 0.8; strong agreement), after which all recordings were scored by one investigator. Visit duration and patient questions were assessed using the audio recordings. Patients had a median (interquartile range) age of 54 (38 to 66) years, 50% were men, 89% were white, 66% had a nontraumatic diagnosis, median (IRQ) years of education was 16 (12 to 18) years, and median (IQR) health literacy score was 5 (2 to 6). Median (IQR) visit duration was 9 (7 to 12) minutes. Cohorts did not differ in important ways. The number of visits per clinician ranged from 14 to 29, and the mean overall communication effectiveness and effort score for the visits was low (8.5 ± 4.2 points of 20 points). A multivariate linear regression model was used to assess factors associated with communication effectiveness and effort.\n\n\nRESULTS\nThere was no correlation between health literacy and clinician communication effectiveness and effort (r = 0.087 [95% CI -0.09 to 0.26]; p = 0.34), nor was there a difference in means (SD) when categorizing health literacy as inadequate (7.8 ± 3.8 points) and adequate (8.9 ± 4.5 points; mean difference 1.0 [95% CI-2.6 to 0.54]; p = 0.20). After controlling for potential confounding variables such as gender, patient questions, and health literacy, we found that longer visit duration (per 1 minute increase: r2 = 0.31 [95% CI -0.14 to 0.48]; p < 0.001), clinician 3 (compared with clinician 1: OR 33 [95% CI 4.8 to 229]; p < 0.001) and clinician 5 (compared with clinician 1: OR 11 [95% CI 1.5 to 80]; p = < 0.02) were independently associated with more effective communication and effort, whereas clinician 6 was associated with less effective communication and effort (compared with clinician 1: OR 0.08 [95% CI 0.01 to 0.75]; p = 0.03). Clinicians communication strategies (the clinician variable on its own) accounted for 29% of the variation in communication effectiveness and effort, longer visit duration accounted for 11%, and the full model accounted for 47% of the variation (p < 0.001).\n\n\nCONCLUSION\nThe finding that the overall low mean communication effectiveness and effort differed between clinicians and was not influenced by patient factors including health literacy suggests clinicians may benefit from training that moves them away from a teaching or lecturing style where patients receive rote directives regarding their health. Clinicians can learn to adapt their communication to specific patient values and needs using a guiding rather than directing communication style (motivational interviewing).Level of Evidence Level II, prognostic study.