Shame in Physician–Patient Interactions: Patient Perspectives

Abstract

This study estimated the extent to which shame is elicited in physician–patient interactions and examined the emotional and behavioral reactions of patients to such interactions. A large adult sample (N = 915) reported on their shame-provoking experiences in interactions with physicians through an anonymous survey. Half of all respondents (n = 456) recalled one or more interactions with a physician that left them feeling ashamed, with significantly more women (59%) than men (39%) reporting this. Forty-five percent of those reporting such experiences stated they terminated treatment with, avoided, or lied to their physician to avoid experiencing further shame. However, 33% believed the shame-provoking interaction provoked useful behavioral changes, and 46% were, on balance, grateful to the physician. Women, relative to men, reported that the interaction led to more negative emotional and behavioral consequences. The type of health issue involved in the interaction also was associated with differential outcomes. Thus, it appears that inducing shame in medical contexts is widespread and may well have both positive and negative effects.

Notes

^a Range = 18–70 years.

^b United States only.

Note. Categories were not mutually exclusive, hence entries sum to more than 100%.

∗p < .05. ∗∗p < .01.

Note. Chi-square analyses compare reactions from those who stated that the shaming event involved the listed topic to those who stated the event did not involve that topic for each type of positive and negative reaction. Within each topic, ∗p < .05, †p < .10.

¹Given that the shaming incident could involve more than one issue, we performed a second set of analyses that excluded participants who had reported that the shaming interaction involved multiple issues. These analyses produced a similar pattern of results as those reported for the whole sample. However, some analyses did not reach significance due to substantially reduced sample sizes. Three results that previously were not statistically significant became significant when participants with multiple issues were excluded: Smoking was related to a decrease in the likelihood of being angry when the incident occurred and improving from the incident. Substance abuse was related to an increase in the likelihood of improving.