Is compassion fatigue a self-fulfilling prophecy? An experimental study of manipulated expectations

Abstract

Compassion is vital in healthcare and has value to patients, clinicians, and the health system. Despite its centrality, however, compassion is not always present in healthcare encounters. Although many factors contribute to compassion in health settings, a disproportionate research focus has been on a possible role for compassion fatigue. However, while the study of compassion fatigue has highlighted a serious issue in healthcare workforces, this work has also contributed to a narrative suggesting that compassion is a finite and depletable resource. The presence of this narrative creates the possibility that negative expectations create a nocebo-like effect in which compassion fatigue becomes a self-fulfilling prophecy. In testing this possibility, the current study experimentally tests whether perceptions regarding compassion can be influenced in both positive and negative directions as well as whether changes in perception lead to differences in self-reported compassionate responding. One hundred and fifty-three physicians and medical trainees aged 20 - 73 years were recruited via online adverting and professional networks to participate in a pre-registered (AsPredicted.org pre-registration number: 105438), anonymous online experimental study. Demographic, professional, and dispositional data (including perceptions of compassion) were collected before participants were randomised to watch a 5-minute video describing compassion as (a) positive or (b) negative or watching (c) a control video before completing further ratings of their perceptions of compassion. Participants then read four patient vignettes designed to systematically vary in terms of patient presentation and responsibility for condition, before making a series of ratings about each. As expected, a factorial mixed model analysis of variance (ANOVA) showed that perceptions of compassion were influenced by the video manipulation. Specifically, those randomised to the positive perception condition reported perceiving compassion as more satisfying and pleasurable and less draining and exhausting, while those randomised to the negative condition reported perceiving compassion as more draining and exhausting. Changes were not seen in the control group. However, this successful change to perceptions did not translate to differences in compassionate responding by experimental group. Consistent with other work there were large effects associated with patient factors and clear evidence for the influence of social desirability. Interpretatively, it seems clear that perceptions regarding compassion are highly malleable although we have some way to go to understand how (and whether) perceptions result in changes to the experience of compassion with respect to patients. Results are discussed in the context of the complex interactions among factors influencing compassion, the small ‘dose’ of the video manipulation, and the primacy of the obligation to care among medical professionals. Notwithstanding some complexities, the finding that perceptions are malleable provides ‘proof of principle’ that narratives surrounding compassion may be influencing perceptions and could be contributing to a self-fulfilling prophecy.