A randomised on-line survey exploring how health condition labels affect behavioural intentions.

• Main Authors: Rae Thomas, Mark T Spence, Rajat Roy, Elaine Beller
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2020-01-01
• Series: PLoS ONE
• Online Access: https://doi.org/10.1371/journal.pone.0240985

<h4>Objectives</h4>We examined the effect of 'labels' versus 'descriptions' across four asymptomatic health conditions: pre-diabetes, pre-hypertension, mild hyperlipidaemia, and chronic kidney disease stage 3A, on participants' intentions to pursue further tests. There were four secondary objectives: 1) assessing confidence and satisfaction in their intention to test further; 2) revealing psychological drivers affecting intentions; 3) exploring whether intentions, confidence and satisfaction differ by label vs. description and health condition; and 4) producing a perceptual map of illnesses by label condition.<h4>Methods</h4>Practitioner validated health-related scenarios were used. Two variants of each condition were developed. Participants were recruited through Qualtrics from Australia, Ireland and Canada and randomly assigned two 'labelled' or two 'descriptive' scenarios.<h4>Results</h4>There was no significant difference in intentions to test between label and description conditions (95% CI -0.76 to 0.33 points, p = 0.4). Confidence and satisfaction were both positively associated with intentions: regression coefficient (β) for confidence β = 0.58 points (95% CI 0.49 to 0.68, p < .001) and for satisfaction 0.67 points (95% CI 0.57 to 0.77, p < .001). Predisposition to seek healthcare (β = 0.72; 95% CI 0.47 to 0.98), attributing illness to bad luck (β = -0.16 points; 95% CI -0.3 to -0.02), and concern about the health condition (β = 0.51; 95% CI 0.38 to 0.65) also significantly predicted intentions.<h4>Conclusions</h4>Unlike studies investigating symptomatic illnesses, the disease label effect on behavioural intentions was not supported suggesting that reducing demand for medical services for borderline cases cannot be achieved by labelling. The average intention to test score was higher in this sample than previous symptomatic health-related studies and there was a positive relationship between increased intentions and confidence/satisfaction in one's decision. Exploratory insights suggested perceptions of the four labelled asymptomatic illnesses all shifted toward greater levels of dread and concern compared to their respective description condition.<h4>Trial registration</h4>ACTRN12618000392268.