Extending the purview of risk perception attitude (RPA) framework to understand health insurance-related information seeking as a long-term self-protective behavior

Abstract

Objective: The purpose of this study is to examine the roles of risk perceptions and efficacy beliefs play in predicting emerging adults' health insurance information seeking behavior based on the risk perception attitude (RPA) framework. In addition, the current study tests a role of worry to elucidate an underlying mechanism of their health insurance information seeking behavior. Participants: A total of 694 participants was recruited from a large public university in Southwest, USA. Methods: Participants completed an online survey questionnaire. Results: Efficacy beliefs influenced health insurance-related information seeking behavior. An intriguing finding is that there were significant differences in worry across the attitudinal groups posited in the RPA framework, which is consistent with findings in prior literature. Conclusions: Information about health insurance helps individuals make good decisions for their health in a long-term perspective, which is the primary interest of the RPA framework and the current study. The findings would provide resources for designing effective communication materials to help human beings improve their well-being and prepare for their well-ending. This communication effort would contribute to ensuring public health safety, reducing health disparities, and enhancing a society's well-being.

Keywords:

• Health insurance
• information seeking
• medical decisions
• risk perception attitude (RPA) framework
• worry

Acknowledgments

Thanks Rajiv Rimal for his wish for my best luck and the same hope with this article.

Conflict of interest disclosure

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed no financial support for the research, authorship, and/or publication of this article.

Notes

1 Part of this data set was used in a publication earlier.⁶

2 As Levene's Test for Equality of Variances was violated (p<.05), the variable ratio was calculated. The value was less than 2.

3 In Mplus, the correlations between the latent independent (predictor) variables are default. When calculating the degrees of freedom (df), it should be considered.

4 The model modification indices suggested to add a direct path from perceived efficacy to health insurance information seeking behavior to improve the model efficacy.

5 Considering prior literature⁴⁰ and the roles of efficacy beliefs in prior literature in the context of health-related issues, it may not be theoretically implausible to posit the direct association of perceived efficacy with health insurance inforamtion seeking behavior.

6 In the initial model, the direct path from perceived efficacy to health insurance information seeking behavior was significant ($\beta$=.21, p <.001) whereas the direct path from perceived risk to health insurance information seeking behavior was not significant ($\beta$=-.03, p=.81). Consistent with the logistic regression results, health insurance information seeking behavior was positively associated with perceived efficacy. The paths from perceived risk ($\beta$=.46, p<.001) and perceived efficacy ($\beta$=-.11, p<.001) to worry were both significant - greater risk perceptions and less perceived efficacy were associated with greater worry. The path between worry and health insurance information seeking behavior was also significantly positive ($\beta$=.19, p=.002). Greater worry was associated with greater health insurance information seeking behavior. There were significant indirect effects from perceived risk ($\beta$=.09, SE=.03, p=.002) and perceived efficacy (b=-.02, SE=.01, p=.02) to health insurance information seeking behavior in addition to the significant direct effect of perceived efficacy on health insurance information seeking behavior (b=.21, SE=.04, p<.001). That is to say, worry mediated the effect of perceived risk (i.e., the indirect-only mediation) and partially mediated the effect of perceived efficacy on health insurance information seeking behavior. Among covariates, age and status significantly influenced worry (age: $\beta$=.10, p=.01; status: $\beta$=-.12, p=.002) and information seeking behavior (age: $\beta$=.26, p<.001; status: $\beta$=-.15, p=.007). Health condition only significantly influenced worry ($\beta$=-.13, p<.001). Sex did not have a significant effect on neither feeling of worry nor health insurance information seeking.

7 For the readers who are interested to know whether there were differences between models with covariates and those without covariates, additional analyses were conducted. Although minor variations were observed between the models with covariates and those without covariates, the overall results were not changed.

8 The author rephrased the sentence according to the very minor suggestion but strong encouragement the anonymous reviewer #1 made in the final revision.