Individuals' Willingness to Talk to Their Doctors About Clinical Trial Enrollment

Abstract

This study investigates whether perceived fairness of doctor–patient interactions relates to individuals' willingness to communicate with their doctors about clinical trial enrollment. It also explores how willingness to talk, the perceived fairness of interactions, and trust in doctors relate to intentions to participate in a future clinical trial. Results from a random digit dial (RDD) telephone survey of U.S. adults (N = 500) measured respondents' willingness to talk to their doctors about clinical trials and intentions to participate in future trials. Perceived fairness of interactions and trust in doctors were associated with willingness to talk about clinical trials. A negative relationship emerged between perceived fairness of interactions and intentions to participate when willingness to talk was introduced into the equation. This relationship suggested that when respondents were more willing to talk to their doctors and perceived these discussions as fair, they were also less likely to express intentions to enroll in future trials. In turn, perceiving these interactions as less fair was related to greater intention to enroll. Fairness of interactions and trust in doctors were less relevant to respondents who were less willing to talk to their doctors; however, these respondents also were more likely to express intentions to enroll in future clinical trials.

Notes

¹There are many reasons why a clinical trial may not be an appropriate course of action for an individual. Neither a fair interaction nor trust in doctors should automatically lead an individual to make the decision to participate. At minimum, the goal should be that increased, balanced discussions (i.e., fair interactions) would lead to more informed judgments on the patient or doctor end.

²Individuals were not preselected by known presence or lack of medical conditions; rather, the RDD method ensured that every telephone household within the United States had an equal chance of being included in the survey, which provided results unlikely to vary more than 4.4 percentage points from answers that would be obtained if all U.S. residents were interviewed.

³The survey was an “omnibus,” meaning that it included a variety of topics including attitudes toward privacy, deception, and television, among other things. It also meant that there were space constraints as to the number of questions we could add to the survey.

⁴According to the American Association for Public Opinion Research's (AAPOR's) standard definitions (2008), response rate is the number of complete interviews with reporting units divided by the number of eligible reporting units in the sample, while cooperation rate is the proportion of all cases interviewed of all eligible units ever contacted.

Note. Items were measured using 5-point, Likert-type scales, from 1 = strongly disagree to 5 = strongly agree.

⁵Hierarchical OLS allowed us to enter variables in a series of blocks with the results at each step indicating the relative influence of the variables on the dependent variable while controlling for variables entered in previous steps (Cohen, Cohen, West, & Aiken, 2003).

⁶Research suggests that interactional justice and trust are conceptually distinct (Bies, 2005); however, some of the underlying notions these two concepts measure could overlap. To ensure that the assumptions for the hierarchical OLS analysis were not violated, we performed multicollinearity tests. All of the tolerance values were much above zero.

Note. Standardized regression coefficients (betas) except where indicated. All variables entered together. Listwise deletion of missing values. ∗p < .10; ∗∗p < .05; ∗∗∗p < .01.

Note. Standardized regression coefficients (betas) except where indicated. All variables entered together. Listwise deletion of missing values. ∗p < .10; ∗∗p < .05; ∗∗∗p < .01.