Smallpox Vaccine Recommendations: Could Trust be a Shot in the Arm?

Download data and study materials from OSF

Principal investigator:

John Parmer

University of Georgia



Sample size: 599

Field period: 4/25/2011-8/11/2011

The successful management of public health threats such as bioterrorism, avian influenza, and H1N1 demand the rapid mobilization of the public to cooperate with government recommendations. The extent of the public's cooperation with governmental directives in a crisis situation may depend heavily on the public's trust in the government and its risk management institutions to effectively manage the crisis. The proposed research study is a post-test only randomized experiment to explore how the trust and confidence one has in an institution responsible for managing a smallpox outbreak influences their willingness to cooperate with government recommendations. A deeper understanding of how trust and confidence lead to cooperation will be useful in guiding the government's communication to the public during a bioterrorism event and will offer practical guidance to risk communicators and public health professionals on how to maximize cooperation with government recommendations during a smallpox outbreak. In addition, this study will explore the relationship between smallpox risk perceptions and social trust of the risk managing institution. If, as some argue, trust is strongly related to risk perception, trust may prove to be a key to the development of more effective risk communication techniques during a bioterrorism attack.

1) Do individuals discriminate between two separate dimensions of institutional trust – one focused on values and the other on performance?

2) Are there significant differences in cooperation with smallpox vaccine recommendations based on experimental groups (trust vs. confidence)?

3) Does trust in CDC influence smallpox risk perceptions?

4) Does knowledge of smallpox influence the relationship between trust and risk perceptions?

Experimental Manipulations
The proposed research study is a post-test only randomized experiment to explore the role of trust and confidence in influencing the public's response to a smallpox outbreak. Participants will begin by reading a short preface statement describing the threat of a smallpox outbreak from a bioterrorist attack and an introduction to a fictional smallpox outbreak scenario that will follow. Following the preface statement, and prior to exposure to experimental conditions, participants will respond to a single item measuring their current knowledge of the CDC.

Study participants will then read short messages describing a smallpox outbreak scenario, the role of CDC in such an event, and vaccine information and recommendations for members of a community believed to have been exposed to the virus. Message content will be structured to contain information that enhances one of the two pathways to cooperation as described by the TCC Model. One message condition will aim to enhance trust by highlighting the shared values dimension of trust. A second message condition will aim to enhance confidence in the CDC to effectively manage the smallpox outbreak by highlighting past performance during recent public health crises (i.e. anthrax and SARS) as well as approaches that were successfully employed in the past to control smallpox outbreaks. After reading one of two messages, participants will then complete surveys measuring shared values with CDC, past performance of CDC, confidence in CDC to manage smallpox, smallpox risk perceptions, knowledge of smallpox, and behavioral intention to cooperate with government vaccine recommendations. Covariates will include general confidence, general trust, and demographics. If 250 respondents are randomized to each condition, this will provide at least 86% power at the two-sided alpha=0.05 significance level to detect a difference of 0.14 between the proportions cooperating with smallpox vaccine recommendations (e.g., 57% versus 43%). With 200 respondents per group, the power is 87% to detect a difference of 0.16 between cooperation proportions.

Social trust in the CDC's handling of the response to a smallpox outbreak will be measured by a scale comprised of six items based on the components of trust identified by previous research. Rather than using scales predetermined by the researcher, Meredith et al. (2007) conducted qualitative research and an inductive analysis strategy to assess the role trust played in participant's reaction to escalating stages of a bioterrorism scenario and identified key components of trust. Applying these dimensions in the context of the H1N1 outbreak, Quinn et al. (2009) developed questions that asked for the level of trust respondents feel regarding the government's openness, honesty, commitment, caring and concern, and competence in addressing H1N1; the extent to which they believe they government's actions in response to H1N1 are in their personal best interest; and how much they believe the government will protect them from H1N1. These items were found to be highly correlated and an exploratory factor analysis indicated that all items loaded on one factor (Chronbach's alpha=0.91). Value similarity will be assessed using a measurement scale similar to one used by Earle and Cvetovich (1997, 1999) and adapted to a smallpox context. Respondents will be asked: With respect to managing a smallpox outbreak, I feel the CDC: (a) shares similar values as me, (b) shares similar opinions as me, (c) thinks in a similar way as me, (d) takes similar actions as I would, and (e) shares similar goals as me.

Risk perception measures are adapted from previous research on influenza vaccination and risk perceptions (Weinstein et al., 2007) and include measures of risk magnitude, perceived severity, and concern. Confidence in the risk management institution (CDC) will be measured by a 3-item scale measuring institutional confidence as judgments of competence and expertise. Items measuring general trust and general confidence were selected from the literature with acceptable internal consistencies, 0.67 and 0.54 respectively (Siegrist, Gutscher, & Earle, 2005). Finally, a measure of participant's general knowledge of the CDC will be administered before exposure to message conditions and included as a covariate.

Summary of Results
Respondents in both groups reported very similar willingness to accept smallpox vaccine recommendations. It is unclear whether message conditions were unable to produce a sufficiently strong enough boost in their relative trust dimension, or if each trust dimension is equally important in influencing decisions about cooperation with official recommendations. Perhaps individual's trust and confidence assessments of DHS are static and resistant to change through messaging.
Two of the six constructs outlined by Trust, Confidence and Cooperation model were found to be significant positive predictors of cooperation with official smallpox vaccine recommendations. General trust, a personality trait rather than a measure of trust in a specific entity, was found to be an important predictor of cooperation. Similar values were found to be a significant predictor of cooperation as well, whereas shared values were not. However, the confidence intervals for the odds ratios of shared values (0.921-1.238) and similar values (1.013-1.420) overlap and therefore these findings remain tentative. That general trust and similar values, and not shared values (tentatively), predict cooperation is an interesting finding. The distinction between similar values and shared values is unclear in the literature. The results of this study suggest that these two constructs are indeed distinct from one another, but that only similar values may influence cooperation during a smallpox outbreak.

Past performance and confidence factors did not appear as significant predictors in this study. It may be that respondents have insufficient experience or knowledge of how DHS has managed past crises for these constructs to influence their willingness to cooperate. Consistent with the literature, risk perceptions were strongly predictive of cooperation with protective measures. Risk perception is a central construct in many theories of health behavior that suggest the risk a person perceives from a health threat will be associated with their willingness to engage in behaviors to protect them from that threat. In the case of vaccine acceptance, it would be interesting to examine the influence of risk perceptions of the vaccine, in addition to those associated with the disease. Future studies on trust and vaccine behavior should measure risk perceptions of both the hazard and vaccine to assess the influence of vaccine risk perceptions on cooperation, independent of risk perceptions related to the hazard.