Disparities in attention to HIV-prevention information

Abstract

Compared to European-Americans, African-Americans have greater probability of becoming infected with HIV, as well as worse outcomes when they become infected. Therefore, adequate health communications should ensure that they capture the attention of African-Americans and do not perpetuate disadvantages relative to European-Americans. The objective of this report was to examine if racial disparities in attention to health information parallel racial disparities in health outcomes. Participants were clients of a public health clinic (Study 1 n = 64; Study 2 n = 55). Unobtrusive observation in a public health waiting room, message reading times, and response-time on a modified flanker task were used to examine attention to HIV- and flu-information across racial groups. In Study 1, participants were observed for the duration of their time in a public health clinic waiting room (average duration: 31 min). In Study 2, participants completed tasks in a private room at the public health clinic (average duration: 21 min). Across all attention measures, results suggest an interaction between race and information type on attention to health information. In particular, African-Americans differentially attended to information as a function of information type, with decreased attention to HIV- versus flu-information. In contrast, European-Americans attended equally to both HIV- and flu-information. As such, disparities in attention yielded less access to certain health information for African- than European-Americans in a health setting. The identified disparities in attention are particularly problematic because they disadvantage African-Americans at a time of great effort to correct racial disparities. Modifying the framing of health information in ways that ensure attention by all racial groups may be a strategy to increase attention, and thereby reduce disparities in health outcomes. Future research should find solutions that increase attentional access to health communications for all groups.

Keywords:

• health disparities
• attention to health information
• HIV-prevention
• African-Americans
• perceived threat

Acknowledgements

We thank Dov Cohen, Sonya Dal Cin, Monica Fabiani, Jesse Preston, and the members of the Health, Attitudes, and Influence Lab (HAILab) at the University of Michigan for detailed comments on an earlier version of this article. Earlier versions of some of the analyses described here were presented as part of a doctoral dissertation by Allison Earl, conducted under the direction of D. Albarracín.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Because the study was designed to examine health disparities between African-Americans and European-Americans, participants of other races were excluded from analyses (n = 15). Furthermore, eight participants (four African-Americans and four European-Americans) were excluded, because no information about duration was reported. As such, although data were recorded for 87 participants, the total sample included is n = 64 (30 African-Americans and 34 European-Americans). Although the clinic typically serves low-income and low socio-economic status clientele, these data were not directly collected from participants.

2. Although coders were privy to the content of the video being played, they had no knowledge of the specific hypotheses regarding attention as a function of information type, or the predicted interaction between race and information type.

3. Pilot testing with 41 clients confirmed that coder's observed race and participant's self-reported race are highly related (χ²= 41.00, p < .001).

4. In addition to alertness and duration, one related concern might be whether different racial groups varied in the degree to which they were accompanied by friends or family versus being alone, although we do not have a direct measure of whether or not participants brought others with them, there are two proxy measures that may shed light on this issue. First, coders recorded the total number of other people in the waiting room. However, neither information type nor race were related to this variable (all ps > .19), and the number of other people in the waiting room did not moderate the interaction between information type and race on attention to the video (p = .18). Second, coders also noted if participants talked to other participants in the waiting room at the same time. However, there was no difference in this measure across racial groups (p > .50). Taken together, these results suggest that other participants in the waiting room did not reliably influence the observed interaction between information type and race.

5. Participants who were paid $40 participated in a larger study that included the measures of interest, but also included additional measures that are not included in this report. All analyses presented below were also run included study group as a factor. However, neither the main effect of study group nor any higher-order interactions with the factors of interest were significant, so the study group factor was dropped from the analyses presented. Two African-American participants (one man and one woman) were excluded from analyses, because technical difficulties prevented recording of some data. Thus, all analyses presented are with a final sample size of 55 (29 African-Americans, 26 European-Americans).

6. The study initially included trials with a non-health control condition (“BOX”). However, the letter “x” differentially interfered with the flanker task across trial types, with more interference in the right-congruent and left-incongruent trials (e.g., BOX >><>> BOX and BOX >>>>> BOX versus BOX <<<<< BOX and BOX <<><< BOX). As such, the BOX condition was dropped from analyses, because the response-times are uninterpretable in these conditions.

Additional information

Funding

This study was supported in part by research grants F31 MH086324 (Earl) and K02 MH075616 (Albarracín) from the National Institute of Mental Health, as well as funding from the Loan Repayment Program of the National Institute on Minority Health and Health Disparities (Earl).