Portal de Periódicos da CAPES

Effects of Communicating Social Comparison Information on Risk Perceptions for Colorectal Cancer

2006; Taylor & Francis; Volume: 11; Issue: 4 Linguagem: Inglês

10.1080/10810730600671870

1087-0415

Isaac M. Lipkus, William M. P. Klein,

Computational and Text Analysis Methods

Abstract People typically believe their health risks are lower than those of others (i.e., optimistic bias). We sought to increase perceptions of colorectal cancer (CRC) risk among adults aged 50–75 who were nonadherent to fecal occult screening (FOBT). 160 participants were randomized to receive information about the following: (1) general CRC risk factors (control), (2) general and tailored CRC risk factor feedback (absolute risk group), or (3) absolute CRC risk factor feedback plus CRC feedback as to how their total number of risk factors compared with that of others (absolute plus comparative risk group). Primary outcomes were perceived absolute and comparative risks, attitudinal ambivalence toward FOBT, and screening intentions; the secondary outcome was return of a completed FOBT. Participants who were told that they had more than the average number of risk factors believed their comparative CRC risk was higher than that of controls and of participants informed that they did not have more than the average number of risk factors. Perceived absolute risk did not vary by group. Participants who received social comparison risk factor feedback expressed greater intentions to screen via a FOBT than participants who received absolute risk feedback and controls; they also expressed less ambivalence about FOBT screening than controls. Although not statistically significant, participants informed they were at lower comparative risk had the highest proportion of completing an FOBT than any other group. These results suggest that providing social comparison CRC risk factor feedback can effectively reduce optimistic comparative risk perceptions. Contrary to findings of models of health behavior change, being informed that one does not have more than the average number of CRC risk factors, while resulting in lower evaluations of perceived comparative risk, did not result in higher ambivalence toward and lower intentions to screen using FOBT or the lowest rate of screening. Acknowledgments This study was funded by NIH grant CA063782. Notes Note. Due to rounding, percentages may not always sum to exactly 100%. Abs. = absolute. Comp. = comparative. 1One person randomized to the social comparison above-average group inadvertently received comparative risk information for the lower risk group. Data from this individual were excluded from the analyses. Further, one participant after going through the intervention realized she had an FOBT within a 2-year period. Data from this person were used for all the analyses except those involving intentions and ambivalence to screening and actual screening. Note. Numbers in parentheses represent the number of participants in the absolute risk only and absolute plus comparative risk condition who mentioned they received a specific risk factor as part of the risk factor feedback. For example, among the 38 participants who were informed that their age was a risk factor for CRC, all 38 mentioned that age may increase their CRC risk. Means with different lettered subscripts within rows differ by p < .05 based on Tukey's post-hoc test Abs. = absolute. Comp. = comparative. Note. Higher means represent greater perceived risk or worry. Means with different lettered subscripts within rows differ by p < .05 based on Tukey's post-hoc test. Abs. = absolute. Comp. = comparative. *p < .009. 2We examined whether there were changes between baseline and the lab in the weighing of CRC risk factors. With the exception of family history, all CRC risk factors were given a stronger association during the laboratory visit relative to baseline; hence, across experimental groups, the CRC risk factors became more salient. These data are available from the first author upon request. Note. Higher numbers represent greater ambivalence and intentions to screen. Means with different lettered subscripts within rows are statistically significantly different by Tukey's post-hoc test. Abs. = absolute. Comp. = comparative. 3We tested whether there was a pattern of greater completion of a FOBT in relation to participants being supplied with more risk information (i.e., controls < absolute risk only < absolute plus comparative risk). The test of a linear trend failed to achieve significance (Mantel-Haenszel X (1) 2=3.77, p=.052).