The Role of Provider–Patient Communication and Trust in Online Sources in Internet Use for Health-Related Activities

Abstract

Provider–patient communication is an important factor influencing patients' satisfaction and health outcomes. This study draws upon the uses and gratification theory to examine how individuals' perception of communication with healthcare providers is associated with their Internet use for health-related activities. Using the data from the 2007 Health Information National Trends Survey (HINTS), we found that as individuals perceived their communication with providers to be less patient-centered, they were more likely to engage in various types of online health activities, such as using websites for healthy lifestyles, searching for healthcare providers, and seeking health information. Trust in online health information was also found to be a significant predictor of online health activities. The results of this study emphasized the important role of provider–patient communication in motivating individuals to turn to the Internet for health purposes.

Notes

¹The nonlinear pattern was caused by the fact that there were no significant associations between perceived patient centeredness of provider–patient communication and outcome variable among those who scored low on a scale of provider–patient communication, whereas the negative association was present with the majority of respondents who were at moderate to high on the scale. By recoding the variable into quartiles, those who showed no association between the variables were grouped together into one level, reducing nonlinear patterns into a tolerable level.

Note. Cell entries are either odds ratios and chi-squares (in binary logistic regression) or standardized betas and R² (in OLS multiple regression). Ns = 4,202 (information search); 4,179 (support groups); 4,177 (websites on lifestyles); 4,174 (health care provider search); 4,179 (total online activities). †p < .06, *p < .05, **p < .01, ***p < .001. df = 11: In addition to the 10 predictors listed above, we also entered a variable (named, SAMPFLAG which was available in the HINTS data) to control for a significant difference in population estimates of outcomes by different modes of sampling (i.e., mail and RDD). The odds ratios of the variable SAMPFLAG in logistic regression were 1.73** (information search), .79 (support groups), 1.01 (Websites on lifestyles), and .93 (health care provider search). The standardized beta of this variable in OLS regression was .06.