The impact of internet health information seeking on COVID-19 vaccination behavior in China – BMC Public Health – BMC Public Health

COVID-19 vaccination and background characteristics

Table1 presents summary statistics of these variables for all samples and separately by whether they were Internet health information seeking. In all samples, the COVID-19 vaccination rate was 73.1%, the proportion of males was 45.8%, the mean age was 51.3 years, 93% were Han Chinese, and most of the respondents marital status (75.1%) were married. In addition, the survey sample was in good health, with an overall sample self-reported health score of 3.47 points. The living area distribution of the overall sample shows that 46.1% of the respondents live in the eastern region, 33.3% live in the central region, and 20.6% live in the western region.

The Internet health information seeking group had a COVID-19 vaccination rate of 84.5%, while the non-Internet health information seeking group had a COVID-19 vaccination rate of 71.6%. This difference was statistically significant (p<0.01). Additionally, there were statistically significant variations in the two groups distributions of age, education level of primary school and below, education level of high school and above, household wealth status of poor, household wealth status of rich, occupational type, and health (p<0.01). The Internet health information seeking group is younger, more educated, wealthier regarding household assets, work in a state-owned enterprise, and healthier than the group not use Internet to seeking health information. There are no significant differences between the two groups regarding gender, ethnicity, marital status, and living area.

The correlation between the independent variables and the control variables and the result variables is analyzed, and the results are shown in Table2. As can be seen from the results table, there is a positive correlation between independent variable internet health information seeking and dependent variable (correlation coefficient r=0.093, P<0.01). Other relevant control variables are also correlated with the outcome variables.

Table3 shows the results that Internet health information seeking has a significant positive impact on COVID-19 vaccination. The average COVID-19 vaccination rate among individuals that non-Internet health information seeking was about 71.6% (Table2, row 13). The COVID-19 vaccination rate among individuals that Internet health information seeking was 4.1% points higher (Table2, row 1). The result is significant at the 1% level. Furthermore, individuals who are younger, belong to ethnic minorities, possess moderate household wealth, are married, maintain better health, and reside in the central and western regions of China exhibit elevated rates of COVID-19 vaccination. This correlation demonstrates statistical significance at the 1% level. Gender and middle-school education were also significantly associated with COVID-19 vaccination rates. Vaccination rates are higher among females and those with middle-school education. The result is significant at the 5% level.

The results of this study thus far indicate a correlation between Internet health information seeking and COVID-19 vaccination rates. To ensure the robustness of our models results, we conduct rigorous validation using the propensity score matching method. Due to the fact that Internet health information seeking is not a random event and can be influenced by a variety of factors, such as age, studies have shown that the younger generation is more likely to use the Internet than the older generation [25], and the level of education is also an important influence, and people with a high level of education are more likely to be proficient in using the Internet [20], there is the possibility of bias in the estimates. This issue, if left unaddressed, could result in significant deviations in the distribution of relevant eigenvalues. Therefore, implementing the propensity score matching method becomes essential for accurately estimating the impact of Internet health information seeking. Given that the population of individuals without Internet health information seeking is seven times larger than those with Internet health information seeking, we employ the one-to-seven matching method. In addition, both caliper matching and kernel matching are employed to ensure the reliability of our analyses.

Table4 shows the matching results. The ATT value derived using one-to-seven matching was 0.050 and significant at the 1% level, indicating that the percentage of vaccination was 5.0% higher with Internet health information seeking than without. The ATT values obtained using Caliper matching and Kernel matching were 0.045(P<0.01) and 0.052 (P<0.01), respectively, which were consistent with the results of one-to-seven matching, and this result verified the robustness of the regression analysis results.

Second, considering the impact of different occupations on COVID-19 vaccination behavior, we replaced our dependent variable with COVID-19 vaccination intention (1=Yes 0=No)(column 2). In the CGSS2021 questionnaire, the specific question about this variables was Which of the following statements best suits your situation?. We define intention to receiving vaccine if their answer is I wanted to be vaccinated.and the results were similar.

In this subsection, we delve into potential mechanisms to gain a deeper comprehension of how disseminating vaccine health information through the Internet influences individuals vaccination behavior. Table5 presents a tripartite breakdown of the pathways through which Internet health information seeking shapes vaccination patterns.

Drawing insights from existing literature, recognizing vaccination benefits is a pivotal driver of vaccine uptake. We embarked on an exploration of the link between Internet health information seeking and individuals perceptions regarding the merits of vaccination, encompassing notions such as the supremacy of vaccination benefits over drawbacks and the establishment of immunity through vaccination. As delineated in Table5, columns 1 and 2, the outcomes revealed a heightened likelihood among Internet health information seeking individuals to acknowledge the superior advantages of vaccination and the potential for immunization through this method. This accurate understanding consequently serves as a catalyst propelling Internet health information seeking individuals towards vaccine adoption.

Approaching the matter from an alternative angle, empirical studies have demonstrated the profound influence of attitudes towards information on individual health behaviors. This investigation explored the impact of Internet health information seeking on individuals recognition of the positive effect of Internet-based information on health behaviors. As depicted in Table5, column 3, the findings underscore a stronger propensity among Internet health information seeking individuals to acknowledge the constructive impact of online information on health behaviors. This acknowledgment engenders a greater acceptance of the veracity of Internet information, thereby fostering the execution of health-conscious behaviors. A visual representation of this impact mechanism is illustrated in Fig.1.

The mediating effects of cognitive factors

We extend our analysis to encompass the heterogeneous effects of Internet health information seeking by incorporating interaction terms. As evidenced in Table6, our investigation reveals limited indications of heterogeneous effects across various demographic and household attributes within our surveyed cohort. These attributes encompass gender, education level, household wealth, health status, and geographical location. Notably, we did identify specific subgroups in which Internet health information seeking seems to have a discernible impact on vaccination likelihood.

Upon closer examination, our findings indicate that older age, Han Chinese ethnicity, and unmarried status correlate with an augmented probability of vaccination due to Internet health information seeking, as indicated in Table6. This pattern can be ascribed to distinct factors that operate within these subgroups. For instance, older individuals might exhibit heightened receptivity to health-related information, rendering them more responsive to the effects of Internet health information seeking on vaccination rates, as detailed in Table6, row 2. Furthermore, the influence of Internet health information seeking on vaccination rates appears to be more pronounced among the unmarried demographic. This can potentially be attributed to the relatively higher availability of time and energy among unmarried individuals, in contrast to their married counterparts who often contend with familial responsibilities. As elaborated in Table6, row 7, those without marital commitments might allocate more resources to safeguarding their physical well-being, making them particularly responsive to the impact of Internet-based health information on their vaccination choices.

See the rest here:

The impact of internet health information seeking on COVID-19 vaccination behavior in China - BMC Public Health - BMC Public Health