This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.
Massive, easily accessible online health information empowers users to cope with health problems better. Most patients search for relevant online health information before seeing a doctor to alleviate information asymmetry. However, the mechanism of how online health information affects health empowerment is still unclear.
To study how online health information processing affects health empowerment.
We conducted a cross-sectional questionnaire study that included 343 samples from participants who had searched online health information before the consultation. Respondents' perceptions of online information cues, benefits, health literacy, and health empowerment were assessed.
Perceived argument quality and perceived source credibility have significant and positive effects on perceived information benefits, but only perceived argument quality has a significant effect on perceived decision-making benefits. Two types of perceived benefits, in turn, affect health empowerment. The effects of perceived argument quality on perceived informational benefits and perceived decision-making benefits on health empowerment are significantly stronger for the high health literacy group than the low health literacy group (t269=7.156,
In the context of online health information, perceived information benefits and perceived decision-making benefits are the antecedents of health empowerment, which in turn will be affected by perceived argument quality and perceived source credibility. Health literacy plays a moderating role in the relationship of some variables. To maximize health empowerment, online health information providers should strengthen information quality and provide differentiated information services based on users' health literacy.
Health empowerment is a cornerstone of a patient-centered approach to healthcare. Empowerment allows patients to take the initiative in making decisions about their own health care and quality of life, rather than passively complying with decisions made by others [
The rise of e-health services has brought new opportunities for promoting health empowerment. Various forms of electronic health services (eg, health information portals, online health communities, consultation platforms, etc) provide the public with abundant and easily accessible health information. Patients can obtain information about the symptoms of the disease, conventional treatment methods, and the treatment experience of others. With that health information, patients can become informed before doctors' visits and participate in health decision-making during the consultation process to enhance their sense of control. And an increasing number of people now obtain health information online. The China Internet Network Information Center pointed out that more than 276 million users in China utilize internet medical services, accounting for 29.4% of all internet users [
Research on health empowerment in the context of eHealth services has become an important research stream. Some scholars explored the logic or dimensions of empowerment in the context of eHealth services [
Previous research provided us with valuable knowledge for understanding health empowerment. Undoubtedly, obtaining health information from online resources to reduce information asymmetry is an indispensable part of patient empowerment [
Overall, we assume that online health information can promote health empowerment during the consultation process, which is the result of the interaction information factors and the health literacy of the recipient. As a popular health resource, online health information can support patients with the ability to participate in the consultation process. Therefore, it is necessary to explore the process by which patients analyze online information and identify the mechanisms by which they contribute to health empowerment. To address this question, based on the elaboration likelihood model (ELM), we conceptualized perceived argument quality, perceived source credibility, and health literacy into online health information processing scenarios and explored their impact on health benefits and health empowerment.
Empowerment theory has been explored by a rich body of research in social work, mainly as it relates to self-esteem, self-worth, self-confidence, and wellness [
Although health empowerment has been one of the core concepts in health promotion research, there is still no unified definition. Past researches have mainly defined health empowerment from three perspectives: process, emergent state, and active behavior [
Due to different definitions and research contexts, previous studies have used multidimensional or single-dimensional assessments of health empowerment. Ouschan et al [
This study considers health empowerment from the perspective of the state of being empowered. Accessible online health information eases the information asymmetry between doctors and patients to a certain extent. The patient is no longer in a completely passive position but can actively participate in health activities. This undoubtedly allows patients to advocate for themselves and increase their sense of control. We define health empowerment as one's belief that they have a significant influence over health outcomes, including the ability to address personal health issues and feel in control over factors that can impact health outcomes.
The elaboration likelihood model (ELM) explains how two types of information persuasion paths affect individuals’ attitude changes, perceptions, and behaviors [
In addition, the ELM generally approaches elaboration likelihood from two influencing dimensions: ability and motivation [
By providing online health information and educational opportunities, information and communication technology (ICT) can empower users to deal with health issues and engage in their own health outcomes [
Online support groups enable patients to learn more about themselves, enhance their social well-being, and thus promote healthy empowerment [
Since involvement in health consultation and decision-making processes is an important element of health empowerment [
Perceived argument quality is reflected in an individual's subjective evaluation of the reasoning that forms the core of presented information. The presentation of information can be strong and convincing or weak and specious. Strong arguments mean that the presented information is reasonable and convincing to the recipient, while weak arguments are doubtful or contradictory [
Perceived source credibility is the evaluation of information from the reliability of information sources. It can be perceived to be credible, acceptable, or untrustworthy by information recipients [
In the health information literature, perceived source credibility is an important topic that relates to individual health outcomes and decision-making behavior. Young people's trust in health information is affected by perceived source credibility. The higher credibility of the information source, the more likely the users are to participate in the information activity [
According to the ELM, the ability to process information can affect the level of elaboration likelihood [
In the process of health information analysis, individuals with adequate health literacy have a greater ability to analyze the arguments presented as part of health information. For individuals whose attitudes or perceptions change based on central route processing, the information influence occurs under conditions of high-end elaboration (ie, content-oriented reasoning). In contrast, for individuals with limited health literacy, information processing is more about evaluating factors other than content, so peripheral cues play a more critical role in processing. These expectations led us to state the following hypotheses:
We assume that the perceived informational benefits and decision-making benefits all contribute to empowerment. However, the two kinds of benefits have different requirements for patients’ health literacy. Informational benefits are the prerequisite for decision-making benefits. Sufficient information can improve the quality of decision-making and reduce risks [
A summary of the conceptual research model is depicted in
Research model.
To test our hypotheses, we administered a self-reported questionnaire to collect data. The questionnaire consisted of two parts: one was designed to investigate the demographic characteristics of the participants, and the other focused on the measurement of the constructs. The research model contained a total of 6 constructs. The measurement scales were developed by drawing on prior literature, and some items were fine-tuned according to the background of this study. We adapted the work of Hur et al [
Since the respondents are Chinese, we need to translate all the items from English into Chinese. All measures were back-translated by another translator who did not know the background of the study to ensure the accuracy of the translation. The two English versions were compared, and potential semantic discrepancies were examined to ensure that the Chinese scales reflected the meaning of all measures accurately. Then 10 postgraduates with experience seeking online health information were invited to participate in a pretest of the scales. Based on their feedback, any ambiguous expressions were amended. The measured constructs and their sources are shown in
We collected data through a questionnaire service website [
Among the valid questionnaires, 47.2% (162/343) were from males, and 52.8% (181/343) were from females. Further, 85.1% (292/343) of respondents’ ages ranged from 18-35 years, implying that the majority of online health information users tend to be younger. In terms of education, 88.9% (305/343) of the respondents had a college degree or above. The majority (234/343, 68.2%) of the respondents had a monthly disposable income in the range of 3000-8999 Chinese Yuan (approximately US $469-1406). As to their occupations, business employees accounted for the largest proportion of participants, reaching 47.5% (163/343). The most popular way to access information was through a health information portal, accounting for 61.8% (212/343) of the respondents, followed by a health consulting platform, accounting for 22.7% (78/343). On average, 63.3% (217/343) of the respondents used online health information sources between 1 and 3 times weekly, and 22.4% (77/343) of the subjects used these sources 4 to 5 times weekly. The specific demographic information of the target samples is shown in
Demographic information of respondents (N=343).
Characteristics | Participants, n (%) | ||
|
|||
|
Male | 162(47.2) | |
|
Female | 181(52.8) | |
|
|||
|
18-25 | 89(25.9) | |
|
26-35 | 203(59.2) | |
|
36-45 | 45(13.2) | |
|
46 and above | 6(1.7) | |
|
|||
|
High school or below | 38(11.1) | |
|
Associate degree | 101(29.4) | |
|
College degree | 176(51.3) | |
|
Master degree or above | 28(8.2) | |
|
|||
|
Under 3000 | 83(24.2) | |
|
3000—5999 | 148(43.1) | |
|
6000—8999 | 86(25.1) | |
|
9000—11,999 | 16(4.7) | |
|
12,000 and above | 10(2.9) | |
|
|||
|
Student | 41(12) | |
|
Business employees | 163(47.5) | |
|
Government and public institutions | 39(11.4) | |
|
Self-employed persons | 45(13.1) | |
|
Other | 55(16) | |
|
|||
|
1-3 | 217(63.3) | |
|
4-5 | 77(22.4) | |
|
6-7 | 25(7.3) | |
|
7 and above | 24(7) | |
|
|||
|
Health information portal | 212(61.8) | |
|
Online patient community | 36(10.5) | |
|
Health consultation platform | 78(22.7) | |
|
Blog or video | 8(2.3) | |
|
Other | 9(2.6) |
aA currency exchange rate of ¥1 = US $0.16 is applicable.
We used variance-based partial least squares structural equation modeling (PLS-SEM) for data analysis. We chose the PLS-SEM method for the following reasons. First, the PLS-SEM method does not require multivariate normal distribution data [
In this study, we used the confirmatory factor analysis process to test the measurement model. As shown in
Furthermore, as shown in
Results of confirmatory factor analysis.
Construct and item | Loading | Cronbach’s α | Composite reliability | AVEa | |
|
0.764 | 0.850 | 0.587 | ||
|
PAQ1 | 0.750 | |||
|
PAQ2 | 0.827 | |||
|
PAQ3 | 0.784 | |||
|
PAQ4 | 0.695 | |||
|
0.802 | 0.870 | 0.627 | ||
|
PSC1 | 0.785 | |||
|
PSC2 | 0.781 | |||
|
PSC3 | 0.793 | |||
|
PSC4 | 0.808 | |||
|
0.756 | 0.845 | 0.578 | ||
|
PIB1 | 0.787 | |||
|
PIB2 | 0.747 | |||
|
PIB3 | 0.708 | |||
|
PIB4 | 0.797 | |||
|
0.717 | 0.823 | 0.539 | ||
|
PDB1 | 0.678 | |||
|
PDB2 | 0.762 | |||
|
PDB3 | 0.724 | |||
|
PDB4 | 0.769 | |||
|
0.786 | 0.854 | 0.539 | ||
|
EM1 | 0.736 | |||
|
EM2 | 0.775 | |||
|
EM3 | 0.702 | |||
|
EM4 | 0.736 | |||
|
EM5 | 0.720 | |||
|
0.895 | 0.916 | 0.578 | ||
|
HL1 | 0.773 | |||
|
HL2 | 0.833 | |||
|
HL3 | 0.787 | |||
|
HL4 | 0.712 | |||
|
HL5 | 0.715 | |||
|
HL6 | 0.758 | |||
|
HL7 | 0.750 | |||
|
HL8 | 0.747 |
aAVE: average variance extracted.
Means, SD, and correlation matrix.
Variable | Mean | SD | PAQa | PSCb | PIBc | PDBd | EMe | HLf |
PAQ | 3.910 | 0.680 |
|
—g | — | — | — | — |
PSC | 3.625 | 0.725 | 0.624 |
|
— | — | — | — |
PIB | 3.918 | 0.722 | 0.634 | 0.553 |
|
— | — | — |
PDB | 3.812 | 0.656 | 0.514 | 0.409 | 0.632 |
|
— | — |
EM | 3.676 | 0.651 | 0.424 | 0.410 | 0.454 | 0.402 |
|
— |
HL | 3.433 | 0.799 | 0.569 | 0.550 | 0.593 | 0.553 | 0.603 |
|
aPAQ: perceived argument quality.
bPSC: perceived source credibility.
cPIB: perceived informational benefits.
dPDB: perceived decision-making benefits.
eEM: health empowerment.
fHL: health literacy.
g—: The correlation matrix is symmetrical; therefore, only the lower-left corner is displayed.
As our data were collected from single respondents, common method variance (CMV) may threaten the validity of the results. To test such bias, first, we used Harman’s single-factor test to assess the 6 constructs in the search model. The results showed that the variance explained by the first factor is 35.4%, which does not exceed 50% [
In this paper, we used SmartPLS 3.0 (SmartPLS GmbH) to test the research model. The path coefficients and significance levels of main effects are shown in
PLS Analysis of main effects. PLS: partial least squares.
A multigroup comparison method developed by Keil et al [
As shown in
The results of moderating effects.
Paths | High health literacy (n=174) | Low health literacy (n=169) | t341 values comparing the two groups | ||
|
Coefficient | SE | Coefficient | SE | |
PAQa→PIB | 0.464 | 0.099 | 0.395 | 0.078 | 7.156 |
PAQ→PDB | 0.358 | 0.107 | 0.341 | 0.099 | 1.526 |
PSCb→ PIB | 0.181 | 0.076 | 0.27 | 0.081 | –10.497 |
PSC→PDB | 0.067 | 0.122 | 0.151 | 0.102 | –6.908 |
PIBc→EMd | 0.190 | 0.094 | 0.266 | 0.126 | –6.344 |
PDBe→EM | 0.292 | 0.089 | –0.003 | 0.141 | 23.240 |
aPAQ: perceived argument quality.
bPSC: perceived source credibility.
cPIB: perceived informational benefits.
dEM: health empowerment.
ePDB: perceived decision-making benefits.
The bootstrap method was used for the analysis of mediating effects [
The results of the mediation effect test.
Indirect path | 95%CI | Direct path | 95%CI | Result |
PAQa→PIBb→EMc | 0.007 to 0.173 | PAQ→EM | –0.011 to 0.256 | Full |
PAQ→PDBd→EM | 0.005 to 0.129 | PAQ→EM | –0.011 to 0.256 | Full |
PSCe→PIB→EM | 0.004 to 0.097 | PSC→EM | 0.058 to 0.285 | Partial |
PSC→PDB→EM | –0.002 to 0.061 | PSC→EM | 0.058 to 0.285 | None |
aPAQ: perceived argument quality.
bPIB: perceived informational benefits.
cEM: health empowerment.
dPDB: perceived decision-making benefits.
ePSC: perceived source credibility.
Based on the ELM model, this paper examined the influencing factors of health empowerment in the context of processing online health information. Our empirical research provided the following results:
First, perceived informational and decision-making benefits are important predictors of users’ health empowerment. Perceived informational benefits accrue when individuals become more informed by browsing online health information. This input allows them to have a more objective understanding of their illnesses and health situations, thereby reducing negative emotions, such as anxiety and panic. Perceived decision-making benefits refer to growth in terms of knowledge and skills gained through seeking online health information. This improvement in decision-making capacity allows individuals to participate more effectively in the consultation process and make reasonable suggestions for treatment. The gain of these two kinds of benefits makes users feel empowered.
Second, the results confirm that perceived argument quality, as involved with the central route, has a positive effect on perceived informational and decision-making benefits, while perceived source credibility, which relies on the peripheral route, only has a significant impact on perceived informational benefits. When getting health information from online channels, the strength of the arguments and credibility of the sources reflect the quality of information. They are the guarantee that users can benefit from the information they receive. Both high-quality arguments and credible sources can enhance an individual's acceptance and approval of the information, thus promoting the perceived informational benefits. Individuals need knowledge and skills to make informed health decisions. The online health information presented with high-quality arguments can provide recipients with health knowledge and treatment experience so they can make informed decisions in medical consultations.
Our results show that the credibility of sources has no significant influence on perceived decision-making benefits. One possible explanation is that the basis for supporting individuals’ participation in decision-making may come more from the information itself, which develops individuals’ knowledge or skills. However, the credibility of information resources as a peripheral cue does not improve knowledge or skill levels and thus cannot support the individual’s participation in the decision-making process.
Third, we also found that the effects of the central route and the peripheral route are different in low and high health literacy groups. For individuals with high health literacy, the effect of central route processing (perceived argument quality) on perceived informational benefits is stronger than the influence of processing using peripheral cues (perceived source credibility). Individuals with high health literacy are more likely to exert cognitive effort when assessing the arguments provided by online information. For these individuals, information source credibility is used as a secondary consideration and has a weaker effect on perceived informational benefits. The opposite is true for individuals with low health literacy. For low health literacy groups, their judgments of online health information rely more on the source credibility.
Finally, the study demonstrates that the effects of the two perceived benefits on health empowerment are different between groups with high and low health literacy. The effect of perceived informational benefits on health empowerment is greater in the low health literacy group than in the high health literacy group. The effect of perceived decision-making benefits on health empowerment is significant in the high health literacy group but not in the low-health literacy group. The results show that there is a higher demand for health empowerment for individuals with high health literacy. Merely information benefits are not enough to promote health empowerment but to further obtain perceived decision-making benefits. For low health literacy groups, health empowerment does not derive from participating in decision-making but from getting enough information to reduce information asymmetry.
There are two theoretical contributions of this study. First, the study provided a profound understanding of the mechanism of health information processing on health empowerment. Previous studies highlighted the convenience and positive health outcomes that can be derived from information technology and online health information [
Second, this study explained the relationship between health literacy and health empowerment from a new perspective that is different from previous literature, which always explores the direct relationship between health literacy and health empowerment [
Based on our theoretical analysis and empirical results, the following practical implications should be noted. First, encouraging patients to search for high-quality online health information is an effective way to promote their health empowerment. The information provider can strengthen information quality management in terms of perceived argument quality and perceived source credibility. Accordingly, to prevent the dissemination of misleading content, online health information providers should establish reasonable evaluation and testing mechanisms. They should strictly scrutinize every piece of health information provided to consumers and ensure that information content is complete, rigorous, sound, and scientific.
Second, information providers should also consider the health literacy of recipients while providing health information. To improve the effectiveness of promoting health empowerment, online health information providers should establish a health literacy assessment mechanism to provide targeted information services to individuals with different health literacy. For individuals with a high level of health literacy, it is an effective strategy to cultivate users' health decision-making ability to promote health empowerment, and providers should highlight the scientific nature of the information. For those with a lower level of health literacy, making them more informed is an effective way to promote empowerment, and providers should highlight the professionalism and reliability of the sources of information.
Although this paper draws some conclusions that cannot be ignored, there are still some shortcomings that should be addressed in the future. First, this study did not consider the impact of the type of online health information service model used by consumers to gather information, such as an online health consultation website or a medical information portal. The unique characteristics of different online health information services may impact health empowerment. Second, our study was based on a static model and cross-sectional data. The processes that affect the promotion of individual health empowerment are likely to be dynamic, so longitudinal research is necessary. Third, we did not involve the measurement of the respondent’s disease and pathology, which may affect a person’s use of online health information. Finally, medical consultation is a process of interaction between patients and doctors. This research only focuses on patient factors. Future research should consider the impact of doctor-related factors (such as empathy and patient-centered communication) on health empowerment.
In this paper, we explored the effect of the central route and peripheral route of online health information on users’ health empowerment. We also considered the moderating role of health literacy in both routes. To test the hypothesis, PLS-SEM was used to analyze the data, and the empirical results supported most of the hypothesis. The findings further confirmed the important role of electronic information technology in promoting health empowerment. In the context of online health information, we must pay more attention to information quality and the interaction effect between individuals’ health literacy and information processing cues. Research results provide practical guidance for health information providers to better serve and maximize individuals’ benefits and empowerment. This study also pointed out the differences in promoting health empowerment besides health literacy. And more research in the future is needed to focus on individualized differences in the promotion of health empowerment.
Measurement Scales.
Common Method Bias Analysis.
average variance extracted
common method variance
elaboration likelihood model
information and communication technology
partial least squares structural equation modeling
This work was funded in part by the National Natural Science Foundation of China (grants 71771219 and 72071213). In addition, we appreciate the academic committee in the “Mobile Health” Ministry of Education-China Mobile Joint Laboratory for reviewing the project proposal and providing ethical approval.
None declared.