“Social impacts of the continuous usage of digital healthcare service: A case of South Korea”

As untact communication is promoted in the era of the COVID-19 pandemic, special attention is paid to remote medical examination and customized healthcare trends. General digital healthcare services among social community members positively af- fect individuals’ healthcare and reduce medical social services’ burden, contributing to the development of society. Accordingly, it is necessary to induce healthcare behaviors through the continuous usage of digital healthcare services among social community members and to examine significant social impact factors in this regard. This study empirically analyzes the impact of three social impact factors – social capital, social support, and social value – on the continuous usage of digital healthcare service with healthcare behaviors and e-health literacy as media. To this end, a survey was conducted among 363 individuals who had used digital healthcare services in Korea, and the statistical data were analyzed. Social capital and social value were found to affect healthcare behaviors, e-health literacy, and continuous usage intentions, but social support did not. Based on this result, it was confirmed that the factors regarded by digital healthcare service users as necessary were the values and perceptions shared in society and the group, information and active communication rather than direct public support.


INTRODUCTION
As advanced technologies such as AI (artificial intelligence), IoT (internet of things), and big data are incredibly converged with medical technology, various IT digital devices and medical services such as customized medical service and precision health service emerge in the medical service market (Saheb & Lzadi, 2019;Badri et al., 2018). Particularly in this age of the COVID-19 pandemic, mobile and easily usable personal medical devices and app-based services increase, along with the increasing demands for digital healthcare services based on the untact service. And this situation is accelerating around the global interest about healthcare in all generations (Papa et al., 2020;Shapiro et al., 2016).
Particularly, digital healthcare service is affected by factors such as personal will and activeness. It is related with users to actively participate in content recommended through healthcare service to enhance such a healthcare service's effectiveness. Personal health promotion and prevention activities with such services finally are essential to maintain and manage a healthy society (Kalem & Turhan, 2015). Sun and Medaglia (2019) pointed out that social community members' health-promoting behaviors positively affect the community's public health. On the other hand, the local community's active support is essential to induce its members' good health management behaviors. And the voluntary and continuous healthcare behavior of community members can have socio-economic impacts (Wald, 2020;Agrawal & Hyrkas, 2020).
In the correlation between social environments and personal health management, health concerns of society members are not merely a matter of an individual's illness but are recognized as related to the entire social community, such as viral infection and environmental contamination (Gereffi, 2020;Harris & Guten, 1989). In addition, as digital media proliferate, the internet, smart-phones, and various other SNS channels are changing the social base. Such environmental changes and the active information exchange in the broadened social network affect individuals' healthcare and medical service consumption behaviors (Mu et al., 2018).
However, there is a limitation of many of the recent studies on digital healthcare because they focus merely on the technical acceptability of digital devices and verification of efficacy among older people. As digital healthcare service and system elements spread worldwide rapidly based on big data and AI (artificial intelligence), the relation between community members' use of digital healthcare service and the society needs to be examined more closely (Lupton, 2013).
Accordingly, this study identified the social impact factors such as social capital, social support, and social value, and developed a research model that is designed to examine the impact of the continuous usage of digital healthcare with healthcare behaviors and e-health literacy as media. By clarifying this relation, this study seeks to verify the importance of their social roles in inducing continuous digital healthcare service usage among community members and suggests specific social activities that can improve healthcare behaviors. First, social capital may be defined as tangible or intangible capital that an individual or group accumulates (Jang et al., 2011). Social capital may be formed through information exchange in social interactions and make possible coordination and cooperation for group members to achieve goals and mutual profits through such factors as trust, norm, and network (Coleman, 1988). Social capital affects information sharing, smooth communication, and community activity with others using digital healthcare services for continuous disease prevention.

LITERATURE REVIEW AND HYPOTHESES
Second, social support means various types of resources that an individual receives in social relations; affection, acknowledgment, information, and material, and includes support from family, relatives, friends, supervisors, or companions within the organization (Cohen, 1983). Such social support positively affects physical and mental health, happiness, and life satisfaction (House & Kahn, 1985). Consumers intending to maintain health with digital healthcare services are affected by perceived threats. A perceived threat means the extent that a patient is affected mentally and physically about his/her disease (Visconti & Morea, 2020). In this regard, social support messages from acquaintances and medical centers reduce perceived threats (Hermes et al., 2020) and positively affect one's conviction about his/her behaviors.
Third, social value contributes to the public good and the development of a community in every area, including society, economy, environment, and culture. Such value is granted by the community and shared with others. Thus, the entire society aims to realize desirable and rightfully promoted values (Balliet et al., 2009). The social value may be divided into economic value, cultural value, and social value (Klamer, 2004). Social value includes the standard complex value elements such as a change in members' self-esteem, psychological stability, community spirit development, and social consensus.

Healthcare behavior and e-health literacy
Healthcare behaviors are acts of people who are assured that a disease can be prevented. Such behaviors include regular exercise, sound living habits, dietary habits, and sufficient rest (Lafferey 1990). Likewise, a series of healthcare behaviors to maintain or recover a healthy state are defined as self-care behaviors. Since disease management requires the patient's sense of commitment and active participation, it is pointed out that self-care needs to continue for disease management, particularly among patients with chronic illness (Asghar et al., 2017).
According to Wood et al. (2014), social learning and perception about immunization lead to immunization's dynamic behavior. As highly educated people and city residents are highly motivated to take immunity and preventive measures, it is highly probable that rather than an individual's value, the social community's value and information sharing through communication affected such dispositions. As suggested by Green et al. (2020), with their health belief model, social support from others in the community is vital to make its members find value from disease prevention behaviors and take the initiative.
This tendency may be observed among digital healthcare service users. In many cases, products and services for healthcare behaviors are purchased not merely as a reflection of one's personal needs but also on learning from society about disease prevention and acquaintances' recommendations ).
E-health literacy is also an extended concept of 'e-health.' It means the ability to pursue, explore, understand, and evaluate health information through the internet or mobile devices, and also the ability to apply and deliver the knowledge acquired in order to solve health problems. As medical information media and access systems are digitized today, an individual's e-health literacy affects the intent of pursuing health information and practicing health-promoting behaviors (Mackert et al., 2014).
Several previous studies also report that health information available on the Internet affects behavioral changes concerning disease prevention. Pursuing information on health through various digital media has a positive and significant impact on personal needs for healthcare (Mathews et al., 2019; Antonio & Antonella, 2020).
Health literacy means the ability to understand information on health maintenance and improvement. Using this ability makes it possible to consider cognitive and social skills based on motivation (Nutbeam et al., 2018). Cognitive skills are related to an individual's subjective perception of health. Individuals whose subjective perception of health is relatively high are more likely to maintain a good health state and healthy behavior than individuals whose subjective perception of health is low (Carroll et al., 2015). Social skills are one of the health policy issues that are considered necessary in addition to a patient's health-related decision-making, safety pursuit, medical cost-saving, and life quality improvement (Nuti et  H2. Social support for digital healthcare services will have a positive (+) impact on healthcare behaviors.
H3. Social value for digital healthcare services will have a positive (+) impact on healthcare behaviors.
H4. Social capital for digital healthcare services will have a positive (+) impact on e-health literacy.
H5. Social support for digital healthcare services will have a positive (+) impact on e-health literacy.
H6. Social value for digital healthcare services will have a positive (+) impact on e-health literacy.
H7. Healthcare behaviors for digital healthcare services will have a positive (+) impact on the intention of continuous service usage.
H8. E-health literacy for digital healthcare services will have a positive (+) impact on the intention of continuous service usage.

Research model and hypotheses development
This study analyzes the effects of social impact factors on the continuous usage of digital healthcare services. Social impact factors are 'social capital,' 'social support,' and 'social value.' Additionally, this study examines the effects of social impact factors on the continuous usage of healthcare behaviors and e-health literacy as media. Regarding this correlation, the study model, as shown in Figure 1, was designed with the presented hypotheses that are based on previous studies.

Demographic information of the data
This study was conducted among digital healthcare service users in Korea. 51.6% of them were male, and 48.4% were female. As for the age groups, 19.8% were in their 20s, 23.9% in their 30s, 25.8% in their 40s, and 30.5% in their 50s, respectively. As for the residential area, the most considerable portion (55.5%) was living in Seoul. 14.6% were living in Gyeongsang-do, 10.2% in Chungcheong-do, and 6.6.% in Gyeonggi-do. As for occupations, the most substantial portion (51.3%) was office workers. 14.6% were professionals, 9.6% students, and 8.0% self-employed persons. As for academic backgrounds, 72.2% were college graduates, and 16.5% completed a graduate school course. Most subjects were highly educated. As for the period of digital healthcare service use, the largest portion (44%) answered '1-3 years,' 32.4% '3-5 years,' and 11.8% '5-10 years.' Most subjects used digital healthcare service use for at least 1 year. As for reasons of use, 67% answered 'healthcare,' and 26.7% answered 'life rhythm management.' Only 4.7% answered 'disease management.' Most subjects were using such a service for healthcare.

Reliability and validity analysis results
As shown in Table 2, it turned out that the factor load was all between 0.604 and 0.880 (0.5 or higher), which was sufficient. As to the internal reliability, the composite reliability level was between 0.798 and 0.872, which was significant. Since the value of t was at least 6.5, it was statistically significant. The average sampling variance (AVE) value was between 0.693 and 0.863, and Cronbach α was between 0.776 and 0.821. Hence, the proper level of composite validity was secured. The correlation coefficient was analyzed to ensure discriminant validity (see Table 3).
As the measurement model's fitness was analyzed, χ²(df ) was 337,940, and χ 2 /degree of freedom was 2,759. The value of Goodness-of-Fit-Index (GFI) was 0.907, that of Adjusted Goodness-of-Fit-Index (AGFI) 0.871, that of Normal Fit Index (NFI) 0.882, and that of Root Mean Square Error of Approximation (RMSEA) 0.036. Thus, the measurement model fitness values were statistically significant.

Structural model analysis results
As shown in Table 4, as the structural model's suitability was analyzed, χ 2 (p) was 290.431(0.000), and χ 2 /degree of freedom was 1.632. The GFI and NFI were 0.938 and 0.882, respectively, and the latter was smaller than 0.9. However, the Root Mean Square Residual (RMR) was 0.025, the AGFI 0.866, and the RMSEA 0.039, respectively. The suitability factors were satisfactory in general, and thus the model suitability was viewed as verified. The CFI value, which is not affected by the sample but represents the model's explanatory power, was 0.911. The value of TLI, which indicates the explanatory power of the structural model, was 0.937. Thus, it was viewed that the basic model was entirely appropriate.
As hypotheses were examined through the structural equation model's path analysis, two out of the eight hypotheses were rejected (see Table 4). It turned out that among the social factors affecting the use of digital healthcare services, social capital had a positive (+) effect on healthcare behaviors as much as 6.236 (p < 0.001), and on e-health literacy as much as 4.176 (p < 0.01). Social value also had a positive (+) effect on healthcare behaviors as much as 6.043 (p<0.001), and on e-health literacy as much as 6.170 (p < 0.001). Thus, this hypothesis was also accepted. However, social support failed to affect healthcare behaviors or e-health literacy, and thus this hypothesis was rejected. Healthcare behaviors had a positive (+) effect on continuous usage by as much as 7.482 (p < 0.001), as well as on e-health literacy by as much as 9.192 (p < 0.001). Thus, this hypothesis was accepted.

DISCUSSION
This study analyzes the correlation between the social impact factors of digital healthcare service and healthcare behaviors and e-health literacy.
It also empirically analyzes their impacts on the continuous service usage. The following are analysis results: First, social capital, a type of intangible capital formed through information exchange in social interactions, affected the continuous usage of digital healthcare service most significantly. It also affected healthcare behaviors and e-health literacy. This result indicates that communications with acquaintances and social networks can significantly affect the continuous digital healthcare  Second, it turned out that social support did not affect either healthcare behaviors or e-health literacy, or continuous digital healthcare service usage. This result may suggest that digital healthcare service users' behaviors to select and maintain such services are often understood as consuming service products according to personal needs rather than behaviors of receiving public goods based on government's support, medical centers, or acquaintances. As pointed out in previous studies, social support may be adequate for society members who need financial and systematic support, such as the elderly and the disabled.
However, disease prevention and digital healthcare services based on a customized management system do not significantly affect society members.
Third, it turned out that social value positively affects digital healthcare service users, specifically regarding their continuous usage, healthcare behaviors, and e-health literacy. This situation means that society members are well-aware that it can positively affect society as they maintain their health properly. In addition, this result suggests that healthcare plays a vital role in terms of social development and citizenship. As emphasized by Hibbard et al. (2003), this result corresponds to previous studies' findings that point out that citizenship, social support, and shared information directly affect citizens' disease preventive behaviors. Healthcare behaviors of society members, mainly through digital healthcare services, can be promoted further when there is a high social consensus and value sharing.

CONCLUSION
Based on the findings stated above, this study shows that the social impact factors such as social capital, social support, and social value are important to improve the people's healthcare behaviors and e-health literacy. When it comes to digital healthcare, organizations operating in the digital healthcare service sector will need to consider the social impacts and environmental issues of healthcare, as well as technical and industrial matters, to ensure effective service improvement in this post-COVID, hyper-technology era.
Furthermore, healthcare behaviors of modern people are closely related to sharing of social values and promoting social capital through active communication in the social network, as evidenced by the results of this study. In other words, both digital healthcare enterprises and governmental institutions that take the lead of healthcare policies need to go beyond the boundary of medical policies focusing on traditional social support and develop services and policies that will raise awareness of the importance of dynamic healthcare behaviors and information sharing.
This study has found that the user's healthcare behaviors and e-health literacy ability on digital healthcare services directly affect continuous usage. It should be explained that the conviction about a society member's self-behaviors based on digital healthcare positively affects continuous disease prevention behaviors in their social commitment. In this respect, this study suggests that digital healthcare services need to lead the community members' participation in existing disease preventive social promotion policies.
Despite the implications stated above, this study is limited to digital healthcare service users only in Korea. Since digital healthcare services are distributed globally and there may be differences between countries depending on their social environments, future research needs to collect samples from various countries in order to expand the research scope and conduct a comparative analysis of data.