Virtual consultation and virtual visit are two primary types of services delivered by internet hospitals in China. An internet hospital is, to a large extent, an equivalent of direct-to-consumer telemedicine. China’s first internet hospital officially opened in 2014 [1,2]; however, since then the internet hospital industry witnessed an initial development stage with ups and downs. Virtual consultation was the primary service type of internet hospitals from 2014 to 2018. The difference between virtual consultation and virtual visit primarily resulted from the evolution of China’s internet hospital regulation schemes. The State Council of China issued the guideline on Internet Plus Healthcare of 2018 [3], and the National Health Commission accordingly issued specific regulation schemes for online medical diagnosis and treatment as well as internet hospitals in 2018 [4]. The issuing of regulation schemes have brought the rapid development of internet hospitals [5], especially for the service type of online medical diagnosis and treatment. Online medical diagnosis and treatment (hereinafter referred to as virtual visit) is very similar to a virtual visit of direct-to-consumer telemedicine/telehealth in many other countries like the United States because physicians are able to diagnose, treat, and prescribe for some common conditions and chronic diseases for non–first-visit patients. However, physicians are not allowed to diagnose, treat, and prescribe in the virtual consultation service.

In addition to the difference in scope of service, virtual visit is also different from consultation in health insurance coverage. Previous research has suggested that health insurance coverage is a significant factor that affects the use of virtual visits [6]. Following the guideline on Internet Plus Healthcare, in August 2019, the National Healthcare Security Administration of China, as the single payer, has accordingly issued specific guidelines aiming to expand insurance coverage for virtual visit services of internet hospitals [7]. However, consultation services of internet hospitals have not been covered by health insurance. Obviously, health insurance coverage would incentivize both providers and consumers to use virtual visits tremendously. However, there is a lack of studies that specifically examine consumers’ intention to use virtual visits in China.

Internet hospitals vary in sponsors. There are primarily brick-and-mortar hospital-sponsored and enterprise-sponsored internet hospitals in China’s direct-to-consumer telemedicine market [5]. Recently, a very small number of local governments have taken initiatives to set up local government internet hospital platforms by pooling local public health care resources. Hence, enterprises, hospitals and governments are 3 major sponsor types of internet hospitals. Han et al [8] demonstrated that different initiators/sponsors of internet hospitals including the government, hospitals, and enterprises have different purposes and scopes of service (ie, target consumer). The virtual visit market is an emerging market in China, and hence a study on consumers’ intention to use virtual visits delivered by different sponsorship types of internet hospitals is critically important to understand the market structure and future development of the industry. However, little is known about consumers’ intention to use different sponsorship types of internet hospitals in China or intention to use virtual visits delivered by different sponsorship types of internet hospitals.

We conducted a cross-sectional survey in Zhejiang province to examine patients’ intention to use virtual visits and their intention to use virtual visits delivered by different sponsorship types of internet hospitals and identify the factors that affect patients’ intention to use virtual visits. Zhejiang takes a leading role in China’s internet hospital industry development because it has the first licensed enterprise-sponsored internet hospital (WeDoctor Group) [9] and the first public tertiary hospital–sponsored one (the First Affiliated Hospital of Zhejiang University) [10], the first direct-to-consumer provincial government internet hospital platform [11].

When our survey was conducted, Zhejiang was the first and only province in China that delivered direct-to-consumer telemedicine services including virtual visits for its residents via its provincial internet hospital platform. Only after the outbreak of COVID-19 did other provincial governments start to deliver direct-to-consumer telemedicine service to their residents; however, the majority of services were specially designed free virtual consultations to contain the COVID-19 epidemic. Zhejiang is currently still the leading province to deliver virtual visits by pooling all local health care resources via its provincial internet hospital platform. Therefore, a survey on residents in Zhejiang Province can provide prospective insights for research questions on Chinese consumers’ intention to use virtual visits delivered by hospital-sponsored, enterprise-sponsored, and government-sponsored internet hospitals.

As two primary service types of internet hospitals, virtual visit and virtual consultation are different from each other in service scope and health insurance coverage. However, there is hardly a clear division between virtual consultation and virtual visit in the current knowledge of internet hospital literature. Previous studies have examined either patients’ intention to use an internet hospital [12,13] or a virtual consultation delivered by internet hospitals [14]. As noted in the introduction, a virtual visit in China is equivalent to a direct-to-consumer telemedicine/telehealth visit in many other countries. However, most previous studies in the literature have examined patients’ intention to use or actual use of direct-to-consumer telemedicine from the perspectives of specific disease-related application [15,16] and specific telemedicine website/app [17,18]. There is a lack of studies on patients’ decision on sponsorship types of internet hospital platforms and determinants of patients’ decision making.

Systematic reviews have revealed that the technology acceptance model (TAM) was the most commonly used model to examine users’ acceptance of health information technologies including telemedicine [19-21]. The TAM beliefs (technological issues: perceived ease of use, perceived usefulness) and consumer trust are two distinct sets of beliefs that contribute in their own right to increase intention to use the website and, through it, transactions with the e-vendor [22]. Through a systematic review and meta-analysis, Tao et al [19] found that trust is significantly correlated with behavioral intention to use consumer-oriented health information technologies including telemedicine. Trust positively influences the patient’s intention to use telemedicine services in developing countries like Pakistan [23]. However, very few previous studies have examined the patient’s trust in a telemedicine service including trust in the care organization, trust in the care professional, trust in the treatment, and trust in the technology [24]. There is a lack of studies that examine patients’ trust in the sponsor/owner. Based on the abovementioned findings, we hypothesized that trust in the sponsor significantly affected Chinese patients’ intention to use virtual visits delivered by different sponsorship types of internet hospitals (Hypothesis 1).

Websites, apps and WeChat public accounts are 3 primary modalities through which Chinese providers deliver direct-to-consumer online health information and services. Health websites and apps in China usually cover provisions of health/medical information and knowledge, in-person visit online appointment, result tracing of laboratory and diagnostic imaging tests, and consultation, etc. For some health websites that have been licensed to deliver internet hospital services, the internet hospital, especially a virtual visit, is also an essential part of these health websites. However, among all types of online health information services, Chinese patients had the highest awareness and use of in-person visit appointments and medical fee payment online [25]. Therefore, internet hospital service is an important part of health websites, but it is still not an essential part of many health websites in China, especially before the outbreak of COVID-19. Accordingly, most Chinese patients only had prior knowledge or experience of health websites, which led to the result that they formed perceived trust in health websites rather than internet hospital platforms.

The classification of health websites is also different from that of internet hospitals in the current literature. The website owner/sponsor has been identified as one of the most widely reported indicators consumers applied to evaluate the quality of online health information [26,27]. Internet users’ perceived trust in online health information and service delivered by the health website varies by its sponsor/owner type. Health websites vary in their content and features, across commercial, governmental, and nonprofit websites, as they must respond to different structural incentives and constraints, motivations, and purposes [28,29]. Although there are some differences in sponsorship categories of health websites in previous studies, health websites are usually categorized as commercial, governmental, organizational (eg, nonprofit medical institutions), educational (universities and academic institutions), and personal [29-31]. Users usually perceived medical institutions, universities, and governments as the more trustworthy health website sponsors/owners [32]. Based on the findings that users’ trust in health websites varied by the website sponsorship type and trust affected behavioral intention, we concluded that patients’ trust in health website sponsor affected their intention to use health websites and patients’ intention to use a health website varied by its sponsorship type. Based on the fact that an internet hospital was an important part of a health website, we hypothesized that patients’ intention to use virtual visits varied by sponsorship types of internet hospitals (Hypothesis 2).

We conducted a cross-sectional survey using stratified sampling in 3 cities of different income levels in Zhejiang Province, China. According to Zhejiang Statistical Yearbook 2019 [33], 5 cities were categorized as high-income cities (per capita gross domestic product [GDP] > US $15,000), 3 cities were categorized as medium-income cities (US $15,000 ≥ per capita GDP ≥ US $10,000), and 3 cities were categorized as low-income cities (per capita GDP < US $10,000). Proportionate sampling was used to determine the weight of each income-level city stratum. The weight of the stratum was the proportion of the population contained in that stratum, and the population was the urban population of Zhejiang Province in 2018 [33]. The sampling weight of high-income cities was 52.72%, that of medium-income cities was 24.24%, and that of low-income cities was 23.04%.

A pretest with a sample of 39 participants was conducted in April 2019. Based on China’s 3-tier health care delivery system, hospitals are designated as primary, secondary (secondary 2A, 2B), and tertiary institutions (tertiary 3A, 3B). This study conducted an in-person structured questionnaire survey in 5 different types of hospitals (tertiary 3A, 3B, secondary 2A, 2B, and primary) in May and June 2019. A total of 1653 participants were then randomly surveyed in the outpatient center of each hospital, and health practitioners were excluded from the survey. Hangzhou was the high-income city (n=871), Jinhua was the medium-income city (n=401), and Lishui was the low-income level city (n=381).

Stata 12.0 software (StataCorp LLC) was used to conduct statistical analyses. Descriptive statistics identified sociodemographic characteristics of the final sample, proportions of participants who had the intention to use virtual visits, and proportions of participants who had the intention to use virtual visits delivered by different sponsorship types of internet hospitals.

A systematic review of end user acceptance of telemedicine use [21] found that logistic regression, structural equation modeling, and linear regression were 3 primary statistical analysis methods conducted in previous studies. It also found that TAM was the most used model. Structural equation modeling was primarily used to examine the relationships between factors like TAM beliefs (perceived ease of use, perceived usefulness) and behavioral intention to use telemedicine. And there were usually several paths through which different factors affected the behavioral intention. However, in this study, there was only one path (ie, the trust in sponsor affected patients’ intention to use virtual visits delivered by different types of internet hospitals). Additionally, the sponsorship type of internet hospital was a qualitative variable (categorical variable). Therefore, logistic regression was the appropriate statistical technique to identify the determinants. Binary logistic regression analysis was used to identify the factors that affect patients’ intention to use the virtual visit. Multinomial logistic regression analysis was conducted to examine the association of patient characteristics with the intention to use virtual visits delivered by different sponsorship types of internet hospitals.

Health information seeking was the patient’s most frequently used eHealth or mobile health (mHealth) activity [37], and health information seeking was also a significant predictor of eHealth and mHealth service use [38,39], therefore we inferred that online medical information seekers were target consumers for internet hospitals. To identify the target market, 349 participants were excluded because they self-reported that they have not sought medical information online before. The study sample for statistical analysis on intention to use virtual visits was 1113. The final sample for statistical analysis on intention to use virtual visits delivered by different sponsorship types of internet hospitals was 894. Figure 1 shows the flowchart of sample selection.

Among adult participants, 76.64% (1145/1494) were online medical information seekers. Among medical information seekers, those who did not answer the question of intention to use were excluded. Table 1 shows the sociodemographic characteristics of the final sample for intention to use virtual visits. There were more female medical information seekers than male seekers, which was possibly because females were more likely to be caregivers and information seekers who accompanied family members to visit hospitals. The majority (767/1113, 68.91%) of seekers had the education level of college graduate. Young seekers aged 18 to 55 years accounted for 98.11% (1092/1113) of participants. The survey was conducted in cities in Zhejiang Province where cities usually had both high stocks and inflows of well-educated young people. The majority (965/1113, 86.71%) of seekers were in good or fair health status. Compared to other variables, monthly income was more equally distributed. A total of 80.41% (895/1113) were active internet users whose major information source was the internet. A total of 87.06% (969/1113) reported that they had intention to use virtual visits of internet hospitals.

The binary logistic regression result of intention to use virtual visits of internet hospitals is presented in Table 2. The female was more likely to have intention to use virtual visits (odds ratio [OR] 1.68, 95% CI 1.13-2.48). Education level was significantly and positively associated with the intention to use virtual visits, senior high school (OR 2.05, 95% CI 1.09-3.86), college graduate (OR 3.44, 95% CI 1.94-6.12), postgraduate (OR 3.09, 95% CI 1.19-8.00). Patients with high income level were more likely than those with low income level to have intention to use virtual visits (OR 3.22, 95% CI 1.13-9.15). Consumers of tertiary B hospitals were more likely than consumers of primary hospitals to have intention to use virtual visits (OR 2.62, 95% CI 1.12-6.13).

Table 3 presents Chinese patients’ intention to use virtual visits delivered by different types of internet hospitals. Public hospital–sponsored internet hospitals were the most prevalent (473/894, 52.9%), followed by the provincial government internet hospital platform (238/894, 26.6%), digital health companies (116/894, 13.0%), medical e-commerce companies (48/894, 5.4%), private hospitals (13/894, 1.5%), and other companies (6/894, 0.7%). This result was consistent with the current market structure of internet hospitals in China. Brick-and-mortar hospital sponsors and enterprise sponsors were 2 primary sponsor types of internet hospitals in China, while hospital-sponsored ones accounted for 83.5% (415/497) of internet hospitals by April 30, 2020. Within the category of hospital-sponsored internet hospitals, public hospitals and private hospitals accounted for 90.4% (375/415) and 9.6% (40/415) [5].

Multinomial logistic regression was used to identify the determinants of intention to use virtual visits delivered by 3 major different sponsorship types of internet hospitals (enterprise-sponsored, hospital-sponsored, and government-sponsored). The final sample for the multinomial logistic regression analysis was 894. Table 4 shows the result of multinomial logistic regression analysis.

Using hospital-sponsored internet hospitals as the base outcome, we had following results. Females were more likely to have intention to use virtual visits delivered by the governmental internet hospital platform and less likely to have intention to use visits delivered by enterprise-sponsored internet hospitals. Participants aged 30 to 40 years were more likely than those aged 18 to 29 years to have intention to use the governmental internet hospital platform. Participants with higher education levels were less likely to have intention to use virtual visits delivered by enterprise-sponsored internet hospitals. Participants in a medium-income city (Jinhua) were less likely than those in a high-income city (Hangzhou) to have intention to use virtual visits delivered by both enterprise-sponsored internet hospitals and governmental internet hospital platform. Consumers of secondary A, tertiary B, and tertiary A hospitals were less likely than consumers of primary hospitals to have intention to use enterprise-sponsored internet hospitals. Participants who trusted commercial health websites most were more likely than those trusted governmental health websites to have intention to use virtual visits delivered by enterprise-sponsored internet hospitals rather than visits delivered by hospital-sponsored ones. Compared to participants who trusted in governmental health websites, participants who trusted all other sponsorship types of health websites were less likely to use virtual visits delivered by the governmental internet hospital platform.

Chinese patients who were online medical information seekers had a high intention to use virtual visits of internet hospitals. Gender, education, monthly income, and consumer type were significantly associated with the intention to use virtual visits. Patients had different intentions to use virtual visits delivered by different sponsorship types of internet hospitals, in which the public hospital–sponsored one was the most prevalent one, followed by the government, digital health companies, medical e-commerce companies, private hospitals, and other companies. Gender, age, education, city income level, consumer type, and trust in health website sponsor were significantly associated with the patient’s intention to use virtual visits delivered by 3 different sponsorship types of internet hospitals (enterprise-sponsored, hospital-sponsored, and government-sponsored).

This study has several limitations. First, this study specifically examined online medical information seekers’ intention to use virtual visits delivered by internet hospitals, which might overestimate the whole group of patients’ intentions to use. Second, there are urban-rural divides at various aspects in China. The survey was conducted in cities of Zhejiang Province. As rural population was not included in this survey, the findings of this study only revealed urban patients’ behavioral intention toward virtual visits. We will extend our study by conducting the survey on the whole population (both urban and rural population, both online and nononline medical information seekers) in the future.

Third, Zhejiang provincial internet hospital platform was initially launched on Alipay (the digital payment platform of Alibaba Group) in January 2019. Our survey was conducted in May and June 2019, and in July 2019, the Zhejiang provincial internet hospital platform was transferred to the all-in-one digital provincial government website/app. This transfer might have some effects on patients’ intention to use the government-sponsored internet hospital platform.

Fourth, this study was conducted before the outbreak of COVID-19, which may have been greatly different from the current situation. The outbreak of COVID-19 has brought tremendous growth of internet hospitals in China and has drastically increased the awareness of virtual visits. The rapid increase of awareness would greatly increase the patient’s intention to use and actual use of virtual visits. We will follow and further the study by continuous survey on patients’ behavioral beliefs, intention to use, and actual use of internet hospitals to measure the impacts of drastic expansion of different sponsorship types of internet hospitals (especially public hospital–sponsored ones) brought by the outbreak of COVID-19.

This study has demonstrated that virtual visits differ from virtual consultations in both service scope and health insurance coverage. The virtual visit market is an emerging market in China. This study has implied that different sponsorship types of internet hospitals (ie, enterprises, hospitals, and the government) have different motivations, target consumers, and online health care resource allocation. This study found that Chinese patients who were online medical information seekers had high intention to use virtual visits. Public hospitals, the government, and digital health enterprises were the top 3 sponsorship types of internet hospitals in which patients had intention to use their virtual visit services.

This study revealed that trust in a health website sponsor significantly influenced the patient’s intention to use virtual visits delivered by different types of internet hospitals, which extended the current knowledge regarding the impact of trust on adoption of direct-to-consumer telemedicine service. This study implied that internet hospitals should pay more attention to consumers with characteristics of being female and a tertiary hospital consumer with higher education level and high income level when developing the virtual visit market. It also implied that different sponsorship types of internet hospitals had different target consumers for virtual visit service. Hospital-sponsored internet hospitals had an advantage over the government-sponsored and enterprise-sponsored ones when developing the virtual visit market in medium-income cities.