The data for this study were taken from the 2023 Chinese National Survey of Telemedicine Development in Hospitals, conducted from September to October 2023, which used online questionnaires. The Telemedicine Informationization Professional Committee of China (TIPC) and the National Telemedicine Center of China (NTCC) were in charge of data collection for the survey.

A multistage sampling method was used to recruit respondents. Initially, 11 provinces were strategically selected from across eastern, central, and western regions for the survey, considering prior collaborative experiences. Subsequently, a TIPC member was designated to manage the survey operations in each province. They were tasked with overseeing hospital coordination and selection. Following receipt of consent from the chosen hospitals, a dedicated staff member was appointed in each facility by the TIPC members to handle the recruitment of participants, all of whom had undergone training by 3 experienced researchers from the NTCC. Ultimately, electronic questionnaires, accompanied by quick response (QR) codes and detailed instructions, were distributed to health care providers engaged in telemedicine services by the staff member in each hospital. Participation was voluntary, allowing participants to withdraw anytime. Data anonymity and confidentiality were guaranteed, and no monetary compensation was offered. Researchers from the NTCC periodically reviewed the data collected during the survey, offered prompt feedback, and carried out centralized data cleansing to ensure data integrity and effectiveness.

At the end of the survey, 996 questionnaires were received from all TIPC members. After excluding questionnaires with repeated information, incomplete responses, inconsistency, and other problems (eg, some providers reported having worked in telemedicine longer than their overall working years), there were 831 (83.43%) valid questionnaires left. Data used in this study were from questionnaires filled by health care providers who had ever practiced telemedicine and has answered the question “How frequently did you offer telemedicine services before/during/after the COVID-19 pandemic explosion?” As a result, only 428 (51.5%) questionnaires were included.

The questionnaires included the following demographic items:

Professional titles included junior, intermediate, and senior. Education recorded doctor, master’s, and bachelor’s degrees, as well as education below the bachelor level. The province was then classified as the eastern, central, or western region of China. Hospital levels were categorized as tertiary or nontertiary. Departments of providers were sorted as clinic and nonclinic departments. How easy it was to acquire test results online and the frequency of telemedicine at 3 time points were scored on a 5-point Likert scale as “always,” “often,” “occasionally,” “rarely,” and “never.”

This study was conducted using an online survey method, collecting data that were limited to demographic information, such as age and gender, with no collection of sensitive or individually identifiable biological information. All procedures were approved by the Medical Ethics Committee (2020-KY-0379-002) of the First Affiliated Hospital of Zhengzhou University. Informed consent was obtained from all individual participants involved in the study. For secondary analyses of research data, we confirmed that the original informed consent or Institutional Review Board (IRB) approval included provisions allowing for secondary analysis without additional consent from the participants.

In this study, we prioritized the privacy and confidentiality of participants. The online survey was designed to collect only nonsensitive demographic information, such as age and gender, without any personal identifiers that could compromise the privacy of the respondents. All responses were collected anonymously, ensuring that no individual could be identified using the data.

Participants were not offered any form of compensation for their participation. No images of individual participants or users were included in the manuscript or multimedia appendices that would allow for identification.

The general frequency of offering telemedicine services was calculated at each time point, and the difference in the frequency between the 3 groups was considered the primary outcome of this study. As the frequency was an ordered categorical variable, it was reassigned a score of 1-5 at each time point. A change in the frequency of offering telemedicine services as the pandemic progressed was calculated, via subtraction, as the difference in the frequencies between the 3 time points for each participant, and the change was categorized as an increase or a decrease, depending on whether the result of the subtraction was >0 or <0. Thus, binary variables showing whether the frequency increased from before the pandemic outbreak to during its explosion, increased from during the explosion to after controls were relaxed, and from before the outbreak to after controls were relaxed were calculated as secondary outcomes of this study.

Age was presented as the mean (SD). Other continuous variables were presented as the median (IQR) as their distribution was not symmetric. Categorical variables were described as counts and percentages. Chi-square tests were performed to compare differences in the frequency of offering telemedicine services between all groups. Single-variate and multiple regression analyses were used to explore correlations between the increase in frequency and other questionnaire responses. Stratified analysis was also performed in using the multiple regression model in different groups. A 2-sided P value of <05 was considered statistically significant in all analyses. Statistical analysis was performed using R version 4.3.1 (R Foundation for Statistical Computing). Figures demonstrating intergroup distributions and advantages and disadvantages were created in Microsoft Excel.

A total of 428 health care providers (n=199, 46.5%, males and n=229, 53.5%, females) filled in the questionnaires and were included in this study. The general characteristics of participants and their responses to the questionnaires are summarized in Table 1. The participants’ mean age was 38.6 (SD 8.51) years. Over half were bachelors (n=257, 60.05%), with a median of 12 (IQR 7-20) years of working experience and 4 (IQR 2-6) years working in telemedicine. The majority were from tertiary hospitals (n=229, 53.5%) and located in central (n=218, 50.93%) and western (n=151, 35.28%) parts of China. Most providers were engaged in telemedicine through platforms operated by their own institutes (n=317, 74.07%), while the rest were engaged through third-party platforms usually operated by internet companies, such as Chunyu Doctor. Participants reported that test and examination results could be generally acquired (n=125, 29.21%, always; n=191, 44.63%, often) during the telemedicine procedure. Most participants believed that the reliability and validity of telemedicine are high, with a median score of 9 (IQR 8-10) and 9 (IQR 7.75-10), respectively, on a scale of 0-10. Regarding the complexity of the diagnosis procedure, 200 (46.73%) participants indicated online virtual visits are quicker than in-person outpatient visits, 99 (23.13%) participants reported the length of virtual medical service is even longer than an offline visit, and the rest reported that the time spent in the 2 methods is not different.

According to the providers’ responses, the frequency of providing telemedicine services, including online visits, generally increased as the pandemic proceeded through the 3 stages: before, during, and after the COVID-19 explosion. As shown in Table 2, approximately 198 (46.26%) providers always or often offered telemedicine before the pandemic outbreak, 249 (58.17%) during the pandemic explosion, and 268 (62.62%) after the pandemic explosion. The differences between these groups, calculated using the chi-squared test, indicated that generally the frequency of providing telemedicine services increased since the pandemic exploded and stayed at similar levels after the explosion (Table 3).

Regression models were performed in single and multiple variate analyses, and whether offering telemedicine services increased during the pandemic explosion versus before the pandemic outbreak, after the pandemic explosion versus during the explosion, and after the pandemic explosion versus before the pandemic outbreak was considered a dependent variable. As demonstrated in Table 4, age, working experience, telemedicine working experience, and providers’ report of the validity of telemedicine were positively correlated or negatively correlated with an increase in the frequency of offering telemedicine services as the pandemic explosion proceeded, with significant correlation but a coeffective value close to 1. The providers’ professional titles were significantly correlated with the frequency increase in the single-variate model: they had an increased frequency compared to those with junior titles. Participants from western regions compared to central regions and from nontertiary hospitals compared to tertiary hospitals were less likely to increase telemedicine services as the pandemic exploded. Compared to participants who felt the time spent between online and in-person visits was not different, those who felt online services take longer (during vs before: odds ratio [OR] 4.27, 95% CI 2.30-8.15; after vs before: OR 3.22, 95% CI 1.79-5.92) or shorter (during vs before: OR 2.09, 95% CI 1.19-3.80; after vs before: OR 2.15, 95% CI 1.28-3.72) increased telemedicine services.

In the multivariate analysis, age, professional title, education, years working in telemedicine, region, type of hospital, platform provider engaged in telemedicine, time spent in online versus in-person service, and validity of telemedicine were included as independent variables in regression models. According to variance inflation factor test results shown in Table S1 in Multimedia Appendix 1, working years (during vs before: 10.85; after vs during: 9.99; after vs before: 10.86) and reliability of telemedicine (during vs before: 6.9; after vs during: 6.72; after vs before: 6.85) were not adjusted in the multivariate regression model. As shown in Table 5, providers with senior professional titles compared to junior titles, who were engaged in telemedicine through third party–operated platforms compared to hospital-run platforms, and who felt online visits are shorter or longer than in-person visits had significant coeffective values. In Tables S2 and S3 in Multimedia Appendix 2, single and multiple regression models involving frequency increases after versus during the pandemic explosion are reported, and no significant correlations were found.

Stratified analysis was performed. The distribution of the online service length between 2 types of platforms (Table S4 in Multimedia Appendix 3) showed that over half of the providers on self-operated platforms (160/317, 50.47%) felt online visits are shorter, while on third party–operated platforms, only 40 (36.04%) of 111 providers felt online visits are shorter. The difference in distribution between the 2 types of platforms was significant. Thus, the correlation of whether online service takes longer and the frequency increase in telemedicine might be different between the 2 types of platforms. As a result, providers were grouped by whether they chose to offer telemedicine on platforms operated by their own hospitals or on third party platforms.

As demonstrated in Table 6, multiple regression analysis was applied, and the variables age, professional title, education, years working in telemedicine, region of hospital, type of hospital, and validity of telemedicine reported by providers were adjusted as covariables in the regression model in each group. In the self-operated platform group, compared to providers who felt the time spent in online and in-person visits is not different, those who felt online visits are longer were significantly more likely to increase offering telemedicine services when COVID-19 exploded (OR 3.84, 95% CI 1.73-8.83) and when pandemic prevention controls were relaxed (OR 3.40, 95% CI 1.61-7.34) compared to before the pandemic. The differences in the telemedicine offer frequency increase for providers who felt online service time is shorter than in-person visits were not significant. In the third party–operated platform group, the increase in telemedicine offer frequency was significant among providers who felt online visits are shorter than in-person visits both when the pandemic exploded (OR 8.16, 95% CI 1.39-68.3) and when prevention measures were relaxed (OR 5.70, 95% CI 1.22-33.6) compared to before the pandemic, while the differences for providers who felt online visits are longer were not significant.

Based on a national survey of telemedicine providers, this study explored the change in the frequency of offering telemedicine services at the time before COVID-19 exploded, during the pandemic explosion, and after prevention measures were relaxed; analyzed the correlation between other factors and the frequency change; and located health care providers who would like to offer more telemedicine services in China even after the pandemic is over.

In general, our study found that the frequency of health care providers offering telemedicine increased both during the pandemic and after prevention measures were relaxed compared to before the pandemic. Specifically, we found that providers’ feelings about the online service length compared to in-person visits were correlated with the telemedicine offer frequency increase as the pandemic proceeded. Among providers offering telemedicine through platforms run by their own hospitals, those who felt online visits take longer were more likely to increase offering telemedicine. Among providers engaged through third-party platforms, those who felt online visits are shorter were more likely to increase offering telemedicine.

Studies worldwide have already proved that telemedicine technologies were helpful tools during the COVID-19 explosion and were generally welcomed after the pandemic was over [1,3,9,23,24]. Considering the condition changed in China, it was significant that the telemedicine offer frequency was still high after pandemic controls were relaxed [22].

In our study, the difference in the telemedicine frequency change was not significant between different types of hospitals but was significant between the 2 types of platforms. As Chinese telemedicine-promoting policies started early, there were various ways of offering telemedicine services at the time the pandemic exploded [17]. In May 2019, there were 158 telemedicine platforms known as “internet hospitals” operated by local public hospitals around the nation and video-based platforms, such as the Huawei telepresence conference system [18]. Private companies, such as Chunyu Doctor (founded in 2011) and Ping An Health (founded in 2015), established and began operating their teleconsulting and diagnosis platforms even earlier. Some physicians said they would still contact patients through phone calls, text messages, or instant messaging software, such as WeChat, but a few responded that they would offer telemedicine services or only consultations through these ways. As a result, common tools, such as WhatsApp, telephone, or email, were not included as a way of offering telemedicine in this study. The difference in frequency between the 2 kinds of platforms was not the same as the platforms or tools used in former studies [25,26].

The 2 types of platforms in this study included providers engaging in telemedicine publicly or privately. As all our participants were from public hospitals, the self-operated platforms were similar to those run by public hospitals. However, the third-party platforms were mainly run by private companies. Former studies on cognition and telemedicine services between public and private providers are not consistent. Galle et al [27] and Cordioli et al [28] suggested that providers in the private sector or private clinics have a higher frequency of offering telemedicine services. Scheffer et al [29] claimed physicians working in both public and private sectors are the most frequent users. Elhadi et al [30] indicated there are no differences in attitude, cognition, knowledge, or skills in providers between work in public hospitals, private hospitals, or both. Mazouri-Karker et al [31] reported that differences between public and private providers are only their preference in tools. In our results, public providers reported that a higher frequency increase after pandemic controls were relaxed might be because public hospitals, especially tertiary public hospitals in China, usually have a better reputation, whereas private third-party telemedicine platforms or physicians who work a “part-time job” on these platforms can hardly acquire patient trust.

We also found that some providers feel that online visits take longer, others believe that online visits are shorter, while still others feel that they take the same time as in-person visits. This is consistent with Ly’s [32] results that some physicians report telemedicine increases their efficiency and lowers the workload, while other physicians report telemedicine takes more time and increases their workload because of having to learn new techniques and because communication on-screen is difficult. Although some researchers believe that one of the advantages of adopting telemedicine during COVID-19 was that it could save time and increase efficiency, it actually saved patients’ traffic-related costs on both time and money and increased efficiency by reducing unnecessary visits through pretriage online [24,30,33]; however, we are still uncertain about whether adopting telemedicine could have saved the providers’ time. Naamani et al [34] suggested that telemedicine could save providers’ time because of pretriage online. Shaarani et al [26] proved that the average time spent in in-person visits reported by former studies is longer than 15 minutes, while providers in their survey reported a shorter average online visit time. Silver et al [35] found out that providers who spent an average visiting time longer than 14 minutes would prefer offering visits in person instead of virtually. Ramsey et al [36] indicated that providers would have to spend more time on documentation, which made the procedure more complicated and online visits took longer.

Providers in our study who reported that online visits are shorter than in-person visits have a good reason to increase the frequency of offering telemedicine services as these have higher efficiency. Providers who reported that online visits are longer than in-person visits, compared to those who believed there is no difference between the 2 service types, would also like to increase the telemedicine usage frequency. This might be because telemedicine has other advantages as well, such as saving patients’ costs, even though it would increase the time spent by providers. As Ramsey et al [36] suggested, the other benefits for clinicians, patients, and quality of care could motivate health care professionals to offer more telemedicine services. In addition, in China, the average outpatient visiting time is much shorter, especially in tertiary hospitals [37]. This indicates that even though online visits take longer, they might still be shorter than average in-person visits in other countries. Providers who feel online visits take more time are not against offering telemedicine according to Silver et al [35], as also shown by our results of stratified analysis. Providers offering telemedicine services through their own hospitals’ platforms might be more responsible and attach importance to their reputation. However, providers using third-party platforms might pay more attention to efficiency. As a result, on self-operated platforms, the increase in the telemedicine offer frequency by providers who felt online visits take longer was significant, while on third-party platforms, the increase in the telemedicine offer frequency by providers who felt online visits are shorter was significant.

Our study showed that the length of online visits is correlated with the frequency change in offering telemedicine by providers, and the frequency increased for both longer and shorter online visits. However, our questionnaire was not designed to acquire the average number of minutes the providers spent in online and in-person visits. Further studies could focus on this to obtain a more detailed and convincing result. The results from this study were all from a cross-sectional survey, though it was nationwide. Thus, the frequencies of offering telemedicine at the 3 time points and the frequency change data were all obtained from providers' self-reports, which may have led to self-report and selection bias. It would be more accurate after regional online prescription–examining platforms in China are established and we can acquire data on them.

Moreover, some questions in this study were rather subjective. As the questionnaire used a 5-point Likert scale, questions regarding the frequency of offering telemedicine was set with 5 levels from “never” to “always,” without specifying the exact time or duration for each level. “What platform do you mainly use to offer telemedicine services?” was set as a single-choice question, which was less suitable for more complicated situations, such as providers using both types of platforms.

The frequency of telemedicine services offered by health care providers has increased since the COVID-19 pandemic exploded and was still high after pandemic prevention controls were relaxed in China. Differences in providers’ professional titles, types of platforms, and length of online visits compared to in-person visits are correlated with the increase in frequency. In using hospital-run platforms, providers with longer online visits are more likely to offer telemedicine services, while on third-party platforms, providers with shorter online visits would like to increase telemedicine services.