Introduction

Since its emergence in the 20th century, telemedicine has played a critical role in global healthcare and has made significant progress in responding to catastrophic events, particularly the COVID-19 pandemic (Hollander and Carr 2020; Mann et al. 2020). Extensive research has documented the considerable potential of telemedicine and thoroughly documented its numerous benefits in preventing and managing disease progression to enhance patient health outcomes. (Rush et al. 2022). As telemedicine is a valuable form of healthcare delivery, its advantages are further amplified during the COVID-19 pandemic. Furthermore, patients with mobility difficulties, those isolated at home, and those residing in poor or remote areas can access health care through telemedicine. Telemedicine also allows healthcare resources to be flexibly mobilized to meet the demand for consultations while relieving pressure on traditional physical hospitals.

Telemedicine in China underwent significant advancement during the COVID-19 pandemic due to a convergence of factors. Firstly, robust policy support has provided a cornerstone for telemedicine development (Yang et al. 2021). The implementation of Chinese government policies, such as the “Notice on Further Strengthening the Construction of Telemedicine Networks,” had underscored the importance of enhancing telemedicine infrastructure and thereby bolstering its accessibility and quality. This regulatory framework incentivised healthcare institutions and professionals to integrate telemedicine technologies into their practices. Moreover, the “Notice on Strengthening Internet Diagnosis and Consultation Services in Epidemic Prevention and Control” further propelled telemedicine adoption by emphasising the necessity of internet-based healthcare consultation services during public health emergencies, such as the COVID-19 pandemic. Secondly, advances in internet technology, including the enhancement of broadband infrastructure, widespread use of mobile connectivity, and integration of artificial intelligence and big data analysis, has continuously grown demand for telemedicine services. Finally, government platforms, hospital platforms, and third-party providers are actively promoting the diversification of service delivery modes, making telemedicine increasingly prevalent. The surge in demand for telemedicine services, especially in regions of western China such as Sichuan Province, highlights their practicality and effectiveness.

The healthcare community possesses an extensive understanding of telemedicine, which is generally regarded as advantageous for patients (Jallal et al. 2023; Ashfaq et al. 2020; Anderson et al. 2017). Most related research has concentrated on the willingness of patients to embrace telemedicine and their satisfaction with it given that they are the primary beneficiaries and funders of telemedicine services (Kruse et al. 2017; Luna et al. 2022); research focusing on the willingness of the physician population to embrace telemedicine has been much rarer. Before the COVID-19 outbreak, physicians primarily embraced telemedicine because of its perceived convenience and usefulness (Shah et al. 2019; Rho et al. 2014). Research since the emergence of COVID-19 has shown that the factors affecting physicians’ uptake of telemedicine are diverse, including the impact of the pandemic (Banks et al. 2021). However, such studies have not fully integrated various aspects influencing the adoption of technology. Therefore, this study constructed an integrated theoretical framework to comprehensively analyse the factors influencing physicians’ willingness to use telemedicine during the COVID-19 pandemic, providing practical insights that would aid healthcare institutions and policymakers in devising strategies to enhance telemedicine utilisation and optimise its integration into the healthcare system. This will, in turn, bolster the ability of the healthcare system to respond and adapt to future public health crises.

Indeed, technological factors play a crucial role in physicians’ acceptance of telemedicine. For example, a lack of adequate technical training was identified as a barrier to physicians’ willingness to use telemedicine in Ethiopia (Ahmed et al. 2021). Similarly, a study of physicians in Spain, Colombia, and Bolivia found that the level of infrastructure development and communications technology implementation, as well as physicians’ use of information technology in their personal lives, were broad factors that significantly influenced their utilisation of telemedicine (Saigí-Rubió et al. 2014). Moreover, other studies have highlighted the significance of technology-related factors in shaping physicians’ willingness to use telemedicine. Notably, the ease with which telemedicine systems could be used was identified as a critical determinant for both physician and patient adoption (Miner et al. 2021). Various methods and models have been developed to investigate the effects of technology-related factors on the adoption of technological systems. In particular, the unified theory of acceptance and use of technology (UTAUT) has emerged as the most effective model because it incorporates variables that significantly impact users’ intention to use a specific technology. These variables include the usefulness of the technology, its ease of operation, and social influence, which are precisely defined in terms of performance expectancy, effort expectancy, social influence, and facilitating conditions. Previous studies have provided extensive support for the strong explanatory power of the UTAUT in predicting the intention to use telemedicine (Rouidi et al. 2022).

Though telemedicine has been available for many years, modern technology has made it more accessible and COVID-19 made it more applicable than ever (Gillman-Wells et al. 2021). Brazilian physicians perceived the primary advantage of telemedicine, particularly since COVID-19, as its capacity to enhance healthcare access in low- and middle-income settings (Scheffer et al. 2022). This suggested that the pandemic may influence physicians’ adoption of telemedicine. Indeed, since the outbreak of COVID-19, telemedicine has been widely used to minimize disease transmission by reducing physical contact between physicians and patients, enhancing safety measures and mitigating viral spread (Altulaihi et al. 2021). As a result, telemedicine has evolved from being simply a convenient form of health care for patients to an infection prevention measure that both physicians and patients can employ to reduce disease transmission (Miner et al. 2021). The use of protection motivation theory (PMT) to assess the risk of disease and its severity has been shown to enhance behavioural intentions and impact health-related decisions (Surina et al. 2021). Therefore, this study combined the PMT and UTAUT approaches to establish a theoretical framework for comprehensively analysing the factors influencing physicians’ willingness to use telemedicine considering the impact of the COVID-19 pandemic. This combined approach was expected to inform a more thorough understanding of the complex interplay among the factors affecting the adoption of telemedicine.

The active participation of physicians is critical for effectively implementing telemedicine as they play a crucial role in maximising its potential and ensuring its success. However, the comprehensive factors influencing telemedicine adoption by Chinese physicians during the COVID-19 pandemic remain uncertain. This study accordingly addressed this gap by examining the factors that affect physicians’ inclination to adopt telemedicine using the PMT and UTAUT models. To achieve this objective, physicians from China were surveyed during the phase of pandemic normalization and control from June 3rd to October 26th, 2022. This survey collected data describing the physicians’ basic knowledge of telemedicine, current levels of acceptance and use, and opinions of factors influencing their willingness to use telemedicine. By identifying these influences, healthcare policymakers and administrators could implement targeted strategies to promote greater telemedicine adoption among physicians, thereby potentially improving healthcare delivery efficiency and accessibility, especially in times of crisis such as the COVID-19 pandemic.

Literature review and hypotheses

Protection motivation theory (PMT)

Generally, PMT provides a theoretical framework for understanding why individuals adopt recommended behaviours to protect themselves from potential threats (Rogers, 1975). It primarily comprises threat and response assessments. A threat assessment considers the perceived vulnerability to a disease, reflecting an individual’s perception of their likelihood of contracting the disease, as well as their perception of the severity of the negative consequences associated with it. Next, individuals conduct a response assessment, in which they evaluate the potential harm or loss presented by the threat and the cost required to avoid it. Individuals subsequently make their choices based on the results of this evaluation. Therefore, individuals are more likely to choose behaviours that avoid perceived threats when the perceived cost is low compared with alternatives that pose greater risks.

Representing a critical framework for explaining health-related behavioural intentions (Nudelman et al. 2023), PMT has been applied to investigate individuals’ disease prevention intentions, including cancer screening behaviours for colorectal and cervical cancers in women (Wei et al. 2022; Bai et al. 2018) as well as preventive vaccination behaviours against emerging infectious diseases, such as COVID-19 (Wang et al. 2021b; Xiao et al. 2021). The perceived severity of and susceptibility to the COVID-19 threat have also been shown to lead to reduced travel behaviours, adversely impacting the tourism industry and accelerating the replacement of brick-and-mortar travel agencies with online platforms (Zheng et al. 2021; Toubes et al. 2021). Similarly, online shopping gained in popularity over offline shopping during the pandemic, and teachers and students adopted online teaching methods to avoid infection (Mishra et al. 2020; Truzoli et al. 2021). Furthermore, employers and employees have embraced virtual meetings for remote work. Finally, the COVID-19 pandemic has prompted an increase in patient use of online consultations, reflecting a broader shift towards digital solutions in professional settings.

In the past, diagnosis and treatment were primarily based on the patient’s need for disease treatment and prevention, with the physician solely responsible for providing healthcare services without preventive protection. However, since the COVID-19 outbreak, the severity and reality of the risk of contracting disease has made in-person activities, such as healthcare consultations, less appealing to both physicians and patients. Objectively speaking, physicians are at greater risk of infection and, as such, have a heightened perception of the risk of disease transmission (Peres et al. 2020; Lin et al. 2021). The critical importance of the perceived vulnerability and severity components of the PMT model in shaping COVID-19 prevention behaviours and intentions cannot be overstated (Surina et al. 2021). As physicians represent a high-risk group for COVID-19 infection, their adoption of preventive behaviours is of the utmost importance in mitigating the spread of the virus (Bashirian et al. 2020). Indeed, the prevention of COVID-19 is crucial for protecting not only physicians but also vulnerable groups and wider society. Physicians who adopted preventive measures during the pandemic sought to protect themselves from infection, thereby reducing the transmission of the virus to others. Given their intimate understanding of the issue, physicians may be more willing to use telemedicine. Therefore, based on the PMT, this study proposed two hypotheses:

H1: Perceived vulnerability positively influences physicians’ willingness to use telemedicine.

H2: Perceived severity positively influences physicians’ willingness to use telemedicine.

Unified theory of acceptance and use of technology (UTAUT)

The COVID-19 pandemic led to a significant shift in the structural patterns of various industries, with many embracing digitalisation (Wang et al. 2021a). Furthermore, rapid advances in information technologies have permeated every aspect of socioeconomic and medical spheres, leading to a growing interest in the factors influencing the adoption and usage of technology. Notably, the UTAUT model, which was developed by Venkatesh’s research team (Venkatesh et al. 2003), is a comprehensive model that integrates several theories, including the innovation diffusion theory, theory of reasoned action, theory of planned behaviour, and technology acceptance model. The UTAUT model has been extensively tested through large-scale empirical studies and has demonstrated a higher explanatory power than other models (Venkatesh et al. 2003). Indeed, the UTAUT model can explain up to 70% of the variance in behavioural intentions (Arfi et al. 2021; Chang et al. 2007). In the field of healthcare technology adoption, numerous studies have highlighted the effectiveness of UTAUT in explaining users’ behavioural intentions towards technology, including user adoption of contact tracking applications (Chopdar 2022), nurse adoption of mobile nursing applications (Pan and Gao 2021) and user adoption of mobile healthcare services (Alam et al. 2020; Schmitz et al. 2022; Alabdullah et al. 2020; Tian and Wu 2022; Khan et al. 2018). The determinant of technology acceptance behaviour is behavioural intention, which is influenced by performance expectancy, effort expectancy, social influence, and facilitating conditions. This study posited that physicians’ behavioural intention to adopt telemedicine is shaped by these factors.

(1) Performance expectancy (PE) refers to the extent to which individuals believe that using a particular technology will enhance their job performance (Venkatesh et al. 2003). Individuals are more likely to use mobile technology frequently and consistently if they perceive a greater number of benefits, indicating that PE is a key influencer of usage intention (Farzin et al. 2023). Furthermore, previous studies demonstrated that PE is the most robust predictor of the behavioural intention to adopt new technologies (Alalwan 2020; Venkatesh et al. 2016). The motivation of physicians to utilise and integrate novel technologies into their practice was largely influenced by their belief that such technologies would assist them in achieving their professional goals (Esmaeilzadeh et al. 2015; Prakash and Das 2021; Kim et al. 2023). Given the unique nature of health care, which benefits both patients and physicians (Grover et al. 2018), a greater perceived usefulness of internet-based diagnosis and treatment was likely to lead to increased adoption of related technologies. Consequently, we hypothesised the following:

H3: Performance expectancy has a positive effect on physicians’ willingness to use telemedicine.

(2) Effort expectancy (EE) refers to the extent to which users find a system easy to effectively operate (Venkatesh et al. 2003). Studies have identified EE as a direct determinant of the intention to use a system, and its influence on technology adoption is widely acknowledged (Li et al. 2022; Leong et al. 2013; Hayat et al. 2022). In the healthcare sector, EE has been identified as a key factor affecting users’ willingness to embrace technology (Hasan and Bao 2022; Hoque and Sorwar 2017). Notably, consulting with patients online is not an entirely novel concept, and physicians possess a certain level of proficiency in conducting online consultations through various modes, such as text, video, and voice. The degree of ease with which physicians can perform these activities online is positively correlated with their willingness to adopt this technology. Therefore, we hypothesised the following:

H4: Effort expectancy positively influences physicians’ willingness to use telemedicine.

(3) Social influence (SI) refers to the extent to which an individual’s use of technology is shaped by others’ opinions and behaviours (Venkatesh et al. 2003). It has been identified as one of the most significant and influential predictors of new technology acceptance (Martins et al. 2014), and is particularly relevant in the healthcare sector (Hussain et al. 2019; Ahmad and Khalid 2017; Cajita et al. 2017; Wang et al., 2020). Research suggested that some users were more likely to embrace telemedicine after witnessing successful experiences among family and friends (Benis et al. 2021). Given the attention paid by the public, patients, and social media to the work of physicians, SI is likely to affect healthcare professionals’ adoption of telemedicine. Therefore, we hypothesised the following:

H5: Social influence has a positive effect on physicians’ willingness to use telemedicine.

(4) Facilitating conditions (FC) are defined by the UTAUT framework as reflecting the degree to which individuals perceive that technical and organisational support exists to facilitate their use of a system (Venkatesh et al., 2003). Previous research has demonstrated that FC can reduce barriers to technology adoption and enhance users’ willingness to use technology (Sebetci and Cetin 2016). In healthcare sector, FC play a crucial role in promoting the use of digital health technologies (Blok et al. 2020). In contrast to other UTAUT constructs, FC directly influence usage behaviour and have been found to represent direct predictors of usage behaviour (Li et al. 2019). There are undoubtedly barriers to the transformation from traditional offline health care to online health care, and FC play a critical role in facilitating the acceptance and use of such digital health technologies (Hussain et al. 2019). The more FC that physicians perceive, the stronger their use behaviour and willingness to adopt the system. Therefore, we hypothesised the following:

H6: Facilitating conditions have a positive impact on the actual use of telemedicine by physicians.

(5) Behavioural intention (BI) is considered in the UTAUT model as a crucial construct that denotes a user’s readiness to perform a particular behaviour (Ajzen 2011). A strong BI to use new technologies has been shown to predict increased actual usage (Kijsanayotin et al. 2009). For physicians, a greater BI to adopt a new technology is likely to result in a corresponding increase in their actual usage behaviour. Hence, we hypothesised the following:

H7: Behavioural intention has a positive impact on the actual use of telemedicine by physicians.

The conceptual research model presented in Fig. 1 was developed for this study, grounded in these theoretical underpinnings and corresponding hypotheses.

Fig. 1: Conceptual model of factors influencing physicians’ intention to use telemedicine.
figure 1

The conceptual model was developed based on the aforementioned hypotheses, proposing that perceived vulnerability, perceived severity, performance expectancy, effort expectancy, and social influence all have a positive impact on physicians’ behavioural intention. Additionally, facilitating conditions and behavioural intention are hypothesized to positively influence use behaviour.

Full size image

Methods

Survey context and subjects

The target population for this study consisted of physicians in Sichuan, China who were selected using convenience sampling across all levels of hospitals to complete a questionnaire between June 3rd and October 26th, 2022. Before filling out the questionnaire, participants were provided with a brief description of the project and informed that their involvement was entirely voluntary. They were also assured that the data were collected solely for research purposes and that no personally identifiable information would be disclosed. This study was approved by the ethics committee of West China Fourth Hospital and West China School of Public Health, Sichuan University in accordance with established scientific and ethical principles. A total of 330 questionnaires were distributed among physicians at tertiary hospitals, secondary hospitals, and township health centres, and 296 valid responses were received, indicating an effective response rate of 89.7%.

Questionnaire

The questionnaire used in this study was designed based on the PMT and UTAUT models and refined through in-depth interviews with 40 doctors to ensure appropriate language and respondent understanding. The questionnaire was divided into three primary components: basic demographic characteristics, perception and use of telemedicine, and willingness to use telemedicine. A five-point Likert scale was used for each question, with 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = strongly agree.

Statistical analysis

A statistical analysis was conducted using structural equation modelling (SEM) in the SPSS 26.0 and AMOS 24.0 software packages. A sample size of approximately 300 individuals was deemed suitable for this method.

Descriptive statistics were obtained using SPSS 26.0, followed by the two-step approach proposed by Anderson and Gerbing (Anderson and Gerbing 1988) to assess SEM. The measurement and structural models were evaluated using AMOS 24.0. The SPSS 26.0 software was also applied to evaluate the reliability and validity of the questionnaire data. Reliability refers to the consistency of the measurement results, and is typically measured using Cronbach’s alpha. Validity refers to the extent to which a test measures what it is intended to measure and was assessed using the Kaiser-Meyer-Olkin (KMO) and Bartlett’s sphericity tests. Additionally, this study used AMOS to read the raw questionnaire data and conduct a validation factor analysis of the variables in the model to determine whether the hypotheses presented in Section 2 were valid. These rigorous statistical techniques provided meaningful insights into the data and informed reliable conclusions drawn from the result.

Results

Demographic characteristics of sample

Table 1 presents the demographic characteristics of the 296 participants, among whom 134 were male (45.3%) and 162 were female (54.7%) and a majority (86.1%) were between 26 and 55 years old. In terms of professional rank, 37.8% of participants were resident physicians, 47.3% were attending physicians, and the rest were associate chief physicians (10.5%) or chief physicians (4.4%). These physicians worked in various departments including internal medicine, paediatrics, gynaecologist, dermatology, and psychosomatic medicine.

Table 1 Demographic characteristics of respondents.
Full size table

The telemedicine utilization among physicians is listed in Supplementary Table S1, indicating that 53.7% had utilised telemedicine services. The vast majority (134 physicians or 84.3%) reported an increase in the frequency with which they provided telemedicine services since the outbreak of the COVID-19 pandemic. The physicians used many different telemedicine platforms, including self-built platforms provided by their hospitals, Doctor Dingxiang, Good Doctor Online, and Doctor Chunyu. Furthermore, 5.7% of physicians provided healthcare services daily through telemedicine platforms. Notably, a substantial majority of the surveyed physicians (88.9%) endorsed the introduction of telemedicine services at their respective healthcare facilities.

Measurement model testing

Cronbach’s α reliability coefficient is extensively utilized to evaluate the reliability of a scale, with an α value of ≥0.7 indicating acceptable reliability. Supplementary Table S2 provides a description of the scale dimensions as well as the results of the reliability analysis, which indicate that the data can be considered reliable. Both the KMO and Bartlett’s sphericity tests were conducted to determine whether the data were appropriate for factor analysis. The results showed that the KMO was greater than 0.7 and Bartlett’s sphericity was less than 0.05, demonstrating the suitability of the data for factor analysis. Finally, Table 2 shows that the model can effectively explain the data.

Table 2 KMO and Bartlett’s test.
Full size table

An exploratory factor analysis was conducted using the principal component extraction method to examine the structure of the model, confirm its validity, and extract the eight factors with eigenvalues greater than 1.0, as presented in Supplementary Table S3. The cumulative percentage of these eight principal components was 69.44%, and the rotated factor loading matrix is shown in Supplementary Table S4 to comprise a total of eight dimensions and 34 items in the measurement model.

Table 3 lists the fitting indices of the validity analysis model, including \({\chi }^{2}/{df}\text{=}2.07\), which is below the recommended value of 3, comparative fit index (CFI) = 0.911, goodness of fit index (GFI) = 0.836, root mean square error of approximation (RMSEA) = 0.06, and root mean square residual (RMR) = 0.037. The satisfactory values for each of these indices suggested a suitable model fit (McDonald and Ho 2002; Jackson and Trull 2001).

Table 3 Confirmatory factor analysis fit indices.
Full size table

The convergent validity of the results was assessed in terms of factor loadings, composite reliability (CR), and extracted average variance (AVE). Supplementary Table S5 lists that the factor loadings of all items in the scale were greater than 0.5. Furthermore, the CR values ranged from 0.77 to 0.91 and the AVE values ranged from 0.52 to 0.71, meeting the acceptance standards and demonstrating acceptable convergent validity (Fornell and Larcker 1981). The discriminant validity was assessed using the heterotrait-monotrait (HTMT) ratio (Franke and Sarstedt 2019). For two distinct latent variables to pass the discriminant validity test, their HTMT values must be less than 0.85. As detailed in Supplementary Table S6, all HTMT values obtained in this study fell below this threshold, satisfying the criteria and indicating the robust validity of the model for hypothesis testing. Note that the potential for common method bias, which refers to variance attributed to the measurement method rather than to the constructs themselves, must be acknowledged. Despite our efforts to mitigate this bias in this study through rigorous survey design and data collection, some influence may persist. Therefore, future research should explore additional methodological safeguards or statistical techniques to further address and control for common method bias.

Statistical results

After evaluating the measurement model, the hypothesised structural equation model was assessed using the maximum likelihood estimation method. The resulting model fit, listed in Table 4, indicated that the structural equation model met the recommended criteria with \({\chi}^{2}/{df}\,=\,2.09\), CFI = 0.907, GFI = 0.893, RMR = 0.037, and RMSEA = 0.06 (Hu and Bentler, 1999). Thus, the hypothesised structural equation model was considered to have achieved an acceptable fit with the measurement model.

Table 4 Structural equation model fit indices.
Full size table

To further evaluate the hypothesised model, it was jointly evaluated with an initial model constructed and fitted using the AMOS software. The standardised path coefficient (β) was used to determine the strength of the relationship between these two models, and the p value (P) was used to assess the statistical significance of these relationships. A p value less than 0.05 indicates statistical significance. The results presented in Table 5 support the hypothesis that perceived vulnerability leads to willingness to use telemedicine (H1, β = 0.27, P < 0.01), perceived severity has a smaller but still significant positive effect on telemedicine usage intention (H2, β = 0.17, P < 0.05), PE supports the willingness to use telemedicine (H3, β = 0.39, P < 0.01), EE supports the willingness to use telemedicine (H4, β = 0.29, P < 0.01), FC positively affect telemedicine usage behaviour (H6, β = 0.25, P < 0.01), and BI has a positive effect on telemedicine usage behaviour (H7, β = 0.41, P < 0.01). However, the structural equation model analysis indicated that SI had no significant effect on physicians’ willingness to use telemedicine (H6, β = 0.001, P > 0.05).

Table 5 Structural equation model hypothesis and results.
Full size table

Discussion

The main aim of this study was to construct a novel theoretical model by integrating PMT and UTAUT to explore physicians’ willingness and intention to use telemedicine during the COVID-19 pandemic. This study made significant progress in addressing a notable gap in the current literature concerning the factors affecting physicians’ adoption of telemedicine following public health crises and the results provided valuable insights into the complexities governing physicians’ attitudes and actions towards telemedicine utilisation during emergencies. Notably, these insights had actionable implications for healthcare institutions and could inform tailored strategies for more effectively navigating similar crises in the future. The detailed understanding of the determinants driving telemedicine adoption among physicians presented in this study could help institutions recalibrate their policy and infrastructure frameworks to bolster telemedicine adoption, foster resilience, and ensure equitable access to healthcare services. Moreover, by elucidating the pathways to telemedicine integration, this study contributed to the broader imperative of advancing healthcare accessibility and operational efficiency, thereby contributing momentum towards optimised resource utilisation and enhanced healthcare delivery. To aid in these efforts, this section discusses the empirical results presented in Section 4 and addresses the research questions and hypotheses formulated in Section 2.

Hypothesis H1 was supported by the results because perceived vulnerability was proved to positively influence physicians’ willingness to use telemedicine, which was consistent with the results of other studies (Singh et al. 2022; Ing et al. 2020; Guo et al. 2015). Physicians who perceived themselves to be susceptible to COVID-19 were more likely to be willing to use telemedicine as an alternative to face-to-face consultations. Indeed, individuals who feel their health was threatened were motivated to use information technology to avoid such threats. In the context of the COVID-19 pandemic, China had implemented various measures to control the spread of the virus, and telemedicine had emerged as a critical tool for reducing the risk of infection for both physicians and patients. By moving healthcare services online, physicians could continue to provide healthcare services while minimising their risk of exposure to the virus. This had not only helped to protect physicians and patients but had also contributed to the overall efforts to control the spread of COVID-19.

Hypothesis H2 was similarly affirmed as perceived severity was a crucial factor determining physicians’ willingness to use telemedicine. The perceived severity of a disease provided a strong impetus for healthcare providers to embrace telemedicine as a viable option. The devastating impact of COVID-19 on human life and economic systems had made the gravity of the disease all too apparent, particularly during the pandemic’s initial stages, which witnessed the tragic deaths of many healthcare physicians (Lupi et al. 2022). These tragedies also contributed to psychological distress among physicians (Huang et al. 2020). Both perceived vulnerability and perceived ease of use were valid bases for threat assessment, and healthcare physicians and governments were typically more thoughtful than the public in assessing threats. Furthermore, as the infection of a physicians not only endangered their own well-being but also put their other patients at risk, they acted more carefully and thoughtfully. Consistent with the general users, the more severe that physicians perceived COVID-19, the more motivated they were to use telemedicine (Rahi et al. 2021).

The results of this study provided compelling support for hypothesis H3, revealing a significant and positive influence of PE on physicians’ willingness to adopt telemedicine. These findings underscored the notion that physicians were more likely to embrace telemedicine when they perceived it as an asset to their practice, as previously suggested (Rahi et al. 2021). Particularly noteworthy was the critical role of telemedicine in promoting the implementation of graded diagnosis and treatment in China, a fact widely recognised by physicians in the country’s less-developed western regions. Given the high expectations placed upon physicians, useful tools were in high demand. Therefore, improving the efficiency and quality of telemedicine services may have a direct impact on physician satisfaction (Kissi et al. 2020) and thereby influence physician readiness to adopt them.

The intention to use telemedicine was significantly promoted by EE, supporting hypothesis H4. Physicians were more inclined to adopt telemedicine when they recognised the ease and convenience of providing diagnostic and therapeutic services to patients via such platforms. In contemporary society, acquiring new knowledge and navigating information technologies were integral components of physicians’ responsibilities. Indeed, physicians perceived a need to invest considerable effort in learning and utilising internet technology. Therefore, minimising the time and energy required to operate telemedicine systems allowed physicians to redirect their focus towards delivering quality care to patients. As a result, EE played a crucial role in facilitating physicians’ adoption of telemedicine, which had the potential to significantly enhance the quality of healthcare services (Kohnke et al. 2014; Molfenter et al. 2021; An et al. 2021).

Notably, hypothesis H5 was largely unsupported as SI was found to have no significant correlation with physicians’ intentions to use telemedicine. In patient-focused studies, SI had been demonstrated to affect individuals’ use intentions as the opinions of family, friends, leaders, or the media could sway their participation in certain activities, even when they were initially hesitant. However, the scope of this study was confined to physicians. We contended that the lack of SI on physicians’ use intention could be attributed to their rigorous and professional nature, which rendered them less susceptible to external influences. Furthermore, physicians possessed a healthcare knowledge base that differed from that of the public, making them more inclined to rely on practical experience or authoritative sources to make judgments. Indeed, prior investigations had observed that SI may inconsistently exert a positive influence on the intention to use a service or technology, particularly within highly professional occupational groups (Sangeeta and Tandon 2021). Therefore, the absence of any correlation between SI and physicians’ intentions to use telemedicine was not unique to this study (Baydas and Yilmaz 2018).

Hypothesis H6 was confirmed by a significant positive correlation between FC and telemedicine utilisation among physicians. These findings indicated that the availability of technical support and organisational infrastructure could significantly influence physicians’ propensity to use telemedicine. In this study, FC were measured by physicians in terms of their access to the resources and support necessary to utilise the telemedicine platform. These results suggested that hospitals and telemedicine platform developers should ensure that physicians have access to all necessary hardware and software facilities, as well as fast and efficient technical support in the case of any system or access-related issues. By proactively addressing these factors, hospitals and telemedicine platform providers could facilitate the integration of telemedicine into clinical practice and maximise the potential benefits of this technology.

Hypothesis H7 was confirmed by the significant association revealed between physicians’ BI to utilise telemedicine and their actual engagement with telemedicine in clinical settings. This finding underscored the pivotal role of physicians’ BI in driving the adoption and integration of telemedicine into their practices. Specifically, physicians who demonstrated a robust intention to embrace telemedicine were more inclined to convert this intention into tangible action and leverage telemedicine tools and platforms to provide effective patient care. This alignment between BI and telemedicine usage highlighted physicians’ proactive stances towards embracing technological innovations to enhance healthcare delivery. Therefore, fostering and reinforcing positive BIs among physicians could catalyse the widespread adoption and utilisation of telemedicine, thereby facilitating the delivery of accessible and efficient healthcare services, particularly in the context of patient care and teleconsultations.

In summary, the results of this study revealed the factors influencing physicians’ willingness to adopt telemedicine in China, particularly in the unique context of the COVID-19 pandemic. Our findings suggested that all factors examined in this study except SI had a positive impact on physicians’ willingness to use telemedicine. Previous studies had examined the factors influencing the adoption and utilisation of telemedicine technology to varying degrees. For example, studies had shown that providing adequate training and enhancing familiarity with technology were key strategies to encourage the adoption and utilisation of telemedicine in developing countries (Ahmed et al. 2021). Similarly, a study conducted in the Republic of Ghana found that physicians were motivated to use telemedicine because of its ease of use and perceived usefulness (Kissi et al. 2020). Another study conducted in Saudi Arabia revealed that one-third of doctors preferred telemedicine as it allowed them to avoid contact with patients suspected of having COVID-19 (Altulaihi et al. 2021), which was consistent with the observed relationships between perceived vulnerability/severity and willingness to use telemedicine in this study. However, while previous studies provided valuable insights into specific factors influencing the adoption and utilisation of telemedicine technology, this study provided a more comprehensive understanding of the various factors that influence physicians’ willingness to use telemedicine during the normalisation stage of pandemic control in China.

Certain disparities had been observed in the adoption of telemedicine among individuals with low digital literacy and limited access to technology (Rodriguez et al. 2021; Gallegos-Rejas et al. 2023), presenting an unfortunate but unavoidable reality. In the face of this challenge, the remarkable strides in telemedicine adoption across China, which have been marked by simplicity and efficiency, must be highlighted. Notably, certain healthcare facilities have implemented dedicated applications and services tailored for vulnerable segments, such as the elderly, aiming to enhance accessibility. Moreover, healthcare providers should proactively mitigate patient impediments to telemedicine adoption during traditional in-person consultations through concerted promotional endeavours and educational outreach initiatives that strengthen patients’ awareness of telemedicine modalities. Finally, contextualising these observations within the distinctive milieu prevalent in the western regions of China (where the geographical isolation of high-calibre healthcare facilities poses a formidable challenge), individuals grappling with low digital literacy, limited technological access, or linguistic marginalisation clearly endure heightened temporal and financial exigencies when endeavouring to access healthcare amenities at physical hospital sites. The relative ease and diminished cost experienced by these individuals when utilising online healthcare services facilitated by grassroots healthcare personnel or individuals well versed in telemedicine represents a tangible alleviation of their healthcare access hurdles.

Limitations

This study was subject to several limitations that should be considered when interpreting its findings. First, as the survey was conducted in a specific region of China, the generalisability of the results to other regions or countries may be limited. Indeed, the varying pandemic conditions across different regions could lead to differences in terms of the motivation for telemedicine adoption. Second, although the use of quantitative research methods allowed for a comprehensive assessment of the factors influencing the adoption of telemedicine among physicians, a qualitative approach incorporating interviews may provide additional insights into their attitudes and perspectives. Future research should accordingly combine both quantitative and qualitative methods to obtain a more comprehensive understanding of this issue. Third, as the data for this study were collected during the pandemic normalisation phase in China, the results of this study might not fully capture physicians’ willingness to use telemedicine after the complete resolution of COVID-19. Finally, this study focused only on the diagnostic and treatment services provided by physicians to patients via the internet and did not consider other services offered by internet-connected hospitals, such as peer-to-peer communication among physicians or healthcare services used by physicians as patients. Future studies should explore physicians’ willingness to use telemedicine in different contexts to provide a more nuanced understanding of the factors influencing its adoption.

Conclusions

This study uses an integrated PMT and UTAUT approach to examine the influence of perceived vulnerability and severity as well as PE, EE, SI, FC, and BI on physicians’ willingness to adopt telemedicine. The results suggested that most of these factors positively affected physicians’ adoption of telemedicine.

However, we observed no substantial association between SI and physicians’ intention to adopt telemedicine. This could be potentially attributed to the distinctive nature of the healthcare physicians, wherein external factors exerted minimal influence on decision-making processes. Furthermore, we observed that the perceived vulnerability and severity components of PMT positively influenced physicians’ willingness to use telemedicine. While this theory has been widely applied in the healthcare field, this is the first time it has been applied to a physician population. This result emphasises the importance that physicians place on prioritising their own safety before providing healthcare services to patients.

These findings provided valuable insights for devising targeted strategies and policies to foster the adoption and utilisation of telemedicine in the healthcare delivery. Such efforts could improve access to healthcare services, particularly during public health emergencies such as infectious disease outbreaks, while creating a more resilient and responsive healthcare system that better meets the needs of patients.