Effects of multi-channel consultation on e-healthcare demand: an empirical study Authors: Yanting Chen, Xinyu Hao, Ying Qian, Jingjing Liu Published: 2026 ABSTRACT BackgroundDiverse e-healthcare channels have emerged with the widespread use of e-healthcare platforms. In particular, the e-healthcare platforms start to provide both graphic-text consultation and video-voice consultation channels, where providers can perform instant or non-instantaneous consultations according to the conditions of the service providers and the patients. While prior research has extensively examined the impact of e-healthcare on offline medical demand or treated online consultation as a unified channel, limited attention has been paid to how different types of online consultation channels on the e-healthcare platforms affect e-healthcare demand. This study aims to address these gaps by investigating the effects of multi-channel consultation on e-healthcare demand.MethodsFirst, we consider extensive potential influencing factors. This study constructed a covariance-based structural equation model (CB-SEM) to investigate the key factors that may influence the e-healthcare demand with multi-channel consultations, including the availability of multi-channel services and prices, etc. Second, relying on the moderating method, we investigate the moderating effects of disease types and the provider’s title on the demand for e-healthcare services with multi-channel consultation. The structural equation model is used for data analysis.ResultsThere were 608 data samples from service providers collected from the “Wedoctor” website. Our results reveal several findings. Firstly, the existence of either a graphic-text consultation service (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=0.146, p<0.001$$\end{document}) or a video-voice consultation service (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=0.383, p<0.001$$\end{document}) for providers increases the demand for e-healthcare services. However, the prices for either graphic-text consultation service (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=-0.081, p<0.001$$\end{document}) or video-voice consultation service (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=-0.64, p<0.001$$\end{document}) reduce the e-healthcare service demand. Secondly, for the graphic-text consultation service, the e-healthcare demands for the patients with chronic diseases exhibit a greater increase compared with those with acute diseases \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:(\beta\:=-0.245, p<0.001$$\end{document}), yet this phenomenon appears insignificant for the video-voice consultation service. Thirdly, for the graphic-text consultation service type, the e-healthcare demand experiences a greater increase for providers with higher professional titles (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=0.136, p<0.001$$\end{document}). On the contrary, regarding the video-voice consultation service type, the e-healthcare demand increases more significantly when the professional titles of providers are low (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=-0.141, p<0.001$$\end{document}).ConclusionsOur findings offer managerial insights on service designs for e-healthcare platforms and enhance their demand by considering the possibilities of multi-channel services. These insights instruct how e-healthcare can be delivered more effectively. Background With the rapid development of information technology, the sophisticated functions that are available on e-healthcare platforms have promoted the usage of e-healthcare platforms (e.g., Teladoc Health and Doctor on Demand (US), wedoctor.com, haodf.com, and alihealth.cn (China)) [1, 54]. The effectiveness of time-saving and transportation-free features has attracted a loyal group of e-healthcare users. E-healthcare consultation is defined as the use of information and communications technologies (ICTs) by remote patients on the basis of instant and non-instant technologies to communicate with doctors and acquire healthcare services via websites that generally include information on consultations, records, community forums, and other information [3, 45]. The offline consultation model usually requires patients to make an appointment, register, visit the hospital, and wait in line. The experience for offline service can be poor, since the patients’ time has been wasted during waiting or commuting [37, 40]. Compared to offline healthcare organizations, e-healthcare platforms are more convenient and time-saving. Patients can consult remotely anytime, anywhere, avoiding queues and long waiting times. E-healthcare platforms help patients and providers effectively overcome time and space constraints [9, 17], which potentially improves patient experience and bridges the disparity in medical resources across geographical areas [16]. In addition, on e-healthcare platforms, patients have access to multi-dimensional information on infectious diseases, which can help reduce the severity of outbreaks of infectious diseases [24]. As a result, e-healthcare platforms have gained considerable popularity and visibility, especially as patients have been reluctant to visit hospitals due to time or distance constraints [2]. In e-healthcare platforms, traditional online consultation functions are purely text-based with limited information. The simple description of symptoms lacks depth and precision, which impedes the provider from offering an efficient diagnosis and recommendations. With the popularization of e-healthcare, patients’ increasing demand for convenient and instant healthcare has prompted the emergence of diverse online consultation channels [48]. In addition to traditional text-based consultations, more online platforms have begun to provide multi-channel consultations such as graphic-text consultation and video-voice consultation [8]. These channels can provide patients with more accurate advice and increase the flexibility of e-healthcare. Providers can perform instant or non-instantaneous consultations according to the conditions of the service providers and the patients. The patients typically use e-healthcare for two main reasons. First, patients seek to gain more information about their condition before an offline in-person consultation or hope to receive online treatment to avoid in-person visits. Second, after an in-hospital consultation, patients may wish to obtain follow-up online treatment or seek e-healthcare assistance if they are dissatisfied with the results of the hospital’s treatment [36]. Research has shown that the growth of e-healthcare would reduce the need for offline providers [50], enhance providers’ reputation [23, 26], and strengthen interaction between providers and patients [26]. Wu and Lu [53] explored how online channel usage affects offline channels, and how the channel effects change with doctors’ online and offline reputation. Bavafa et al. [4] used an analytical model to indicate that electronic visits (e-visits) are effective in reducing the need for traditional face-to-face visits by patients. The existing research explored the impact of different online consultation services on offline appointments [7]. However, there remains a research gap regarding how the diversification of online consultation channels themselves—particularly the existence of graphic-text consultation and video-voice consultations—affects patients’ demand for online healthcare services. Exploring the differentiated effects of multi-channel consultations can deepen the understanding of patient behavioral mechanisms in the e-healthcare context, revealing how the characteristics of different channels interact with patients’ medical conditions and physician attributes to influence patients’ behavior. As e-healthcare platforms invest heavily in multi-channel capabilities, understanding which channels are more effective for both patients and physicians can guide the platforms to optimize the designs, pricing strategies, and the allocation of physician resources. In the research field of e-healthcare platforms, scholars have explored the core factors affecting the number of consultations by providers. Research indicates that the individual characteristics of providers and patients are key variables determining the number of consultations on e-healthcare platforms. Cao et al. [6] adopted an empirical research approach, extensively analyzing the influence of service quality and electronic word-of-mouth (eWOM) on patients’ decision-making processes by collecting and analyzing substantial patient decision-making data. Their findings have laid a solid foundation for subsequent exploration of the interactive relationship between patient behavior and medical services. Lu et al. [34] further expanded the research on this basis, focusing on the role of electronic word-of-mouth (eWOM) published by patients on e-healthcare platforms in patients’ decision-making. The research found that positive eWOM posted by patients regarding providers’ online service quality has a positive impact on providers’ online consultation demand. There are literatures focusing on patients’ intention to use in teleconsultation. Ray et al. pointed out the trust, perceived risk, contamination avoidance, psychological predispositions, habitual behavior, and perceived severity of illness, which play a significant role in the engagement of users to virtual healthcare services [42]. He et al. revealed that perceived usefulness and perceived ease of use are the primary drivers affecting the behavioral intention of older adults to use telemedicine, followed by cost value, system quality, trust, and self-efficacy. In contrast, Technological anxiety has an adverse effect on the older adults’ intention to use telemedicine [20]. Tse et al. indicate that the government subsidy also has an impact on the intention to use the teleconsultations [45]. Osei et al. suggest that the technological advancements have a positive impact on the intention to use [38]. Vannelli et al. conduct a survey to investigate Continuance Intention for Telehealth Visits in Children’s Hospitals. They show that the variable information quality positively influenced the variables continuance intention and perceived usefulness and that perceived usefulness positively influenced continuance intention [46]. Most of the current research focuses on the impact of patients’ individual characteristics (e.g., gender, age, health status, etc.) on their e-healthcare demand [58]. Few studies have delved into how patients’ preferences for choosing multi-channel consultations interact with disease type. Different types of diseases have different natures, severities, and treatment plans, which can directly affect patients’ choice and demand for multi-channel consultation. By exploring the role of disease types, we can understand the needs of patients with different diseases and then provide them with more accurate services. E-healthcare providers, who are fundamental to e-healthcare, continue to be a key focus of research [26]. Provider-specified characteristics have a significant impact on patients’ willingness to use e-healthcare services. Several studies have explored how factors such as provider ratings, titles, and hospital ranks affect providers’ online and offline appointment bookings. Liu et al. obtained online appointment data from e-healthcare platforms, developed a model to analyze the factors influencing appointment numbers, and examined the impact of various factors on appointment numbers. The study found that factors such as provider ratings, hospital rank, and the provider’s professional title have an impact on the provider’s in-person appointment number [32]. Fan et al. [13] conducted a study on the factors influencing providers’ in-person consultation numbers. They used a difference-in-differences (DID) approach to test the impact of providers offering online consultation services on their in-person consultation numbers [23]. The empirical results show that offering online consultation services positively influences providers’ in-person consultation demand. Li et al. [30] found that patients’ trust in the online platform, the competence of the platform’s providers, and the integrity of the platform’s providers positively influence patients’ willingness to use e-healthcare services. Existing studies have mostly focused on factors such as physician ratings and hospital rankings. This study specifically focuses on how the characteristic of the provider’s title plays a role in multi-channel consultations. By analyzing the influence of doctors with different titles on patients’ choice of online consultation channels, this study reveals that the influence of a provider’s professional title on multi-channel consultation is different. By exploring the moderating role of providers’ professional titles, different services available from providers with different professional titles can enhance their e-healthcare demand. The role of disease type and provider characteristics as moderators is grounded in task-technology fit theory and patient engagement models. Specifically, the nature of the disease dictates the extent to which e-healthcare can substitute or complement offline visits: chronic conditions typically require continuous monitoring and follow-up, making them more sensitive to the benefits of e-health services, whereas acute conditions rely heavily on physical examinations, thus weakening the effect [25, 42]. Concurrently, provider characteristics (e.g., professional titles) signal clinical authority and reputation, which moderate patient trust and the perceived value of online consultations. For instance, lower-tier providers may experience weaker effects from service integration due to lower patient acceptance [23]. Therefore, both variables are posited to moderate the outcomes in this context significantly. Through e-healthcare platforms, patients can choose multi-channel consultations such as graphic-text consultation and video-voice consultation, significantly enhancing accessibility, particularly in special circumstances like living in remote areas. However, with the diversification of platform user groups and the disparities in providers’ backgrounds, professional titles, specialties, and patients’ individual needs, a single form of consultation service can no longer fulfil the requirements of all patients. By deploying a discrete choice model for empirical analysis, Snoswell et al. [43] find that for patients with various diseases, costs, and waiting times, etc., the different preferences for online consultation types exist. Li et al. [29] developed an intelligent matching algorithm based on patient condition descriptions, medical history, and physicians’ expertise. Empirical results show that customized online consultation channel recommendations can significantly improve consultation efficiency and patient satisfaction. For chronic diseases, Ramdani et al. [41] observed the superiority of optimally combining a multi-channel service design framework (including text/graphics, video, and regular follow-ups). Liu et al. [31] empirically find that the diversity of the patients leads to different demands for online service channels. Therefore, it is of importance to first identify the key elements which play roles in multi-channel e-healthcare services, and then clarify how these key elements influence the e-healthcare demand with multi-channel consultation. Therefore, how to provide differentiated consultation functions based on varying physician characteristics and patient needs has become a critical challenge for reoptimizing e-healthcare platform services. However, three key research gaps remain in addressing this challenge. First, while prior studies have examined e-healthcare’s impact on offline channels, the influence of multiple types of online consultation functions (graphic-text vs. video-voice) on e-healthcare demand itself remains underexplored. Second, the differentiated demand patterns for online consultation channels among patients with different disease types (e.g., acute vs. chronic conditions) have not been systematically verified. Third, although patient and provider characteristics are recognized as important factors, their moderating roles in the relationship between multi-channel consultations and online demand are not well understood. In our study, we collected data on a prominent e-healthcare platform in China, which includes providers’ titles, specialties, and the provision of online consultation services. We deploy a structural equation model (SEM) to analyze the impact of multi-channel online consultation functions on the e-healthcare demand. Furthermore, we explore the different moderating factors for the main effects of online consultation functions on the e-healthcare service providers (i.e., the providers’ professional title and disease type). Our study has addressed these research gaps and has provided three key contributions to the literature on e-healthcare platforms and multi-channel service design. First, while prior research has primarily focused on the impact of e-healthcare on offline demand or treated online consultation as a unified channel, our work zooms into the more specified channels for e-healthcare (e.g., graphic-text and video-voice consultation), and provides empirical evidence on how these multi-channel consultation functions affect e-healthcare demand. This addresses the underexplored area of how differentiated online services influence the demand for e-healthcare services. Second, our research delves into how patients’ preferences for choosing multi-channel consultations interact with disease type. By exploring the role of disease types, we systematically verify the differentiated demand patterns between acute and chronic conditions, thereby helping to understand the needs of patients with different diseases and providing a foundation for more accurate services. Third, we extend the understanding of provider-specified characteristics by examining the moderating role of the provider’s professional title in the relationship between consultation channels and demand. Our findings reveal how the characteristic of the provider’s title plays a role in multi-channel consultations, offering insights into how different services available from providers with different professional titles can be optimized to enhance their e-healthcare demand. The specific three objectives are (1) preliminarily determining the factors that affect multi-channel consultation on e-healthcare demand, (2) building a theoretical model to clarify the interaction between the influencing factors, and (3) empirically validating the mechanism of action among the various factors. The theoretical model is conducted by the following factors influencing multi-channel consultation on e-healthcare demand. The remaining sections of this paper are organized as follows. Section Theory foundation and research hypotheses provides the theoretical foundation and research hypotheses. Section Materials and methods describes the data collection and variable measurement. Section Data analysis and results presents the data analysis and the hypothesis testing results. Finally, we include discussions on key findings, implications, and future research directions in Sect. Discussion. Theory foundation and research hypotheses Based on Information Richness Theory [10, 57], the graphic-text consultation service provides the lowest level of information richness, because patients and providers need to communicate in the form of discontinuous texts. Video-voice consultation service offers a medium level of information richness, as it provides non-stop communications for patients and providers over a certain period of time. Holden and Karsh [21] pointed out that in studying health information technology, “perceived usefulness” should be instantiated as improvements in health outcomes or enhancements in the convenience of healthcare access. This implies that different types of online services (such as graphic-text or video-voice consultations) may differentially influence patients’ channel choices and ultimately affect the demand [11]. Therefore, the variables we consider are: multi-channel consultation services, consultation price, disease type, and the provider’s title. Multi-channel consultation features refer to the diverse technological modalities (e.g., text, voice, or video) that facilitate patient-provider interaction. When e-health platforms offer comprehensive consultation channels, patients perceive higher accessibility and flexibility, which significantly reduces barriers to seeking care [51]. Previous studies have demonstrated that the richness of communication channels can enhance the e-healthcare demand [22]. Furthermore, the availability of these features allows providers to address patient needs more efficiently, leading to an increase in their total service volume. Therefore, we propose the following hypotheses: Hypothesis 1 Graphic-text consultation service will increase e-healthcare demand for providers. Hypothesis 2 Video-voice consultation service will increase e-healthcare demand for providers. In e-healthcare, price is a critical factor. Yuan et al. [56] explicitly identified “price value” as a core factor, demonstrating that users’ evaluation of “price value” significantly influences their continuance intention. This provides theoretical support for directly incorporating service pricing into our model. We can regard the pricing of different channels as external variables that affect patients’ choices, which directly influence patients’ demand for online healthcare services. Consultation price represents the monetary cost incurred by patients to obtain medical advice online. According to demand theory, price is a critical determinant of consumer choice; higher service costs often lead to a decrease in demand as patients weigh the perceived benefits against the expenditure. In the e-health context, pricing strategy not only reflects the provider’s professional value but also acts as a competitive signal. Extant research suggests that reasonable pricing can lower the “entry threshold” for patients, especially those sensitive to costs, thereby expanding the provider’s patient base [55]. Consequently, the following hypotheses are formulated: Hypothesis 3 The price of graphic-text consultation will decrease e-healthcare demand for providers. Hypothesis 4 The price of video-voice consultation will decrease e-healthcare demand for providers. Moderation describes a situation in which the relationship between two constructs is not constant but depends on the values of a third variable, referred to as a moderator variable. In a moderation model, the moderator variable influences the direction or strength of the relationship between an independent and a dependent variable [19]. The role of disease type and provider characteristics as moderators is grounded in task-technology fit theory and patient engagement models [23, 25, 42]. Therefore, Disease types potentially moderate the e-healthcare demand with multiple consultation channels. Therefore, we conceptualize disease type as a moderator that alters the strength of the relationship between channel availability and e-healthcare demand. Disease type (chronic/acute) serves as a critical moderating variable influencing providers’ online consultation demand. Kim and Park [27] indicated that patients’ perceptions of disease influence their health-related behavior decisions. This provides a basis for this study to explore how disease type moderates patients’ choice of different online service channels. Patients with chronic conditions may be more inclined to consistently use online graphic-text or video-voice consultations, thereby contributing to a stable online consultation demand; whereas patients with acute conditions, even if they initiate consultations online, are more likely to have their needs redirected to offline healthcare services, thus exerting a relatively limited effect on increasing providers’ online consultation demand. We consider the types of diseases diagnosed in patients, as various medical conditions necessitate distinct management approaches. Generally, chronic diseases (e.g., asthma, cancer, diabetes, and psoriasis) require planned, periodic interactions and continuous follow-up with healthcare providers. Unlike acute illnesses, chronic conditions often cannot be fully cured through short-term treatment, and their symptoms tend to persist over extended periods [33]. For certain types of diseases, such as chronic diseases and minor diseases, graphic-text consultation may be more attractive because patients with such diseases have a higher demand for a simple and quick form of consultation. Therefore, differences in disease types may impact patients’ e-healthcare demand to use graphic-text consultation, thereby increasing the e-healthcare demand for providers. For acute diseases, patients usually require immediate, face-to-face medical intervention, in which case the demand for video-voice consultation service is relatively low, as treatment of acute diseases usually relies on emergency visits. For chronic diseases, patients usually require long-term management and follow-up, which makes video-voice consultation more attractive for chronic diseases. As a result, the impact of the video-voice consultation service on e-healthcare demand is weaker among healthcare providers focused on acute conditions and stronger among those focused on chronic conditions. Therefore, we investigate how disease type moderates the relationship between graphic-text/video-voice consultation and e-healthcare demand, and propose the following hypothesis: Hypothesis 5a The effects of graphic-text consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases. Hypothesis 5b The effects of video-voice consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases. Provider professional title serves as a critical moderating variable influencing providers’ online consultation demand. Empirical research on online trust by Gefen et al. indicates that institutional cues (such as qualifications and professional titles) are central to establishing initial trust and reducing users’ perceived risk [15]. This provides a basis for this study to explore how physician titles moderate patients’ choice of different online service channels. The professional title of a provider signifies their clinical experience and seniority, which may lead to variations in demand after offering online consultation services. A significant challenge in e-health is information asymmetry, which is often more pronounced than in other industries [39]. Consequently, patients in e-health environments tend to seek treatment from providers with higher professional titles. This behavioral tendency leads to the “cold start” problem for less experienced providers, which may, in turn, exert a negative impact on their online consultation volume [18]. Furthermore, e-health providers with lower titles typically attract fewer patients through online medical communities [53]. In contrast, providers with higher titles usually possess a well-established reputation, enabling them to attract a larger patient base [12]. Therefore, patients may exhibit a stronger preference for these high-titled providers when selecting graphic and text-based consultation services. However, the situation may be different for the video-voice consultation service. Providers with higher titles have very busy daily routines, especially with high clinical workloads, and video-voice consultation requires longer real-time interactions, which may limit their ability to accept more patients. In contrast, the non-real-time nature of graphic-text consultation allows providers to respond according to their schedules. So, they may be more inclined to provide traditional face-to-face medical service. Patients may feel that providers with higher titles are less likely to consult via video-voice, as they prefer to provide more direct and specialized treatment. As a result, the impact of video-voice consultation on e-healthcare demand is weaker among providers with higher titles. In contrast, providers with lower titles may be more reliant on video-voice consultations because they may face less demand for face-to-face visits, and they may have more time for instant online consultations. Therefore, we investigate how provider’s title moderates the relationship between graphic-text/video-voice consultation and e-healthcare demand, and propose the following hypothesis: Hypothesis 6a Provider’s title will be a significant moderator, such that the effects of graphic-text consultation service on the e-healthcare demand will be stronger for providers with higher titles. Hypothesis 6b Provider’s title will be a significant moderator, such that the effects of video-voice consultation service on the e-healthcare demand will be weaker for providers with higher titles. A summary table of the research hypotheses is shown in Table 1: Table 1Research HypothesesNo.HypothesesDetails Description1H1Graphic-text consultation service will increase e-healthcare demand for providers.2H2Video-voice consultation service will increase e-healthcare demand for providers.3H3The price of graphic-text consultation will decrease e-healthcare demand for providers.4H4The price of video-voice consultation will decrease e-healthcare demand for providers.5H5aThe effects of graphic-text consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases.6H5bThe effects of video-voice consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases.7H6aProvider’s title will be a significant moderator, such that the effects of graphic-text consultation service on the e-healthcare demand will be stronger for providers with higher titles.8H6bProvider’s title will be a significant moderator, such that the effects of video-voice consultation service on the e-healthcare demand will be weaker for providers with higher titles. Table 1Research HypothesesNo.HypothesesDetails Description1H1Graphic-text consultation service will increase e-healthcare demand for providers.2H2Video-voice consultation service will increase e-healthcare demand for providers.3H3The price of graphic-text consultation will decrease e-healthcare demand for providers.4H4The price of video-voice consultation will decrease e-healthcare demand for providers.5H5aThe effects of graphic-text consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases.6H5bThe effects of video-voice consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases.7H6aProvider’s title will be a significant moderator, such that the effects of graphic-text consultation service on the e-healthcare demand will be stronger for providers with higher titles.8H6bProvider’s title will be a significant moderator, such that the effects of video-voice consultation service on the e-healthcare demand will be weaker for providers with higher titles. Research Hypotheses We build an empirical research model as shown in the following Fig. 1. Fig. 1Conceptual Model of the Effects of Multi-channel Consultation on e-Healthcare Demand Conceptual Model of the Effects of Multi-channel Consultation on e-Healthcare Demand Materials and methods Study setting and data collection Our study context resides in one of the pioneering e-healthcare platforms in China (Wedoctor.com). A screenshot of a provider’s homepage on the e-healthcare platform has been provided in Fig. 2. On the e-healthcare platform, patients can view the provider’s qualifications, areas of expertise, and patient reviews. When using online consultation, patients can choose to have a graphic-text consultation or a video-voice consultation, depending on their condition. Patients need to briefly describe their symptoms or medical history and upload relevant examination reports, photos, or images of symptoms, and the provider will make a diagnosis and recommendation based on the patient’s description and relevant information. At the end of the consultation, the provider will provide treatment recommendations based on the symptoms, issue an e-prescription, or recommend further tests. Fig. 2Screenshot of a provider’s homepage on the e-Healthcare platform Screenshot of a provider’s homepage on the e-Healthcare platform Graphic-text consultation The consultation process for a patient with a graphic-text consultation is shown in Fig. 3. The first step is to submit detailed symptoms and clearly describe the symptoms of the disease and upload photos related to the condition, such as photos of the affected area or examination reports, to provide more basis for the doctor’s diagnosis. Next, the patient pays for the service online according to the different services. Finally, the doctor will combine the patient’s symptom description, photos, and reports, and conduct a professional diagnosis and assessment within 24 hours. Apparently, the doctor does not have to respond instantly. A delayed reply within 24 hours is allowed. After diagnosis and recommendation, if the patient’s problem has been clarified and no follow-up treatment is required, the consultation process will be completed. If the doctor believes that the problem cannot be effectively solved online. Then the patient will be strongly suggested to visit a medical institution for further comprehensive examination, diagnosis, and treatment. Fig. 3Graphic-text consultation workflow Graphic-text consultation workflow Video-voice consultation The consultation process for a patient with a video-voice consultation is shown in Fig. 4. First, the patients need to set up an appointment with the doctor. Secondly, a brief description of the symptoms and supporting materials will be uploaded. And the doctor will begin the 15-minute-long appointment online. During the appointment, the communication between the patient and the doctor is uninterrupted. During the process, patients can add further details of their condition, and the doctor will combine the pre-submitted documents with the real-time communication content to diagnose and give diagnosis and treatment recommendations. Similarly, if the patient’s problem has been clarified and does not require follow-up treatment, the consultation process will be completed. If the doctor assesses that the problem has not been effectively solved and cannot be treated online, the patient will need to visit an offline medical institution for more in-depth examination, diagnosis, and targeted treatment. Fig. 4Video-voice consultation workflow Video-voice consultation workflow Variable measurement In this study, we refer to the factors influencing the e-healthcare demand for providers selected from previous literature and combine them with the characteristics of objective data collected from the e-healthcare platform of “Wedoctor” to select and measure the variables for the proposed hypotheses. Note that all variables in our study are observed, which are not latent; therefore, preliminary analyses such as confirmatory factor analysis (CFA) were not required [5]. And the specific measurement indexes of each variable are as follows: E-healthcare demand This variable refers to the total number of online consultations. Usually, it is the number of online consultation services conducted through internet platforms where patients communicate with doctors or health experts via video, voice, or text to obtain health advice, diagnosis, or treatment options [52]. The e-healthcare demand is measured by the number of consultations on the provider’s homepage. Given the wide numerical range of this variable, it was transformed using the natural logarithm (ln) in subsequent modeling to stabilize the analysis and enhance the robustness of the results [35]. Graphic-text consultation This variable refers to whether or not the provider provides a graphic-text consultation service. Graphic-text consultation is a process in which patients describe their symptoms, medical history, and other information through text. Patients upload pictures of their symptoms (e.g., skin diseases, wounds, etc.) or other relevant medical examination reports to help the provider diagnose more accurately [10, 57]. And the provider responds in text based on the symptoms, pictures, or videos provided by the patient with suggestions, diagnosis, or advice to the patient to do further examination. This variable takes the value of 1 if the provider provides a graphic-text consultation service, and takes the value of 0 if she/he does not provide this service. Video-voice consultation This variable refers to whether the provider provides the video-voice consultation service. For video-voice consultation, the patient communicates with the e-healthcare service provider via voice and video call to discuss symptoms, conditions, and treatment recommendations [10, 57]. The provider can observe the patient’s symptoms and conduct detailed questioning and examination via voice and video. This variable takes the value of 1 if the provider provides video-voice consultation service, and takes the value of 0 if she/he does not provide this service. The price of graphic-text consultation This variable refers to the fee that the patient has to pay for the graphic-text consultation service. Graphic-text consultation is a way for a patient to remotely consult with a provider through text, pictures, and other graphic materials (e.g., photos of symptoms, medical reports, etc.). The price of a graphic-text consultation is the cost of the service, which is the amount the patient will need to pay for a graphic-text consultation [56]. The price may vary depending on the provider’s qualifications, the platform’s policies, the number of communications, and other factors. The price of graphic-text consultation is measured by the price of graphic-text consultation displayed on the provider’s homepage. The price of video-voice consultation This variable refers to the fee that the patient pays for video and voice consultation services. Video-voice consultation is a form of online consultation in which the patient communicates with the doctor in real-time via video and voice call to describe symptoms and conditions and receive a diagnosis and advice from the doctor. This variable indicates the price of the service, which is the amount the patient would have to pay for a video-voice consultation [56]. The price may vary depending on the provider’s qualifications, platform policies, the duration of communication, and other factors. The price of video-voice consultation is measured by the price of video-voice consultation displayed on the provider’s homepage. Disease type This variable refers to different categories of diseases, categorized according to different medical criteria, pathological features, causes of pathogenesis, or clinical manifestations. When measuring the disease type, this study classified the diseases into two categories [26]. Acute diseases such as acute enteritis and febrile convulsions would take the value of 1. Moreover, diabetes, coronary heart disease, hypertension, and other chronic diseases would take the value of 0. Provider’s title This variable refers to the title of the provider in the healthcare field [15]. When measuring the provider’s title, this study chose the data of 0 and 1 to represent the provider’s title, with a value of 1 for the senior and associate titles, such as chief physician and associate chief physician, and a value of 0 for the junior and middle titles, such as attending physician and resident. Table 2 shows the details concerning the measurements of all the variables. Table 2Variable measurementVariableDescriptionMeasurementIndependent VariableE-healthcare demandThe number of online consultations for providersShown on the provider’s homepage of the platformGraphic-text ConsultationWhether or not the provider provides graphic-text consultation service1: Provide graphic-text consultation service0: Not provide graphic-text consultation serviceDependent VariableVideo-voice ConsultationWhether or not the provider provides video-voice consultation service1: Provide video-voice consultation service0: Not provide video-voice consultation serviceThe price of Graphic-text ConsultationThe price of graphic-text consultation for providersShown on the provider homepage of the platformThe price of Video-voice ConsultationThe price of video-voice consultation for providers.Shown on the provider homepage of the platformModerator VariableDisease TypeDifferent categories of diseases1: Acute disease0: Chronic diseaseProvider’s titleThe title of the provider in the healthcare field1: Chief physician or associate chief physician0: Otherwise Table 2Variable measurementVariableDescriptionMeasurementIndependent VariableE-healthcare demandThe number of online consultations for providersShown on the provider’s homepage of the platformGraphic-text ConsultationWhether or not the provider provides graphic-text consultation service1: Provide graphic-text consultation service0: Not provide graphic-text consultation serviceDependent VariableVideo-voice ConsultationWhether or not the provider provides video-voice consultation service1: Provide video-voice consultation service0: Not provide video-voice consultation serviceThe price of Graphic-text ConsultationThe price of graphic-text consultation for providersShown on the provider homepage of the platformThe price of Video-voice ConsultationThe price of video-voice consultation for providers.Shown on the provider homepage of the platformModerator VariableDisease TypeDifferent categories of diseases1: Acute disease0: Chronic diseaseProvider’s titleThe title of the provider in the healthcare field1: Chief physician or associate chief physician0: Otherwise Variable measurement 1: Provide graphic-text consultation service 0: Not provide graphic-text consultation service 1: Provide video-voice consultation service 0: Not provide video-voice consultation service 1: Acute disease 0: Chronic disease 1: Chief physician or associate chief physician 0: Otherwise Data analysis and results This study employed path analysis, a specific form of structural equation modeling, to systematically examine the direct and complex mediating effects of patient characteristics, service types, and their pricing on E-healthcare Demand. This analytical approach estimates hypothesized relationships among multiple variables based on their covariance matrix as a whole and effectively addresses potential issues of multicollinearity, serving as a robust tool for analyzing such multifaceted causal relationships [14]. Moreover, this approach allows for the explicit examination of both direct relationships and indirect, mediated relationships among variables. Path analysis and SEM have been widely adopted in health services and technology adoption research [44, 49]. While a full structural equation model comprises a measurement model and a structural model, the present study focuses on path analysis with observed variables, estimating the system of direct and indirect effects within the structural model. As all variables in our model are observed, no preliminary analyses such as confirmatory factor analysis (CFA) were required [5]. The analysis proceeds directly to path analysis, estimating the relationships among these manifest variables. The data were analyzed using IBM SPSS Statistics v26.0 for preliminary analyses. For data analysis, this study employed covariance-based structural equation modeling (CB-SEM) using IBM AMOS v26.0. CB-SEM was chosen over partial least squares SEM (PLS-SEM) because our primary objective is theory testing and confirmation [19, 28, 47], which aligns with the strengths of covariance-based approaches, rather than prediction or exploration. Descriptive statistics The data used in this paper are all from the platform of ‘Wedoctor’, collected in December 2024. The collection steps are as follows: Firstly, the Python crawler is used to crawl the data samples of the providers on the website of ‘Wedoctor’. Then, we leave out the providers who don’t have online consultations and the number of visits to their pages is zero, and finally obtained a total of 608 data samples of providers who meet the conditions. The descriptive statistics are shown in Table 3. Table 3Descriptive StatisticsVariableObsMeanSDMinMaxE-healthcare demand6086176.08128179.9241.000395,492Graphic-text Consultation5901.0000.0001.0001.000Video-voice Consultation3071.0000.0001.0001.000The price of Graphic-text Consultation590166.339159.2336.0001000.000The price of Video-voice Consultation307346.518327.42420.0002000.000Disease Type6080.6740.4690.0001.000Provider’s title6080.8110.3920.0001.000 Table 3Descriptive StatisticsVariableObsMeanSDMinMaxE-healthcare demand6086176.08128179.9241.000395,492Graphic-text Consultation5901.0000.0001.0001.000Video-voice Consultation3071.0000.0001.0001.000The price of Graphic-text Consultation590166.339159.2336.0001000.000The price of Video-voice Consultation307346.518327.42420.0002000.000Disease Type6080.6740.4690.0001.000Provider’s title6080.8110.3920.0001.000 Descriptive Statistics The descriptive statistics table shows that the e-healthcare demand varies greatly among providers, with 97.04% of providers having provided graphic-text consultation and 50.49% having provided video-voice consultation. The mean value for disease type is 0.674, indicating that the number of providers treating acute diseases is slightly higher than the number of providers treating chronic diseases. The mean value of the provider’s title is 0.811, indicating that there are more providers with high titles providing online consultation services. The mean value of hospital rank reaches 0.99, indicating that more providers who provide online consultations are from high-ranking hospitals. Structural equation model and hypothesis testing Fig. 5Structural equation model and path. Note: *** means significant at the 0.001 level, R2 is the squared multiple correlations Structural equation model and path. Note: *** means significant at the 0.001 level, R2 is the squared multiple correlations The structural equation model path is shown in Fig. 5. The results support all four direct effect hypotheses. Specifically, offering graphic-text consultation significantly increases the e-healthcare demand (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=0.146, p<0.001$$\end{document}), confirming H1. An even stronger positive effect emerges for video-voice consultation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=0.383, p<0.001$$\end{document}, providing support for H2. Turning to pricing effects, both consultation channels exhibit negative price elasticities, though with markedly different magnitudes. The price of graphic-text consultation shows a modest negative effect on demand \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=-0.081, p<0.001$$\end{document}, validating H3. In contrast, video-voice consultation price exerts a substantially larger negative influence \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=-0.64, p<0.001$$\end{document}, supporting H4. We first investigate the moderating role of disease type, which is at the patient level. The interaction term between graphic-text consultation and disease type is significant (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=-0.245, p<0.001)$$\end{document}, indicating that the disease type moderates the relationship between graphic-text consultation and e-healthcare demand. Since \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:\:$$\end{document}is negative, the positive effect of graphic-text consultation on e-healthcare demand is weaker for acute diseases than for chronic diseases, supporting H5a. However, the interaction term between video-voice consultation and disease type is insignificant (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=-0.004, p=0.829$$\end{document}), suggesting no moderating effect of disease type on the relationship between video-voice consultation and e-healthcare demand. We then examine the moderating effects of professional title, which is at the provider level. The interaction term between graphic-text consultation and professional title is significant (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=0.136, p<0.001)$$\end{document}, indicating that the professional title moderates the relationship between graphic-text consultation and e-healthcare demand. Since \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} is positive, the positive effect of graphic-text consultation on e-healthcare demand becomes stronger as the professional title advances from the lower title to the higher title, supporting H6a. Moreover, the interaction term between video-voice consultation and professional title is also significant (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=-0.141, p<0.001$$\end{document}). However, a negative \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} indicates that the positive effect of video-voice consultation on e-healthcare demand becomes weaker as the professional title changes from the lower title to the higher title, supporting H6b. The structural equation model path coefficient results are shown in Table 4. Table 4Structural equation model path coefficient resultsHypothesesVariable relationshipStandardized path coefficientStandard errorCritical ratioP valueH1Graphic-text Consultation\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand0.1460.2187.369***H2Video-voice Consultation\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand0.3830.7419.338***H3The price of Graphic-text Consultation\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand-0.0810.037-4.091***H4The price of Video-voice Consultation\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand-0.640.037-32.307***H5aGraphic-text Consultation* Disease Type\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand-0.2450.037-12.368***H5bVideo-voice Consultation * Disease Type\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand0.0040.0780.2160.829H6aGraphic-text Consultation* Provider’s title\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand0.1360.0916.861***H6bVideo-voice Consultation * Provider’s title\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand-0.1410.037-7.102*** Table 4Structural equation model path coefficient resultsHypothesesVariable relationshipStandardized path coefficientStandard errorCritical ratioP valueH1Graphic-text Consultation\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand0.1460.2187.369***H2Video-voice Consultation\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand0.3830.7419.338***H3The price of Graphic-text Consultation\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand-0.0810.037-4.091***H4The price of Video-voice Consultation\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand-0.640.037-32.307***H5aGraphic-text Consultation* Disease Type\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand-0.2450.037-12.368***H5bVideo-voice Consultation * Disease Type\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand0.0040.0780.2160.829H6aGraphic-text Consultation* Provider’s title\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand0.1360.0916.861***H6bVideo-voice Consultation * Provider’s title\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\to\:$$\end{document}E−healthcare Demand-0.1410.037-7.102*** Structural equation model path coefficient results Table 5 is a summary table of hypothesis tests, listing the results of all hypothesis tests. Table 5Summary of Hypothetical ResultsNo.HypothesesDetails DescriptionResults1H1Graphic-text consultation service will increase e-healthcare demand for providers.√2H2Video-voice consultation service will increase e-healthcare demand for providers.√3H3The price of graphic-text consultation will decrease e-healthcare demand for providers.√4H4The price of video-voice consultation will decrease e-healthcare demand for providers.√5H5aThe effects of graphic-text consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases.√6H5bThe effects of video-voice consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases.×7H6aProvider’s title will be a significant moderator, such that the effects of graphic-text consultation service on the e-healthcare demand will be stronger for providers with higher titles.√8H6bProvider’s title will be a significant moderator, such that the effects of video-voice consultation service on the e-healthcare demand will be weaker for providers with higher titles.√ Table 5Summary of Hypothetical ResultsNo.HypothesesDetails DescriptionResults1H1Graphic-text consultation service will increase e-healthcare demand for providers.√2H2Video-voice consultation service will increase e-healthcare demand for providers.√3H3The price of graphic-text consultation will decrease e-healthcare demand for providers.√4H4The price of video-voice consultation will decrease e-healthcare demand for providers.√5H5aThe effects of graphic-text consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases.√6H5bThe effects of video-voice consultation service on the e-healthcare demand will be weaker for acute diseases, compared with chronic diseases.×7H6aProvider’s title will be a significant moderator, such that the effects of graphic-text consultation service on the e-healthcare demand will be stronger for providers with higher titles.√8H6bProvider’s title will be a significant moderator, such that the effects of video-voice consultation service on the e-healthcare demand will be weaker for providers with higher titles.√ Summary of Hypothetical Results Discussion Principal findings and interpretations Through the Structural Equation Modelling (SEM) method, we have identified various factors influencing e-healthcare demand, which allows us to evaluate the impact of different consultation services and moderators on the demand for e-healthcare providers. From Hypotheses H1 and H2, our study finds that graphic-text consultation services and video-voice consultation services both have a significant positive impact on the e-healthcare demand. For managerial operations, this suggests that by providing diversified online consultation services, providers are able to better meet patients’ demand and thus increase the number of online consultations. This is true because graphic-text consultation services help providers respond conveniently to patients’ inquiries through text and picture communication, while video-voice consultation services respond to patients instantly through more interactive real-time video exchanges, thus attracting more patients to online consultation. Therefore, the managerial insights indicate that hospitals or medical institutions should encourage providers to provide various types of online consultation services, which not only help to improve the efficiency of providers but also increase the convenience and satisfaction of patients. From Hypotheses H3 and H4, we see that the price of graphic-text and video-voice consultation had a significant negative impact on e-healthcare demand. This suggests that higher consultation prices may inhibit patients’ willingness to choose online consultations, which in turn reduces the e-healthcare demand. The managerial insights suggest that the prices for various channels in e-healthcare should be properly designed. From Hypotheses H5a and H5b, our exploration of the moderating factors revealed that the effects of the graphic-text consultation service on the e-healthcare demand will be weaker for acute diseases. This may be due to the fact that acute illnesses tend to be characterized by a high degree of urgency and the need for timely face-to-face management, and patients are more inclined to seek quick and direct medical interventions. For patients with chronic diseases, they may prefer the convenient and flexible graphic-text consultation service, so the e-healthcare demand for chronic disease providers is more significantly facilitated by graphic-text consultation services. The managerial insights indicate that the e-healthcare service providers may further improve or even enrich the available channels, especially for chronic diseases. From Hypotheses H6a and H6b, we also find that the provider’s title is a significant moderator. Specifically, the effects of the graphic-text consultation service on the e-healthcare demand will be stronger for providers with higher titles. In contrast, the effects of the video-voice consultation service on the online demand will be weaker for providers with higher titles. Higher professional title providers may be better able to engage and increase e-healthcare demand due to their professional background, academic credentials, and authority. Patients may be more inclined to choose providers with higher titles for graphic-text consultation, as they believe that higher title providers may have richer knowledge and experience and can provide more reliable service. Higher professional title providers usually have busier work schedules, and they do not have enough time to provide enough video-voice consultations. Such providers usually prefer offline healthcare, focusing on complex and difficult cases, so video consultations may not be a priority for them. In contrast, lower title providers may have more flexible work schedules, so they have more time for video consultations. The managerial insights indicate that the e-healthcare service providers should find a balance between the provider’s titles and service availability among multi-channel services. Existing literature investigates the intention to use in e-healthcare extensively. With the development of technologies, more forms, especially more consultation channels, have become available in the e-healthcare platform. However, the investigation on the impact of multi-channel consultations on e-healthcare demand and core features in the design of multi-channel consultations is still absent. From this perspective, our findings extend the current research by providing theoretical support in designing multi-channel consultations in e-healthcare platforms. In particular, our findings indicate that the launch of the multi-channel consultation and its corresponding price have a significant effect on the e-healthcare demand. Patient’s disease type and the e-healthcare provider’s title have significant heterogeneous roles in the influence of multi-channel consultation on the e-healthcare demand. These findings provide efficient managerial insights for e-healthcare service providers with multi-channel consultations to improve the demand for e-healthcare. Theoretical implications and practical implications This study makes several contributions to prior literature on multi-channel consultation, healthcare operations management, and e-healthcare design. First, our study is one of the first to explore the impacts of multi-channel consultations on e-healthcare demand. The literature on the consultation functions in the field of e-healthcare platforms emphasizes the importance of different consultation functions and explores their impact on the offline demand. Recent empirical studies have demonstrated that various online consultation methods have significant complementary effects in online health communities. Our research has explored the differentiated impact of multi-channel consultation on the e-healthcare demand, thus offering new managerial insights. Practically, we have revealed how the introduction of the multi-channel has significantly increased the e-healthcare demand on the platform. This change not only reflects the growing demand for online consultations but is also closely related to the accessibility, convenience, and diversity of the platform. At the same time, we have also found that price differences have a significant impact on the frequency and willingness of patients to opt for online consultations. A more reasonable pricing strategy can effectively reduce patients’ costs and enhance their willingness to use e-healthcare, thereby promoting the growth of platform consultations. Therefore, this study fills the research gap in related fields in explaining the influence mechanism of physician behavior, platform function, and price setting on patient decision-making, and provides a valuable empirical reference for optimizing the operation strategy of online medical platforms in the future. Moreover, we have identified that the different moderating effects of multi-channel consultations on disease type and the provider’s title have an impact on e-healthcare demand. Our study has indicated that there are significant differences in the use of graphic and text consulting services between patients with acute and chronic diseases. Patients with acute diseases tend to demand more direct face-to-face healthcare, while patients with chronic diseases prefer convenient graphic-text consultation services. The practical meaning of the graphic-text consultation process lies in its unique characteristics as a non-instantaneous consultation channel. For patients with chronic diseases, this channel provides convenient access to follow-up care, allowing them to describe symptoms and upload relevant examination reports without the need for instant interaction. For providers, graphic-text consultation offers flexibility to respond within 24 hours, accommodating their busy clinical schedules while still delivering professional diagnoses and recommendations. Moreover, the non-instantaneous nature allows providers to research complex cases before responding, potentially improving diagnostic quality. For patients with acute diseases, however, the non-instantaneous nature may be less suitable, as they typically require immediate medical intervention. For graphic-text consultation, platforms should actively recruit and promote high-title providers, as their professional authority significantly enhances demand for this non-instantaneous channel. Conversely, for video-voice consultation, platforms should encourage and support lower-title providers, who have more flexibility to offer instant consultations. At the same time, providers’ professional titles also have different impacts on the e-healthcare demand. Providers with high professional titles have e-healthcare demand, but they are less likely to participate in video-voice consultations due to their busy work schedules. Providers with low professional titles are more flexible in time and can better meet the immediate needs of patients. By analyzing these moderating factors, this study provides support for the practical design and promotion of personalized services in the field of e-healthcare and promotes the digital transformation of medical services. Limitations and future research Not all healthcare providers have the ability or resources to offer both graphic-text and video-voice consultation services, so the broad applicability of this variable may be limited. In addition, the moderating variables such as disease type and provider’s title mentioned in the study, while presenting a theoretical basis, may be affected in practice by sample selection and the scope of data collection, resulting in lower external validity of the findings. In addition, the impact of price factors on online demand, while theoretically expected to have certain effects, the magnitude of price changes, in reality, may be limited by factors such as policy, market competition, and patients’ budgets, which limit the generalization of the findings at this moment. Apart from provider factors (channel availability, pricing) and provider/patient characteristics observable on the platform (disease type, provider’s title), we may also include the widely addressed variables such as patients’ intention to use, patients’ perceived value, etc. Since multi-channel consultations, which emerge as the technology develops, have recently been equipped in most of the platforms. More available data for these new service channels in the future would enable us to enrich the investigations for the designs of the e-healthcare services. We may incorporate a wider range of patient heterogeneity factors (e.g., age, health literacy, and digital literacy) to explore their moderating effects on channel preferences. Moreover, we may generalize our investigations by validating our findings across various domestic and international e-healthcare platforms with different cultural backgrounds and service operational schemes. More importantly, as the technology develops, more channels in e-healthcare may appear. The investigations of the potential emerging channels would be beneficial both for the e-healthcare service providers and users (patients).