Primary Care Physicians’ and Specialists’ Experiences on acceptance and use of technological innovation: Successful electronic consultation service initiative in Quebec, Canada

Background: Electronic Consultation (eConsult) is an eHealth service that allows primary care providers (PCPs) to electronically consult specialists regarding their patients’ medical issues. Many studies have demonstrated that eConsult services improve timely access to specialist care, prevent unnecessary referrals, improve PCPs’, specialists’ and patients’ satisfaction, and therefore have a large impact on costs. However, no studies have evaluated PCPs’ and specialists’ acceptance of eConsult services in Canada, and worldwide. Objective: This exploratory study aimed to identify factors affecting eConsult service acceptance by PCPs and specialists in urban and rural primary care clinics across three regions in the province of Quebec, Canada, by integrating the Unified Theory of Acceptance and Usage of Technology (UTAUT) and Task-Technology Fit (TTF) models, and user satisfaction. This research was designed to broaden and assist in scaling up this effective eHealth service innovation across the province. Methods: A cross-sectional web-based survey was sent to all PCPs (N=263) and specialists (N=62) who used the eConsult Quebec Service between July 2017 to May 2021. We proposed a unified model integrating the UTAUT model and TTF model, and user satisfaction by endorsing eleven hypotheses. The partial least squares (PLS) was used to investigate factors influencing the acceptance of the eConsult Quebec Service. Results: Of the 325 end users, 41.8% (N=101 or 38% PCPs and N=35 or 56% specialists

1) Would you like to publish your submitted manuscript as preprint?Please make my preprint PDF available to anyone at any time (recommended).
Please make my preprint PDF available only to logged-in users; I understand that my title and abstract will remain visible to all users.Only make the preprint title and abstract visible.No, I do not wish to publish my submitted manuscript as a preprint.2) If accepted for publication in a JMIR journal, would you like the PDF to be visible to the public?
Yes, please make my accepted manuscript PDF available to anyone at any time (Recommended).
Yes, but please make my accepted manuscript PDF available only to logged-in users; I understand that the title and abstract will remain v Yes, but only make the title and abstract visible (see Important note, above).I understand that if I later pay to participate in <a href="http

Introduction Background
Access to specialized services remains a significant issue in Quebec, Canada.The Quebec Ministry of Health and Social Services' (MSSS) 2019-2023 strategic plan identified the improvement of access to specialized services, specifically consultations with a medical specialist, as one of the major objectives on which to focus over the next few years [1].Health care systems face constant pressure to control health care costs while improving access and providing patients with safe, highquality care.Developing a more efficient and cost-effective health care system is essential to providing better services and a better patient care experience.As part of the 2017-2027 Québec Life Sciences Strategy, the province increased its investment in health research and innovation in order to accelerate the adoption of new and innovative practices [2].In line with the government's digital transformation strategy, the MSSS has already begun this transformation within the health and social services network (HSSN), by implementing digital services that will facilitate access to care as well as through rapid and efficient management of patients' health.

eConsult Services Worldwide
Today, technological advances and the integration of new practices in the health sector offer interesting possibilities to solve the problem of excessive wait times and equitable access to specialists, namely through digital health solutions [3][4][5][6][7].Digital health solutions can help overcome barriers to health care access and patient care management.This is especially true for patients living in rural or remote areas who often experience inequitable access to services compared to those living in urban areas [8][9][10][11][12][13].
Traditionally, patients have in-person consultations with their doctor.Virtual care is an alternative method that can be used to improve access and better utilize specialized resources [14].eConsult or electronic consultation provides an effective and efficient alternative method of assessing a clinical situation without having the patient meet with the specialist in person [3,[15][16][17][18]. eConsult services are delivered through secure web-based applications that facilitate asynchronous communication between primary care providers (PCPs) and specialists, allowing a primary care provider to submit a clinical question to a specialist and to get an answer within a week [16,[19][20][21].The specialist responds with advice on how to treat the patient, referral recommendations, or a request for additional information about the case.The primary care provider and specialist continue to communicate until the case is resolved.
However, no study has ever been conducted on the perceptions and experiences of PCPs and specialists regarding the factors influencing adoption, acceptance and use of eConsult services.
The aim of this exploratory study is to fill this gap, integrating variables from two models, the United Theory of Acceptance and Use of Technology (UTAUT) and Task-Technology Fit (TTF), as well as user satisfaction to explain user behavior regarding the adoption of the eConsult Quebec Service.

Research Model and Hypotheses
With the development of information technology (IT), many theories and models of the acceptance and use of new technology have been developed or used in order to better understand the factors that influence the acceptance and use of technology [96][97][98][99][100][101].Several health sector studies are based mainly on the Technology Acceptance Model (TAM) [93,[102][103][104][105][106][107][108][109].This model was developed by Davis, Bagozzi and Washaw in 1989 [96], drawing from the theory of reasoned action to model user acceptance of new information technology (IT).In the TAM, intention to use is determined by the user's attitude toward the technology, which is itself determined by two factors: 1) perceived usefulness and 2) perceived ease of use of a technology.However, the presence of external factors or determinants can potentially influence these two perceptions [100].Several studies have extended the TAM model by adding other explanatory variables [98,[110][111][112].Other works have enhanced the TAM model by creating integrative models that combine the TAM with other explanatory models of user behavior [95,[113][114][115].In view of the multiplicity of existing models on the acceptance and use of technology, Venkatesh et al. (2003) proposed a unified explanatory model of user behavior which is the basis of the Unified Theory of Acceptance and Use of Technology (UTAUT) [100], integrating variables drawn from eight models [116].
The UTAUT model is considered to be one of the most effective [117].This model has been tested and empirically validated in several fields and in different contexts such as health care in order to study the determining factors in acceptance and use of a technological innovation.Moreover, the UTAUT model can explain up to 70% of the variance in intention to use technology and about 50% in actual use of the latter, representing a high predictive power [100,101,104].
The approach postulates that technology use is a function of intention to use, which itself is influenced by four determinants, which are: 1. Performance expectancy, "The degree to which an individual believes that using the system will help him or her to attain gains in job performance" (Venkatesh et al., 2003, p. 447) [100].2. Effort expectancy, "The degree of ease associated with the use of the system" ( And finally, the intention to use a technology is an individual's motivation and willingness to make an effort to achieve the target behavior, namely the use of the technology [100].Technology use is the performance of the behavior itself, namely adoption of the technology [100]. At the same time, "fit" is an important notion in IT.It is defined as "the degree to which a technology assists an individual in performing his or her portfolio of tasks" (Goodhue and Thompson, 1995, p.218) [118].Task-technology fit (TTF) is another theoretical model that has been studied to explain how a new technology leads to performance, to evaluate the impact of adoption, and to assess the relationship between task and technology characteristics.TTF seeks to assess whether the technology's functionality is well aligned, ie, compatible with the work done by end users.Studies have shown that technology will be more readily accepted and have a positive impact on individual performance if the technological characteristics match expected tasks.Based on a systematic review, authors were interested in identifying the fields of interest of studies that applied the TTF model and the majority were conducted in the field of health [119][120][121][122][123][124][125][126].
As Goodhue and Thompson's (1995) TTF model assesses the fit between the task and the technology, it appeared highly relevant to combine the TTF and the UTAUT models because they have the potential to further explain PCPs' and specialists' acceptance of the eConsult Quebec Service.Additionally, certain studies integrated UTAUT and TTF to understand and predict end-user behavior and intention regarding acceptance and adoption of new health-related technology [127][128][129][130][131][132][133].
Some studies focused on explaining user behavior based on other individual variables such as satisfaction.Resistance to the implementation of new IT can have a negative impact on user satisfaction [134,135].Resistance may be caused by the perceived risk of performance loss and dissatisfaction following use of the new IT [136].The user of an IT seeks to improve their performance by improving the way they work, perform their tasks and achieve their objectives [122,137,138].Other studies in the field of health have focused on the application of the TTF model based on end-user satisfaction [126,139].
The models' diversity reflects the existence of several factors influencing IT use behavior, which is a complex and multidimensional phenomenon.Consequently, our study integrated and adapted the UTAUT and TTF models and end-user satisfaction with the eConsult Quebec Service.
The integrated research framework is presented in Figure 1, and the 11 hypotheses are listed in Table 1.The proposed integrated model includes 10 constructs that have been refined and adapted to the context of the study.The relationships between the variables were hypothesized by referring to previous studies.

H1
Task characteristics (TAC) have a positive impact on task-technology fit (TTF).H2 Technology characteristics (TEC) have a positive impact on task-technology fit (TTF).H3a Task-technology fit (TTF) has a positive impact on performance expectancy (PE).H3b Task-technology fit (TTF) has a positive impact on adoption (ATT) of the eConsult Quebec Service.H3c Task-technology fit (TTF) has a positive impact on satisfaction (SAT).H4 Performance expectancy (PE) has a positive impact on satisfaction (SAT), H5a Performance expectancy (PE) has a positive impact on behavioral intention (BI) to use eConsult Quebec Service.

H5b
Effort expectancy (EE) has a positive impact on behavioral intention (BI) to use of the eConsult Quebec Service.

H5c
Social influence (SI) has a positive impact on behavioral intention (BI) to use of the eConsult Quebec Service.

H6
Facilitating conditions (FC) has a positive impact on adoption (ATT) of the eConsult Quebec Service.H7 Behavioral intention (BI) to use eConsult Quebec Service has a positive impact on adoption (ATT) of the eConsult Quebec Service.

Development of the eConsult Quebec Service
In Ontario, Canada, the Champlain eConsult Base TM (Building Access to Specialists through eConsultation) service was initially launched in 2009 to address the issue of long wait times for patients requiring non-urgent care and specialist advice [14,16,140].Given its success in Ontario and in other parts of Canada, the eConsult Quebec Service was based on the Champlain Base TM business model and was replicated on an existing telehealth platform within the Quebec health network [3].
In addition, the Quebec team was part of the Canadian Foundation for Healthcare Improvement (CFHI) Connected Medicine collaborative, in partnership with Canada Health Infoway, the College of Family Physicians of Canada and the Royal College of Physicians and Surgeons of Canada, to roll out remote consultation services to improve access to specialist medical advice in primary care settings.This project was carried out over a period of 18 months from 2017 to 2018 in seven Canadian provinces (British Columbia, Alberta, Saskatchewan, Manitoba, Quebec, New Brunswick, Newfoundland and Labrador) [141].The collaborative project provided an opportunity to deploy a proven Ontario-based innovation in the three targeted regions of Quebec to improve access to specialist advice by enabling PCPs to submit questions to medical specialists.

Study Design and Setting
Given our objective, we adopted a cross-sectional data collection approach.Kumar (2014) explains that cross-sectional design is adequate to study the prevalence, or the occurrence, of a phenomenon, situation or attitude within a subset of a given population at a certain point in time [142].In light of this, the study of users' perceptions, attitudes and continuance intention with regard to technological innovations lends itself to a cross-sectional study approach.
We conducted a survey-based multicenter cross-sectional study across three regions in the province of Quebec (Outaouais, Abitibi-Témiscamingue and Mauricie-et-du-Centre-du-Québec) as part of the eConsult Quebec pilot project.The online survey was sent to all PCPs (N=263) and specialists (N=62) who used the service between July 7, 2017 and May 17, 2021 in order to assess acceptance and use of the service.

Instrument
To ensure measurement validity, all items of each model variable were developed based on UTAUT and TTF as well as on end-user satisfaction.Variables were measured using reliable items that had already been used by previous studies; we adapted certain items to our study context.The first version of the survey was translated from English into French by a group of researchers and validated by a professional translator to ensure that the content had not lost any of its original meaning.The second and final version of the survey was validated with four PCPs and two specialists who had already used the eConsult Quebec Service.They were asked to identify any issues that might lead to confusion.

Technology Characteristics (TEC) FMT1
The way elements are arranged on the eConsult Service screen makes it easy to read the information.FMT2 The information in the eConsult Service is clear.FMT3 Overall, the information is presented in a useful format.

SEC1
The risk of an unauthorized third party accessing the eConsult Service is low.SEC2 I believe that only the appropriate people have access to information.SEC3 I believe that the eConsult Service is secure enough to handle sensitive information.

Task-Technology Fit (TTF) ACC1
The patient information received through the eConsult Service is accurate.ACC2 I am satisfied with the accuracy of the information in the eConsult Service.

ACC3
Overall, I believe the information provided is free of errors.

Performance Expectancy (PE) PU1
Using the eConsult Service helps me make clinical decisions or offer advice faster.PU2 I believe that using the eConsult Service enables me to make safer decisions or offer safer advice.PU3 I believe that using the eConsult service enables me to make more accurate clinical decisions or offer more accurate advice.

PU4
Using the eConsult Service makes my job easier.PU5 Overall, I find the eConsult Service to be useful in supporting my clinical decision-making or when offering advice.

Adoption (ATT) ATT1
Using the eConsult Service is a smart idea.

ATT2
Using the eConsult Service is a safe experience.ATT3 I am in favor of using the eConsult Service.

Satisfaction (SAT) SAT1
Using the eConsult Service meets my needs.SAT2 I am happy with my use of the eConsult Service.SAT3 I am extremely satisfied with my use of the eConsult Service.

Data Collection
The survey was developed using online survey software Survey Monkey.The survey was distributed to all users on May 19, 2021 as an electronic letter through the eConsult Quebec Service electronic mailbox system.The letter included the context, study objective, guarantee of confidentiality, duration and the link to access the online survey.A reminder letter was sent via the electronic mailbox system two months after the initial invitation to maximize the response rate.All participants gave their consent electronically before beginning the survey.Participation was anonymous and voluntary.
This study was approved by the Institutional Review Ethics Committee of the Outaouais Integrated Health and Social Services Centre (ref.number 2016-183_88), Québec, Canada.

Participants
In the end, 136 respondents from the three regions of Quebec took part in the survey, representing a response rate of 41.8%.The sample is composed of 101/263 PCPs and 35/62 specialists.

Data Analysis
Several authors argue that the partial least squares (PLS) method is appropriate for exploratory projects and data in various fields [143][144][145][146][147]as well as for the formative measurement of variables that require the operationalization and conceptualization of various concepts [148] in order to: "Predict and explain a key target construct and/or to identify its relevant antecedent constructs.In other words, this approach generates latent variable scores that maximize within-sample prediction in terms of the dependent latent variable's R 2 value.As such, the estimated coefficients depict the relevance of constructs in a certain model that directly, indirectly and totally contribute to the explanation of a target construct of interest" (Chin et al., 2020, p.2162) [149].
In addition, PLS is appropriate to evaluate relatively new measurement models.Thus, PLS is a promising method for research projects related to information systems and emerging technologies [146], as is the case here.Moreover, this method is especially useful for smaller samples [144] where it may be more difficult to obtain a large number of respondents and where "the goal is to predict and explain the key target constructs or identify the key driver constructs" [150].This data analysis method is commonly used in research projects based on UTAUT [151], as well as in those combining UTAUT and TTF due to the complexity of the model with several constructs [132,152].For this reason, this study uses PLS to examine the research model and test the 11 hypotheses.Data was analyzed using SmartPLS 4.0 software.
To begin, as per recommendations from Bagozzi and Phillips (1982) [153], and Venkatraman and Grant (1986) [154] concerning the preliminary verification of constructs, an evaluation of their validity and reliability was established.Then, the measurement model was evaluated by examining reliability, internal consistency, and convergent validity of measurements.For exploratory research, a Cronbach's alpha coefficient of 0.50 is considered an acceptable value [155][156][157].As for the composite reliability (CR) index, it must be equal to or greater than 0.70 [158].Although Hair et al.
(2010) recommend a factor loading greater than 0.50 to be considered significant [156].A factor loading greater than 0.40 in the exploratory phase was considered to be a significant contribution [159].During the reliability test, two items (SEC1 and SF1) were removed from our research model for data analysis.The validity of constructs was verified using the average variance extracted (AVE) index, the value of which must be equal to or greater than 0.50.[160].
Lastly, the structural model was evaluated by coefficients of determination (R 2 ), and path coefficients and their significance by running 5000 bootstrap subsamples.Finally, the research hypotheses must be statistically proven with a t value greater than 1.96 and a P value less than 0.05.

Research Ethics
Ethical approval was obtained from the Research Ethics Committee of the Outaouais Integrated Health and Social Services Centre before the beginning of the study (ref.number 2016-183_88), in Quebec, Canada.

Measurement Model Evaluation
As shown in Table 3, for our research model, all factor loading values for all items are greater than 0.40 (0.413 to 0.931), all Cronbach alpha values are greater than 0.50 (0.591 to 0.891), CR values exceed 0.70 (0.783 to 0.928), and all AVE values are greater than 0.50 (0.550 to 0.812).These results indicate good reliability and validity of the constructs.

Hypothesis Testing
As mentioned, to test our hypotheses, we used the PLS method because it is widely used to test complex causal models, incorporating several latent variables.Sample size constraints are also more flexible and measurement scales do not require broad approval.Thus, the PLS method is well suited to exploratory analyses.In our structural equation model, path coefficients (β), t values, and P values are examined in order to distinguish the relationships between the constructs of our research model.Additionally, we examined the variance in the dependent variables explained by the independent variables to evaluate the explanatory and predictive power of the structural model (R 2 ) [161].
According to Falk and Miller (1992), the coefficient of determination (R 2 ) should be greater than 0.10 [162].
The results of the hypothesis tests on our model integrating UTAUT, TTF and user satisfaction with the eConsult Quebec Service are summarized in Table 4 and depicted in Figure 2. Based on Table 4, the statistical result of each path in the research model indicated that most of the hypotheses were supported, except H5c effect of social influence (SI) on behavioral intention to use (BI) and H6 effect of facilitating conditions (FC) on adoption (ATT) of the eConsult Quebec Service because the P value was greater than 0.05 and the t value was less than 1.96.
Both TAC and TEC were found to positively influence TTF, supporting H1 (t value=2.243,P<.05) and H2 (t value=6.119,P<.001).The calculated R 2 values (Figure 2) showed that 36.5% of the variance in TTF was explained by TEC and TAC, with TEC having the strongest influence (t value=6.

Theoretical support
To our knowledge, this study is the first to validate both models, namely United Theory of Acceptance and Use of Technology (UTAUT) and Task-Technology Fit (TTF), along with user satisfaction to explain user behavior regarding the adoption of the eConsult Quebec Service.The aim of our exploratory study is to identify the factors that predict acceptance of the service by PCPs and specialists in urban and rural primary care clinics across three regions in Quebec, Canada.Results of the PLS analyses indicate that 9 of the study's 11 hypotheses are supported.To explain the adoption of the service by PCPs and specialists, the direct relationships uniting the various constructs of the model highlighted the importance of several key constructs and predominant correlations, thus confirming the majority of research hypotheses.The research model's variables influencing each endogenous variable explain 36.5% of the variance in TTF, 13.4% of the variance in PE, 45.5% of the variance in BI, 50% of the variance in ATT, and 52.7% of the variance in SAT with regard to the adoption of the eConsult Quebec Service.
First, results suggest that satisfaction (SAT) is the key driver behind the use of the eConsult Quebec Service.Similarly, previous studies in the field of health have shown that satisfaction is a key factor in predicting individuals' willingness to adopt technology [95,163,164].From the perspective of the Information Systems Success Model (ISSM), individuals' level of satisfaction can significantly influence the acceptance and use of a particular system [95,117,[165][166][167].Another study demonstrated that user satisfaction is the strongest predictor of perceived benefits and technology continuance usage intention [164,167,168].Our model shows that this construct is powered by two other constructs that explain 52.7% of the variance: Task-technology fit (TTF) and performance expectancy (PE).Studies have shown that PE has a significant impact on satisfaction [169].According to Bhattacherjee (2001), when user expectations are confirmed, they will be satisfied [113].Thus, PE will affect user satisfaction with a system.Individual productivity is defined as a respondent's belief in their effectiveness and efficiency, and in the quality of their work [122,170].A user's perceived feeling of performance in making more accurate and safer clinical decisions or offering more accurate and safer advice (PU2 and PU3) can influence the adoption of the eConsult Quebec Service, how they use it and the level of satisfaction resulting from their experience with the service.People are happier and more productive when the technology they are using is adapted to their daily tasks.Thus, the better the technology meets the PCPs' and specialists' information-related needs when making clinical decisions related to the health of their patients, the more satisfied they are.This correlation shows that the role of task-technology fit affects end-user satisfaction.Results suggest that with high levels of task-technology fit (TTF) and performance expectancy (PE), PCPs and specialists experience higher levels of satisfaction (SAT).
Looking at the task-technology fit (TTF) model, results demonstrate that task characteristics (TAC) and technology characteristics (TEC) explain 36.5% of the variance in TTF.The factor with the most direct and significant influence on users is TEC (H2, P<.001).TAC have a less significant influence (H1; P<.05).These results are not surprising.During the implementation of the eConsult Quebec Service, the steering committee and its subcommittees ensured the modeling of the eConsult process, established IT-related needs and supervised the technological solution.To ensure the success of the initiative, the team defined the organizational model supporting the workflow of an electronic consultation (eConsult).To reduce any risk associated with the development of a new technological solution, the pilot project relied on an existing and already proven platform in the Quebec health network.The eConsult Quebec Service was adapted to the platform as an additional trajectory among other similar digital consulting trajectories already in operation since 2006.This made it possible to comply with Quebec's context and security requirements, and to offer a simplified user experience, especially for users who were already active on a telehealth platform and did not need to adapt to an additional tool.The new trajectory was integrated into the PCPs' and specialists' existing clinical workflow.
With regard to the UTAUT model, performance expectancy (PE), effort expectancy (EE), and social influence (SI) explain 45.5% of the variance in behavioral intention (BI).BI is the third positive and significant predictor of eConsult Quebec Service use.The most significant direct correlation is between PE (H5a, ***P<.001) and intention to use the technology.The most widely studied relationship in the field of health is that between performance expectancy and the intention to use a technology [104,[171][172][173][174][175][176][177][178][179].Thus, when clinicians' performance expectancy was achieved, it positively influenced their intention to use the system.Based on our empirical observations, in the context of voluntary adoption, the correlation between performance expectancy and intention to use the technology is the strongest.The second most frequently measured correlation concerns the impact of effort expectancy on intention to use the system, which has proven to be positively significant in several studies [109, 172-175, 178, 180-183].In this study, the correlation is weaker, but remains positive and significant (H5b, *P<.05).In other words, when considered alone, the userfriendliness of a technology does not totally influence the intention to use a technology, it must also be perceived as useful.Based on our field data, we expected to see a weaker correlation between effort expectancy (EE) and behavioral intention (BI) to use the eConsult Quebec Service than between performance expectancy (PE) and behavioral intention to use the service.Regarding the role of social influence (SI) in relation to intention to use the system, the reviewed studies tested a positive correlation between these two constructs [184].Note that social influence (SI) has no direct effect on this factor (H5c) in our study.As a first line of thought, it is possible that physicians as selfemployed workers within the public health system in Quebec are less sensitive to social pressure when using technology.As a second explanation, it is possible that this finding reflects the voluntary context of the adoption and use of the eConsult Quebec Service by PCPs and specialists.
Adoption (ATT) is the second predictor of the use of the eConsult Quebec Service.Our model demonstrates that this construct is powered by three other constructs that explain 50% of the variance: Behavioral intention (BI), facilitating conditions (FC) and task-technology fit (TTF).The most significant direct correlation is BI (H7; ***P<.001).Some studies have noted that a person's emerging intention represents a critical point during the adoption of a technology, which may be considered as the moment when the decision is made to accept the change and modify their behavior [104,171,173,185].Thus, attitude toward technology lies in the intention to use the technology.The second direct but weaker correlation is TTF (H3a; *P<.05).Several studies have confirmed the importance of TTF in the adoption of a new technology [119-121, 127, 186].Generally, the compatibility of the IT with the preferred work style and current work practices would influence adoption of the IT [130].As for the last construct, facilitating conditions (FC) have no effect on the adoption of a system (H6: 0.114).

User Adoption
This construct was defined as the user's perception of the factors in the work environment that promote adoption and use of the service, namely with regard to the available support, coaching and training provided.This means that all the barriers hindering the adoption of a technology were eliminated by the team during the roll out of the eConsult Quebec Service.Initially, the pilot project was launched in the Outaouais region with one clinic; other clinics were added during the deployment phase in this region.This approach made it possible to control recruitment and, above all, it facilitated PCPs' and specialists' ownership of the use of the service as well as enabling validation of monitoring mechanisms.Secondly, the project was rolled out in several clinics simultaneously in two other regions (Mauricie-et-Centre-du-Québec and Abitibi-Témiscamingue). Operation of the service is very simple and user-friendly.Given the user-friendliness of the platform and available tools (video, PowerPoint presentation and test platform), no specific training is required.Users contact the resources responsible for the project in their region, as needed.Therefore, the system itself requires little user support.As part of the pilot project, the team made sure that users could operate with maximum autonomy.Thus, PCPs and specialists were comfortable using the eConsult Quebec Service.

Integrating theoretical models
The relationship between the TTF model and the UTAUT model is demonstrated by the influence of TTF on PE.Indeed, TTF has a positive and significant effect on PE (H3b; ***P<.001),and TTF explains 13.4% of the variance in PE.This finding is similar to other studies [130,187].Thus, the authors found that a technology can have a positive impact on performance, when it is used and aligned with the supported task [118].
In the context of this study, the higher the TTF level, the more PCPs and specialists perceived the eConsult Quebec Service as useful and conducive to making work easier.Thus, the technology's fit to user requirements would directly influence their expectations.During the pilot project, integration of the eConsult Quebec Service into health care practices was identified by PCPs and specialists as an important issue likely to affect clinical practices, as well as the safety and the quality of care provided to patients.Studies carried out in health care settings clearly indicate that during its adoption, the compatibility of a technology influences performance expectancy [122,138,172,175,179,[188][189][190][191].
To this end, deployment of the service made it possible, namely, to optimize the consultation process following an eConsult and to provide the patient with optimal care management following an eConsult, and in the majority of cases, the PCP's action plan was improved.Overall, PCPs and specialists found the eConsult Quebec Service to be useful in supporting their clinical decisionmaking or when offering advice (PU5).Moreover, the correlation between the two models is presented by the influence of TTF on ATT.The greater the alignment between the technology and the task at hand, the more likely the user will be to adopt and use the technology.As mentioned, results indicate that TTF contributes little to explaining the attitude toward actual use of the eConsult Quebec Service (H3a: P<.05).Note that user perception can be affected by various factors, namely the voluntary nature of adopting a technology.Also, as part of the pilot project, adopting the eConsult Quebec Service was not mandatory, so PCPs and specialists used it on a voluntary basis.

Limitations and Future Research
This study has some limitations that can be addressed in future studies.First, the results of this study may be difficult to generalize as the study was conducted in only three regions across Quebec.It would be relevant to extend the scope of the research project to a larger population of PCPs and specialists in Quebec and Canada.This would help confirm the model and its research variables relating to the use of the eConsult Service on a larger scale.
Second, this study did not compare the user groups, ie, PCPs and specialists, because the objective was to study the applicability of the model integrating UTAUT and TTF, and end-user satisfaction with the eConsult Quebec Service.However, perception of the service may be different from one group to another.
Third, the research design was cross-sectional, capturing PCPs' and specialists' overall experience at a specific point in time.A longitudinal study would assess perception over time in order to make the results of the study more robust and validate our research model.
Finally, all the reliability and validity analyses demonstrated satisfactory results, with the exception of two items (SF1 and SEC1) that were withdrawn due to low factor loading.

Conclusions
Based on an extended model integrating UTAUT and TTF as well as user satisfaction, this study provides a better understanding of the factors influencing the intention to adopt the eConsult Quebec Service by PCPs and specialists in three regions across Quebec.In addition, this study tests a research model and a technology that has not previously been explored in Quebec, Canada.In light of the results of this study and of the fact that most of the research hypotheses have been confirmed, we can say that our integrated research model is particularly well suited to the context of the study.In this regard, the eConsult Quebec Service is well suited to the users' need to improve access to specialized medical advice.
This study has demonstrated the significance of certain factors that contribute heavily to actual use of the service.Indeed, our results indicate that task characteristics, technology characteristics, tasktechnology fit, performance expectancy, effort expectancy, behavioral intention, adoption and satisfaction have an effect on users of the eConsult Quebec Service, whereas social influence and facilitating conditions have no significant impact.Note that satisfaction is a significant predictor, essential to evaluating acceptance of the eConsult Quebec Service.PE and EE can positively predict BI, and BI can impact ATT.In addition, TTF has an influence on PE, ATT and SAT.
The significant conclusion of this study is that identifying the determining factors in the adoption of the eConsult Quebec Service was necessary to successfully scale up the service.The results of our study have made a valuable contribution to the implementation of the service by policy-makers in order to maximize its acceptance, use, adoption and success.In fact, after four successful years, the eConsult Quebec pilot project is now the Conseil Numérique (CN) digital consultation service.This new service, launched in the spring of 2021, has been rolled out across Quebec by the MSSS.The Conseil Numérique (CN) digital consultation service promotes swift communication between PCPs and specialists by providing timely access to specialist medical advice.Thus, it contributes to better case management by the PCP.So far, more than 30,000 cases have been processed.In over 40% of electronic consultation cases, an in-person visit to the specialist was considered but deemed unnecessary following the electronic consultation.This reduced unnecessary wait times for patients and freed up specialist resources for patients who needed them most.When a patient needed a visit with a specialist, those visits were often more efficient and productive due to the steps or treatment initiated through the electronic consultation prior to the in-person visit.

Figure 1 .
Figure 1.The proposed research model based on UTAUT and TTF, and user satisfaction.
The survey was administered in French to all users of the eConsult Quebec Service.The survey comprised two parts.The first part consisted of demographic information (area of origin, profession, and specialty group [for specialists]).The second part included 10 constructs and 38 items, including task characteristics (TAC), technology characteristics (TEC), task-technology fit (TTF), performance expectancy (PE), effort expectancy (EE), social influence (SI), facilitating conditions (FC), behavioral intention (BI), adoption (ATT) and satisfaction (SAT), to measure PCPs' and specialists' perception regarding the use of the eConsult Quebec Service.All items were measured on a 5-point Likert scale (1.Strongly disagree, 2. Somewhat disagree, 3. Neutral, 4. Somewhat agree, 5. Strongly agree).
Service impacts my work performance.JF2 Using the eConsult Service can significantly improve the quality of the results of my work.JF3 Using the eConsult Service can improve my level of efficiency in the performance of my tasks.
Learning to use the eConsult Service is easy for me.PEOU2 My interactions with the eConsult Service are clear and understandable.PEOU3 I think that the eConsult Service meets my needs.PEOU4 Overall, I find the eConsult Service easy to use.Social Influence (SI) SF1 I use the eConsult Service because the majority of my colleagues use it.SF2 My organization facilitated the use of the eConsult Service.SF3 My association supports my use of the eConsult Service.SF4 The organization generally supported my use of the eConsult Service.Facilitating Conditions (FC) FC1 I have the knowledge required to use the eConsult Service.FC2 My organization supports the use of the eConsult Service.FC3 I have the necessary resources to use the eConsult Service.Behavioral Intention (BI) BI1 I will continue to use the eConsult Service.BI2 I plan to use the eConsult Service frequently.BI3 Overall, I think using the eConsult Service is beneficial.BI4 I would recommend the eConsult Service to my colleagues.
PE: performance expectancy PLS: partial least squares SAT: satisfaction SI: social influence TAC: task characteristics TAM: technology acceptance model TEC: technology characteristics TTC: Telehealth Coordination Centre TTF: task-technology fit UTAUT: Unified Theory of Acceptance and Usage of Technology

Figures
Figures

Table 2 .
Table 2 presents the items of each construct.Items used in the research model.

Table 3 .
The measurement model evaluation.