This multiple case study takes place in the Province of Quebec and focuses on two specific applications: telehealth and electronic health record. These applications have been selected based on their key position in current federal policies as regards computerisation of the health care system in Canada [
10] as well as their importance at the provincial level in Quebec [
35]. Specific methods and strategies are being used to explore factors influencing the utilisation of scientific knowledge to support e-health implementation at each decisional level.
Political level
A critical incidents analysis will be conducted to identify key decisions that have influenced the implementation of telehealth and EHR in the Quebec health care system. This method has been used in previous studies on the impact of health technology assessment on political decisions [
13,
22]. It consists in identifying, through document analysis and/or contact networks, the 'incidents' that represent specific events or decisions having affected the implementation of the technology. A summary of these critical incidents will be prepared and validated among a purposive sample of key stakeholders who have been involved in these critical incidents.
Then, factors that supported or constrained utilisation of scientific knowledge in the decisions identified will be explored through interviews with key stakeholders who have been involved in these decisions. A semi-structured interview guide will be used to explore specific questions that will be adapted to the nature of the decision and the position of the interviewed stakeholder. Approximately 40 keys informants, including policy-makers, e-health projects managers, evaluators and representatives of various lobbies (professional associations, technological companies, etc.) will be interviewed. Nevertheless, this number may vary according to the data saturation criteria [
36]. Interviews are planned to last about one hour and will be audio recorded after obtaining participant's consent.
At the political level, decisions regarding the implementation of telehealth and EHR are analysed according to three principle themes: 1) funding, diffusion and sustainability of e-health projects, 2) evaluation of e-health projects and its impact on decision-making; 3) other sources that have influenced decisions regarding the implementation of e-health applications. The characteristics and position of stakeholders will be analysed based on Monnier's classification [
37], and the creation of coalitions and alliances will be studied based on the frameworks of Gamson [
38] and Lemieux [
39]. These frameworks will allow identifying stakeholders who have been involved in important decisions regarding e-health implementation and how their specific position and characteristics have influenced knowledge utilisation in these decisions.
The N*Vivo software will be used to perform qualitative analyses. A thematic content analysis will be carried out according to the method described by Huberman and Miles [
40] which implies coding, organizing and linking the material collected from interviews. The codification will be based on concepts from theoretical frameworks relevant to health care policy analysis [
37‐
39]. To ensure the internal validity of the analytical process, codification of the interviews will be made independently by two researchers of the team. The results will then be shared and consensus will be sought for final codification. Interview material will be categorised using this codification tree. Constant comparison and iterations methods will be used to identify the factors that have facilitated or constrained the utilisation of knowledge to support decisions surrounding e-health implementation according to stakeholders' characteristics and role.
Organisational level
At the organisational level, one telehealth and one EHR project implemented since at least one year will be selected for the case study. For telehealth, a telehomecare project in Gaspésie – Îles-de-la-Madeleine, has already be identified. A primary care clinic that has implemented an EHR has also agreed to participate.
Semi-structured interviews with key decision makers (regional and local decision makers, project managers and health care professionals), and document analysis will be used for data collection. Participants will be identified through the contacts network method[
36] and will then be invited to take part to the study. A total of 20 interviews is planned (10 for each project), but this number could change according to the data saturation criteria [
36]. A major element to consider in selecting participants will be the representation of all points of view, namely the presence of decision-makers from the various stakeholders groups involved in the project (professionals, managers, promoters, evaluators). Interviews – of approximately one hour duration – will be audio recorded with participants' consent.
Theoretical models from organisation sciences, such as the neo-institutional theory [
41], the theory of communities of practice [
42] as well as professional theories [
43,
44] provide a framework to examine the contribution of knowledge in decision-making pertaining to e-health projects. The neo-institutional theory proposes the concept of isomorphism, according to which the borders between the organisations become less and less rigid because of associations that are formed between members of various organisations. This concept is highly relevant given the networking purposes of e-health applications. Furthermore, the addition of the concepts of 'communities of practice' and 'occupational boundaries' to the institutional theory, as suggested by Aanestad and al [
45], allows conceiving knowledge utilisation in decisions on e-health as a phenomenon of networking between various organisations and professional groups. These concepts will be used to analyse the application of knowledge in organisational decision-making during the implementation of e-health projects.
Data collected will be analysed qualitatively using the N*Vivo software. An iterative analytical method will be employed starting with concepts found in the theoretical models selected [
41‐
45]. These categories could be adjusted through interactions with the field. The first interviews will be independently coded by two researchers of the team and will be discussed with other members of the team in order to reach a consensus on a final codification tree. Moreover, a participative approach encouraging feedback from representatives of each project will ensure the validity of the analyses [
34] and is likely to favour a greater appropriation of the results by the stakeholders [
46].
Clinical level
At the clinical level, the psychosocial determinants of health care professionals' intention to use e-health in their practice are explored using the Theory of Interpersonal Behaviours (TIB) [
47]. The TIB is considered as an exhaustive psychosocial theory since it integrates most of the construct found in other theories as well as dimensions such as values, social roles and norms that are not taken into account in other models. Moreover, the TIB regards the culture or the subculture as a factor influencing behaviour [
48]. According to this theory, human behaviour is formed by three components: intention, facilitating conditions, and habit.
Intention refers to the individual's motivation regarding the performance of a given behaviour.
Facilitating conditions represent objective factors that can make the realisation of a given behaviour easy to do.
Habit constitutes the level of routinisation of a given behaviour, i.e. the frequency of its occurrence.
In the TIB, the behavioural intention is formed by attitudinal as well as normative beliefs. Attitudinal beliefs comprise affect and perceived consequences.
Affect represents an emotional state that the performance of a given behaviour evokes for an individual. It is considered as the affective perceived consequences of the behaviour, whereas
perceived consequences refer to the evaluation by the individual of the possible consequences of the behaviour. The TIB also distinguishes between two normative dimensions: social and personal. Perceived social norms are formed by normative and role beliefs.
Normative beliefs consist of the internalisation by an individual of referent people or groups' opinion about the realisation of the behaviour, whereas
role beliefs reflect the extent to which an individual thinks someone of his or her age, gender and social position should or should not behave. With respect to the other normative components of the TIB, the
personal normative belief represents the feeling of personal obligation regarding the performance of a given behaviour, whereas
self-identity refers to the degree of congruence between the individual's perception of self and the characteristics associated with the realisation of the behaviour. Previous work on clinicians adoption of health care technologies [
49] confirm the utility of the TIB to understand the behaviours of health care professionals.
Based on results from previous work [
49] and on our recently completed Cochrane systematic review on "
Interventions for promoting information and communication technologies adoption in healthcare professionals" [
50], a questionnaire will be developed. The questionnaire will be built according to a consensus of experts in social psychology [
51] and will follow the methodology recommended by Gagné and Godin [
52] This questionnaire will assess the psychosocial factors influencing the adoption of telehealth and EHR by health care professionals. It will also assess the role of scientific evidence to support clinicians' decision-making regarding their use of these e-health applications.
The etic-emic approach [
53] will be used in the development of the questionnaire since it allows adapting the theoretical constructs (etic dimension) to specific context of the culture studied (emic dimension) according to the studied population. A convenient sample of about 20 clinicians will be selected to complete an open-ended questionnaire. Qualitative analyses will be carried out in order to extract participants' beliefs corresponding to each of the TIB constructs. The frequency of each belief will be compiled in order to identify the modal salient beliefs in the population. Identified beliefs will then be used as question items for measuring each of the theoretical constructs in the study questionnaire. The internal consistency of theoretical constructs and the temporal stability of their measurement [
54,
55] will be checked using the test-retest method by asking 30 clinicians to complete the questionnaire twice within a two-weeks interval. A final version of the questionnaire will be prepared following the test-retest.
The sample size necessary depends on the number of theoretical variables (7) and studied external variables (5). According to Cohen [
56], a sample of 547 participant is necessary to perceive the effect of two groups of variables in the regression of the dependent variable: the first formed group of the seven variables of the TIB (R
2 equal to 0.10); the second group made up of five additional variables with an increase in R
2 equalizes to 0.02 (alpha = 0.05 and power of 0.80 for the two groups of variables). Thus, by estimating the participation rate at 40%, which corresponds to the average found in similar studies [
57], it will be necessary to target 1368 physicians in order to recruit 547 of them. Strategies aiming at increasing response rate will be used such as the participation of the Head of Department and sending recall letters according to the procedure suggested by Dillman [
58].
The following analyses will be carried out using the SAS software: 1) distribution of the variables in terms of percentage and average; 2) Pearson or Spearman coefficients of correlation; 3) multiple regression of the factors (independent variables) predicting clinicians' intention to use evidence on e-health (dependent variable). The independent variables that will be evaluated are: affect, perceived consequences, normative beliefs, role beliefs, personal normative belief, self-identity, and facilitating conditions; 4) the influence of the external variables (sociodemographic and professional) on the intention by comparing R
2 of the model including all variables to the model containing only the psychosocial variables [
59]; and 5) a multiple variance analysis in order to determine the theoretical constructs allowing to distinguish the subjects intending or not to use knowledge to support their decisions.