7. DEUXIÈME ARTICLE

Telehealth adoption in hospitals: An organisational perspective

Gagnon M-P., Lamothe L., Fortin J-P., Reinharz D., Godin G., Gagné C., Cloutier A. (2003). Journal of Healthcare Organisation Management, submitted for publication .

Abstract

Human and organisational factors are central to the adoption of telehealth and influence its diffusion into integrated networks. A review of the literature on the adoption of information and communication technologies in the healthcare sector has identified relevant concepts from different theoretical frameworks. These variables were combined to propose a comprehensive framework of potential factors affecting telehealth adoption by hospitals. Variables were operationalised and adapted to the characteristics of the telehealth network under study. A questionnaire was administered via telephone interviews to the 32 hospitals involved in the Extended Telehealth Network of Quebec. Contingency analyses were performed to determine which organisational factors have influenced telehealth adoption. Subsequently, a multiple case study was conducted among 9 hospitals representative of different categories of telehealth adopters. In-depth interviews with various actors involved in telehealth activities have permitted to deepen our understanding of the impact of clinical and administrative contexts on telehealth adoption.

Keywords: Telehealth adoption, Healthcare organisations, Organisational theory, Structural characteristics, Cultural characteristics.

Résumé

Les facteurs humains et organisationnels influencent l’adoption de la télésanté et sa diffusion dans des réseaux de services intégrés. À partir d’une revue de la littérature, un cadre conceptuel des facteurs d’adoption organisationnels de la télésanté a été développé en s’inspirant de différentes théories. Les variables proposées ont été opérationnalisées et adaptées à la nature de l’objet à l’étude. Un questionnaire a été administré par le biais d’entrevues téléphoniques dans les 32 centres hospitaliers impliqués dans le Réseau québécois de télésanté élargi. Des analyses de contingence ont permis d’identifier les facteurs ayant influencé l’adoption de la télésanté par ces établissements. Ensuite, une étude de cas multiple s’est déroulée dans neuf hôpitaux représentant différentes caractéristiques associées à l’adoption de la télésanté. Des entrevues en profondeur avec différents acteurs impliqués dans les activités de télésanté ont permis une compréhension plus riche des facteurs structurels et contextuels ayant influencé l’utilisation de la télésanté.

Introduction

Rapidly emerging information and communication technologies (ICT) have spurred the recent escalation of various telehealth applications (Lehoux, Battista & Lance, 2000). However, telehealth adoption in healthcare organisations is considered a prerequisite for its integration as an alternative mode to support healthcare delivery services. Telehealth has been defined as the use of advanced telecommunication technologies to exchange health information and provide health care services across geographic, time, social, and cultural barriers (Reid, 1996). This technology has the potential to increase quality and access to healthcare and to lower costs (Bashshur, Reardon & Shannon, 2000; Ohinmaa, Hailey & Roine, 1999).

In Canada, several telehealth networks have been implemented over the last two decades with relative success (Noorani & Picot, 2001). However, the diffusion of this technology in the Canadian healthcare system is still limited. Telehealth is considered as a major innovation not only at the technological level, but also at the cultural and social levels (Bashshur, Reardon & Shannon, 2000; Klecun-Dabrowska & Cornford, 2001). As Bangert and collaborators (1999) have stated, telehealth represents a ‘paradigm shift’ and is expected to impact all levels of healthcare organisations. Major barriers impeding the integration of telehealth networks in healthcare services organisations have been identified with respect to financial and technological dimensions, but other factors are also limiting the diffusion of telehealth applications, such as physicians’ unfamiliarity with the technology and ineffective change management (Sheng et al., 1999). Furthermore, some institutional barriers related to organisation cultures, medical norms, and operating systems may limit the diffusion of telehealth into integrated networks across regional and national boundaries (Robinson, Savage & Sydow-Campbell, 2003). Ultimately, the success or failure of the implementation of telehealth mainly depends upon human and organisational factors (Perednia & Allen, 1995; Aas, 2001).

Specifically, the foreseen benefits of telehealth technology with respect to healthcare access, quality, and continuity are unlikely to materialise without its integration into clinical practice (Bashshur, Reardon & Shannon, 2001; Perednia & Allen, 1995). Physicians are the main end-users of telehealth; thus their acceptance of this technology represents a major challenge for the sustainability of telehealth networks. Furthermore, the introduction of telehealth blurs the delimitation between healthcare organisations and induces a redefinition of traditional professional roles (Bangert, Doktor & Warren, 1999; Aas, 2001; May et al., 2001). Hence, telehealth introduction lies upon the conjunction of several factors. At the individual level, some studies have investigated the psychosocial determinants of telehealth adoption by physicians (Croteau & Vieru, 2002; Hu et al., 1999). Furthermore, Chau & Hu (1999) have proposed a hierarchical model of telehealth acceptance that includes dimensions at the individual, technological, and organisational levels. In this model, variables at the organisation level were related to peer influence on physicians’ decision to adopt telehealth and on the compatibility of the technology with physicians’ current work practice. However, these variables were measured at the individual level from the physicians’ perspective.

Review of the literature

Little research has been done to understand the impact of hospitals’ characteristics on telehealth adoption from an organisational perspective. For instance, studies have investigated organisational characteristics associated with telehealth adoption in the Hong Kong healthcare system (Sheng et al., 1999; Hu, Chau & Sheng, 1999). The availability of in-house technological support, the presence of telehealth champions in the hospital, and the management of change at the individual and organisational levels were identified as facilitators for telehealth integration. Moreover, these analyses suggest that telehealth networks that were based upon and respected each of the involved organisations’ competencies were more likely to be efficient.

According to Whitten & Allen (1996), the principal factors of success of a telehealth network in Kansas were the public-funding of the project by the state, the promotion of telehealth in organisations and the presence of administrative structures supporting the use of the technology. However, other factors, such as the lack of a central leadership and difficulties related to telehealth services reimbursement, have limited the diffusion of telehealth. These authors also argue that the context in which each telehealth program is implemented has an important role to play and should be carefully examined instead of assuming that there is only one ‘best’ organisational structure that fits all programs.

The evaluation of two telehealth networks in Michigan (Whitten & Adams, 2003) has revealed many organisational aspects affecting programs success. For instance, commitment of telehealth leader, logistical support from the organisation, as well as autonomy and self-sustainability of one of the program were found to be major conditions for success. Oppositely, lack of financing, technical, and human resources were identified as the principal factors limiting success of the other telehealth program.

In Quebec, various telehealth programs have been evaluated over the last decade. For instance, Fortin et al. (2003) indicate that a telehealth network that is based upon the needs of the requestor hospital and respects the functioning of healthcare services in place is more likely to be successful. Their evaluation has also pointed out the importance of adapting organisational structures to this new mode of healthcare delivery as well as the need for supporting innovation at all management levels. In the telehealth network evaluated by Sicotte and collaborators (1999), factors related to participating organisations, such as administrative restructuring and the availability of adequate expertise in requestor centres have seriously limited telehealth services utilisation. Conversely, important geographical distance between one of the requestor centres and the reference centre has contributed to a greater telehealth utilisation.

Such reports are of great value to understand why a given telehealth program has been successful or not. However, the results of these studies apply to unique experiments and are thus hard to generalise. Furthermore, most of the studies on the organisational dimensions affecting telehealth adoption were not anchored in explicit theoretical models. Therefore, some of the dimensions pertaining to healthcare organisations’ characteristics that could possibly influence telehealth adoption might have been overlooked in past studies.

The field of management has a long research tradition on the adoption of technological innovations by different types of organisations. Among the theoretical models that have been used to investigate the organisational characteristics influencing technology adoption, Mintzberg’s configuration theory (1979) and institutional theory (DiMaggio & Powell, 1983; Meyer & Rowan, 1977) propose relevant concepts to analyse the relationships between hospitals’ organisational structures and the process of telehealth integration.

Among the studies that have examined the influence of organisations’ characteristics on the adoption of innovations, Kimberly & Evanisko (1981) used a multidimensional model to investigate factors influencing administrative and clinical innovations adoption by healthcare organisations. Their framework included variables at the individual and organisational levels. Autonomy (decentralization of decision making), specialisation (number of medical specialities), organisation’s size (number of beds) and functional differentiation (number of work subunits) had a positive correlation with adoption of technological innovations. Conversely, external integration, that is the intensity of communication and liaison mechanisms, did not significantly influence adoption of technological innovations. With respect to contextual variables, high competition (measured by the number of hospitals in the region) and the size of the community were not significant predictors of innovativeness, whereas age of the hospital had a positive effect on adoption of technological innovations.

In Quebec, a multidimensional qualitative study (Lapointe, 1999; Lapointe & Rivard, 1999) has analysed individual, professional and organisational factors that have influenced adoption of a hospital information system. Horizontal specialisation, measured by the number of medical specialities in the hospital, and a weak formalisation of procedures were negatively associated with system adoption. Rarity of technological resources, complexity of external interventions and financing, associated with the structure of the healthcare system, had also a negative influence on adoption. However, the formation of coalitions around specific issues had a positive or a negative influence on adoption, depending on which group of professionals was considered. Ultimately, the power exerted by physicians, who resisted the information system, constituted an important obstacle to adoption.

Damanpour (1991) has conducted a meta-analysis of the organisational dimensions involved in the adoption of innovations across various types of organisations. For non-profit public organisations, the analysis suggests that specialisation, vertical differentiation, and managerial attitude towards change are positively associated with innovation adoption, whereas centralisation and formalisation have negative influence.

Finally, the study of Prasad & Prasad (1994) has underlined the predominant influence of the ideology of professionalism on the adoption of information systems by healthcare professionals. According to these authors, technology adoption in healthcare organisations is not only influenced by instrumental considerations such as efficiency, performance, and profitability. They have adopted an institutional theory perspective that has allowed for considering non-instrumental factors, such as symbolic, cultural and political aspects involved in the processes of work computerisation in hospitals.

This study is part of a larger research project aiming at exploring individual, professional, organisational, and contextual dimensions that have influenced the adoption of telehealth technology within the context of a large provincial network in Quebec (Canada). The present paper focuses on the study of the organisational characteristics only. This study involved three stages. Firstly, a framework of organisational determinants of telehealth adoption was developed from a combination of theoretical concepts. Secondly, based upon this framework, an exploratory survey was conducted among the 32 hospitals involved as services requestors in the Extended Telehealth Network of Quebec to identify organisational characteristics influencing telehealth adoption. Finally, nine hospitals were selected to represent the various categories of telehealth adopters and a multiple case study was conducted via in-depth interviews with various actors involved in telehealth decisions and activities. Results from these multiple sources were then triangulated to bring out some key conditions for the diffusion of telehealth into integrated networks.

In this paper, a presentation of the telehealth project under study is first provided. Second, operationalization of theoretical concepts and research hypotheses are presented. The methodology of the exploratory study and the multiple case study is described in the third place. Fourth, the results of both studies are presented and discussed, followed by a triangulation of these results. Finally, this paper identifies some implications of the results for the integration of telehealth in healthcare systems.

Description of the project

In Canada, healthcare falls under the jurisdiction of provincial governments. In the Province of Quebec, the healthcare system is publicly-funded and organised around a three-levels (local, regional, and supra-regional) structure. The telehealth network under study was launched in 1998 in order to provide specialised services in paediatric cardiology for local and regional medical centres of the Province of Quebec (Canada). This network was funded by the Quebec Ministry of Health and Social Services. Installation of telehealth equipment in hospitals and their connection to the RTSS (the provincial health telecommunication network of Quebec) took place during 1999. Based upon a previous experiment in Eastern Quebec, the first clinical application covered by the telehealth network was distance diagnosis of heart pathologies among new-borns and children (Cloutier, 2000). Four supra-regional university medical centres were involved as telehealth services providers for 32 requestor centres of different health regions of the province. Figure 1 depicts the localisation of the different healthcare centres involved in the network at the onset of the project. In order to encompass a broader diversity of applications, this network has become the Extended Telehealth Network of Quebec (French acronym RQTE) in 2001. Experimentation of various applications in different medical speciality areas was thus anticipated.

The total number of telehealth transmissions performed between January 2000 and December 2002 was 328. However, levels of telehealth utilisation vary considerably between hospitals. Some of the centres did not use telehealth equipment at all during this period, while other hospitals made several telehealth consultations to the reference centres. Thus, from a provincial perspective, integration of telehealth services is far from being optimal. In reaction to this situation, the RQTE management team has decided to remove telehealth installations in some hospitals and to upgrade the equipment in others. The high variation observed in the level of telehealth utilisation suggests that, among the factors influencing telehealth adoption, some may be related to the organisational characteristics of hospitals. Thus, the purpose of this study was to explore the influence of hospitals’ organisational characteristics on telehealth adoption by healthcare centres involved in the RQTE.

Figure 7.1. The Quebec Provincial Extended Telehealth Network

Research model

A synthesis of the literature on the adoption of innovations by hospitals has allowed to propose a comprehensive framework of the organisational dimensions that could influence telehealth adoption. The conceptual model underlying this study was thus developed from a combination of theoretical concepts used in previous studies; they were adapted to the specific nature of the innovation under study (telehealth), the particular context in which this innovation was implemented (healthcare organisations), and the larger context in which the experimentation took place (the socio-political environment). Furthermore, following interviews with telehealth actors, other concepts were found that added to the understanding of the factors influencing telehealth adoption.

Theoretical concepts and research hypotheses

Results from the consulted studies were coupled with organisational characteristics proposed by Mintzberg’s configuration theory (1979). The structural components of the professional bureaucracy, the type of configuration usually found in healthcare organisations, are defined in Table 1a. Furthermore, concepts pertaining to the context in which telehealth was introduced, inspired by the institutional theory (DiMaggio & Powell, 1983; Meyer & Rowan, 1977), are described in Table 1b. Based upon results of previous studies (Kimberly & Evanisko, 1981; Damanpour, 1991), research hypotheses on the expected influence of each dimension on telehealth adoption are also presented for each of the structural and contextual variables.

Table 7.1.a) Structural variables and research hypotheses

Variable

Description

Hypothesis

Horizontal specialization

The division of work is negotiated between the various specialties rather than on a hierarchical basis.

1. Horizontal specialization has a negative influence on telehealth adoption [24-25].

Functional differentiation

Differentiation, that is how the work is divided, is based upon production units, or fields of expertise.

2. The influence of functional differentiation on telehealth adoption depends on groups’ values towards the system [24-25].

Size of units

The size of units depends on the clientele size because professionals are grouped together according to their expertise.

3. The size of units has an undetermined influence on telehealth adoption.

Planning and control systems

Professionals try to exert a collective power on administrative decisions. For this reason, physicians often hold administrative positions in hospitals and central control is thus limited.

4. Few planning and control systems have a negative influence on telehealth adoption [24-25].

Internal communications

Information exchanges are mostly informal and the use of formal communication mechanisms is limited in the operational core. Permanent committees and taskforce groups represent a form of communication mechanism.

5. Internal communications mechanisms have a variable influence on telehealth adoption [24-25].

Decentralization of power

Informal power is both vertically and horizontally decentralized. Power is dispersed towards the bottom of the hierarchical chain and professionals exert a control over decision processes.

6. Decentralization of power has a variable influence on telehealth adoption, depending on physicians’ values towards the technology [24-25].

Table 7.1.b) Contextual variables and research hypotheses

Variable

Description

Hypothesis

Age

The number of years of the hospital’s existence under its current name.

7. Older hospitals are more likely to adopt telehealth [23, 24-25].

Size

The relationship between hospital’s size and adoption of innovations depends on the type of innovation considered. In the case of telehealth, viewed as a tool to support healthcare delivery in undeserved communities, a negative relationship is expected.

8. Smaller hospitals are more likely to adopt telehealth.

Competition

The number of hospitals in the health region.

9. Hospitals in regions where there are few other hospitals are more likely to adopt telehealth.

Localization

Hospitals in the Province of Quebec are located in urban, outlying, remote or isolated regions. Since telehealth is usually viewed as a tool to support healthcare delivery in undeserved areas, remote and isolated regions are considered as the first telehealth users.

10. Hospitals located in remote and isolated regions are more likely to adopt telehealth.

Interorganisational relationships

Relationships are already existing between the organizations involved in the telehealth network because physicians from university hospitals practice as visiting specialists in many regional hospitals.

11. Hospitals receiving visiting physicians from university medical centers are more likely to adopt telehealth.

Methods

Exploratory study

The study protocol and data collection instruments were reviewed and approved by the ethical committee of the university where the research was conducted. A short open-ended questionnaire was selected as the first data collection method since it enabled the compilation of specific information on each hospital’s structural and cultural characteristics. Questionnaires used in previous studies helped operationalise theoretical concepts; they were though adapted to the specific context of the telehealth network under study. Two of the researchers (MPG and LL) proceeded to the elaboration of the questionnaire that was subsequently validated by two of the co-researchers involved in telehealth networks (JPF and AC).

The study questionnaire comprised a total of 22 short questions. Horizontal specialisation was assessed by calculating the number of different medical specialities in the hospital. The three aspects measuring functional differentiation were the number of departments, units, and services in the hospital. Size of units was measured by asking the respondents to indicate the mean number of employees working in each department, unit, and service of the hospital. The influence of planning and control systems was assessed by means of three questions: the first pertained to the presence of formalism in the organisation of medical work in general, the second dealt with the presence of formalism with respect to telehealth utilisation, and the third measured the number of administrative positions held by physicians. Internal communications were assessed by asking if committees or taskforce groups involving physicians were set up to participate in hospital decision-making in general and in telehealth decision-making. Three questions were used to measure decentralization of power: “Do physicians participate actively in administrative decisions?”; “Do physicians participate actively in decisions concerning the clinical work?”; and “Are physicians directly involved in decisions pertaining to telehealth?”.

Some contextual factors were also measured in the questionnaire. Age of the organisation was calculated by asking respondents since when their hospital existed under its current name. Size of the hospital was calculated in two different ways: the total number of beds and the number of admissions during the last year. Two questions assessed competition: the total number of hospitals in the health region and the fact of being a reference healthcare centre for the region. Interorganisational relationships were measured by the number of visiting physicians (physicians who provide services for certain specialities uncovered locally) coming at the hospital annually. Furthermore, the number among those visiting physicians who came from the telehealth reference centre was calculated. Localisation was measured by determining if the hospital was in one of these three categories of region: 1) isolated or remote; 2) outlying; 3) urban.

Finally, three questions dealt with the level of telehealth adoption in the hospital. The first one measured the total number of telehealth transmissions from January 2000 to December 2002. The second question assessed if an upgrading of the telehealth equipment was planned and the third question referred to whether telehealth equipment had been removed or was planned to be removed from the hospital.

Study population and data collection

All the contacted hospitals (32) agreed to participate in the study. Medical Directors of the 32 healthcare centres involved as telehealth services requestors in the RQTE were selected as key informants for data collection since they hold a position at the conjunction of the hospital’s administrative and clinical spheres. A copy of the questionnaire was sent by fax and/or electronic mail before the telephone interviews. For the purpose of the exploratory study, questionnaires were completed via semi-directed telephone interviews. Interviews aimed at completing the information gathered in the questionnaires and at documenting the specific context of each hospital with respect to its participation in telehealth activities. Furthermore, these interviews have allowed for identifying the principal actors involved in telehealth activities in each hospital. Length of interviews varied from 5 to 25 minutes and they were not recorded since the responses were usually short and could be inscribed directly on a form. Data of the 32 completed questionnaires were compiled and analysed via SPSS.

In 25 of the 32 surveyed hospitals, the Medical Director was interviewed. In other cases, the Director’s assistant was interviewed. At some occasions, key informants’ knowledge on telehealth activities was limited and thus, other informants, such as the Chief of Information Systems, were consulted in order to complete the questions dealing with this topic.

Exploratory analyses

Descriptive qualitative analyses of the completed questionnaires were first performed in order to create response categories for each of the theoretical concepts. Secondly, simple descriptive statistics such as mean and mode were calculated to determine the range of each category. Variables were then coded into these emergent categories in order to proceed to statistical analyses. Finally, measures of association were performed via separate chi-square tests (Fisher’s exact test was used when one cell was smaller than 5) to identify which organisational characteristics were associated with telehealth adoption.

Multiple case study

In order to enrich the understanding of the organisational dimensions affecting telehealth adoption in hospitals involved as services requestors in the RQTE, a multiple case study was conducted. Case study constitutes an appropriate method of inquiry to take into account the contextual conditions of a given phenomenon and relies on multiple sources of evidence (Yin, 1993; Stake, 1995). In-depth semi-structured interviews with key informants is considered an appropriate data collection method for case studies (Reidy & Mercier, 1996). However, the large number of hospitals in the network limited the feasibility of conducing in-depth interviews with a significant number of actors involved in telehealth activities within each hospital. Thus, a selection of hospitals, representing the different “cases”, was made based on the following specific criteria. The results of the exploratory study have allowed for the identification of categories of telehealth adopters among the studied hospitals. Hence, hospitals have been categorised as being either adopters or non-adopters of telehealth. Furthermore, the size of hospitals as well as their localisation were used as selection variables because of their significant effect on telehealth adoption, as indicated from the exploratory study results. In addition, participation of physicians in telehealth decision-making was also considered to discriminate hospitals. However, the number of units as well as the number of physicians holding administrative positions in the hospital were not considered as selection variables since they were associated with hospital’s size.

Triangulation of the results

According to Stake (1995), protocols that are used to ensure accuracy and alternative explanations of research findings are called triangulation. The need for triangulation arises from the ethical need to confirm the validity of the study processes. Triangulation has been employed in the study of information technology adoption in hospital by Kaplan & Duchon (1988). In their study, mixing qualitative and quantitative data from different sources has brought new insights and modes of analysis that would have been omitted with a single method.

In this research, a triangulation of the results from the exploratory and multiple case studies was thus performed in order to improve the validity of the findings and to provide a richer understanding of the organisational dimensions that have influenced telehealth adoption by hospitals. A qualitative method was applied to triangulate the results from both studies. To do so, the content of interviews was classified according to the different organisational dimensions affecting telehealth adoption identified from the exploratory study. Convergence and divergence between these observations were noted and new insights gained from interviews were highlighted.

Selection of hospitals for case study

Nine hospitals were thus selected from the analysis of the exploratory survey results. Among those hospitals, five were considered as being telehealth adopters because they have had at least one telehealth consultation during the experimentation period. Three out of these five hospitals were located in a remote or isolated region (each one from a different geographical area of the Province of Quebec), the fourth hospital was located in an outlying region, whereas the last one was located in an urban area. With respect to size, three of these hospitals had less than 200 beds, while the other two were considered as large hospitals. The two larger hospitals had physicians participating in telehealth decision-making.

Telehealth non-adopters were represented by four hospitals of the sample. One of them was a large hospital located in an urban area. The second hospital had also more than 200 beds, but was located in an outlying region whereas the last two were small hospitals located in remote or isolated regions. Finally, in three of these hospitals, physicians were participating to decisions related to telehealth. Table 7.2. provides information on the characteristics of the selected hospitals.

Table 7.2. Characteristics of hospitals in the multiple case study

Hospital

Status

Localisation

Size

Physicians’ participation in telehealth decisions

Type and number of respondents

Hospital 1

Adopter

Remote region

Small

No

2 Physicians

1 Administrator

Hospital 2

Adopter

Remote region

Small

No

2 Physicians

1 Administrator

Hospital 3

Adopter

Remote region

Small

No

1 Physician

1 Administrator

Hospital 4

Adopter

Outlying region

Large

Yes

1 Physician

2 Administrators

Hospital 5

Adopter

Urban region

Large

Yes

2 Physicians

1 Administrator

Hospital 6

Non adopter

Remote region

Small

No

1 Physician

1 Administrator

Hospital 7

Non adopter

Remote region

Small

Yes

2 Physicians

1 Administrator

Hospital 8

Non adopter

Outlying region

Large

Yes

2 Physicians

1 Administrator

Hospital 9

Non adopter

Urban region

Large

Yes

1 Physician

1 Administrator

Selection of key informants for interviews

The exploratory survey has permitted to identify the principal actors involved in telehealth activities in each hospital. In the selected hospitals of the multiple case study, all of the identified persons were contacted and were asked to indicate if other persons in their hospital were also involved in telehealth decisions or activities, following the snowball sampling method (Goodman, 1961). This method solicits referrals from initial subjects to generate additional subjects. As presented in Table 2, the number of interviewees in each of the sampled hospitals varied from two to four and consists of administrators as well as physicians.

Each potential respondent was contacted by the investigator in charge of the interviews (MPG) and a meeting was planned at the respondent’s location, excepted for two interviews that were conducted via telephone. A total of 24 persons were contacted and all agreed to participate. Interviews lasted from 45 minutes to one hour and 45 minutes and were tape recorded, with the consent of all participants.

Analysis of interviews with key informants

The method proposed by Huberman & Miles (1994) was followed in order to analyse the qualitative data that consisted of the verbatim transcripts of interviews, together with researcher’s field notes and results from the exploratory study. The Nu*dist Vivo software (Qualitative Research Solution, Pty Ltd, Australia) was used for data codification, classification and treatment. In a first step, all interviews were read to extract some general impressions. In a second step, codification nodes were created and organised in a logical tree, according to the variables of the theoretical framework. Free nodes were also employed to classify interviews’ content in emerging themes. Then, the content of interviews was grouped into categories and linked to the theoretical concepts. Finally, a triangulation of the results [38] from the exploratory survey and the multiple case study was carried out to obtain a more comprehensive understanding of the organisational dimensions affecting telehealth adoption by hospitals.

Results

Description of the surveyed hospitals

A description of the structural characteristics of the hospitals involved as services requestors in the RQTE is presented in Table 7.3.

Table 7.3. Structural characteristics of hospitals

Structural Characteristics

Number of hospitals

Specialisation

Low (1 – 20)

High (21 – 36)

18

14

Functional differentiation

Small (< 34)

Large (≥ 35)

18

14

Size of units

Small (< 40)

Large (≥ 40)

14

18

Planning and control (general)

Yes

No

32

0

Planning and control (telehealth)

Yes

No

19

13

Physicians in administrative positions

1

2 – 3

4 and more

13

14

5

Internal communications (general)

Yes

No

32

0

Internal communications (telehealth)

Yes

No

7

25

Decentralization (administrative)

Yes

No

31

1

Decentralization (clinical)

Yes

No

32

0

Decentralization (telehealth)

Yes

No

23

9

The number of different medical specialities in the hospitals varies from one to 36. Hospitals having between one and 20 specialities are considered as having a low degree of specialisation; on the opposite, hospitals having between 21 and 36 specialities are considered as having a high degree of specialisation. The functional differentiation score is the total number of work subunits (department, services, and units) in the hospital. Two categories are distinguished: small (between one and 34 subunits) and large (35 subunits or more).

With respect to size of units, it was difficult to obtain the mean number of employees working in each department, service or unit because of considerable differences between the size of each subunit in a single hospital. Thus, a mean global score was computed by dividing the total number of employees in the hospital by the total number of subunits. The size of units in each hospital is categorised as being large (40 employees per unit or more) or small (less than 40 employees per unit). The degree of planning and control is considered for these two specific domains: organisation of the medical work in general and telehealth utilisation. Most of the hospitals are described as being rather formal with respect to the organisation of medical work in general. However, there is less formalism with respect to telehealth utilisation. Furthermore, the number of physicians holding administrative positions in the organisation is presented as a measure of medical control over hospital’s decisions. The concept of internal communications is divided in two dimensions: general and telehealth-specific. Decentralization of power is also measured for different types of decision-making: administrative, clinical in general, and telehealth-specific.

Contextual variables are presented in Table 7.4. For hospital’s age, categories are: “less than 10 years”; “between 10 and 29 years”; and “30 years and over”. With respect to size, two measures are presented: the number of beds and the annual number of admissions. Large hospitals are those having 200 beds or more, while small hospitals have less than 200 beds. The annual number of admissions is categorised in two: 10,000 or less and more than 10,000. Nevertheless, these two measures are treated as distinct variables since they could have a different effect on telehealth adoption. More specifically, a high volume of patients in a hospital with a limited number of beds could stimulate the utilisation of telehealth. Furthermore, both measures of the competition concept are also analysed separately since they could influence telehealth adoption in a different way. The total number of hospitals in the region assesses the population density while the “reference centre status” is related to the concepts of specialisation and size. Hospitals are considered has having interorganisational relationships if they receive at least five visiting specialists annually. The strength of existing relationships with telehealth reference centres is also calculated by the proportion of visiting physicians coming from the reference centre. A high proportion is found when 20% of the visiting specialists or more are practising at a telehealth reference centre. With respect to hospitals’ localisation, the three categories represented are urban, outlying, and remote or isolated.

Finally, hospitals are categorised as telehealth adopters if they have made at least one telehealth transmission over the observation period (January 2000 – June 2003). Moreover, the number of hospitals for which an upgrade of telehealth equipment is foreseen and those for which a decision to remove telehealth equipment has been made are given in Table 7.4.

Table 7.4. Contextual characteristics of hospitals

Contextual Characteristics

Number of hospitals

Hospital’s age

Less than 10 years

Between 10 - 29 years

30 years and above

10

10

10

Number of beds

Small (< 200)

Large (≥ 200)

13

19

Annual number of admissions

10 000 or less

More than 10 000

19

13

Number of hospitals in the region

4 or less

5 and more

15

17

Number of visiting specialists

4 or less

More than 4

8

24

Visiting specialists from reference centre

Less than 20%

20% and more

25

7

Regional reference centre status

Yes

No

21

11

Localisation of hospitals

Urban

Outlying

Remote or isolated

6

12

14

Telehealth adoption

Yes

No

19

13

Decision to upgrade telehealth equipment

Yes

No

19

13

Decision to remove telehealth equipment

Yes

No

10

22

Associations between organisational dimensions and telehealth adoption

A series of contingency analyses were performed using the chi-square (χ2) statistics in order to identify significant relationships between hospitals’ structural and contextual variables and telehealth adoption. Table 7.5. shows the χ2 value for each of the predictor variables and indicates which predictors have a significant impact on telehealth adoption by hospitals. For the variables with less than five observations for one cell or more, the Fisher’s exact test was used since this measure is not affected by cell size.

Statistical tests were not conducted for some of the structural and contextual variables measured. Specifically, the impact of general internal communications could not be assessed since committees and taskforce groups involved in hospital decision-making in general are present in every hospital. Thus, only internal communication mechanisms pertaining to telehealth are considered. Moreover, since the degree of formalism with respect to work in general was equivalent between hospitals, only formalism related to telehealth utilisation is analysed. Decentralization of administrative decision-making and clinical decision-making in general are neither considered because all but one hospital have reported physicians’ participation in those decisions. However, the effect of decentralization with respect to telehealth decision-making is reported.

According to Table 7.5., significant associations are observed between telehealth adoption and the following structural characteristics of hospitals. A smaller number of subunits (functional differentiation) is positively associated with telehealth adoption, whereas the number of physicians in administrative positions (planning and control) and the participation of physicians in telehealth decision-making (decentralization) are negatively associated with telehealth adoption. With respect to contextual dimensions, smaller hospitals and hospitals with less than 10,000 annual admissions (size) are more likely to adopt telehealth. Hospital’s localisation is also significantly associated with telehealth adoption: most of the hospitals located in remote or isolated regions are telehealth adopters. A significant association is also found between telehealth adoption and the decision to upgrade or to remove telehealth equipment. Comparison of results with research hypotheses is presented in the Discussion section.

Table 7.5. Associations between telehealth adoption and hospitals’ characteristics

Structural and contextual characteristics

Adopter

Non-adopter

χ2

Specialisation

Low (1 – 20)

High (21 – 36)

13

6

5

8

2.82

Functional differentiation

Small (< 34)

Large ( ≥ 35)

14

5

4

9

5.78*

Size of units

Small (< 40)

Large (≥ 40)

11

8

3

10

3.80

Planning and control (telehealth)

Yes

No

10

9

9

4

0.88

Physicians in administrative positions

1

2 – 3

4 and more

10

9

0

3

5

5

9.11*

Internal communications (telehealth)

Yes

No

3

16

4

9

1.01

Decentralization (telehealth)

Yes

No

11

8

12

1

4.52*

Hospital’s age

< 10 years

10 – 29 years

≥ 30 years

5

4

10

5

6

2

4.78

Number of beds

Small (< 200)

Large (≥ 200)

2

13

11

6

8.72**

Annual number of admissions

≤ 10 000

> 10 000

14

5

5

8

3.97*

Number of hospitals in the region

4 or less

5 and more

10

9

5

8

0.62

Number of visiting specialists

0 to 4

More than 4

4

15

7

6

3.68

Visiting specialists from reference centre

Less than 20%

20% and more

15

4

10

3

0.02

Regional reference centre status

Yes

No

13

6

8

5

0.16

Localisation of hospitals

Urban

Outlying

Remote/isolated

1

7

11

5

5

3

6.68*

Decision to upgrade equipment

Yes

No

17

2

2

11

17.57**

Decision to remove equipment

Yes

No

4

18

9

1

14.70**

* p < .05

** p < .01

Observations from the multiple case study

This section presents the major observations pertaining to organisational dimensions that have influenced telehealth adoption drawn from the multiple case study. Whenever possible, content of interviews conducted with clinicians and administrators involved in telehealth activities was categorised in accordance with study hypotheses. However, some observations gained from the interviews could not be linked to any of the theoretical dimensions and are thus presented as additional variables to incorporate into the conceptual framework.

As demonstrated with the exploratory study results, the size as well as the localisation of the hospital have greatly influenced the decision of requiring telehealth services from a reference centre. When a hospital was located at a reasonable driving distance from the reference centre, the use of telehealth for paediatric cardiology consultations was considered more complex than transferring the patient directly. In larger hospitals, located in the suburbs or in urban areas, the specialised resources were also sufficient to provide most of the services on site.

Another aspect to consider was the availability of the necessary equipment and human resources to perform distance echocardiography for new-borns. Indeed, some of the hospitals did not have the appropriate probe for paediatric echocardiography. In other cases, technicians were not adequately trained for the manipulation of this probe, or the volume of echocardiography for new-borns was too low to keep the technician’s skills up to date. As one physician has mentioned: “ The resolution of images is not adapted for real-time consultations... I can see the images but it’s someone else who manipulates. So it’s hard... Also, one must trust the technician on the other side. It’s operator-dependant ... ” (physician, Hospital 7). However, the lack of resources in a hospital with respect to some medical specialities could also limit telehealth utilisation. More specifically, in hospitals where there was no paediatrician, the telehealth network has not been used since problematic cases were automatically transferred.

In some of the hospitals, telehealth was considered a priority by top management. The Medical Director was often the most important telehealth decision-maker. However, as the Medical Director from an adopting hospital has stated, financial constraints on the hospital have limited the availability of resources to support telehealth development: “ We have a demand from the paediatricians [ ... ] and we need to give them support. However, an organisation has to face many choices and our budget is not infinite, so sometimes we have to make painful decisions. If we decide to put an amount of money to support telehealth, it’s someone else or another service that won’t benefit from it. ” (administrator, Hospital 4). In other cases, administrators that promoted telehealth utilisation have faced physicians resistance: “ On many occasions, I have taken actions to stimulate physicians, but I really had the impression of preaching in the desert! Physicians of the hospital do not see what they can gain from using telehealth versus the efforts they would need to make in order to become familiar with this technology and to develop ways of functioning with it. They don’t want to make this investment because they don’t perceive any benefit to do so. ” (administrator, Hospital 9).

Participation to the telehealth network was clearly the result of a need from the paediatricians in most of the adopting hospitals. However, even if telehealth was adopted, physicians’ participation in decision making related to telehealth was still limited in those hospitals. As one administrator (Hospital 3) pointed out: “ Now we know who are the key actors. We wanted to set up a committee with paediatricians, but sometimes there are conflicts between them and the cardiologists... That can be a sign of the fact that the project is not going as well as expected [ ... ] . This project has started because of management’s will who was convinced of its utility [ ... ] but not necessarily from users’ demand.

Conversely, in hospitals where the technology was not used, physicians appeared to have had influence. In some cases, physicians resisted telehealth utilisation because of a negative previous experience with immature telehealth technology: “ People remain with a kind of bitter taste: you know when you start something and it turns bad, people always recall where you’ve failed. They don’t recall what worked well. ” (physician, Hospital 6). Some physicians perceived that top management did not adequately address their interests when implementing the new technology: “ They [the management] have consulted us, but did not take our concerns into account... The organisation did not put in place the needed infrastructure. Who is responsible for the equipment, for its functioning?... ” (physician, Hospital 8).

Similarly, administrators were not consulted on their needs for a telehealth network at the onset of the project: “ At one point in time, we received equipment, that we had never ordered. We have unpacked the boxes and the instruction book arrived a week or two later, stating that we were part of the Quebec Paediatric Telehealth Network, one of the 34 or 36 chosen hospitals! That’s how it all started [ ... ] Later, they have written or phoned us to see how we were using it and there was also almost a formal demand because some hospitals were not using it at all. So, we had already started using it and we have sustained utilisation, but I think that we are the only one in [the health region] . ” (administrator, Hospital 4).

Technology quality, performance, ease of use, and conviviality were other major themes emerging from the interviews with clinicians and administrators as well. Even in hospitals with a high level of telehealth utilisation, technology improvements are needed to ensure the integration of telehealth into clinical practices. As one physician has indicated: “ Telehealth utilisation requires logistics and physicians are not eager to organise it themselves. For physicians, it needs to be convivial, easy to use, and functional. If you don’t have that, physicians get fed up and don’t want to use it anymore ”. (physician, Hospital 1).

Another important observation drawn from the interviews was the large gap between telehealth logistical support requirements and human resources affected to the scheduling of telehealth consultations, installation of the technical devices, and supervision of the equipment during consultations. Also, clinicians as well as administrators have mentioned the need for a list of services providers and a schedule for teleconsultations.

According to two administrators, one of the major limits to telehealth utilisation was the lack of support from the regional health authority. Conversely, as one of the administrators of an adopting hospital has pointed out, an integrated development of telehealth at the regional level was a key element of the project’s success: “ So, [the telehealth network] is interesting for our patients, for the general population, because our orientation as an institution is to reach the largest autonomy possible, while maintaining the critical level that allows us to offer high quality services, and this is supported by our regional health authority. ” (administrator, Hospital 5). Moreover, some of the respondents perceived telehealth as a means to rationalise services on a regional basis. One physician (Hospital 2) has even changed his referral patterns since the introduction of telehealth: “ Because of the good service in cardiology, we are now referring to [the supra-regional reference centre]. I think that’s an advantage for health care rationalisation because each hospital’s mission is respected .”.

The potential impact of telehealth on clinicians’ workload also deserves attention. Even if a potential increase in workload has been mentioned by many interviewed physicians, those who actually use telehealth on a routine basis have a different opinion. According to this physician: “ Medical work in a remote region is far more laborious than elsewhere. We are often more innovative in many ways, so we are used to it. When one really wants something, one has to work for it. ” (Hospital 1). In remote hospitals, telehealth can be a tool for reducing uncertainty and reassuring physicians because transfer decisions are often their responsibility: “ The few times where it has not been perfectly precise, we knew at least if something serious was going on or not, so we were able to wait [ ... ] . We never remained uncertain. ” (physician, Hospital 3). The following section discusses some implications of these results for the management of telehealth networks.

Discussion

Verification of research hypothesis

With respect to baseline research hypotheses, findings from the exploratory study provide support for Hypotheses 2, 4, 6, 8, and 10. However, Hypotheses 1, 3, 5, 7, 9, and 11 were not supported by the results. The structural variables that seem to have a determinant effect on telehealth adoption are functional differentiation, planning and control, and decentralization with respect to telehealth. Thus, Hypothesis 2 is confirmed since there is an association between the number of subunits in a hospital and telehealth adoption that depends on professional groups’ values towards the technology. In the present study, this association is negative because hospitals with a few number of units are more likely to be telehealth adopters. Telehealth utilisation must receive support from the specialists of all clinical units in order to be adopted broadly by the hospital. Thus, the technology could be integrated more easily when the number of clinical departments and services was rather small. In a study on hospitals’ adoption of clinical information system (Lapointe, 1999; Lapointe & Rivard, 1999), the division of work by medical specialities was found to have a variable effect, depending on the willingness of each group of specialists to use the system.

The number of physicians in administrative positions, a dimension related to planning and control systems, is significantly associated with telehealth adoption. Thus, Hypothesis 4 is supported since the limited managerial control in hospitals has a negative influence on telehealth adoption. As a professional group, physicians influence general decision-making in hospitals but they also exert a direct control over administrative decisions because some physicians hold administrative positions (Lapointe, 1999) In this study, telehealth adoption is negatively associated with physicians’ control over administrative decisions since the totality of hospitals having four physicians or more in administrative position are non-adopters of telehealth. Succi & Lee (1998) have mentioned that the direct involvement of physicians in hospital management decisions constitutes an opportunity for them to advocate their interests. The present study suggests that telehealth applications in larger hospitals may be inconsistent with physicians’ interests. Thus, in hospitals where physicians can exert directly their power by holding strategic positions, such as in larger hospitals, resistance to telehealth might be harder to overcome.

Similarly, results of the exploratory survey support Hypothesis 6 concerning the negative effect of decentralization of power on telehealth adoption. Most of the hospitals where physicians are directly involved in telehealth-specific decision making have not adopted telehealth. In those hospitals, current telehealth applications could be inconsistent with physicians’ interests and participating to telehealth decision-making could be a means for them to voice their resistance. Therefore, physicians have an important role to play in decisions related to telehealth and the different groups of specialists involved in telehealth experiments should be consulted when telehealth decisions are made to ensure that their various needs and interests are considered. As Sheng et al. (1999) have suggested, local multidisciplinary telehealth committees could play an important role in the diffusion of the technology to different clinical applications by allowing groups of professionals to express their preference in decisions concerning telehealth developments.

The structural characteristics affecting telehealth adoption are consistent with previous studies on adoption of innovation in hospitals (Kimberly & Evanisko, 1981; Lapointe & Rivard, 1999). When the technology introduced is perceived as threatening professionals’ autonomy and modifying their roles and responsibilities, resistance is expected. Change in the professional bureaucracy does not sweep in from new administrative rules, but seeps in by the slow process of changing professionals’ norms, skills and knowledge (Mintzberg, 1979). However, effective change management strategies are necessary to address physicians’ worries concerning modifications to their professional roles and responsibilities (Sheng et al., 1999). Furthermore, a symbolic support from hospital managers is not sufficient to assure telehealth integration into daily clinical practice; they have to exert actively their power in order to promote change and to facilitate professionals’ utilisation of the technology (Linderoth, 2002).

With respect to contextual characteristics, Hypothesis 8 is verified since a negative relationship was observed between the number of beds and telehealth adoption as well as between annual admissions and telehealth adoption. Furthermore, a correlation of .78 (p < .000) was found between these two variables, indicating that both are similar measures of size. As expected, smaller hospitals were more likely to have adopted telehealth. A similar relationship was found between localisation and telehealth adoption, supporting Hypothesis 9. Specifically, telehealth adoption is more frequent in hospitals located in remote or isolated regions.

Small hospitals are often located in remote areas and do not have a large number of specialists. Therefore, they must rely on the support of other healthcare organisations to deliver specialised services. Telehealth is perceived as an alternate mode of healthcare services delivery to visiting specialists from regional or supra-regional medical centres in remote hospitals. Thus, for hospitals depending on services from other centres, telehealth utilisation is more coherent with usual ways of functioning. Conversely, hospitals that are independent of other organisations to provide specialised services may perceive telehealth as a threat to their autonomy. In that sense, telehealth networks should respect the mission of each institution involved (Sheng et al., 1999; Hu, Wei & Cheng, 2002). Telehealth adoption is also associated with a decision to upgrade or to remove telehealth equipment. This indicates that the decisions made by the telehealth network managers are consistent with the actual levels of utilisation. However, reasons for non-adoption may differ from one hospital to the other and should be carefully investigated in order to suggest other solutions instead of removing the equipment.

Insights from the triangulation of results

In the present study, triangulation of results was relevant to support, complement, and exemplify exploratory findings with specific situations and experiences that have occurred in the different hospitals. Furthermore, interviews have suggested potential variables to incorporate into the theoretical framework.

From these results, it appears that the specific context of organisations influences the success of a telehealth network. Telehealth is more consistent with the values, purposes and ideologies of smaller hospitals from remote areas. The results of the multiple case study are consistent with this observation. For smaller hospitals from outlying and remote regions telehealth is perceived as an acceptable solution since specialised services in paediatric cardiology necessitated costly and risky emergency transfers. However, prior to using telehealth services, a given hospital needs to have the required equipment and qualified staff, conditions that were lacking in many of the non-adopting hospitals. Nevertheless, other contextual variables, such as the number of visiting specialists and the reference centre status did not appear to influence telehealth adoption neither in the exploratory study nor in the case study. Some factors related to the environment have also emerged from interviews, such as the lack of support from the regional health authority and the impact of telehealth on the rationalisation of services. However, the questionnaire did not address issues related to the involvement of regional health authorities in the process . Thus, more research is needed to explore the relationships between these contextual factors and telehealth adoption.

Interviews with administrators and physicians have indicated a lack of consultation when the network was implemented. Thus, some hospitals have received the equipment without having expressed a need for it. Even in hospitals where the need for telehealth services was explicit, administrators as well as physicians’ involvement in decision making was limited. These observations underline the necessity of taking into account the perceived need for the telehealth services being implemented. The degree of priority given to telehealth by the Medical Director and the physicians should also be taken into account when deciding of a hospital’s involvement in a telehealth network. Furthermore, as observed in the majority of hospitals, no formal consultation was made with administrators and physicians during the implementation process. This could have greatly limited the integration of telehealth as a tool to support healthcare services delivery since the concerns of the most important stakeholders have not been addressed at the different phases of the process (Succi & Lee, 1998). As the results of the exploratory study have shown, physicians’ involvement in telehealth decision making was more frequent in hospitals where participation to the network was not considered a priority, resulting in non-adoption of telehealth.

Like any other information and communication technology, telehealth needs to be perceived as user-friendly in order to be adopted in practice. Undoubtedly, the quality and conviviality of technology is of great importance, but it is equally important to consider the necessary infrastructure to support organisation of telehealth sessions. In interviews, clinicians as well as administrators have mentioned the need for having a list identifying telehealth services and providers available. The fact that clinicians know the consulting specialists and are sure of their willingness to provide a consultation is considered as a condition to the smooth delivery of telehealth services. This aspect has also been pointed out in previous studies of telehealth adoption and underscores the relevance of affecting specific resources to the organisation of teleconsultations for the sustainability of an integrated telehealth network (Whitten & Adams, 2003; Fortin et al., 2003).

The impact of telehealth on medical professionals’ roles and responsibilities was not assessed by the exploratory study. However, interviews with physicians revealed that telehealth was likely to have different effects on their tasks. For instance, telehealth has the potential to decrease uncertainty by providing timely expertise but, at the same time, could increase uncertainty by giving the responsibility of more acute cases to the physician of the remote hospital. Also, telehealth could change the referral patterns of physicians and thus, support the rationalisation of health care delivery on a regional basis. The lack of infrastructure and technical support were also identified as elements increasing the complexity of telehealth utilisation. Nonetheless, the relatively low volume of telehealth activity until now has limited the effects on clinical practice. Subsequent research should include a measure of telehealth impact on the perception of clinicians’ roles and responsibilities, as well as the potential effects on their professional autonomy.

Implications for telehealth diffusion

Some implications for telehealth diffusion may be drawn from this study. As Hu et al. (2002) have stated, the diversity of services provided is a key element of success for telehealth programs. At the onset, selection of hospitals involved in the RQTE was based upon data concerning needs for specialised services in paediatric cardiology. However, the specific characteristics of hospitals and their individual needs with respect to the types of technology and clinical applications to be implemented were not taken into account. In 2001, a needs assessment has been conducted in each of the RQTE hospitals in order to identify speciality areas for which telehealth could be used as a mode of services delivery. However, only a few applications are currently being experimented in addition to paediatric cardiology. Since each of the hospitals involved in the RQTE has specific characteristics, the extension of services offered via the telehealth network should be adapted to their needs. In a previous telehealth experiment in the Magdalene Islands, a remote region of the Province of Quebec, the selection of telehealth services provided was based on the needs of the community and on the complementarity with services available in the local hospital (Fortin et al., 2003). During this experiment, a total of 14 different medical specialities were involved in the telehealth service. Moreover, Sheng et al. (1999) indicate that telehealth networks involve multiple organisations and that contextual and organisational heterogeneity must be identified and addressed by management teams. Thus, considering the specific needs of healthcare organisations involved in the telehealth network seems central for the diffusion of a variety of clinical applications within the RQTE.

Interviews with physicians have also indicated that telehealth could modify professional roles and responsibilities. In spite of a relatively low volume of utilisation, concerns have been expressed with respect to telehealth impacts on clinical practice. Utilisation of the technology is still perceived as being complex and cumbersome. There is a need to address these apprehensions when implementing telehealth clinical applications. Effective technological and logistical support are thus needed to facilitate telehealth utilisation. Besides, telehealth could potentially influence physicians referral patterns; hence, it is important to acknowledge the effects of this technology on professional autonomy, continuity of care, and rationalisation of services.

The results from both the questionnaire and interviews support the observation made by Whitten & Adams (2003) that telehealth programs are not isolated, but located within larger health organisations. Moreover, health care organisations are also positioned in a larger geographical, economical and socio-political environment. Therefore, it is important to investigate the context in which telehealth projects are taking place prior to experimentation.

Conclusion

This exploratory study has permitted to investigate telehealth adoption from an organisational standpoint. Moreover, the concepts used to explore the relationships between telehealth adoption and hospitals’ characteristics were anchored in an integrated theoretical framework. The operationalisation of concepts was carefully adapted to the nature of the technology under study as well as the settings in which it was implemented and the larger socio-political context. Qualitative and quantitative approaches were combined in data collection to get a broader and richer understanding of telehealth introduction in each healthcare organisation. Moreover, a multiple case study was conducted to explore the influence of organisational characteristics on telehealth adoption with more depth. Triangulation of data sources and methods has permitted to support the findings of the exploratory study and has also suggested avenues for further research.

Nevertheless, additional qualitative studies are needed to explore with more depth the dynamics of telehealth introduction in healthcare organisation. Furthermore, studies with advanced quantitative techniques should be used to investigate telehealth adoption in a larger number of hospitals across different provinces or countries in order to analyse organisations’ characteristics more precisely and to explore their impact on telehealth adoption. Finally, this study has highlighted the relevance of considering the characteristics and the dynamics of healthcare organisations at each stage of telehealth implementation in order to take their specific needs into account.

Acknowledgements

The study on which this paper is based was substantially supported by a grant form the Canadian Institutes of Health Research (Project No. 49452). The realisation of this research was also made possible with the support of a doctoral scholarship from the FCAR/FRSQ to Marie-Pierre Gagnon.

References

Aas, I.H.M. (2001). A qualitative study of the organizational consequences of telemedicine. Journal of Telemedicine and Telecare , 7 (1), 18-26.

Bangert, D., Doktor, R. & Warren, J. (1999). Introducing telemedicine as a strategic intent. 32nd Hawaii International Conference on System Sciences , Maui : IEEE Computer Society.

Bashshur, R.L., Reardon, T.G. & Shannon, G.W. (2000). Telemedicine : A new health care delivery system. Annual Review of Public Health , 21, 613-637.

Cloutier, A. (2000). Qu’est-ce que le RQTE ? Bulletin du Réseau québécois de télésanté de l’enfant , 1 (1), 1-2.

Croteau, A-M. & Vieru, D. (2002). Telemedicine adoption by different groups of physicians. 35th Hawaii International Conference on System Sciences , Maui : IEEE Computer Society.

Damanpour, F. (1991). Organizational innovation: a meta-analysis of effects of determinants and moderators. Academy of Management Journal , 34 (3), 555-590.

Denzin, N.K. (1989). The research act : A theoretical introduction to sociological methods, 3rd Edition .Englewook Cliffs, N.J. : Prentice Hall.

DiMaggio, P. & Powell, W. (1983). The iron cage revisited : Institutionnal isomorphism and collective rationality in organizational fields. American Sociological Review , 48, 147-160.

Fortin, J.P., Gagnon, M-P., Cloutier, A.& Labbé, F. (2003). Evaluation of a telemedicine demonstration project in the Magdalene Islands. Journal of Telemedicine and Telecare , 2003, 9 (2), 89-94.

Goodman, L.A. (1961). Snowball sampling. Annals of Mathematical Statistics , 20, 572-579.

Hu, P.J. & Chau, P.Y.K. (1999). Physician acceptance of telemedicine technology : An empirical investigation. Topics in Health Information Management , 19 (4), 20-35.

Hu, P.J., Chau, P.Y.K., Sheng, O.L. & Tam, K.Y. (1999a). Examining the technology acceptance model using physician acceptance of telemedicine technology. Journal of Management Information Systems , 16 (2), 91-112.

Hu, P.J., Chau, P.Y.K. & Sheng, O.L. (2000). Investigation of factors affecting healthcare organization‘s adoption of telemedicine technology. 33rd Hawaii International Conference on System Sciences , Maui : IEEE Computer Society.

Hu, P.J., Wei, C.P.& Cheng, T.H. (2002). Investigating telemedicine developments in Taiwan : Implication for telemedicine program management. 35th Hawaii International Conference on System Sciences , Maui : IEEE Computer Society.

Huberman, A.M. & Miles, M.B. (1994). Qualitative data analysis : An expanded sourcebook, 2nd Edition . Thousand Oaks : Sage Publication.

Kaplan, B. & Duchon, D. (1988). Combining qualitative and quantitative methods in information systems research : A case study. MIS Quarterly , 4, 571-586.

Kimberly, J.R. & Evanisko, M.J. (1981). Organizational innovation : The influence of individual, organizational and contextual factors on hospital adoption of technological and administrative innovations. Academy of Management Journal , 24 (4), 689-713.

Klecun-Dabrowska, E. & Cornford, T. (2001). Evaluation and telehealth : An interpretative study. 34th Hawaii International Conference on System Sciences , Maui : IEEE Computer Society.

Lapointe, L. (1999). L’adoption de systèmes d’information cliniques par les médecins et les infirmières : Une étude des variables individuelles, socio-politiques et organisationnelles . Thèse de doctorat, École des HEC, Université de Montréal.

Lapointe L. & Rivard, S. (1999). L'implantation d'un système d'information clinique. ASAC 99 .

Lehoux, P, Battista, R. & Lance, J.M. (2000). Telemedicine : Passing fad or lasting benefits? Canadian Journal of Public Health , 91 (4), 277-280.

Linderoth, H. (2002). Implementation and evaluation of telemedicine : A Catch 22 ? 35th Hawaii International Conference on System Sciences , Maui : IEEE Computer Society.

May, C.R., Gask, L., Atkinson, T., Ellis, N., Mair, F.S. & Esmail, A. (2001). Resisting and promoting new technologies in clinical practice : the case study of telepsychiatry. Social Science & Medicine , 52 (12), 1889-1901.

Meyer, J. W. & Rowan, B. (1977). Institutionalized organizations : Formal structures as myth and ceremony. American Journal of Sociology , 83 (2), 340-363.

Mintzberg, H. (1979). The structuring of organizations : A synthesis of the research . Englewood Cliffs, NJ : Prentice Hall.

Mitchell, J.G. (2000). Increasing the cost-effectiveness of telehealth by embrasing e-health. Journal of Telemedicine and telecare , 6 (Suppl. 1), 16-19.

Noorani, H.Z.& Picot, J. (2001). Evaluation of telehealth videoconferencing in Canada. Canadian Coordination Office of Health Technology Assessment, Ottawa.

Ohinmaa, A., Hailey, D., & Roine, R. (1999). The assessment of telemedicine : General principles and a systematic review . Edmonton: Finnish Office for Health Care Technology Assessment.

Perednia, D.A. & Allen, A. (1995). Telemedicine technology and clinical applications. Journal of the American Medical Association , 273 (6), 483-487.

Prasad, P. & Prasad, A. (1994). The ideology of professionalism and work computerization: An institutionalist study of technological change. Human Relations , 47 (12), 1433-1458.

Reid, J. (1996). A telemedicine primer: Understanding the issues . Billings, Montana : Artcraft Printers.

Reidy, M. & Mercier, L. (1996). La triangulation. In M-F. Fortin (Dir.) : Le processus de la recherche de la conception à la réalisation . Montréal : Décarie Éditeur.

Robinson, D.F., Savage, G.T. and Sydow-Campbell, K. (2003). Organizational learning, diffusion of innovation, and international collaboration in telemedicine. Health Care Management Review , Vol. 28, No. 1, 68-78.

Sheng, O.L., Hu, P.J., Wei, C-P. & Ma, P-C. (1999). Organizational management of telemedicine technology : Conquering time and space boundaries in health care services. IEEE Transactions on Engineering Management , 46 (3), 265-278.

Sicotte, C., Champagne, F., Farand, L., Lacroix, A., Rousseau, L., Ayé, M. and Reinharz, D. (1999). Analyse de l’expérimentation d’un réseau inter-hospitalier de télémédecine . Research report 99-03. Montreal, GRIS, University of Montreal.

Stake, R.E. (1995). The art of case study research . Thousand Oaks : Sage Publications.

Succi, M.J. & Lee, S.D. (1998). Trust between managers and physicians in community hospitals : The effect of power over hospital decisions. Journal of Healthcare Management , 43 (5), 397-414.

Whitten, P. & Allen, A. (1996). Organizational structure in telemedicine programs. Telemedicine Today , 4 (2), 15-22.

Whitten P, & Adams, I. (2003). Success and failure: A case study of two rural telemedicine projects. Journal of Telemedicine and Telecare , 9 (3), 125-129.

Yin, R.K. (1993). Applications of case study research. Applied Social Research Methods Series , 34. Newbury Park: Sage Publications.