Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

Investigation of Factors Affecting Healthcare Organization’s Adoption of Telemedicine Technology
Paul Jen-Hwa Hu University of South Florida Patrick Y.K. Chau University of Hong Kong Olivia R. Liu Sheng University of Arizona

Abstract
Recent advances in information and biomedicine technology have significantly increased the technical feasibility, clinical viability and economic affordability of telemedicine-assisted service collaboration and delivery. The ultimate success of telemedicine in an adopting organization requires the organization’s proper addressing both technological and managerial challenges. Based on Tornatzky and Fleischer’s framework, we developed and empirically evaluated a research model for healthcare organizations’ adoption of telemedicine technology, using a survey study that involved public healthcare organizations in Hong Kong. Results of our exploratory study suggested that the research model exhibited reasonable significance and classification accuracy and that collective attitude of medical staff and perceived service risks were the two most significant factors in organizational adoption of telemedicine technology. Furthermore, several implications for telemedicine management emerged from our study and are discussed as well

1. Introduction
Telemedicine is essentially about use of information and biomedicine technology to support, facilitate or improve collaboration and delivery of healthcare services among geographically dispersed parties, including physicians and patients [1]. Healthcare organizations have become increasingly aware of and knowledgeable about telemedicine in recent years. In effect, many organizations have exhibited considerable interest in adopting telemedicine technology to support practices of member physicians or extend services of the organization, as manifested by a fast growing number of telemedicine programs established around the world. The ultimate success of telemedicine requires an adopting organization to address both technological and managerial challenges effectively [2]. In this

exploratory study, we investigated the decision factors important for healthcare organizations’ adoption of telemedicine technology. We took a factor modeling approach and specifically employed the organizational adoption framework proposed by Tornatzky and Fleischer [3], who suggested that an organization’s adoption of a technological innovation should take into account several essential contexts, including the environmental, the organizational and the technological. This framework conceptually describes organizational innovation adoption phenomena and, at the same time, provides a necessary foundation upon which relevant adoption factors can be identified within the respective contexts. Anchoring our analysis of telemedicine technology adoption by healthcare organizations within this framework, we identified relevant adoption factors jointly leading to the development of a research model, which was evaluated using a survey study that involved public healthcare organizations in Hong Kong. Results of the study provided a desirable point of departure for subsequent research of organizational adoption of telemedicine technology in health care. The organization of the remainder of the paper is as follows. Section 2 reviews previous telemedicine research and relevant prior innovation/technology adoption studies, highlighting our motivation. Section 3 describes our overall research framework and the resulting research model, together with the specific hypotheses to be tested by the study. Section 4 details our research approach, design and data collection methods, followed by discussion of data analysis results in Section 5. The paper concludes in Section 6 with some implications for telemedicine management, readily derived from our findings.

2. Literature Review and Motivation
Broadly, technology adoption can be understood as an organization’s decision to acquire a specific technology and make it available to target users for their task performance. Contrary to a common dichotomous view, we considered technology

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

1

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

adoption to consist of a series of distinct and consecutive stages that lead to ultimate technology acquisition and use or final rejection. Specifically, we focused on healthcare organizations’ adoption of telemedicine technology, a significant technological innovation that has the potential to bring about a paradigmatic shift in health care. The concept of telemedicine emerged nearly four decades ago, when futurist physicians teamed up with technologists to experiment with use of telecommunications technology to support remote patient care or service collaboration [4-5]. Propelled by long-standing problems in contemporary health care in such areas as service access, quality and costs, recent information and biomedicine technology advancements have engineered a strong resurgence of telemedicine around the globe, which Bashshur designated the second generation of telemedicine [1] Most previous telemedicine research has concentrated on technology developments and their clinical applications [2]. A few studies have examined issues related to technology adoption, especially at the individual level [6]. For instance, Mairinger et al. [7] surveyed individual physicians in different European regions or countries about their perceptions toward telemedicine. Gschwendtner et al. [8] investigated medical students’ general assessments of telemedicine. However, most of these handful of studies have concentrated on individual technology adoption or acceptance; thus, offering limited, if any, discussion of technology adoption that takes place at the organizational level. Adoption of telemedicine technology in many organizations has been driven by legitimate motivations, including service improvement, patient market extension, and organizational performance and competitiveness enhancement. However, not all program drivers are pragmatic or sustainable. In effect, healthcare organizations can and do adopt telemedicine technology without proper motives or due consideration. As Liu Sheng et al. commented [9], the adoption of telemedicine technology by a healthcare organization may result from compromises between physicians and management or proceed without due consideration of important decision factors. Technology adoption has been an important issue for IS research and practices [10-11]. Many previous studies have built their theoretical premises around

Rogers’s innovation adoption theory [12], which essentially states that observed adoptions are largely prompted and determined by key innovation attributes that have been communicated to potential adopters. This theory encompasses an innovation (technology) emphasis and has primarily arisen to explain or predict innovation (technology) adoption by individuals, making its applicability or utility questionable in situations where adoptions take place at an organization level. For instance, Brancheau and Wetherbe [13] concluded that Rogers’s innovation adoption theory did not provide a complete explanation for technology adoption and implementation in organizations. Fichman [14] reviewed prior information technology (IT) innovation studies and noted that classical innovation attributes by themselves are not likely to be strong predictors of organizational technology adoptions, suggesting additional factors are needed. Prior empirical studies anchored in innovation adoption theory have produced findings of considerable inconsistency. To a large extent, the observed discrepancies may have been in part attributable to several reasons, including failure to differentiate individual and organizational adoption and neglecting other essential contexts. As Zmud [15] commented, much of the inclusiveness of prior research can be partially attributed to a failure to recognize that innovation attributes can be perceived significantly differently according to the organizational context involved. In addition to the technological context, several other contexts have been identified as essential to organizational technology adoption. For example, Bretschneider [16] compared the implementation of management information systems in public and private organizations and suggested that the organizational context was important to technology adoption. Copper and Zmud [17] investigated IT implementation in organizations and concluded that both organizational and task considerations were essential. Kimberly and Evanisko [18] examined innovation adoption in health care, singling out the importance of individual, organizational and contextual variables. Furthermore, Tornatzky and Fleischer [3] investigated innovation adoption processes in organizations and proposed a fairly comprehensive and well received framework that suggests innovation adoptions by organizations can

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

2

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

be determined by combination of the organizational, technological and environmental contexts. In response to the growing significance of telemedicine and limited previous research in organizational adoption, we investigated adoption of telemedicine technology by healthcare organizations in Hong Kong. We proposed to develop a research model that included factors important for organizational technology adoption and empirically to examine the model, using a survey study targeting public healthcare organizations in Hong Kong. Detailed descriptions of the research framework and model follow.

3. Research Framework and Model
In this study, we employed the organizational innovation adoption framework suggested by Tornatzky and Fleischer [3] for several reasons. First, the framework is largely consistent with and supported by results of most previous research, including Brancheau and Wetherbe [13], Fichman [14], Zmud [15], Bretschneider [16], Copper and Zmud [17], Kimberly and Evanisko [18]. In addition, the framework appeared to encompass most of the important adoption issues identified in a previously conducted case study [9] as well as from our prestudy interviews with clinical managing physicians from multiple organizations. According to Tornatzky and Fleischer [3], technology adoption that takes place in an organization is influenced by factors pertaining to the technological context, the organizational context, and the external environment. The technological context essentially describes the technology to be adopted and can be in part depicted by its important attributes. The anticipated results of technology use are another locus of the technological context and often have significant effects on an adopting organization, within its existing organizational and environmental contexts. Thus, the technological context can be jointly described by important technology attributes and the anticipated results of technology use. We replaced Tornatzky and Fleischer’s [3] organizational context with organizational readiness, which refers to the availability of the conditions needed for an organization’s adopting telemedicine technology

[20]. An organization usually has considerable influence on its internal condition with respect to a particular technology adoption. However, desirable changes to and cultivation of organizational readiness may require considerable time or resources and are often subject to various existing constraints, internal and external. The external environment defines the external world in which an organization operates. In most cases, an organization has limited influence or control over its external environment and thus needs to take the context as it is, striving for an optimal fit with and rapid adaptability to the context. Choice of the research framework provided a foundation upon which specific factors essential for adoption of telemedicine technology by healthcare organizations were identified. Figure 1 depicts our research model. As shown, the technological context includes both technology attributes and anticipated results of technology use. Perceived ease of use [19] is an important technology attribute. As a group, physicians are not particularly known for technological competence and have a tendency to consider technology as tools for supporting their practices. Thus, a technology that is difficult to use or operate is not likely to be well received by physicians. In turn, as the organization proceeds in the adoption process, this concern may represent a backward pressure on the adoption decision making. The organization consequently needs to properly evaluate a technology’s ease of use as perceived by physicians. Effects of perceived ease of use have been inconsistent in previous research, suggesting its significance on technology adoption might be technology-specific or generally of a moderate level.
T e c h n o l o g ic a l C o n t e x t
T e c h n o l o g y A t tr ib u te s :
- P e rc e iv e d E a s e o f U s e - P e rc e iv e d T e c h n o l o g y S a f e ty

R e s u lts o f T e c h n o lo g y U s e :
- P e rc e iv e d B e n e f its - P e rc e iv e d R i sk s

T e c h n o lo g y A d o p t io n

O r g a n i z a t io n a l R e a d in e s s
- C o l le c tiv e A tt itu d e o f M e d ic a l S t a ff

E x te r n a l E n v ir o n m e n t
- S e rv ice N ee d s

Figure 1: Research Model for Organizational Adoption of Telemedicine Technology

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

3

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

Thus, we hypothesized that perceived ease of use has a positive effect on an organization’s likelihood of adopting telemedicine technology. H1: Higher levels of perceived ease of use have a positive effect on an organization’s likelihood of adopting telemedicine technology. Perceived technology safety may be another important technology attribute. Broadly, most telemedicine technologies have as yet to mature, as indicated by limited evidence of their clinical efficacy. To a great extent, physicians are cautious about the safety of the technology used in their providing or delivering needed care and services. This paramount safety concern of physicians can be summarized by the first principle of their practice: Do no harm! Accordingly, we posited that perceived technology safety has a positive effect on telemedicine technology adoption by healthcare organizations. H2: Higher levels of perceived technology safety have a positive effect on an organization’s likelihood of adopting telemedicine technology. Perceived service benefits and perceived service risks are two essential aspects of anticipated results of technology use. Support for perceived benefits as a crucial technology adoption factor has been strong [20]. In this study, perceived service benefits were largely comparable to Rogers’s [12] relative advantages and refers to the degree to which telemedicine technology is perceived as being better than or superior to existing service arrangements for patient care and services. Accordingly, we hypothesized that perceived service benefits have a positive effect on organizational adoption of telemedicine technology. On the other hand, healthcare organizations are concerned with the service risks that might result from the use of an innovation, including technology, protocol, procedure and treatment plan. Understandably, an organization may have concerns about the potential risks of telemedicine in such service areas as physician-patient relationships, patient privacy and service effectiveness. As such, we postulated that perceived service risks have negative effects on technology adoption. H3: Higher levels of perceived service benefits resulting from use of telemedicine technology

have a positive effect on an organization’s likelihood of adopting the technology. H4: Higher levels of perceived service risks resulting from use of telemedicine technology have a negative effect on an organization’s likelihood of adopting the technology. In most cases, physicians are arguably the most important users of telemedicine technology. Based on findings from a previously conducted case study [9] as well as comments made by clinical managing physicians in pre-study interviews, the attitude of medical staff toward telemedicine and its enabled services is important to organizational technology adoption. “The bottom-line is that my staff would use the technology,” commented the chief-of-service of a surgery department where a previously acquired computer-based patient record system had seldom been used by his fellow surgeons. Previous research has also suggested that attitudes of key personnel are an important factor for organizational technology adoption [21-22]. Specifically, attitude assessment should proceed at a collective rather than an individual level. Thus, we hypothesized that the collective attitude of medical staff has a positive effect on organizational adoption of telemedicine technology. H5: Stronger levels of collective attitude of medical staff toward telemedicine and its enabled services have a positive effect on an organization’s likelihood of adopting telemedicine technology. Service needs are an important factor for the external environment. Healthcare organizations’ purpose is to provide services to those in need and thus they need to explore and evaluate alternatives when existing delivery arrangements cannot meet service demands, measured by service volume or quality. In their investigation of organizational adoption of computer-aided software engineering (CASE) technology, Rai and Yakuni [23] concluded that needs-pulled factors were important to organizational adoption decision. Accordingly, we posited that service needs have a positive effect on telemedicine technology adoption by healthcare organizations. H6: Higher levels of service needs have a positive effect on an organization’s likelihood of adopting telemedicine technology.

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

4

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

4. Research Approach, Design and Data Collection
This section describes our research approach, target organizations, instrument development, and data collection methods.

4.1 Research Approach
We took a factor modeling approach, aimed at advancing the general understanding of telemedicine technology adoption by healthcare organizations. As Kimberly and Evanisko [18] commented, “concentration of the research focus can help to identify and isolate factors that clarify the nature of phenomena in a particular sector and, at the very least, can suggest hypotheses that can be generalized beyond that sector and tested in other sectors, even though the finding may have limited direct applicability.” With the factor modeling approach taken, we identified factors that may affect organizational adoption of telemedicine technology and empirically evaluated their significance, generating results that will be needed for subsequent confirmatory studies and research model development or refinement. We took a key informant approach to data collection. Specifically, we used responses of clinical managing physicians contacted at participating organizations (including clinical departments) to assess technology adoption that had taken place in their respective organizations. The particular target informants included hospital executive officers, clinical departments’ chiefs-of-service, and care center directors. Use of key informants to obtain information about the investigated organizations is justifiable and common [21,24]; its application in our study was appropriate and desirable for several reasons. First, key informants presumably have a fairly comprehensive understanding of the external environment and the internal condition of the their respective organization. By and large, our target informants had better knowledge about the overall (big) picture and therefore were considered to be more qualified information providers than individual physicians. At the same time, key informants were physicians themselves and therefore were able to analyze and evaluate technology adoption from the perspective of care providers as well. The dual role

of target informants as administrator/manager and clinician was essential and desirable for our examination of telemedicine technology adoption taking place at the organizational level. The dependent variable was organizational technology adoption, which was measured using a continuum consisting of 7 distinct and consecutive current adoption levels that approximate the likelihood of telemedicine technology adoption by individual organizations. That is, the probability of an organization’s adopting telemedicine technology will increase as its current adoption stage moves up the adoption continuum. For instance, an organization that has already submitted a formal adoption proposal currently under an external funding agency’s review is more likely to adopt telemedicine technology than organizations that have thought about potential adoption but decided not to pursue it at present. Use of the adoption continuum not only described organizational technology adoption in increasing detail but also permitted dichotomous classification in data analysis. This process-oriented treatment is intuitive and logical because technology adoption taking place in an organization usually progresses through several distinct but consecutive latent stages before reaching an observable state, including actual technology acquisition and use. In this light, absence of observable adoption manifestations does not necessarily suggest that an organization has decided not to adopt a particular technology. On the contrary, the technology adoption under consideration may have been steadily progressing through the necessary intermediary latent stages and soon become observable. From the data analysis perspective, use of the adoption continuum is pragmatic and desirable. As described, the consecutive adoption levels can represent or indicate the likelihood of technology adoption and thus are appropriate for our hypothesis testing. As such, the logistic regression technique can be applied to classify the dependent variable using some appropriate adoption stage threshold. Furthermore, the adoption continuum is effective in coping with existing real-world constraints. To a great extent, telemedicine developments in Hong Kong are largely in an early phase and actual technology use currently is not widespread but is expected to grow rapidly. Thus, use of the adoption continuum allowed our investigation of current

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

5

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

adoption stages of various organizations in spite of their overall limited technology use and, at the same time, enabled the identification of potential adoption barriers.

4.2 Target Organizations
We targeted public healthcare organizations in Hong Kong, including hospitals, rehabilitation care centers and specialized clinics. Clinical departments were considered as independent units of analysis because of their autonomy and specialization. The decision to concentrate on public healthcare organizations was made primarily because of the relative likelihood of their adopting the technology. To a certain extent, public healthcare organizations are more likely to adopt telemedicine technology than their counterparts in the private sector for several reasons [9]. First, public healthcare organizations represent the major care providers in Hong Kong and thus have greater service needs that may be effectively addressed by telemedicine than private care providers. Secondary and tertiary care, in particular, have considerable under-addressed service demands. Second, public organizations have relatively greater access to the resources necessary for adopting new technologies, compared with private clinics and hospitals. Third, public organizations have reasonable technical support, from both in-house technology bases and the Health Authority (HA), which probably has the most sophisticated technology capability in the Hong Kong healthcare system. Thus, these organizations tend to be more technologically ready for telemedicine than private clinics and hospitals.

4.3 Instrument Development
The development of our survey instrument proceeded as follows. First, we reviewed relevant prior research to identify and select appropriate candidate measurement inventories, which were then supplemented with additional items derived based on findings of pre-study interviews and discussion with a focus group that consisted of 3 chiefs-of-service from different specialty areas. The resulting preliminary question items were examined by the same focus group, which evaluated their content validity at face value. Based on group feedback, several minor modifications, including wording

choice, were made to enhance the question items’ communicability in the healthcare context. The question items were subsequently tested using a card sorting method [25] that involved one chiefof-service, one hospital medical executive and one director of a long-term care center. None of these pre-test physicians had participated in the focus group and, like the focus group physicians, they were excluded from the subsequent formal study. The question items were printed on 8x6 index cards, which were shuffled and presented randomly to the pre-test physicians, each of whom was asked to sort the cards into appropriate categories. Results from the card-sorting test were largely satisfactory; the physicians were able to categorize the question items correctly with an accuracy rate of 83 percent or better. A 7-point Likert scale was used for all question items except the one that measured the dependent variable, with 1 being strongly agree and 7 being strongly disagree. To ensure desired balance of the items in the questionnaire, half of the question items were properly negated to invite the attention of respondents who, as a result, might become increasingly alert to manipulated question items. In addition, all the question items were arranged randomly to minimize the potential ceiling (floor) effect that could induce monotonous responses to question items designed to measure a particular underlying construct. To anchor the responses properly [26], we provided in the questionnaire an explicit working definition of telemedicine and included in each packet selected general references of telemedicine and some example technologies. The dependent variable, current adoption level, was defined at 7 distinct and consecutive levels, each of which provides a necessary foundation for the level above it. For data analysis purposes, we considered an organization as an “adopter” if it has at least submitted a formal technology adoption proposal that was currently under external review. Thus, as we defined them, adopters also included organizations that had already implemented telemedicine technology and used it for clinical purposes or had located and secured the necessary financial resources and technology source. Organizations whose current adoption stages had not reached the particular adoption threshold were then considered as ”non-adopters” in our data analysis. We chose proposal submission for external review as

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

6

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

the minimum adoption criterion because it was observable, signaling that formal adoption had begun. Furthermore, the distinction between proposal submission for external review and its succeeding stage (i.e., have or about to complete an adoption plan) was more defined and explicit than that between the succeeding stage and its precursor (i.e., have designed a task force or individual to investigate potential adoption).

questionnaire were faxed to the target respondents who had not yet responded at the end of the extended response period. Finally, a terminating response window of one week was indicated in a faxed reminder to the remaining non-respondents, who were asked a final time to mail in their completed questionnaires.

5. Data Analysis Results
In this section, we highlight data analysis findings in terms of respondent profile, measurement reliability and construct validity, and logistic regression results, described as follows.

4.4 Data Collection
We collected the data using a self-administered questionnaire survey. Contact information for the target informants was obtained from an internal medical staff directory published by the HA. Before distributing the questionnaires, each target respondent was informed by a faxed introductory letter that briefly stated the study’s purpose and its anticipated results and significance. Questionnaire packets were sent by postal mail. Each contained a cover letter that explicitly stated the purpose and intended use of the data to be collected, endorsement letters from the Hong Kong Telemedicine Association (HKTA) and the IT division of the HA, selected general references of telemedicine and sample technologies, the questionnaire and a selfaddressed stamped envelop. Use of the HA internal medical staff directory allowed coding and tracking for individual informants, enabling the identification of non-respondents to be contacted in subsequent follow-up procedures. Each target informant was given approximately 2 weeks to complete the questionnaire, dated from the estimated packet arrival. A reminder letter was faxed to all target respondents a week after their estimated receipt of the questionnaire. A second reminder letter was faxed each respondent’s secretary 2 or 3 days before the specified response time window expired, asking her to remind the subject to complete the questionnaire and return it using the stamped return envelop provided. Reminders and additional questionnaires were sent by mail to those who failed to return the completed questionnaire within the initial response period. Late respondents were given another 10 days to complete the questionnaires and their secretaries were telephoned to inform them about the incoming questionnaires and ask them again to remind the subjects to complete the questionnaires. A second reminder and another

5.1 Respondent Profile
Of the 188 questionnaires distributed, 113 were completed and returned, showing a 60.1 percent response rate. Among them, nine were partially completed and therefore were excluded from the subsequent data analysis, making the effective response rate 50%. As a group, the responding organizations had an average of 34.8 physicians or specialists and employed 142.1 nurses and 34.3 technicians. Most of our respondents were male (84.1%), held the post of chief-of-service (66.4%), and had received their basic medical education in Hong Kong (81.4%). On the average, respondents were 43.5 years of age and had had 17.7 years of post-internship clinical practice. From the perspective of medical specialty areas, distribution of the responding organizations was fairly balanced. A total of 18 specialty areas were represented by the collected data, which showed relatively higher levels of participation from Internal Medicine, Pediatrics, Radiotherapy and Oncology, Surgery, Obstetrics and Gynecology, and Pathology than other specialty areas. Primary care, long-term care and rehabilitating care were also included and accounted for 3.5, 0.9 and 6.2 percent of responses, respectively. A total of 62 responses were completed and returned within the initial response window, accounting for 65.5 percent of the effective responses. These respondents were considered early respondents, whereas the remaining ones were late respondents. Comparative analysis of the early and late respondents suggested no significant differences in their respective organizations, as measured by the

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

7

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

number of physicians, nurses and technicians. Similarly, these two respondent groups were largely comparable in several areas, including age, postinternship clinical experience, and the respective distribution of gender, post, and country where medical school was attended. Jointly, findings from the comparative analysis suggested little or reduced non-response bias.

coefficient analysis and factor analysis suggested that our instrument encompassed adequate construct validity, as manifested by satisfactory measurement convergent and discriminant validity.

5.4 Logistic Regression Results
Logistic regression was used to evaluate our research model and hypotheses. Choice of this particular data analysis technique was based primarily on its flexibility in assumption requirements and our intended dichotomous analysis of the dependent variable. Specifically, we examined the significance and classification accuracy of the research model. Regression results showed that the research model was not significantly different from a perfect model that can classify all responding organizations to the correct adopter or non-adopter category, as indicated by its goodness-of-fit statistic with chi-square being 73.17 and significance being 0.8551. The classification accuracy or discriminant power of the model was also examined. Judged from the adoption threshold used in the study, our data consisted of 19 adopters and 75 non-respondents. Thus, random guesswork in theory would result in a classification accuracy of 67.74%; that is, (19/94)2 + (75/94)2 = 0.6774. On the other hand, the classification accuracy accomplished by the research model was 84.04, considerably higher than that of random choice. The observed superiority suggested that the research model encompassed desirable classification accuracy or discriminant power. Support for individual hypotheses was examined using the respective regression coefficients and their significance. As summarized in Table 2, collective attitude of medical staff, perceived service risks and perceived ease of use were found to have significant effects on organizational technology adoption, with p-values being 0.005, 0.006 and 0.013 respectively. However, perceived service benefits, service needs and perceived technology safety appeared to be insignificant in determining whether or not a healthcare organization would adopt telemedicine technology.

5.2 Measurement Reliability
Use of multiple items to measure a specific construct requires examination of the reliability or internal consistency among the measurements [27]. We evaluated measurement reliability using the Chronbach alpha derived from the question items for the same construct. As shown in Table 1, all investigated constructs exhibited an alpha value of close to or greater than 0.7, a common reliability threshold for exploratory research [28]. Thus, data analysis results suggested that the measurements used in the study encompassed reasonable reliability.
Factor Perceived Benefits (5 Items) Perceived Risks (4 Items) Service Needs (2 Items) Collective Attitude of Medical Staff (3 Items) Ease of Use (2 Items) Perceived Technology Safety (2 Items) Cronbach’s Alpha 0.75 0.80 0.85 0.78 0.70 0.68

Table 1: Analysis of Measurement Reliability

5.3 Convergent and Discriminant Validity
Construct validity of the survey instrument was evaluated in terms of convergent and discriminant validity [27]. Specifically, we performed inter-item correlation analysis and factor analysis. Based on the analysis results, correlation coefficients were considerably higher among question items designed to measure the same construct than among those intended for different constructs. The observed higher levels of correlation among measurements for the same than different constructs suggested that our instrument exhibited adequate convergent and discriminant validity. A principal component factor analysis was also performed, using varimax rotation method with Kaiser normalization. A total of 6 components were extracted, precisely matching the number of constructs included in our research model. Question items designed for the same construct exhibited prominently and distinctively higher factor loadings on a single component than on others, suggesting satisfactory convergent and discriminant validity of the measurements. Jointly, results of correlation

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

8

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

Factor Perceived Service Benefits Perceived Service Risks Service Needs Collective Attitude of Medical Staff Perceived Ease of Use Perceived Technology Safety

Coefficient -0.42 -0.99 0.46 1.48 -0.71 0.38

Wald Statistic 1.66 7.72 2.30 8.00 6.12 1.37

Significance 0.1975 0.0055 0.1292 0.0047 0.0133 0.2421

Table 2: Logistic Regression Results

6. Implications for Telemedicine Management
Several implications for telemedicine management can be readily drawn from our study findings. First, attitudes of medical staff toward telemedicine and its enabled services are essential to technology adoption and thus management may need to ensure that a favorable collective attitude has been cultivated and solidified before proceed with actual technology acquisition and implementation. In our study, collective attitude of medical staff toward telemedicine and its enabled services was the most significant factor differentiating adopters and nonadopters. This may in part have resulted from the professional nature of health care, in which physicians often have relatively autonomous decision making and are able to preserve individual independence in determining whether or not to adopt a technology. Compared with end-users in a business-organization context, physicians appeared to have more influence on adoption decisions that may affect their practices, making their attitude assessment and management an increasingly important issue in organizational technology adoption. Our study results suggested that perceived service risks were the second most significant factor for organizational adoption of telemedicine technology. The propensity to resist change can be considerable in health care, especially when change may bring about significant service uncertainty or adverse ramifications to physicians’ practices. Conceivably, physicians may be concerned about incorporating telemedicine in their practices in the light of potential legal liability and service disputes or degradation. However, not all perceived service risks are verifiable and their removal requires evidencebased information exchange and communication. Effective means for removing or reducing undue perceived service risks may include pre-adoption technology experimentation and trial use as well as making arrangements for interacting with peers routinely providing telemedicine-assisted services. Perceived ease of use of telemedicine technology was also found to have a significant effect on the likelihood of an organization’s adopting technology. However, the coefficient obtained from the regression analysis was negative. This suggests that an organization currently in an adoption stage close

to reaching actual technology implementation and use may not consider perceived ease of use as important actor as would an organization currently in a primitive adoption stage. The observed discrepancy is in agreement with some prior research that has suggested perceived ease of use to be an insignificant or relatively weak predictor of technology adoption. In the context of telemedicine, the finding suggests that an organization highly anxious about the technology’s ease of use may become less concerned about this particular issue when it has moved to a more advanced adoption stage (e.g., beyond formal technology investigation and evaluation). That is, perceived ease of use might be a relatively less important technology adoption factor but its significance can be over-appraised by organizations not familiar with telemedicine or situated in an early stage of the adoption process. Undue perceived service risks may result from knowledge barriers which can be reduced or removed with detailed technology assessment and proper communication of evaluation results. Our study findings also suggested that effects of perceived service benefits on technology adoption were not significant. One plausible explanation may be that telemedicine remains a novelty to many organizations whose technology adoption or intention is primarily driven by considerations other than specific service benefits, including envisioned professional status enhancement and clinical technology experimentation or exploration. Similarly, service needs were not found to be an important adoption factor, suggesting that telemedicine should not be viewed as a “panacea” for all unmet service needs but rather as facilitating or supporting some but not all physicians’ tasks and services. Desirable sustainability requires an organization to identify explicitly in its technology adoption plan the target services of telemedicine that cannot be satisfactorily addressed by existing service arrangements. In addition, effects on perceived technology safety were not significant, either. This may in part reflect that many organizations tend to consider technology adoption from the perspective of service experimentation or clinical trials.

References:
[1] Bashshur, R.L, Sanders, J.H., and Shannon, G.W. (eds.), Telemedicine: Theory and Practice, Charles Thomas, Springfield, IL, 1997.

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

9

Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000

[2] Perednia, D.A., and Allen, A., “Telemedicine Technology and Clinical Applications,” Journal of American Medical Association, Vol.273, No.6, 1995, pp.483-488. [3] Tornatzky, L.G. and Fleischer, M., The Process of Technological Innovation, Lexington Books, Lexington, MA, 1990. [4] Jutra, A., “Teleroentgen Diagnosis by Means of Videotape Recording,” American Journal of Roentgenology, Vol.82, 1959, pp.1099-1102. [5] Wittson, C.L., Afflect, D.C., and Johnson, V., “Twoway Television Group Therapy,” Mental Hospitals, Vol.12, 1961, pp.22-23. [6] Mitchell, B.R., Mitchell, J.G., and Disney, A.P., “User Adoption Issues in Renal Telemedicine,” Journal of Telemedicine and Telecare, Vol.2, No.2, 1996, pp.81-86. [7] Mairinger, T., Gable, C., Derwan, P., Mikuz, G., and Ferrer-Roca, O., “What Do Physicians Think of Telemedicine? A Survey in Different European Regions,” Journal of Telemedicine and Telecare, Vol.2, No.1, 1996, pp.50-56. [8] Gschwendtner, A., Netzer, T., Mairinger, B., and Mairinger, T., “What Do Students Think About Telemedicine?” Journal of Telemedicine and Telecare, Vol.3, No.3, 1997, pp.169-171. [9] Liu Sheng, O. R., Hu, P.J., Wei, C., Higa, K., and Au, G., “Adoption and Diffusion of Telemedicine Technology in Healthcare Organizations: A Comparative Case Study in Hong Kong,” Journal of Organizational Computing and Electronic Commerce, Vol.8, No.4, 1998, pp.247-75. [10] Niederman, F., Brancheau, J.C., and Wetherbe, J.C., “Information Systems Issues for the 1990s,” MIS Quarterly, Vol.15, No.4, December 1991, pp.475500. [11] Brancheau, J.C., Janz, B.D., and Wetherbe, J.C., “Key Issues in Information Systems Management: 1994-95 SIM Delphi Results,” MIS Quarterly, Vol.20, No.2, June 1996, pp.225-242. [12] Rogers, E.M., Diffusion of Innovations, 4th Edition, Free Press, New York, NY, 1995. [13] Brancheau, J.C. and Wetherbe, J.C., “The Adoption of Spreadsheet Software: Testing Innovation Diffusion Theory in the Context of End-User Computing,” Information Systems Research, Vol.1, No.2, 1990, pp.115-143. [14] Fichman, R.G., “Information Technology Diffusion: A Review of Empirical Research,” Proc. of the Twelfth International Conference on Information Systems, Dallas TX, December 1992, pp.195-206. [15] Zmud, R.W., “Diffusion of Modern Software Practices: Influences of Centralization and Formalization,” Management Science, Vol.28, No.12, 1982, pp.1421-1431. [16] Bretschneider, S., “Management Information Systems in Public and Private Organizations: An Empirical Test,” Public Administration Review, Vol.50, No.5, 1990, pp.536-545. [17] Cooper, R. and Zmud, R., “Information Technology Implementation: A Technological Diffusion Approach,” Management Science, Vol.36, No.2, 1990, pp.156-172. [18] Kimberley, J.R. and Evanisko M.J., “Organizational Innovation: The Influence of Individual,

Organizational, and Contextual Factors on Hospital Adoption of Technological and Administrative Innovations, Academy of Management Journal, Vol.24, No.4, 1981, pp.689-713. [19] Davis, F.D., A Technology Acceptance Model for Empirically Testing New End-user Information Systems: Theory and Result, Doctoral Dissertation, Sloan School of Management, Massachusetts Institute of Technology, 1986. [20]Iacovou, C.L., Benbasat, I., and Dexter, A.S., “Electronic Data Interchange and Small Organizations: Adoption and Impact of Technology,” MIS Quarterly, Vol.19, 1995, pp.465-485. [21]Nickell, G.S. and Seado, P.C., “The Impact of Attitudes and Experience on Small Business Computer Use,” American Journal of Small Business, Vol.10, No.4, 1986, pp.37-47. [22] Thong, J. and Yap, C.S., “CEO Characteristics, Organizational Characteristics and Information Technology Adoption in Small Business,” Omega: The International Journal of Management Science, Vol.23, No.4, 1995, pp.429-442. [23] Rai, A. and Yakuni, R., “A Structural Model for CASE Adoption Behavior,” Journal of Management Information Systems, Vol.13, No.2, 1996, pp.205234. [24] Chau, P.Y.K. and Tam, K.Y., “Factors Affecting the Adoption of Open Systems: An Exploratory Study,” MIS Quarterly, Vol.21, No.1, March 1997, pp.1-24. [25] Moore, G.C. and Benbasat, I., “Development of an Instrument to Measure the Perception of Adopting an Information Technology Innovation,” Information Systems Research, Vol.2, No.3, 1991, pp.192-223. [26] Hufnagel, E.M. and Conca, C., “User Response Data: The Potential for Errors and Biases,” Information Systems Research, Vol.5, No.1, 1994, pp.48-73. [27] Straub, D.W., “Validating Instruments in MIS Research,” MIS Quarterly, Vol.13, No.2, June 1989, pp.147-169. [28] Nunnally, J.C. Psychometric Theory, 2nd Edition, McGraw-Hill, New York, 1978

0-7695-0493-0/00 $10.00 (c) 2000 IEEE

10