i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33

journal homepage: www.intl.elsevierhealth.com/journals/ijmi

Knowledge barriers to PACS adoption and

implementation in hospitals

Guy Paré a,∗ , Marie-Claude Trudel b,1

a Canada Research Chair in Information Technology in Health Care, HEC Montréal, 3000 Côte-Ste-Catherine Road,
Montréal, Que., Canada H3T 2A7
b IT Management Department, HEC Montréal, 3000 Côte-Ste-Catherine Road, Montréal, Que., Canada H3T 2A7

a r t i c l e i n f o a b s t r a c t

Article history: Purpose: Drawing on the classical theory of diffusion of innovations advanced by Rogers [E.M.
Received 14 September 2005 Rogers, Diffusion of Innovations, 4th ed., Free Press, New York, NY, 1995] and on the theory
Received in revised form of barriers to innovation [P. Attewell, Technology diffusion and organizational learning: the
3 January 2006 case of business computing. Organ. Sci. 3 (1992) 1–19; H. Tanriverdi, C.S. Iacono, Knowledge
Accepted 8 January 2006 barriers to diffusion of telemedicine. Proceedings of the 20th International Conference on
Information Systems, Charlotte, NC, 1999, pp. 39–50; S. Nambisan, Y.-M. Wang, Roadblocks
to web technology adoption? Commun. ACM, 42 (1) (1999) 98–101], this study seeks a better
Keywords: understanding of challenges faced in PACS implementations in hospitals and of the strate-
PACS gies required to ensure their success.
Adoption Methods: To attain this objective, we describe and analyze the process used to adopt and
Implementation implement PACS at two Canadian hospitals.
Innovation process Results: Our findings clearly demonstrate the importance of treating any PACS deployment
Knowledge barriers not simply as a rollout of new technology but as a project that will transform the organiza-
tion. Proponents of these projects must not lose sight of the fact that, even if technological
complexity represents a significant issue, it must not garner all the project team’s attention.
This situation is even more dangerous, inasmuch as the greatest risk to the implementation
often lies elsewhere. It would also appear to be crucial to anticipate and address organiza-
tional and behavioral challenges from the very first phase of the innovation process, in order
to ensure that all participants will be committed to the project.
Conclusions: In order to maximize the likelihood of PACS success, it appears crucial to adopt a
proactive implementation strategy, one that takes into consideration all the technical, eco-
nomic, organizational, and human factors, and does so from the first phase of the innovation
process.
© 2006 Elsevier Ireland Ltd. All rights reserved.

1. Introduction of improved diagnostic technologies and increased funding

for diagnostic imaging services. This increase in investment
The use of digital imaging devices such as MRI, CT, and ultra- has resulted in greater accuracy, speed, and efficiency in diag-
sound has expanded in hospitals worldwide over the past few nostic imaging services, as well as improvements to quality
years due to a number of factors, including the availability of care [5]. However, diagnostic imaging devices alone cannot

∗
Corresponding author. Tel.: +1 514 340 6812; fax: +1 514 340 6132.
E-mail addresses: guy.pare@hec.ca (G. Paré), marie-claude.trudel@hec.ca (M.-C. Trudel).
1
Tel.: +1 514 340 6476; fax: +1 514 340 6132.
1386-5056/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.ijmedinf.2006.01.004
 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33 23

deliver full value in service and cost reduction. Second-order our empirical approach, Section 3 describes the methodology
technologies such as PACS (picture archive and communica- employed to attain our research objectives, Section 4 explains
tions systems) and RIS (radiology information systems), which the two PACS implementation projects and provides obser-
support the acquisition and communication of critical infor- vations based on the theoretical frameworks selected, and
mation, must also be applied. Section 5 presents our findings from the comparative case
PACS has become an important component of services analysis. Finally, the conclusion presents the practical impli-
offered by many radiology departments and hospitals around cations of our findings.
the world. To fully understand the potential benefits of PACS,
one must first comprehend how the technology works. A PACS
is an information system that enables diagnostic imaging ser- 2. Theoretical foundations
vices to manage, store, and distribute digital images from CT,
MRI, X-ray, and ultrasound imaging devices. The main compo- As shown in Table 1, two conceptual frameworks were selected
nents of a PACS system include a server; acquisition interfaces and combined to frame our scientific approach. The first
that digitally capture radiology images, convert these images model, derived from work by Rogers [1], illustrates the inno-
to a standard format, and then store them on the PACS server; vation adoption process in organizations. Under this model,
display workstations that retrieve the radiology image from the adoption of an innovation should not be seen as a single
the PACS server and display it to the technologist, radiolo- act of selection, but rather as a series of steps leading towards
gist, and clinician; storage devices that provide permanent continuous utilization of the innovation. In fact, this process
long-term storage for radiology images; a communications comprises two major phases that are carried out sequentially:
infrastructure or network that allows various computer com- initiation and implementation. The initiation phase begins
ponents to exchange information; and a RIS interface through with a variety of information collection, design, and planning
which data can be transferred between the RIS and the PACS activities that support adoption of the innovation. The very
[5]. Although the RIS is not an integral component of PACS, first step of the process, agenda-setting, is mainly directed at
it is considered a related technology. RIS is used to perform identifying and prioritizing the organizational problems that
functions such as patient registration, exam ordering, and form the basis of the need to acquire an innovation [16]. There
scheduling. It is also used for the entry, storage, and distri- are times when just learning of the existence of an innovation
bution of the radiologist’s diagnostic report [6]. will trigger an adoption process, without a clear need hav-
There are many advantages associated with the use of PACS ing been first identified. In the second stage, commonly called
and RIS in a medical imaging department. Among the more matching, project proponents are primarily concerned with
important benefits are more rapid diagnostic readings, a sig- the fit between a need identified in the previous stage and the
nificant reduction in the number of lost images, more patients qualities of the innovation itself. The proponents then attempt
examined, fewer rejected images (and rescheduled exams), to determine the extent to which the innovation in question
accelerated improvements in the productivity of radiologists can solve the organizational problem or problems previously
and technologists, the elimination of films and the chemi- identified. This process sometimes results in an initial review
cal products needed to develop them, and improved patient of the problems that may arise when the innovation is imple-
care [7–15]. In large hospitals, the reduction in expenditures mented [1]. At the conclusion of this stage, the organization
related to the elimination of films and to higher productivity decides whether or not to adopt the innovation. In the event
is sufficient to finance a good portion of system implementa- that the required investment is approved, the second phase of
tion costs, which, depending on the scale of the operations, the process, implementation, begins. This phase includes all
can reach several million dollars. According to the Canadian actions and decisions related to the innovation’s deployment,
Association of Radiologists, about 25% of all diagnostic imag- assimilation, and integration within the organization [1].
ing exams performed in Canada were managed with PACS In the third, redefining and restructuring stage, the inno-
technology in 2003 [5]. In 2004, the United States PACS mar- vation is modified and reinvented in order to adapt it to the
ket grew by over 25% and, by 2009, it will generate $3.0 billion organization’s specific operating environment. In other cases,
in revenues.1 According to Frost and Sullivan, 36% of North organizational processes and structures will be altered to
American hospitals will have deployed PACS by 2008.2 accommodate the specific nature of the innovation. When the
In this article, we examine the process used by hospitals organization fails to modify its processes to accommodate the
to adopt and implement this technology. Based on the the- innovation, oftentimes the implementation itself will fail, and
ory of diffusion of innovations [1] and the theory of barriers to the adoption process may come to a halt once the investment
innovation that came out of work by Attewell [2], Tanriverdi is approved by senior organizational leaders. The clarifying
and Iacono [3], and Nambisan and Wang [4], this study aims stage begins with the innovation being more and more widely
to develop a better understanding of the challenges faced in used in the organization, causing the many uncertainties that
carrying out this kind of project, the conditions for success, originally surrounded the project to gradually fall away. The
and the strategies required to fully take advantage of PACS. social and human aspects of the implementation are cru-
The following section presents the theoretical background for cial at this stage, since it is through daily interactions and
exchanges with their peers that system users develop a com-
mon understanding of the innovation. Finally, the routinizing
1
Millenium Research Group, June 2004. stage begins when the innovation loses its distinct identity
2
Frost and Sullivan (www.frost.com), site consulted on January and using it becomes a natural part of the organization’s cur-
5, 2005. rent activities. At this point the adoption process has ended,
24 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33

Table 1 – Conceptual framework

“X” marks indicate at which stages of the innovation process the various types of knowledge barriers are expected to be addressed in order to
avoid project abandonment or failure.

as members of the organization no longer perceive the tech- who should lead the PACS project, the radiology department
nology as an innovation per se [1]. or the computer department? How should the resources com-
The second conceptual framework to serve as a theoretical mitted to the project be shared? How can this investment be
foundation for this study highlights various barriers to inno- financed, and over how many years? Who should participate
vation. This model is mainly derived from work by Attewell in the development of a business plan? Under the theoretical
[2], Tanriverdi and Iacono [3], and Nambisan and Wang [4]. framework presented in Table 1, these types of problems must
The extensive literature on implementing information tech- be addressed very early in a project, preferably in the matching
nologies reveals that the organization’s intention to adopt stage [4].
a technology is influenced by a number of factors, includ- Second, facing technological barriers refers to not having the
ing the cost of the desired innovation, its user-friendliness, knowledge needed to make many critical decisions concern-
the technology’s compatibility with existing systems, and its ing issues such as the selection of a technological infrastruc-
alignment with the organizational structure [17]. These factors ture, the hardware, and the software, as well as security deci-
are particularly relevant in the case of technologies consid- sions. For example, several options involve digitizing analog
ered “simple”, such as electronic mail (e.g. [18]). This model images created by traditional radiographic devices. Hospitals
nevertheless reveals its limits when one seeks to explain or may digitize films with a laser digitizer, use computed radio-
predict the adoption of technologies considered “complex” or graphy (CR) devices that are compatible with analog devices,
“advanced”, such as PACS. Because these technologies require or install digital radiography (DR) devices [20]. Each of these
specialized knowledge for their parameterization, installa- options has advantages and disadvantages, and the project’s
tion, and integration, their diffusion is strongly determined proponents must arrive at an informed choice, i.e. one that
by each institution’s ability to adopt them, independent of the gives due consideration to the organizational environment.
degree of interest shown for the technology. In other words, A lack of technical expertise in a hospital can therefore lead
the adoption of complex technologies is mainly influenced to serious problems in the redefining/restructuring phase and
by the organization’s capacity to lower or even remove the the clarifying phase.
various knowledge barriers associated with these same tech- Third, organizational barriers reflect the difficulties encoun-
nologies [2]. tered integrating and inserting a technology into existing prac-
Knowledge barriers associated with the adoption of PACS tices and structures and the challenges inherent in learning
in hospitals can be broken down into four general categories. how to adequately support regular use of the technology [3,4].
First, there is the notion of project barriers, which refers in par- Given the specific type of knowledge required at this level, i.e.
ticular to the financing problems faced when acquiring an how to manage the impact of PACS on existing processes and
innovation, and whether or not there are financial incentives structures [19,21], organizational barriers are also very impor-
linked to its use [3]. These barriers also include the knowledge tant in the redefining/restructuring and clarifying stages.
required for project management and for management of the Finally, behavioral barriers are related to resistance to change
resources allocated to the project [19]. In short, the issues or among individuals affected by the implementation as well
questions faced by hospitals that want to adopt PACS include: as problems related to organizational power dynamics [3].
 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33 25

Our analysis of the data included a general coding of the

Table 2 – Site profile
interview transcripts based on the constructs of the concep-
Alpha hospital Beta hospital tual framework (see Table 1). The content of the interviews and
Number of beds 571 343 the documentation was then analyzed, providing the infor-
Average number of 20000 14000 mation required to build a timeline of events leading to the
patients adoption and implementation of PACS at each site and a per-
hospitalized each sonal recounting of events as they were experienced by each
year
of the various stakeholders. Reflective remarks were directly
Average number of 80000 140000
acute-care
entered into the transcript. These remarks were ways of get-
admissions ting ideas down on paper and of using writing as a way to
Number of 267 187 facilitate reflection and analytic insight.
physicians As suggested by Miles and Huberman [34], we used chrono-
Number of 11 9 logical and schematic matrices to synthesize the information
radiologists
collected (see Tables 3 and 4). This synthesis of elements con-
Number of 40 45
sidered critical to project success enabled us to draw up a
technologists
Average number of 110000 110000 portrait of the general situation at each site. In order to ensure
examinations per construct validity [33], a draft of the case report (story asso-
year ciated with the implementation project) was reviewed by key
Approximate total CAN$3.5 million CAN$2 million informants at each site, i.e. those in a position to reflect on the
cost of the PACS case’s “big picture”. Next, in line with Eisenhardt’s [35] recom-
project
mendation to compare findings with existing literature, we
Expected payback 5 years 8 years
compared our findings with the general principles under the
period
conceptual framework presented in Table 1.
A key criterion for judging the quality of case study research
Resistance to information technologies among physicians is is that of reliability. The objective is to be sure that, if a later
well documented [22–31], and it represents a major barrier. researcher follows exactly the same procedures as described
As indicated in Table 1, problems of resistance to change are by an earlier investigator and conducts the same case study all
often “managed” in a reactive mode, so they are addressed in over again, the later investigator will arrive at the same find-
the clarifying and routinizing stages [32]. ings and conclusions [33]. In other words, the goal of reliability
is to minimize the errors and biases in a study. Two tactics were
3. Methods adopted in this study to minimize errors and biases. First, a
case study protocol containing the interview guides was used.
In order to attain the research objectives discussed above, Second, reliability was also increased through the mainte-
we carried out a retrospective, multiple-case study [33] in nance of a case study data base. A data base organizes and
two Canadian hospitals that had adopted PACS. The first site, documents the data collected for each case. Each of the two
which we have called Alpha, is an acute care hospital with 267 data bases (one for each case) contains the following elements:
physicians. This 571-bed facility is a university hospital, and (1) raw material (including interview transcripts, researcher’s
its medical imaging department employs 11 radiologists and field notes, and documents collected during data collection);
40 technologists and conducts an average of 110,000 radiolog- (2) coded data; (3) coding scheme; (4) data displays; (5) general
ical examinations per year. The Beta site is also an acute care chronological log of data collection.
hospital. It counts 343 beds and 187 physicians, and its medical
imaging department employs 9 radiologists and 45 technolo-
gists. The Beta site processes as many radiological exams as 4. Case description
the Alpha site, the only difference being that over half of the
exams conducted at the Beta hospital are external referrals. 4.1. PACS implementation project at the Alpha
Table 2 presents general profiles of the two sites. hospital
The raw data used to build the case histories were collected
through observation, interviews and available documentation. It is important to specify from the outset that the PACS adop-
Two days of additional observations were required of PACS use tion process adopted at the Alpha hospital was not created
by radiologists, technologists and specialists. This approach by any one individual, but rather came out of the work of a
resulted in a dozen informal interviews with the physicians we team of three people who had direct involvement in the oper-
were observing. As part of the formal interviews, we met indi- ations of the radiology department: two radiologists, including
vidually with the key stakeholders involved in the PACS adop- the head of the department, and the technologist in charge of
tion process in each hospital. All the comments we collected administrative services.
were transcribed from recordings made of more than a dozen
interviews with stakeholders. Finally, we consulted many doc- “The success of the PACS project can be explained in part
uments: in particular, we reviewed the annual reports of each by the sustained involvement of three individuals, includ-
institution, internal newsletters and press releases making ing myself. Each one of us was responsible for a key fea-
mention of the PACS project, log books, minutes of meeting, ture of the project, namely, the quality of the equipment
site visit reports, and a variety of cost-benefit analyses. and images, the financial aspects, and the lobbying and
26 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33

Table 3 – Strategy implemented at the Alpha hospital to lower barriers to PACS adoption and implementation

Material in bold indicates where strategies and tactics were deployed although not expected according to the conceptual framework presented
in Table 1.

negotiation efforts with hospital administrators and exter- become better acquainted with system functionalities but also
nal partners” [Head of radiology department]. to gain a better appreciation of potential benefits to the depart-
ment and, more generally, to the hospital itself. The expe-
In January 1995 the medical imaging department was using rience reinforced their interest in digitizing operations and
films, like the great majority of Canadian health care facil- provoked more serious thoughts about how the department
ities. The department had the authority to choose its film was approaching the issue: in particular, the importance given
supplier, since it had to pay for these supplies out of its oper- to supplier selection. As a result, in early 1996 the hospital
ating budget. The 1990s were characterized by significant bud- asked a second PACS supplier to submit a bid, for purposes of
get cutbacks in the public service across Canada. The goal comparison.
of balanced budgets led to massive cutbacks in health care The first official proposal to acquire PACS technology was
and in hospitals, which had to work with equipment acquisi- submitted to hospital administrators in April 1996. It was pre-
tion budgets that were virtually nonexistent and which had pared by the three individuals who initiated the project and
to run on reduced operating budgets. In the face of these comprised two distinct phases. The first phase dealt with sys-
difficult financial conditions, the radiology department initi- tems that were already digitized, such as angiography and
ated discussions about a PACS ultrasonography acquisition. In transaxial tomography systems, as well as the ultrasonogra-
February 1995 its film supplier proposed providing a demon- phy system, which could easily be digitized by adding frame
stration of its PACS technology, one that would, according grabbers. The second phase, which would be more complex
to the supplier, enable the institution to realize substantial and, above all, more costly, addressed the analog systems:
annual savings. The offer was accepted, and several weeks exposure imaging and fluoroscopy. The key element in this
later the technology was demonstrated to hospital adminis- implementation plan was the efforts that would be made to
trators as well as to all the heads of clinical units. Following minimize disturbances to hospital operations and manage
this presentation, the supplier was asked to submit a formal the transfer as a gradual change. In general, ultrasonographic
bid. images were rarely read by clinical physicians, whose needs
In the fall of the same year, the medical imagery depart- were met by consulting the radiologist’s report. The same was
ment’s head of clinical services and administrative director true of transaxial tomography and angiography images, which
attended the annual meeting of the Radiological Society of were only consulted by specific types of specialists, such as
North America, where a large number of conferences and sem- neurologists. In short, according to the plan, the first phase
inars were devoted to PACS. A complete PACS system was of implementation would mainly target the imaging depart-
also being demonstrated, providing an opportunity to not only ment.
 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33 27

Table 4 – Strategy implemented at the Beta hospital to lower barriers to PACS adoption and implementation

Material in bold indicates where strategies and tactics were deployed although not expected according to the conceptual framework presented
in Table 1.

Due to the severe budget cuts that had been in effect the hospital’s general expectations concerning PACS use and
for several years, the first condition that the project’s pro- the areas that could be affected by its deployment. The hospi-
ponents had to meet if PACS was going to be considered by tal sought a 95% reduction in the amount of film used and the
the hospital’s administration was that the project had to pay related costs, rapid access to images, a user-friendly system,
for itself. All the savings afforded by PACS would need to be and integration with the RIS already in place. This request was
estimated in order to be considered as acquisition financing. sent to six potential suppliers.
Direct savings included the cost of acquiring films and other Between February and May 1997, several site visits were
radiology supplies (for example, chemical solutions and X- conducted in hospitals where systems from the proposed sup-
ray envelopes), the cost of maintaining film processors, and pliers were already in use. The hospitals were in Canada,
labor costs (fewer technicians and file management clerks). the United States, Austria, and Germany, and each visit was
Indirect savings were much more difficult to quantify and attended by two of the three project proponents. The visits
would be less easily accepted by the administration. Among provided answers to several questions on the savings that
the types of indirect savings identified were the gains associ- could be achieved and the difficulties encountered integrat-
ated with improved productivity in each professional category, ing PACS technology into a hospital the size of Alpha.
the impact that delays in reading images have on a patient’s On the strength of information collected during these visits
hospital stay, and the costs related to not replacing equipment and based on the detailed budgets prepared during this period,
and supplies, such as lights in light boxes and shelving in the the project team came to the conclusion that, contrary to what
film library. had been proposed earlier, the best approach was a single-
The cost of acquiring and maintaining PACS technology is phase implementation.
not a fixed amount; it varies according to a number of factors,
“We believed that this strategy would maximize savings on
such as the amount of memory on the central server, the num-
film and related supplies, since 65% of such materials were
ber of workstations required, the resolution required on each
used in traditional radiography, and our radiography sys-
station monitor, and the amount of network cabling required
tem was scheduled for replacement as part of the project’s
to install the system. Since the project managers were unable
second phase” [Technologist on project team].
to determine a configuration of the equipment and software
that would be ideal for their specific needs, the traditional In addition, having only one functional system (electronic
method of preparing detailed project specifications could not filing) would accelerate the savings achieved in staffing levels
be used in the tendering process. Hence a request for proposals (estimated at 10%). Space savings would also constitute a sig-
was developed in January 1997. The request mainly identified nificant advantage, since 2000 square feet would be liberated
28 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33

once the PACS technology came online. On the other hand, a Two implementation committees were created at the
single-phase approach presented one disadvantage: the work beginning of 1998. The first and smaller committee consisted
habits of physicians throughout the hospital would be dis- of the three project proponents, a technologist who would be
rupted, even if only temporarily. responsible for managing the future system, the assistant to
A detailed action plan was prepared for an integral PACS the director of professional and hospital services, and a project
implementation and submitted to hospital administrators in manager appointed by the supplier. This committee met on
May 1997. The plan described the key issues in each of the a weekly basis to plan the deployment process, coordinate
employee groups that would be affected by the implementa- work on PACS installation and implementation, and monitor
tion and targeted key individuals in each of these groups who project progress. A second, larger committee was also estab-
could be called upon to promote the project in their depart- lished. It comprised the members of the smaller committee;
ments. This report also detailed how the acquisition would be a representative from the council of physicians, dentists, and
financed and presented a draft operational schedule. When pharmacists; a representative from the table of department
the plan was tabled, 2 months had passed since the request for heads; a representative from the technical and financial ser-
proposals was issued and three proposals had been received. vices department; and a computing representative. This com-
Shortly after the action plan was presented, one of the pro- mittee met on a biweekly basis, and its principal mandate was
posals was approved. One of the main reasons for selecting to plan, draft, and implement a detailed communications plan
this supplier was that it was already providing almost 70% of that would foster rapid appropriation of the system by users.
the institution’s systems, so administrators hoped that stay- For example, a leaflet entitled PACS Snapshots was published
ing with the same supplier would minimize integration prob- once a month. In addition to keeping the medical commu-
lems. Other points raised in the recommendation included the nity informed about the status of the project, the leaflet was
user-friendliness of the software, the fact that it could be cus- mostly used to demonstrate the administration’s support for
tomized to meet the needs of physicians, its file management the project.
software was considered superior to what was offered by com- Before acquiring its PACS, the Alpha hospital had a RIS,
petitors, the proposed high-contrast monitors were of a very although their RIS did not comply with the HL7 standard,
high quality, and overall the system was functionally well inte- so it would not be easy to establish communication between
grated. Total project cost was estimated at approximately $3.5 the two systems. Budgetary limitations prevented solving the
million, and the project payback period at 5 years. problem with a new RIS, so it was decided to upgrade the
The supplier had everything to gain with this contract, existing system instead. The RIS supplier therefore upgraded
since the Alpha hospital’s PACS implementation would estab- its system so that it would recognize basic PACS operating
lish a baseline for hospitals across the province. The supplier information. The hospital’s radiologists, technologists, and
installed several demonstration workstations of varying qual- physicians received their training several days before the PACS
ity to show future users how the images would appear on deployment. Due to the small size of the radiology depart-
monitors, and an invitation was sent out to all the radiolo- ment, most of the sessions were offered on a one-on-one basis.
gists, technologists, physicians, and heads of care units to see The physicians, however, received their training at depart-
the equipment for themselves. By the end of October 1997, mental meetings in order to minimize its impact on their busy
over 60 of them had viewed the various qualities of image on schedules.
display and had given their comments and discussed their The PACS deployment, which corresponded with the start
reactions with the main interested party, the supplier. Gener- of the clarifying stage, was undertaken the following fall and
ally speaking, the implementation problem, particularly as it consisted of three parts. The system was initially deployed in
concerned the reaction of future users, was taken very seri- the ultrasonography, transaxial tomography, angiography, and
ously. For example, communication efforts began well before radioscopy rooms. Two days later, the system was functional
the contract to acquire the PACS was signed (an event that in all the imaging rooms except those in the ER, where it went
coincided with the beginning of the redefining/restructuring online the following month. Technical support was provided
stage). around the clock by the PACS supplier. There were many posi-
tive comments made during the first few days that the hospital
“We never hided any information or detail to our staff. Sev- operated without films. People remarked on the very high
eral months prior to the “go live” everyone knew exactly quality of the images, and it was said that this aspect would
which jobs would be abolished or transformed and which become the hub of the system because it would be widely used
ones would not. Some people have thus decided to quit in the hospital. In ultrasonography, the technologists quickly
the department before the layoffs. As a result, the transi- remarked on the lighter task load. In transaxial tomography,
tion process happened pretty smoothly” [Technologist on the PACS visualization tool was judged very effective and the
project team]. accessibility it gave to images was much appreciated. Finally,
according to the project post-mortem report, 6 weeks of tech-
Moreover, from the start of 1996, the hospital’s adminis- nical adjustments were required to completely stabilize the
trative director attended numerous meetings and multidisci- system, and the first gains in productivity were registered after
plinary assemblies where she made a point of giving updates only 3 months.
on the project. The two radiologists also helped bring medical Twelve months after system deployment an in-house poll
personnel on board by devoting many hours to making pre- was conducted of PACS users. This fell during the routinizing
sentations and discussing the advantages of PACS, often in an stage. According to the survey, 100% of users preferred using
informal setting, with their clinical colleagues. PACS to using films, the ease of access to images saved the
 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33 29

average physician 28.5 min per day, and the radiologists and tee was made up of four influential members of the institution.
technologists perceived significant improvements in produc- One was the radiology department’s director of administra-
tivity. Table 3 summarizes the Alpha hospital’s approach to tive services, whose primary responsibility was to prepare the
PACS adoption and implementation. economic arrangements that would ensure that the project
would be self-financing and to assume a leadership position
4.2. PACS implementation project at the Beta hospital in the bid evaluation process. The department’s clinical leader
had been practicing radiology since 1975 and had managed
Compared to the average Canadian hospital, the Beta hospital the department for several years. A second radiologist on the
is known as an institution with cutting-edge technology. For steering committee had actively participated in the evalua-
example, the Beta hospital was among the firsts to acquire tion of many imagery equipment acquisition projects in the
a computer-controlled CT scanner (CCCT) in the 1980s and past. He was a computer enthusiast and was considered by
a magnetic resonance device in the 1990s. According to the his colleagues to be well versed in cutting-edge technology.
head of the radiology department: “This strategic approach Since the mid-1990s he had read everything he could find on
[innovation] has helped us achieve client retention and growth PACS. Rounding out the committee was the director of biomed-
objectives”. It should also be mentioned that the great major- ical engineering and computing, someone who had extensive
ity of technological innovations acquired by the Beta hospital technical expertise and had managed several large comput-
in recent years have been financed through very successful ing projects in the hospital, including the deployment of a
public fundraising campaigns. large network. Collectively, the members of the steering com-
For several years, the radiology department’s clinical leader mittee represented a broad base of experience in diagnostic
and its director of administrative services attended the annual equipment acquisitions, and all of them were innovators in
meeting of the Radiological Society of North America in order the sense given by Rogers [1].
to monitor the types of diagnostic devices and digital systems The tendering process was officially launched on Novem-
available in the marketplace. During the 1997 meeting, they ber 20, 1997, and a public opening of the bids was held on
conducted a technology scan that eventually led to the acqui- January 23, 1998. Two sets of specifications were developed,
sition of a new CCCT, a new digital fluoroscopy room (with the first for diagnostic devices and the second for the PACS.
three digital systems) and laser cameras (dry processors for The steering committee met to analyze each of the tenders.
digital systems). The total estimated cost of this equipment The committee recommended accepting the lowest bid on the
was CAN$7 million. It was at this same meeting that they diagnostic device tender because it met all the requirements
became interested in PACS technology, as several PACS solu- of the tender package. As for the PACS tender, the commit-
tions were being demonstrated in the trade exhibition areas. tee retained the two lowest tenders for further consideration.
Since the hospital had already digitized the angiography ser- Visits of the tenderers’ manufacturing facilities and of clini-
vice and since the ultrasonography service could easily be cal sites were conducted in February and March 1998. Finally,
converted by adding frame grabbers, they realized that, with the committee selected the second-lowest bidder, based on
their new acquisitions, over half of the Beta hospital’s imaging several factors including the superior quality of the proposed
systems would be ready to be connected to a PACS. In addition, monitors, the versatility of the PACS consoles for the CT scan-
the acquisition of PACS technology would eliminate the need ner and magnetic resonance, the possibility of integrating
for new laser cameras. reports, and the capacity for personalized set-ups in the exam-
Shortly after the two specialists returned from the meet- ination consoles. In addition to these factors, mention was
ing, a steering committee was struck and a decision made to made of the supplier’s established track record, both in Canada
integrate a PACS acquisition into the current project to reor- and internationally. Finally, another, non-negotiable element
ganize the imaging department. In fact, the Beta hospital was of the bid was that the system would be highly integrated with
at that time planning a significant physical reorganization of the new diagnostic equipment, to be provided by the same
the department. The first part of this reorganization involved supplier.
acquiring digital imaging equipment (a magnetic resonance The committee’s recommendations, along with the pro-
device, a digital fluoroscopy device, a new CCCT). The second posed economic arrangements, were submitted to Beta’s
part consisted of digitizing all the department’s operations, administration at the beginning of April. One month later
thereby minimizing if not eliminating the use of film. Acquir- the hospital’s board of directors approved both recommen-
ing a PACS was an integral part of meeting this objective. At dations, effectively marking the end of the initiation phase.
this point in time, it was expected that PACS would soon be All of the interviewed stakeholders felt that the influence of
deployed in virtually all the province’s health care facilities, individuals on the steering committee had played a key role
and the Beta hospital wanted to be among the first to try it. in the board’s decision to approve the PACS project. The finan-
The third part of the reorganization would require a physical cial plan proposed a project with an 8-year payback period,
reorganization of the facilities. More specifically, the hospital financed through successive operating budgets of the medical
wanted to bring together the various activities of the imaging imaging department. The total cost of the PACS project was
department and build new rooms for the magnetic resonance estimated at approximately $2 million.
device, the fluoroscopy device, and the electronic filing equip- The mandate of the steering committee ended as the
ment. It also planned to build new waiting rooms for patients implementation phase began, and once the contract was
and new office space for the radiologists. signed with the supplier, its four key members quickly gave
The creation of a steering committee considerably eased up their responsibilities in the file. This event represented
the Beta hospital’s process for acquiring a PACS. The commit- an important shift in the story of the Beta hospital’s PACS
30 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33

implementation; the project’s proponents considered the rest local practices (the routinizing stage) was still incomplete,
of the project to be more operational in nature, and they would and no gain in productivity had yet been registered. Films,
hand project management to a small group consisting of a however, were no longer being used. Table 4 summarizes the
technologist, who would administer the PACS, a technician RIS/PACS adoption and implementation process employed at
provided by the supplier, and another technician from the hos- the Beta hospital.
pital’s computer department.
Following this transition, work focused essentially on the
technical aspects of the project, leaving aside all of the human 5. Analysis and discussion
and organizational considerations normally associated with
this type of project. The PACS was parameterized and inte- This section analyzes the process that led to PACS adoption
grated with the RIS, and the planned office facilities and and implementation in each of the two cases described above
new rooms were built. Several initiatives were undertaken to as well as the success of each. The main goal of this com-
develop interfaces for communication between the RIS, which parative analysis is to better understand the nature of the
had been in operation for several years, and the PACS. No sig- challenges faced in this type of project, the conditions for suc-
nificant work was devoted to managing organizational change: cess, and the strategies that can be implemented in order to
a single demonstration of the system was organized by the maximize the benefits derived from PACS.
supplier shortly after the contract was awarded. Several physi- The first element that these projects had in common was
cians only learned of the imminent arrival of the system in a unquestionably the presence of a group of individuals, lim-
notice sent by the hospital’s executive director. ited in number, who wielded enough decision-making power
The PACS administrator took a week of intensive training to oversee the PACS acquisition process. The professional
in the United States and trained another technologist upon bureaucracy, a structure very typical of hospitals, certainly
her return. The three new project “managers” provided train- played a role in this outcome, since professional bureaucracies
ing sessions to future system users several days before the confer greater decision-making power to the professionals
technology was deployed. Approximately 30% of the invited working in an operational center: in this case, the radiologists.
physicians came to the sessions, a clear indication of the Mintzberg [36] has indicated that within such a structure, pro-
importance they gave to the project. fessional groups exercise a significant amount of control, not
only over the nature of their own work but also over the admin-
“Many problems arose shortly after the deployment in early
istrative decisions that could affect them, particularly those
2000. Several physicians moaned about the poor quality of
related to resource allocation. A second condition that had a
the images, and these complaints disrupted activities in
significant influence over the PACS adoption in both hospitals
the radiology department for several weeks. Eventually the
was that economic barriers fell when it could be shown that
head of our department had to ask the hospital’s executive
the project would cover its own costs. In order to make their
director to intervene and force all the physicians involved
way through the relatively complex minefield represented by
to use the system” [Radiologist on project team].
this budgeting process, key actors in each of the two establish-
Furthermore, there were many and frequent problems ments undertook a detailed analysis of all the costs that could
encountered in integrating the PACS with the RIS. The prob- be eliminated or reduced by a PACS deployment, not only in
lems disrupted use of the PACS and, as a result, affected activ- the radiology department but throughout the hospital.
ities in the radiology department. Given the seriousness of The presence of a group of influential actors and proof
these problems, almost a year after the initial PACS deploy- of financial self-sufficiency represent necessary conditions for
ment it was decided to issue a new call for tenders in order project success, but they in themselves are not sufficient. The
to acquire a new RIS. This time the acquisition project was innovation process adopted at each site had major reper-
managed by the computing department, which provided the cussions on both the approach taken in the respective PACS
necessary funds from its operating budget. For obvious rea- project and its success. The most classic way to differentiate
sons, the hospital decided to acquire the RIS from the PACS innovation styles is “technology versus people” [37]: the tech-
supplier. Some re-engineering work was required, however, to nology style implies a planned and rational strategy centered
render the new RIS compatible with departmental procedures. on technological considerations, with a relative exclusion of
Other problems were encountered as the new RIS was wider psychosocial and organizational concerns, while the
deployed. The RIS server crashed on its first day of opera- people style implies explicit consideration of the experiences
tion. Now that the two systems, the RIS and the PACS, were of end-users as well as organizational impacts and opportu-
completely integrated, a crash of the RIS system paralyzed nities. As illustrated below, these styles differ according to the
the entire department. Among other things, the supplier had objectives of the innovation, the strategies used to introduce
apparently underestimated the amount of clinical data that the technology, and the effects of technology implementation
would move between the two systems. The old RIS there- on organizations and individuals.
fore had to be re-installed long enough for the supplier to The PACS project at the Alpha hospital was motivated by a
resolve a series of technical issues. The department made a clear need to reduce operating costs in the radiology depart-
second attempt at deployment several weeks later. This time, ment and improve the performance of the imaging depart-
the decision was taken to proceed incrementally, modality by ment by changing processes. From the outset, the project team
modality. looked at the innovation from the points of view of all the
As we were conducting this study, 15 months after the PACS concerned authorities and located it in a comprehensive plan
technology was first deployed, the system’s integration with for transforming hospital procedures for reading radiological
 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33 31

images. In the process, the team was quickly able to lower the majority of the challenges that arose in the implementation
technological, economic, behavioral, and organizational bar- phase.
riers inherent in this type of project. In stark contrast, the innovation process adopted at the
The project’s technological barriers were considerably low- Beta hospital was primarily motivated by a desire to make
ered by their choice of supplier. The successful bidder was improvements to the technological platform of the medi-
a competent supplier, and the hospital’s business relation- cal imaging department and, in the process, underscore the
ship with it had been based on mutual trust over a number department’s cutting-edge image within the hospital. Once
of years. The quality of input from the supplier’s represen- these objectives were established, a situation was created
tatives had a positive effect, starting with the first phase where almost all efforts would be concentrated on lowering
of the adoption process when there were many uncertain- technological barriers: the facility’s computer network was
ties surrounding the innovation. Expectations on the part of upgraded, the PACS was integrated with diagnostic equip-
various stakeholder groups were then well managed when ment, the RIS was parameterized, and new premises were
a preliminary proposal for PACS adoption was submitted to built. After they had demonstrated the project’s financial fea-
the hospital and numerous site visits were conducted. The sibility and acquired the technology from the chosen supplier,
goal of the site visits was to demystify the type of impact the project proponents decided to pull out of the file, leav-
this technology could have on the organization of work and ing all the responsibility for the PACS implementation in the
the organizational structure. Once the economic feasibility of hands of a small team consisting of two computer technicians
the project had been demonstrated and the technology had and one technologist. This decision was clearly a serious mis-
been acquired, more energy was devoted to lowering orga- take. In contrast to the Alpha hospital’s approach, after the
nizational and human barriers through a series of activities. Beta hospital had made their acquisition practically no effort
There was a proposal to reorganize the radiology department, was made to address organizational and behavioral barriers.
and two multidisciplinary implementation committees were This type of technological deployment strategy is based on
created. Indeed, in the Alpha hospital’s PACS project a sig- the kind of “magic thinking” described by Markus and Ben-
nificant number of decisions and strategies concerning the jamin [38]. Project proponents believed, wrongly, that on the
implementation were in place even before an official finan- strength of its intrinsic qualities and, almost like magic, the
cial commitment had been signed with the supplier. As for PACS would be matched to current hospital practices and
behavioral barriers, a significant effort was expended on com- have the desired effects on the imaging department and the
munications from the very start of the matching stage, and rest of the hospital. When users signaled their dissatisfaction
this effort continued through three post-acquisition stages shortly after the system came online, rather than listening to
with a series of activities and initiatives, including letting their complaints and attempting to make some adjustments,
users test the new technology before it was deployed, strik- project managers simply tried to force the system on them. As
ing two multidisciplinary implementation committees, grad- explained above, several months after PACS deployment at the
ually deploying the PACS in units and departments, and pro- Beta hospital, films had disappeared from use (the key project
viding technical support to users once the technology was objective), yet uncertainties surrounding the impact of the
in their hands. With this proactive approach, the hospi- PACS on the radiology department and resistance from users
tal was able to anticipate and adequately address the great remained.

Table 5 – Summary of conclusions and lessons learned for practitioners

Main conclusions Lessons learned for practitioners

Merely deciding to adopt PACS does not guarantee success; effective To demonstrate the financial viability of the project is a necessary,
PACS implementation is also necessary but not sufficient condition for success
Share the project vision with all the concerned authorities and
parties
Conduct an early assessment of the project context and derive a
substantive plan describing potential key challenges

A planned and rational implementation strategy centered on Treat any PACS deployment not simply as a rollout of new
technological considerations, with a relative exclusion of wider technology but as a project that will transform the organization
organizational and human concerns, is most likely to lead to Do not believe in “magic thinking”
project failure Rather, adopt a proactive strategy that takes into consideration all
the technical, economic, organizational and human factors and
that does so from the very first phase of the innovation process

The quality of the implementation strategy can largely be predicted The active and sustained involvement of highly motivated actors
by the key actors involved in the process, given their backgrounds, with complementary skills and interests is likely to favor project
commitment, and levels of motivation success
Key actors must exert enough decision-making power to oversee
the PACS acquisition process
32 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33

[7] S. Bryan, T.D.C. Weatherburn, J.R. Watkins, M.J. Buxon, The

6. Conclusion benefits of hospital-wide PACS: a survey of clinical users of
radiology services, Br. J. Radiol. 72 (1999) 469–472.
[8] J. Pilling, Problems facing the radiologist tendering for a
Drawing on the classical theory of diffusion of innovations and
hospital wide PACS system, Eur. J. Radiol. 32 (2) (1999)
on the theory of barriers to innovation, this study has sought a
101–105.
better understanding of challenges faced in PACS implemen- [9] R.O. Redfern, H.L. Kundel, M. Polansky, C.P. Langlotz, S.C.
tations in hospitals and of the strategies required to ensure Hoii, P.N. Lanken, A PACS shortens delays in obtaining
their success. To attain this objective, we have described, ana- radiographic information in a medical intensive care unit,
lyzed, and contrasted the process used to adopt and imple- Crit. Care Med. 28 (4) (2000) 1006–1013.
ment PACS at two Canadian hospitals. The main conclusions [10] J. Watkins, G. Watherburn, S. Bryan, The impact of a PACS
upon an intensive care unit, Eur. J. Radiol. 34 (1) (2000) 3–8.
derived from our study along with some key practical lessons
[11] B. Reiner, E. Siegel, M. Scanlon, Change in technologist
are summarized in Table 5. productivity with the implementation of an enterprisewide
What we learned is that merely deciding to adopt PACS PACS, J. Digit. Imaging 15 (1) (2002) 22–26.
does not guarantee success; effective implementation is also [12] S.G. Langer, Impact of tightly coupled PACS/speech
necessary. PACS are not turnkey systems. In other words, the recognition on report turnaround time in the radiology
results from this research clearly demonstrate the importance department, J. Digit. Imaging 15 (1) (2002) 234–236.
[13] B. Cox, N. Dawe, Evaluation of the impact of a PACS system
of considering a PACS deployment as a project of organiza-
on an intensive care unit, J. Health Organ. Manage. 16 (2/3)
tional transformation, not simply as a deployment of tech-
(2002) 199–205.
nology. Although there are no universal normative models of [14] L. Lepanto, G. Paré, A. Gauvin, Impact of PACS deployment
PACS adoption and implementation, the proponents of these strategy on dictation turnaround time of chest radiographs,
projects must keep in mind the fact that the project’s techni- Acad. Radiol., in press.
cal complexity, even when very real, must not garner all the [15] L. Lepanto, G. Paré, D. Aubry, P. Robillard, J. Lesage, Impact of
attention and efforts of the project team. In order to maxi- PACS on dictation turnaround time and productivity, J. Digit.
Imaging., in press.
mize the likelihood of success, it would appear to be crucial to
[16] J.W. Dearing, E.M. Rogers, Agenda-setting, Sage, Thousand
adopt a proactive implementation strategy, one that takes into
Oaks, CA, 1996.
consideration all the technical, economic, organizational, and [17] R.G. Fichman, The diffusion and assimilation of information
human factors, and does so from the first phase of the adop- technology innovations. in: R.B. Zmud (Dir.), Framing the
tion process. As illustrated in the Beta case, one of the major Domains of IT Management: Projecting the Future through
mistakes made in this type of situation is to underestimate the the Past, Pinnaflex Educational Resources Inc., Cincinnati,
critical importance played by a PACS in the redesign of depart- OH, 2000, pp. 105–127.
[18] C. Romm, N. Pliskin, The role of charismatic leadership in
mental and hospital-wide workflows [39]. Finally, evidence
diffusion and implementation of E-mail, J. Manage. Dev. 18
from the two cases reveals that the nature and quality of the (3) (1999) 273–298.
implementation strategy can largely be predicted by the key [19] N.M. Lorenzi, R.T. Riley, Organizational issues = change, Int. J.
people involved in the process, given their backgrounds, com- Med. Inform. 69 (2/3) (2003) 197–203.
mitment, and levels of motivation. Therefore, it would appear [20] K.J. Dryer, A. Metha, J.H. Thrall, PACS: A Guide to the Digital
that the active and sustained involvement of key actors with Revolution, Springer-Verlag, New York, NY, 2002.
[21] C. Nøhr, S.K. Andersen, S. Vingtoft, K. Bernstein, M.
complementary skills and interests, as prescribed by Law and
Bruun-Rasmussen, Development, implementation and
Zhou [40], is likely to favor the emergence of a higher-quality
diffusion of EHR systems in Denmark, Int. J. Med. Inform. 74
implementation strategy and, consequently, encourage the (2–4) (2005) 229–234.
rapid absorption of an innovation into a hospital’s ongoing [22] A.F. Dowling, Do hospital staff interfere with computer
activities and routines. system implementation? Health Care Manage. Rev. 5 (1980)
23–32.
[23] P.J. Hu, P.Y.K. Chau, O.R. Liu Sheng, K. Yan Tam, Examining
references the technology acceptance model using physician
acceptance of telemedicine technology, J. Manage. Inf. Syst.
16 (2) (1999) 91–112.
[1] E.M. Rogers, Diffusion of Innovations, 4th ed., Free Press, [24] N.M. Lorenzi, R.T. Riley, Managing change, J. Am. Med.
New York, NY, 1995. Inform. Assoc. 7 (2000) 116–124.
[2] P. Attewell, Technology diffusion and organizational [25] T.W. Lauer, K. Joshi, T. Browdy, Use of the equity
learning: the case of business computing, Organ. Sci. 3 implementation model to review clinical system
(1992) 1–19. implementation efforts, J. Am. Med. Inform. Assoc. 7 (2000)
[3] H. Tanriverdi, C.S. Iacono, Knowledge barriers to diffusion of 91–102.
telemedicine, in: Proceedings of the 20th International [26] M.-P. Gagnon, G. Godin, C. Gagné, J.-P. Fortin, L. Lamothe, D.
Conference on Information Systems, Charlotte, NC, 1999, Reinharz, A. Cloutier, An adaptation of the theory of
pp. 39–50. interpersonal behaviour to the study of telemedicine
[4] S. Nambisan, Y.-M. Wang, Roadblocks to web technology adoption by physicians, Int. J. Med. Inform. 71 (2/3) (2003)
adoption? Commun. ACM 42 (1) (1999) 98–101. 103–115.
[5] CAR (Canadian Association of Radiologists), PACS for [27] T.A.M. Spil, R.W. Schuring, M.B. Michel-Verkerke, Electronic
Canadians, 2003. prescription system: do professionals use it? Int. J. Healthc.
[6] O. Ratib, M. Swiernik, J.M. McCoy, From PACS to integrated Technol. Manage. 6 (1) (2004) 32–55.
EMR, Comput. Med. Imaging Graph. 27 (2002) [28] D. Short, M. Frischer, J. Bashford, Barriers to the adoption of
207–215. computerised decision support systems in general practice
 i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 7 6 ( 2 0 0 7 ) 22–33 33

consultations: a qualitative study of GPs’ perspectives, Int. J. [34] M.B. Miles, A.M. Huberman, Qualitative Data Analysis: An
Med. Inform. 73 (4) (2004) 357–362. Expanded Sourcebook, 2nd ed., Sage, Beverly Hills, CA,
[29] K. Zheng, R. Padman, M.P. Johnson, H.S. Diamond, 1994.
Understanding technology adoption in clinical care: [35] K.M. Eisenhardt, Building theories from case study research,
clinician adoption behavior of a point-of-care reminder Acad. Manage. Rev. 14 (4) (1989) 532–550.
system, Int. J. Med. Inform. 74 (78) (2005) 535–543. [36] H. Mintzberg, The Structuring of Organizations: A Synthesis
[30] J.M. Schectman, J.B. Schorling, M.M. Nadkarni, J.D. Voss, of the Research, Prentice-Hall, Englewood Cliffs, NJ, 1979.
Determinants of physician use of an ambulatory [37] F. Blacker, C. Brown, Alternative models to guide the design
prescription expert system, Int. J. Med. Inform. 74 (9) (2005) and introduction of the new information technology into
711–717. work organizations, J. Occup. Psychol. 59 (1986) 287–313.
[31] L. Lapointe, S. Rivard, A multilevel model of resistance to IT [38] M.L. Markus, R.I. Benjamin, The magic bullet theory in
implementation, MIS Q. 29 (3) (2005) 461–491. IT-enabled transformation, Sloan Manage. Rev. 38 (2) (1997)
[32] K. Lewin, Group decision and social change, in: E. 55–68.
Newcombe, R. Harley (Eds.), Readings in Social Psychology, [39] E. Seigel, B. Reiner, Work flow redesign: the key to success
Henry Holt, New York, 1952, pp. 459–473. when using PACS, Am. J. Roentgenol. 178 (2002) 563–566.
[33] R.K. Yin, Case Study Research, Design and Methods, 3rd ed., [40] M.Y.Y. Law, Z. Zhou, New directions in PACS education and
Sage, Beverly Hills, CA, 2003. training, Comput. Med. Imaging Graph. 27 (2002) 147–156.