journal homepage: www.e

Available online at www.sciencedirect.com

nsystems Mirroring a wide spectrum

$

Ulf Melinn, Karin Axe

Department of Management and ELinkoping, SwedenAvailable online 13 November 2013

KEYWORDSHealthcare

on and use. This istation is linked to

healthcare organizations as professional bureaucracies, being highly politicized and institutionalizedso discussed in this

reserved.

s (cf. [26]) of anof technologicalprocess are also

in managing healthcare costs and in improving the quality ofSafe public e-services an issue of trust and organization,

Health Policy and Technology (2014) 3, 26352211-8837/$ - see front matter & 2013 Fellowship of Postgraduate Medicine. Published by Elsevier Ltd. All rights reserved.care [9]. Simultaneously, HIS implementation processeshave been evaluated by many IS researchers. Heeks [12]indicates that a majority of these studies highlights success-ful implementation processes, while failures to implement

http://dx.doi.org/10.1016/j.hlpt.2013.11.001

nanced by: Swedish Civil Contingencies Agency.nCorresponding author. Tel.: +46 13 28 44 37.E-mail addresses: ulf.melin@liu.se (U. Melin),

Karin.axelsson@liu.se (K. Axelsson).that can be related to the implementation of a healthcare illustrated. Integrated HIS are increasingly implementedglobally in the care sector [27,32] and have much to offerand to the IT artefact as an integrated HIS. Viewing images as rational myths is alpaper as an original approach to understand HIS implementation.& 2013 Fellowship of Postgraduate Medicine. Published by Elsevier Ltd. All rights

Introduction

This paper deals with the spectrum of interpretations

information system (HIS). Different imageimplementation project as well as imagesand organizational changes in a complexillustrates several images of HIS implementation and use. Highlighting images is onreluctance, support, complexity and power that are present in HIS implementatione important contribution from this article. The complexity in the implemenrelation between IS and organization. This challenge literature on critical success factors. This studye way to illustrateinformation system;Structuration theory;Implementation;IT project;Resistance;Imageslsson

ngineering, Information Systems, Linkoping University, SE-581 83

AbstractThe main purpose of this paper is to discuss the spectrum of interpretations that can be related to theimplementation and use of a healthcare information system (HIS). The empirical part of this paper isbased on a qualitative case study of a Swedish healthcare provider, called Alpha, where a HIS wasimplemented. By studying how different actors interpret technological and organizational changes ina healthcare case, we mirror different images of the implementation project. Put together, thisdiversity of images provides an illustration of the complexity associated with the process ofimplementing a HIS. We apply an adjusted version of Orlikowski's practice lens, with its roots inStructuration Theory, in order to study technology in organizations (focusing inertia, application andchange). The implementation process of a HIS is much too complex to be judged as being eitherentirely positive or negative; instead it offers an illustration of the multi-faceted and reciprocalof images of an IT projectImplementing healthcare ilsevier.com/locate/hlpt

formation

27Images of HIS implementation and useHIS are rarely discussed, even though HIS implementationcan be perceived by [] change resistant stakeholdersas disruptive or even potentially life threatening. ([1],p. 177). By analyzing previous HIS studies from a failure (andsuccess) perspective, Heeks [12] explores a gap betweendesign of a new HIS and reality (present state) at a careunit. The gap can relate to different aspects; e.g. informa-tion, technology, processes, objectives and values, stafngand skills, management systems and structures. How thisgap is handled seems to be a parameter that implies thelevel of success or failure in the implementation process(ibid.). This gap is therefore relevant to study how it isshaped in terms of images of a HIS before, during and afterimplementation and use. This is the point of departure forthis paper. A HIS has several characteristics in common withenterprise systems (ES) in general (cf. [21]). An ES is, like aHIS, an answer to several problems with ordinary infor-mation systems (IS), such as low level of integration,disparate data formats and separated databases (cf. [7]).A HIS is also often standardised and provided by a supplieron a market. In HIS, like other packaged IS, it is obvious thatactions that constitute the IS are consequently separatedfrom the actions that are constituted by the system [28].HIS contains more or less standardized processes offered asbest practice and a high degree of integration (cf. [40]).Existing nuances in a healthcare context also provide a richenvironment from which to learn more of existing IStheories and their application [5,32]. The healthcare con-text, being highly politicized and institutionalized [27],is also considered to be particularly problematic in termsof realizing the benets of IS (ibid.) and still struggling withall kinds of development on a structural level (cf. [37]). Thisdimensions of the area makes it even more interesting tostudy the spectrum of interpretations that can be related tothe implementation of a HIS outlined above.

In order to capture different images of the implementa-tion and use of an HIS in this paper, the subjective andobjective aspects of social structures, human actions, and IS[28] will be used as a point of departure to analyze the HIScase in a Swedish public health provider (called Alpha).Following this view, an IS is a social product of subjectivehuman interpretation [3] and action, and they have aconstitutive role. An IS embodies interpretative schemes,provides coordination facilities and is deeply implicated inlinking social action, structure and interaction. By using agenerative practice lens for studying technology in Alpha,we will examine the institutional, interpretive, and tech-nological conditions which shape the on-going constitutionof different structures and vice versa [28]. Using thepractice lens, with its roots in Structuration Theory (ST)make it possible to capture the dimensions of institutiona-lization, interpretation and interaction characterizing thespecic empirical domain [3,27,37].

The purpose of this paper is to discuss the spectrum ofinterpretations that can be related to the implementationand use of a HIS. By studying how different actors interprettechnological and organizational changes in a healthcare case,we mirror different images of the studied implementationproject. Put together, this diversity of images provides anillustration of the complexity associated with the process ofimplementing a HIS. The paper contributes with understandingof HIS complexity by discussing these simultaneous images usingthe structuration theory. Increased understanding of HIS com-plexity has both theoretical and practical implications, asdiscussed in the concluding section of the paper.

The paper is arranged in the following sections; wediscuss the theoretical background in section two. We dothis by viewing HIS as a special case of ES. We also discussresearch on IS implementation and change. In the thirdsection we describe the research approach, followed by thecase description in section four. The case is analyzed anddiscussed in the fth section, where we draw attention toimages of the project and the HIS. The paper is concludedwith contributions and further research.

Theoretical background

Here HIS are introduced together with implementationand change in a general IS context, and a practice lens oftechnology.

Health Information Systems

The strategic importance of integrated HIS, as introducedabove, is obvious since it is used as a tool to improveservices and decrease medical mistakes [22]. Simulta-neously, HIS implementation processes have been evaluatedas such in many recent studies (e.g. [14,18,35,38]) and alsoearlier studies on e.g. computer tomography (CT) scanners[3]. Heeks [12] indicates that a majority of case studiesabout HIS implementations highlights successful implemen-tation processes, while failures to implement HIS are rarelydiscussed. By analyzing previous HIS studies, Heeks exploresa knowledge gap regarding what we can learn from failures(ibid.).

Evaluating success is a challenging activity in this con-text. There are many authors suggesting and discussingcritical success factors (CSFs) (e.g. [11,44]). Berg [4] claimssuch CSF lists to be problematic since success can be judgedin many dimensions; such as effectiveness, efciency,organizational attitudes and commitment, employee andpatient satisfaction. This makes the situation very complexand CSF lists often offer a more simplied solution. In orderto illustrate the complexities of HIS implementation pro-cesses, Berg (ibid.) investigates three myths related to suchprocesses; implying that HIS implementation is a technicalrealization of a planned system in an organization, that itcan be left to the IS department, and that the implementa-tion including the required organizational redesign can beplanned (ibid.). By scrutinizing these myths, he concludesthat HIS implementation instead is a mutual process whereorganization and technology inuence each other (furtherelaborated below). The management of a HIS implementa-tion process also implies a balancing act between organiza-tional change and using the HIS as a change agent (ibid.).Identifying and discussing the spectrum of interpretationsthat can be related to the implementation and use of a HIS,as in this paper, is one way of trying to broaden the scoperegarding simplied sets of CSF and investigate situationalaspects further.

An important aspect when discussing HIS implementa-tion is to acknowledge the involved actors' expecta-tions and perceptions. It is not feasible to announce an

success or failure of an HIS depends on whether physicians

U. Melin, K. Axelsson28accept or resist its implementation (ibid). Sallas et al. [35]present an assessment approach that focuses on users incontext. The study indicates that focusing on users incontext is a way to avoid or minimize negative reactionsupon IS implementation among highly skilled professionalslike physicians (ibid.). Jensen and Aanestad [15] use asense-making perspective to study how professionals reactupon implementation of new HIS. They argue that managersmust clarify the needs, values, and preferences of the usersin order to understand how professionals interpret andrespond to HIS implementation. This is line with a widenedperspective on the reluctance to use HIS among medicalprofessionals reported by e.g. Lapointe and Rivard [17]. Thepresent study follows the perspective on resistance high-lighted by Lapointe and Rivard [19,20], that resistance isinteresting and useful in order to understand implementa-tion of HIS and its reciprocal relationship with actors andstructure. A systematic analysis of different models ofresistance is elaborated on by Lapointe and Rivard [19]. Inthe images highlighted below, resistance is one dimensionamong others.

Enterprise systems are enterprise-wide application pack-ages that tightly integrate enterprise functions [7]. Suchsystems have some aspects in common with integrated HISlike the one studied in this article. Khoumbati et al. [16]argue for the healthcare sector's need to integrate applica-tions organizational-wide, which is in line with our view ofintegrated HIS. ES, as well as HIS, are created in differentsteps: the development of the system, the modication ofthe system, and the use of it. The point, partially followingOrlikowski [28], is that the ES is constituted in at least twosteps: by the supplier and by the system implementer,followed by constituting user action.

Information systems implementation, change andstructures

The relation between IS and organization is full of nuances andcontradictions. We investigate the multi-faceted and reciprocalrelation in line with Barley [3], Howcroft et al. [13] andOrlikowski [28], moving away from an over-simplied under-standing of it [23,34]. The role and inuence of IS in organiza-tions [30,41] and the emergence and improvisation is importantto investigate when trying to understand IS implementation.Exploring the institutional context is also important when tryingto understand HIS and its relation to organizational structures,processes and outcomes (cf. [4]). There can be many reasonsimplementation a success without stating from whoseperspective it is successful, cf. Berg [4]. Actors can belongto a certain profession; e.g. physicians, nurses, medicalsecretaries or an organizational unit; e.g. a hospital ward ora health center. It is, thus, important to identify involvedactors in the beginning of an implementation project [22]and also to assure that the actors are involved and infor-med [18]. A professional group's reaction upon the imple-mented HIS is, among other things, inuenced by the waythe HIS is perceived to improve work situations or whetherthe HIS is interpreted as a control system (ibid., [14]). Therole of medical professions is also highlighted in a study ofthree implementation processes [20]. They claim that thefor implementing an HIS, not only to improve medical carequality. To gain and retain legitimacy can be even more impor-tant (cf. [25]). Structurational perspectives focus on humanaction and have a potential to help us understand emergenceand change in technologies and the use of it (cf. [3]). IS, as atechnology, is developed through a social and political processresulting in structures embedded in the present technology.Following that, the use of technology is heavily inuenced bythe users' interpretation of the functionality and by: []images, descriptions, rhetorics, ideologies, and demonstrationspresented by intermediaries such as vendors, journalists,consultants, champions, trainers, managers, and powerusers. ([28], p. 409).

The images mentioned above can be regarded as ratio-nalized myths (cf. [25]). Claims put forward from inter-mediaries (e.g. journalists and consultants) are oftenpersuasive and users tend to take them for granted [29].Technologies never become complete or fully stabilized(ibid.) and identical technologies can lead to differentconsequences [3]. We analytically regard technology asbeing possible to redene and modify meaning of. Proper-ties and application of technology can also be modied afterinitial development. It is usually common that users thathave similar work, common training, shared socialization,and on-the-job experiences, enact similar technologies inpractice. That technologies-in-practice may become insti-tutionalized and tends to predetermine prescriptions foraction. Technologies-in-practice may also impede change(ibid.). Different images (cf. [26]) of an implementationproject as well as images of technological and organiza-tional changes can be regarded as a connection betweenthinking and action within organizations. Understandingimages carried by members of an organization can be aplatform to help us nd new ways of organizing, to under-stand and change perspectives and create shared under-standing (ibid.). Images is also linked to metaphors byMorgan (ibid.) and used to create powerful illustrations ofe.g. organizations as machines. Metaphors can also be a wayof not seeing since they are based on certain ways ofthinking.

By using a practice lens for studying technology in thecase study, we examine the institutional, interpretive, andtechnological conditions which shape the on-going constitu-tion of different structures within Alpha. This lens, and theadherent perspective, is primarily based on Giddens's [10]ST. Structuration is viewed as a social process that involvesthe reciprocal interaction of human actors and structuralcharacteristics of organizations. Structures are viewed ashaving two sides (the duality of structure), enabling andconstraining human action. At the same time structures areproducts of human action. When humans act in organiza-tions, they also create and recreate the elements of socialinteraction: meaning, power, and norms. Power is linked tostructures of domination by resources. Moral sanction islinked to structures of legitimation by norms. Interpretiveschemes are stocks of knowledge and form the core ofmutual knowledge in the production and reproduction ofinteraction. Interpretive schemes also serve as a constraint.Meaning, power, and norms are dependent on technology.Elements from ST enactment (cf. [43]) will be discussedin terms of using three different types [29]; (1) inertia(no evident change in process, technology or structure

occurred), (2) application (some change has occurred in oneor more of process, technology and structure), and (3)change (change in one or more of process, technology andstructure occurred) (ibid.). The three types of enactmentare based on the conditions, actions and consequences.

Research approach

The empirical part of this paper is based on a case study ofAlpha where a HIS was implemented. The case study wasconducted during 20082011. We had the opportunity to studythree care units on site which had recently got the HISimplemented; two public health centers (one located in acity center [C] and the other located in a suburb [S]) and ahospital clinic. We also studied the Care Process Center (CPC)which managed the project. CPC is Alpha's central, strategic,resource for organizational, quality and IS development andmaintenance of care supporting processes. The case study isbased on 26 interviews within Alpha as well as studies ofdocuments. Examples of documents are Alpha's website

mentioned before). Some of the interviews also consisted ofsystem studies (Table 1), when the interviewee/system userdemonstrated a certain characteristic or pattern of use linkedto their work in the system.

This work corresponds to central concepts and ideals ininterpretive and qualitative research, such as interpreta-tion, pre-understanding and the use of multiple methods fordata generation [42]. The interpretative tradition in the ISeld is suitable here because of the kind of researchpurpose and questions raised; e.g. how the actors in Alphastruggle, both strategically and in daily work, with the HISand how they enact and interpret the system and theorganizational change. Orlikowski's framework has beenused as a guide for analyzing the interviewees' interpreta-tions of the HIS and as a part of an iterative process of datacollection and analysis [42].

The case of an integrated health informationsystem

ata

teryste

ter

Medical secretary

teryste

29Images of HIS implementation and useInterview andsystem study

Healthcenter [S]

Nurse, responsible for rstpatient contacts

Ins

Medical secretaryinformation to patients, internal project documentation,budgets, external evaluation reports, and media's coverageof the project. The different, multiple, sources of empiricaldata as well as the data collection techniques were used inorder to get a more varied and truthful view of actors'perspectives, the implementation and use of the HIS. Dataand method triangulation [17] is used as an attempt togenerate a rich picture of descriptions and impressions ofthe implementation process, its institutional context andtechnology in practice. Two interview rounds were done,one initial and one with follow-up interviews. The interviews(Table 1) were audio recorded and the average interview lastedfor 1 h. Additional notes were also taken covering impressionsand opinions of the studied actors. Qualitative interview guideswere used, with a mix of pre-dened open and open-endedquestions, topics and informal communication [31]. On-going,upcoming, themes generated by the context and by theinterviewees were also included (e.g. when an intervieweementioned new dimensions in the implementation process, not

Table 1 Interview/respondent information.

Data collection Unit Role D

Interview CPC Organization developer Ins

Project leaderCPC manager

Interview andsystem study

Clinic/surgery

Physician 1 and 2 InAssistant nurse andadministrative coordinatorview andm study

Health center[C]

NursePhysicianAssistant nurse 1 and 2Medical secretary; nurse,health center managerDistrict nurse; nurse; healthcenter manager

view Clinic/ward Assistant nurse and admin.coordinator; nurse

Physician, assistant nurse andadmin. coordinator

view andm study

Health center[S] (cont.)

Physician, health centermanagerNurse, health centercoordinatorThe HIS implementation is studied in a large Swedishorganization providing medical care. Alpha is a democrati-cally run organization and their task is to promote publichealth according to the needs of the population in theregion. Local healthcare is organized into three geographi-cal areas. Specialist healthcare operates from ten centerspositioned county wide. Alpha has 11.000 employees. Alphaoperates 40 care centers and four hospitals, which includeshighly specialized medical healthcare.

The Swedish IS vendor Cambio has developed the HIS Cosmicthat is marketed as a fully integrated and organizational-widesystem under the slogan One journal one system for allhealthcare. Cosmic has about 10,000 users in Alpha, reachingfrom four different hospitals, regional and local healthcareunits, administration, etc. Cosmic includes both a back-ofcesystem (e.g. supporting the storing of medical drugs prescribed,laboratory test results, and statistics) and a present artefact, atdesktops, when patients are meeting physicians, nurses andsecretaries (e.g. having a dialog of medical records in care

collection Unit Role

situations). Alpha purchased the HIS, consisting of severalapplication modules for healthcare, e.g. patient administrationand clinical care support, from Cambio. An important function-ality in Cosmic is the integrated Electronic Medical Record(EMR) built around a common database (cf. [21]). Alpha hadearlier bought other applications from Cambio, which was areason for choosing this vendor again. The project was initiatedin October 2005. After a planning and preparation phase, a pilotstudy started in February 2007, which was followed by a roll-outphase in 2008. The implementation project, ofcially, ended inDecember 2008 when all care units within Alpha had imple-mented the HIS. The project was run by CPC which completedthe project on time and below budget. The implementationproject's purpose was to dene and implement new processesand corresponding system support. Key aspects in the projectwere accessibility, security and efciency. The HIS comprises awidely integrated EMR for all care units in Alpha which impliedan important change compared to the prior situation. The

been in the HIS implementation project, in project manage-ment, in CPC, or in organizational development activities ingeneral at Alpha. The meanings also depend on eachrespondent's personal associations and experiences of IS ingeneral and EMRs in particular.

Several of the images (i.e. a way of thinking and seeing,cf. [26]) (summarized in Figure 1) relates to the project inthe role of image-builder for the organization and signi-cant actors in it. Prexes such as visionary, impress,and paradigm shift indicate intentions of wanting to be aforward-looking, modern organization supported by a mod-ern HIS and processes (cf. [2]). Image-building processes areimportant for the self-image within the organization as wellas in situations where there is a need to reposition Alphaagainst other public or healthcare organizations. Severalaspects also affect the scope, i.e. the volume, of theproject. This concerns the entire range from being a huge(the largest) project in Alpha's history to what, from an

of past actions (such as previous decisions to implement

ges

U. Melin, K. Axelsson30project was characterized as the largest change project thatAlpha had ever initiated. For the local healthcare centers theintegrated HIS replaced existing local IS solutions for medicalrecords. The centers located at the hospitals in the region didnot have any EMR before, so digitalized medical records werenew for them. They have had disparate IS before, handlinge.g. schedules and lab results, but no integrated IS.

Analysis and discussion

In the two sections below we analyze and discuss theempirical ndings in order to mirror various images of thestudied implementation project. The explored diversity ofimages provides an illustration of the complexity associatedwith the HIS implementation process. Then we apply anadjusted version of Orlikowski's [29] practice lens for study-ing technology in organizations.

Images of the project and the HIS

The studied project and HIS implementation were givenmany different meanings by the actors we have been incontact with. In what way the meanings differ, partlydepends on how closely involved the individual actor has

Figure 1 The many imaseveral modules from the same vendor). Thus, earlier madeIS implementation choices have already mapped out thedirection for this implementation and alternative routes, tothe one now chosen, were few, if existing at all. Alpha hasalso stated that they are not willing or even capable ofacting in the role of a system architect or integrator.

[] and we cannot pay the wages and we are not sentto keep on with it. This is what software vendors aregood at and get to work. [Interview, CPC Manager,January, 2009]

The choice of a standard solution from one or a fewvendors is based on the above intentions. The choice toproceed with several HIS modules from the same vendor asearlier implemented modules, rather than choosing othersolutions, is also motivated by education, skills and resourcereasons.

of the studied project.individual employee's and HIS user's perspective, can beseen as yet another change of IS. At the same time as theproject is considered to be forward-looking and proactive bysome respondents, it is also interpreted as an imposedproject or a consequential project by others.

It is seen as an imposed decision made as a consequence

31Images of HIS implementation and useThe implementation of Cosmic is also justied by theneed for employees to be able to move from old, paper-based routines to the integrated use of IS for multipletasks. Remarkable many of the images above also touchupon the organizational dimension of the HIS implementa-tion in terms of management, standardization and rationa-lization (i.e. to reduce costs, save time, etc.). Suchincreased opportunities for administrative functions withinAlpha to plan, coordinate and follow-up healthcare activ-ities are possible thanks to the HIS implementation (includ-ing increased information access, measure data, etc.).These opportunities are expressed in many of the identiedproject prexes. This is a reason for the expressed image ofthe project as being a civil servant project. Concerns arereported by medical staff that the control over healthcarewill increase due to the HIS. Explicit health related aspects(e.g. increased quality of healthcare) have not been ascommon as the organization related images in our empiricaldata. In contrast, expected negative effects on quality ofhealthcare have frequently occurred. The project and thesystem have been given the role as scapegoat not leastin media coverage of the process. It seems less delicate toassign an artefact negative attributes and to be an excusefor weak administrative processes or lack of leadership thanto criticize the organization of care processes, actors andadministrative processes. If we look at the non-project label(Figure 1) this aspect represents an image of the project as yetanother change of IS as one nurse at a health center expressesand interprets the implementation. For her it was the third EMRthat was implemented in a rather short time. This time moreintegrated and complex but still another system. This imagechallenges the view of the project as a huge project from anoverall organizational perspective and a CPC project manage-ment perspective.

Types of enactment and institutional aspects

In the analysis of types of enactment and institutionalaspects of the implementation process and the HIS as such,we illustrate our analysis by giving examples of the differentdimensions based on Orlikowski [29]. We use the main typesof enactment (inertia, application and change) as a basiswhen structuring our analysis. If we take a look at thedifferent types of consequences (processual, technologicaland structural) (ibid.) we refer to the different changes (ifany) that is a result of the technology-in-practice. Theprocessual consequences refer to changes in performanceand outcome of work practices; the technological conse-quences refer to changes in the technology available to theusers; and the structural consequences refer to changes instructures of a larger social system. The technical condi-tions are the same for every type of enactment in ouranalysis. We have therefore omitted this in Table 2. Tech-nical conditions are the HIS available for all users in Alpha.

In the case we have identied inertia as one type ofenactment. There is a limited technical knowledge amongdifferent professionals (assistant nurses, nurses and physicians);even if physicians tend to express their norms and values moreoften regarding the HIS in use based on medical competence.

Somebody should measure the time we spend in frontof the hourglass everyday it must be 15 minutes eachday, for each physician in the organization. [Interview,Physician, September, 2009]

The far-reaching role of medical professions highlightedby Lapointe and Rivard [19,20] is echoed also in thisimplementation study. The HIS role in physicians spendingmore time managing patient records in front of the compu-ter is also present in the study by Lapointe and Rivard [20].The images of the success or failure of the present HISdepends on strong professional groups and e.g. whetherphysicians support or resist its implementation (ibid.;[14,15]) and which rationalized myths that are spread inthe organization and its environment.

Several running commentaries in the regional media havealso been published, exposed for and discussed by personnelat the different healthcare units at Alpha.

Cosmic is dangerous for the patients and Cosmic jeo-pardizes the patients security and creates a bad workingenvironment (Regional Newspaper, October, 2008)

This is an illustration of the HIS as potentially life threat-ening (cf. [1,20]). This is also an example of external actors(journalists) that have an inuence on the users norms,interpretation and use ([29], p. 409f.) of the present HIS.We can also see that strong stakeholders use media as theirmouthpiece when expressing their views and thoughts ofCosmic, putting pressure on the project management team.These images are as a part of the rationalized myths [25,33]and also a question of power distribution. Such myths arepersuasive and users tend to take them for granted. Theimmediate interpretation of Cosmic has often been that thesystem is slow and complex, containing too many clicks;but when we continued the dialog a more multi-facetedview emerged. The overall picture is that the system isquite usable. Shared meanings (rationalized myths) are inline with a sociological understanding of institutions and therelation to IS use. What is interesting in the present case isalso that comments like this and rationalized myths startedeven before the HIS was rolled-out in the organization.It started already in the pre-implementations phase basedon limited pilot testing. In some sense the resistance startedbefore the real implementation project begun, and alsopartially in media. This is a complementary results com-pared for example with Lapointe and Rivard [19] thatidenties that the resistance starts when an implementationstarts. If we take a look at the technology-in-practicecategory within inertia we have identied, in some sense,a limited use of the HIS. Our identication is that:

[] users choose to use their new tool rarely andperfunctorily, and show little or no interest in integrat-ing its use into their ongoing work practices ([29],p. 421).

Perfunctorily use was identied in the case study con-cerning handling letters of referral where users use thesystem, but use the old paper based routines in paralleleven one year after system implementation. This is moti-vated by the users (assistant nurses) as a way of keepingtrack of the referral ow, in order to physically identify theletters in the mailbox to construct personal security andhandle stress. The use of the HIS in terms of frequency is not

d c

chrac

imiterfot rf th

ollaolleroblvidivrodroc

ollandroblvidivrodrocpr

U. Melin, K. Axelsson32Table 2 Case analysis enactment conditions, actions an

Type ofenactment

Interest inusing thetechnology

Interpretiveconditions

Institutionalconditions

Tep

Inertia Low Limited technicalknowledge amongdifferentprofessionalgroups

Hierarchical L(pno

Individualistic

Application Moderateto high

Moderatetechnicalknowledge (users)

Hierarchical CIndividualistic C

psoMore detailed

technicalknowledge (super[power] users)

CollaborativeInpP

Change Moderateto high

Moderatetechnicalknowledge (users)technicalknowledge (superusers)

Collaborative CapsoInpPImlimited; the use is mandatory regarding basic functionality(e.g. using the note function in the medical record).

The institutional conditions, within the inertia category,surrounding the HIS are rather regulated and work tasks arerelated to professional groups. In line with Orlikowski's [29]experiences we have identied a rigid career hierarchyhaving an inuence on the user of the HIS in the daily work.Hospitals have strong technical and strong institutionalcontrols [36] which explain this situation as well as thedivision of labor. Some changes in division of labor occurredwhen physicians use the note tool in order to (very brief)document diagnoses in the EMR modules, instead of voicedictation and handing over transcription to medical secre-taries. However, this change in division of labor is not anorganizational phenomenon; it is more up to the individualphysician to decide. It is also a question of norms, aquestion of age and a question of how one interprets one'sroles in Alpha (valid both for the physicians and the medicalsecretaries).

In the case study we have identied IS use providingrenement of existing actions. Examples are: to schedulepatient-physician meetings, to issue letters of referral, toprescribe medical drugs, and to document in a medicalrecord. The application of the HIS provides an overallprocess support across unit borders not provided beforeand a workow based functionality. The overall interpreta-tion of the case is that the interest in using the technology ismoderate to high. It is moderate among most of the usersinterviewed. They usually regard Cosmic as yet anothertool but more complicated. Several users identify thestructural potential in the system. The interest in using theonsequences.

nology-in-tice

Processualconsequences

Technologicalconsequences

Structuralconsequences

edunctorily;arely) usee HIS

Minor(duplication ofwork occurs)

None Reinforce andpreserve statusquo concerningsome routines

boration Increasedeffectiveness incommunicationand care

Adaptationsto the tool

Reinforce andenhance statusquo

ctivelemng Change in the

toolidualuctivity

Improvedcollaboration

ess support Increasedeffectiveness inproblem solving

borationcollectivelemng

Redened workdistribution

Change in thetool

Transformstatus quo

Shift in type ofcollaboration

Adaptationsto the tool

idualuctivityess supportovisationtechnology is higher among the super users. They aretrained to be ambassadors and to support early and dailyuse with a more detailed technical knowledge compared tothe average user.

The institutional conditions are identied in the hier-archical, individualistic and collaborate level. The hierarch-ical and control dimensions are commented above regard-ing inertia. Personal (individualistic) productivity gains areidentied e.g. regarding the fact that several people do nothave to search, or even hunt, for the paper based medicalrecord anymore. However, the opposite is also identied(productivity loss) regarding e.g. physicians searching forinformation and getting an overview of a lot of material inthe HIS (compared to the earlier fast dip into the paperbased le containing a certain patient's medical record).The HIS implementation, as analyzed above, also have anoverall efciency intention (in terms of direct medical care,administration of medical care and overall management ofthe healthcare units). The collaborative dimension is pre-sent in the integrated character of the HIS. The systemprovides the opportunity to work over healthcare unitboundaries (e.g. to solve problems together) that have notbeen the case before (without the EMR) (i.e. a processualconsequence). This has proven to make communication andcollaboration among actors easier (in terms of providingpatients with e.g. medical care overviews, previous appoint-ments, letters of referrals, and lab tests).

Changes in the HIS (a technological consequence) has alsobeen made during upgrades of the system. This has beenpartially based on comments and requests from users andsuper users within the organization and also based on the

tional one. Using Orlikowski's terms (2000) it is considered

stating that new HIS are considered as supporting for

33Images of HIS implementation and useto be more reinforcement, or modest transformation, thansubstantial transformation.

Alpha has the interpretation of the HIS as a technologythat never becomes complete or fully stabilized, even ifthey choose to treat the system as temporary stabilizedwhen nishing the implementation project. The projectmanagement team at CPC launches an initiative labeled as asecond wave that is supposed to deal with modicationsof the system after initial development [29] in order tochange processes and activities connected to the use of thesystem. This is made in order to achieve substantialorganizational change. Challenges connected to the socalled second wave can be that the technology-in-practice (ibid.) has become institutionalized and tends topredetermine prescriptions for action in the organization.Technology-in-practice may impede change (ibid.) and theobjectives from CPC associated with the initiative cancreate a new wave of inertia or even resistance tochange. These challenges are in line with a view of amulti-faceted and reciprocal relation between IS and orga-nization [13,42].

The third type of enactment (ibid.) is labeled change.Here people choose to use new technology to [] sub-stantially alter their existing way of doing things. Suchenactment results in transformation of the structural statusquo, and signicant modications to users' work practices aswell as the technological artefact. ([29], p. 423). The factthat Cosmic is a standardised HIS leaves not much space forimprovisation, emergence or customization (cf. [30,42]) ofthe system as such, for most users in Alpha. The strongtechnical and institutional controls present at hospitals andcare centers can also impede change (cf. [36]). New ordeveloped/changed ways of working have, however, beenidentied. Some of these new ways of working are relatedto the functionality concerning the EMR.

It is so easy to do the same action as before [when youhave all the history and information available in theHIS]; when you actually need to do a new examination ofthe patient. It should not be too easy [Interview,Physician, September, 2009]

The balance between routine and reection is an impor-tant aspect in the expression above. The physician thinksthat it is too easy to take certain actions when the HIS is inuse. The physician takes one example when the systemdeals with this complexity in a proper way; every medicaldrug is prescribed separately from a list of a patientsearlier use it is not possible to prescribe more than onedrug without signing each time. This creates time forvaluable reection rather than to get into a rut.vendor's knowledge and general ES development process. Atthe same time, users have adapted to the HIS regarding forexample the language used for several functions androutines implemented in Cosmic. Even if some changes areidentied in the processual and structural consequencecategory (application and change enactment below), theoverall impression is that Cosmic reinforces and enhancesstatus quo. It is, so far (also in follow-up interviews from2010), more a technological innovation than an organiza-administrative activities like monitoring and potentiallythreatening the power of the medical professional groups[14,15,39]. We have also commented upon the fact thatcollaboration among different units has been facilitated bythe HIS. It is hard to classify if this is an example ofincreased collaboration (application) or a shift in type ofcollaboration (change), though.

Conclusions, limitations and further research

An enterprise-wide HIS implementation engages and some-times provokes. But there also supporters. This is illustratedin the present case study. Different actors have theirinterpretations, images and opinions about both the IS,organizational changes to come and the results of it onseveral levels (their own work practice, their organizationalunit, Alpha as an entire organization and e.g. the patient).Some express positive interpretations and expectations,based on individual or group values and norms; the health-care practice will be more effective, more accessible andsafer. Other actors or groups of actors fear the systemimplementation, the system as an artefact and the organi-zational changes it will bring and are reluctant. Profes-sionals and groups have their attitudes, media providesother views of the process, and patients might have yetother interpretations. The present HIS is interpreted aseverything from a scapegoat to a savior for healthcare, anda way of being modern (cf. [2]) on an organizational level.The spectrum of interpretations and images that can berelated to the implementation and use of a HIS has been amain theme in this paper, and is an important contribu-tion from this research. The case study analysis illust-rates several images that can be regarded as rationalizedmyths (cf. [25]). Such images are important expression ofRedened work distribution has been identied to someextent between physicians and medical secretaries (men-tioned above). We have also identied that particularhealthcare units needs to document e.g. patients' medicalstatus thinking of the fact that everyone in Alpha can getaccess to it, read it and even base their decisions on thatpiece of information. This means that documentation by acertain unit is done for the benet of other units. This alsomeans that there is a distribution of (documentation) workamong units. Achieving change can also be restrained by theimages of the project with prexes as measurement,management control, standardization or rationaliza-tion expressing resistance from medical professionals [20].These interpretations exist among nursing staff within Alphaand create a gap between administrators and medicalprofessionals concerning motives, objectives and normssurrounding the implementation, use and effects of Cosmic.This is also a question of power between medical profes-sionals and administrators identied also by e.g. Lapointeand Rivard [16]; administrators threatening the power ofmedical professionals though the new integrated HIS withall possibilities to monitor, control, and manage resources(like with any integrated IS, such as enterprise systems). Ingeneral our case study also identies that administrators aremore positive towards the new HIS than medical profes-sionals. This is in line with previous studies (e.g. [39])

U. Melin, K. Axelsson34the medical professions' [20] view of the HIS and its imple-mentation. Such images heavily inuence the success or failureof an HIS (ibid). The results from the case study conrmsconclusions from previous HIS studies like Lapointe and Rivard[19,20] and [14,15], but also expand the body of knowledgethrough the focus on discussing images from a structurationtheory perspective. The present study also shows that theresistance and the rationalized myths starts even before the fullscale implementation process begun, which for example not ishighlighted in the study by Lapointe and Rivard [19].

The case study is also an illustration and analysis of thecomplexity in HIS implementation linked to healthcareorganizations as professional bureaucracies, being highlypoliticized and institutionalized (cf. [27]). At the same timeelements and pockets of diversity in healthcare processesare present in the organization regarding patients, disci-plines, treatment options, and delivery processes [9] chal-lenging a standardized HIS. The tension between theroutinized work and the need for exibility magnies thecomplexity surrounding HIS implementation and use and thesituational nature of it (ibid.). This tension is also high-lighted in general IS studies and studies focusing integratedIS in terms of enterprise systems and labeled as anorganization's administrative paradox [24].

In this article we have used the main types of enactment(inertia, application and change) as a basis when structuringour analysis and illustrated the different types of conse-quences (processual, technological and structural) based onOrlikowski [29]. By providing multiple images of the imple-mentation process, the project and the artefact, we haveaimed to focus the challenges that a large-scale implemen-tation process in the healthcare sector has to manage.Identifying and acknowledging images can be a way tohandle the gap between design and reality in HIS imple-mentation projects, mentioned by Heeks [12]. This has severalpractical and theoretical implications. These diverse imageshave to be identied, acknowledged and constructively dealtwith from a practical perspective and not dismissed asdysfunctional and destructive [19,20]. Jensen and Aanestad[15] also argue that it is important clarify the needs, values,and preferences of the users in order to understand howprofessionals interpret and respond to HIS implementation.However this is a challenge on a strategic level, as well on atactic and operative level of an organization. The challengeconcerns implementation strategy and pace, information andcommunication, norms and values within groups and indivi-duals, of conditions and expected consequences, as well as therelation between organizational and IS change. Theoreticallywe need to identify and use perspectives, frameworks and waysof being sensitive in order to capture multiple and temporalimages of implementation processes, artefacts and structures.Our use of structuration theory as a meta-theory is an attemptto focus such dimensions that are well known and reported in ISliterature [28,30]. When identifying diverse images, theinvolved actors are focused. In order to succeed with this kindof implementation projects, concerned professionals have to beinvolved ([22]) and informed [18]. By acknowledging variousimages of the HIS implementation actor awareness is alsoreached, which is an important aspect according to otherstudies in the HIS eld [14,15,35].

The implementation process of a HIS is much too complexto be judged as being either positive or negative it is anillustration of the multi-faceted and reciprocal, complexrelation between IS and organization (cf. [13,42]) which isillustrated through the images in the present article. Ourndings show that the images of the implementationprocess and the HIS depend on what dimensions that arein focus, which professionals that are interviewed, andwhen (cf. [4]). Our results therefore challenge the norma-tive denitions of standardised IS providing a kind ofshortcut to operational excellence (cf. [40]). Multipleimages of the project as well as the artefact as such (e.g. asa scapegoat and a savior) show the span of interpretationsin Alpha. This challenge presentation of CSF's, the reasoningof what best practice design is (ibid.) and rationalapproaches towards system implementation and change ingeneral, based on e.g. a technological imperative [23].These ndings also have managerial implications; from aproject management perspective it is important to acknowl-edge different images of a project, discussing e.g. impres-sions of shortcuts vs. long-term change. Time is animportant factor challenging the norms of trying to realizeshort term benets of HIS (cf. [27]). In the case study thereis also evidence that more radical change on a structurallevel [37] is not a result of the HIS implementation so far.

This paper uses a single case study approach in order toin-depth explore a HIS implementation process. This is aresearch design decision that of course has both pros andcons. We have been able to follow this implementation caseclosely and longitudinally and, thus, gained good access toAlpha. However, further case studies will be performed inorder to validate our ndings. A possible next step is toanalyze the links from this project to the national policylevel; a Swedish action plan for e-health has been launchedrecently. To focus on the institutional logics [6] and thepolicy implementation from a national level to a regionaland local level can also be rewarding. Analytically, institu-tional theory [25,33] can be used as a more distinct plat-form to further analyze the empirical data or adaptivestructuration theory (AST) [8] as a way to be even morespecic about technology changes, for whom and for whichprocess this makes a difference. There are also severaldimensions of resistance in the empirical data that can behighlighted and studies more carefully in line for examplewith Lapointe and Rivard [19]. In the present analysis theinterpretation of empirical data is guided by a particulartheoretical lens, with its merits and shortcomings.

Acknowledgments

The authors would like to thank PhLic Ester Andrasson forher cooperation when performing the interviews in the casestudy. This study is funded by the Swedish Civil ContingencyAgency and a part of the "Safe Public e-services - an issue oftrust and organization" project at the Department ofmanagment and engineering, Linkping University, Sweden.

References

[1] Abraham C, Junglas I. From cacophony to harmony: a casestudy about the IS implementation process as an opportunityfor organizational transformation at Sentara Healthcare. JStrateg Inf Syst 2011;20(2):17797.

[2] Abrahamson E. Management Fashion. Acad Manag Rev 1996;21(1):25485.

[24] Melin U. The enterprise system as a part of an organization'sadministrative paradox. J Enterprise Inf Manag 2010;23(2):181200.

[25] Meyer JW, Rowan B. Institutionalized organizations: formalstructure as myth and ceremony. Am J Sociol 1977;83:34063.

[27] Newell S. Editorial: special section on healthcare informationsystems. J Strateg Inf Syst 2011;20(2):15860.

[28] Orlikowski W, Baroudi J. Studying information technology in

35Images of HIS implementation and usezations: myths and challenges. Int J Med Inf 2001;64:14356.[5] Chiasson MW, Davidson E. Pushing the contextual envelope:

developing and diffusing IS theory for health informationsystems research. Inf Organ 2004;14(3):15588.

[6] Currie WL, Guah MW. Conicting institutional logics: a nationalprogramme for IT in the organisational eld of healthcare. JInf Technol 2007;22(3):23547.

[7] Davenport TH. Putting the enterprise into the enterprisesystem. Harv Bus Rev 1998;76(4):12131.

[8] DeSanctis G, Poole MS. Capturing the complexity in advancedtechnology use: adaptive structuration theory. Organ Sci1994;5(2):12146.

[9] Fichman RG, Kohli R, Krishan R. The role of informationsystems in healthcare: current research and future trends.Inf Syst Res 2011;22(3):41928.

[10] Giddens A. The constitution of society: outline of the theory ofstructuration. Polity Pr., Cam; 1984.

[11] Gil-Garca JR, Pardo TA. E-government success factors: map-ping practical tools to theoretical foundations. Gov Inf Quart2005;22(2):187216.

[12] Heeks R. Health information systems: failure, success andimprovisation. Int J Med Inf 2006;75:12537.

[13] Howcroft D, Newell S, Wagner E. Understanding the contextualinuences on enterprise system design, implementation, useand evaluation (Editorial). J Strateg Inf Syst 2004;13(4):2717.

[14] Jensen TB, Aanestad M. Hospitality and hostility in hospitals: acase study of an EPR adoption among surgeons. Eur J Inf Syst2007;16(6):67280.

[15] Jensen TB, Aanestad M. How healthcare professionals makesense of an EPR adoption. Inf Syst Manag 2007;24:2942.

[16] Khoumbati K, Themistocleous M, Irani Z. Evaluating theadoption of enterprise application integration in health-careorganizations. J Manag Inf Syst 2006;22(4):69108.

[17] Klein HK, Myers MD. A set of principles for conducting andevaluating interpretive eld studies in information systems.Manag Inf Syst Quart 1999;23(1):6794.

[18] Kohli R, Kettinger WJ. Informating the clan: controllingphysicians costs and outcomes. MIS Quart 2004;28(3):36394.

[19] Lapointe L, Rivard S. A multilevel model of resistance toinformation technology implementation. Manag Inf Syst Quart2005;29(3):46191.

[20] Lapointe L, Rivard S. Getting physicians to accept newinformation technology: insights from case studies. Can MedAssoc J 2006;174(11):15738.

[21] MacKinnon W, Wasserman M. Integrated electronic medicalrecord systems: critical success factors for implementation. In:Hawaii international conference on information systems(HICSS-42), IEEE Computer Society, January 58, Big Island,HI, USA, http://10.1109/HICSS.2009.264; 2009.

[22] Mantzana V, Themistocleous M, Irani Z, Morabito V. Identifyinghealthcare actors involved in the adoption of informationsystems. Eur J Inf Syst 2007;16(1):91102.

[23] Markus ML, Robey D. Information technology and organiza-tional change: casual structure in theory and research. ManagSci 1988;34(5):58398.organizations: research approaches and assumptions. Inf SystRes 1991;2(1):128.

[29] Orlikowski WJ. Using technology and constituting structures: apractice lens for studying technology in organizations. OrganSci 2000;11(4):40428.

[30] Orlikowski WJ, Robey D. Information technology and thestructuring of organizations. Inf Syst Res 1991;2(2):14369.

[31] Patton MQ. Qualitative evaluation methods. Beverly Hills,California: Sage Publications; 1980.

[32] Payton FC, Pare G, Le Rouge CM, Reddy M. Health care IT:process, people, patients and interdisciplinary considerations.J Assoc Inf Syst 2011;12(2/3):ixiii.

[33] Powell W, DiMaggio P. The new institutionalism in organiza-tional analysis. Chicago: University of Chicago; 1991.

[34] Robey D, Boudreau M. Accounting for the contradictoryorganizational consequences of information technology: theo-retical directions and methodological implication. Inf Syst Res1999;10:16785.

[35] Sallas B, Lane S, Mathews R, Watkins T, Wiley-Patton S. Aniterative assessment approach to improve technology adoptionand implementation decisions by healthcare managers. Inf SystManag 2007;24:4357.

[36] Scott WR. Organizations rational, natural, and open sys-tems.4th ed. N.J: Prentice Hall; 1998.

[37] Spil TAM, Stegwee RA. Strategies for healthcare systems. In:Stegwee RA, Spil TAM, editors. Strategies for healthcaresystems. Idea Group Publishing; 2001. p. 111.

[38] Sunyoung C, Mathiassen L, Nilsson A. Contextual dynamicsduring health information systems implementation: an event-based actor-network approach. Eur J Inf Syst 2008;17(6):61430.

[39] Venkatesh V, Xiaojun Z, Sykes TA. Doctors do too littletechnology: a longitudinal eld study of an electronic health-care system implementation. Inf Syst Res 2011;22(3):52346.

[40] Wagner EL, Newell S. Best for whom?: the tension betweenbest practice ERP packages and diverse epistemic cultures ina university context J Strateg Inf Syst 2004;13(4):30528.

[41] Walsham G. Interpreting information systems in organizations.Chichester: John Wiley; 1993.

[42] Walsham G. Interpretive case studies in IS research: natureand method. Eur J Inf Syst 1995;4:7481.

[43] Weick KE. The social psychology of organising. Addison-Wesley;1969. (Reading Mass).

[44] vretveit J, Scott T, Rundall TG, Shortell SM, Brommels M.Improving quality through effective implementation of infor-mation technology in healthcare. Int J Qual Health Care2007:18.[4] Berg M. Implementing information systems in health care organi- [26] Morgan G. Images of organizations. Newbury Park: SagePublications; 1986.[3] Barley SR. Technology as an occasion for structuring: evidencefrom observations of ct scanners and the social order ofradiology departments. Adm Sci Quart 1986;31(1):78108.

Implementing healthcare information systems Mirroring a wide spectrum of images of an IT projectIntroductionTheoretical backgroundHealth Information SystemsInformation systems implementation, change and structures

Research approachThe case of an integrated health information systemAnalysis and discussionImages of the project and the HISTypes of enactment and institutional aspects

Conclusions, limitations and further researchAcknowledgmentsReferences