([SORULQJ WKH XVH RI GHVLJQ SDWWHUQ PDSV IRU … · The study reported here is done as a methodolo- ... (Bakhtin 1986) and on speech act theory (Searle 1969, ... Here they use the

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

Exploring the use of design pattern maps for aligning new technical support to new clinical teammeeting routines

Carl Ake Walldius* and Ann Lantz

School of Computer Science and Communication, Royal Institute of Technology, KTH, Stockholm, 100 44 Sweden

(Received 1 June 2010; final version received 6 January 2011)

We propose the collaborative activity of mapping design patterns against stakeholder values as a viable method forgrounding conceptual design of information and communication technology (ICT) services for heterogeneous sets ofstakeholders. Preliminary experiences from designing a case book service for video-mediated gastro-medical teammeetings are presented. A diverse set of stakeholders and the challenge to apply novel technologies in a demandingenvironment have placed more responsibility on the Human Computer Interaction (HCI) team to conceptualise newwork practices and their expected e!ects than what traditional participatory design projects typically entail. Bycombining the methodologies of genre analysis and pattern languages, design pattern maps have been used toconceptualise solutions that span both work and interface aspects of the solution and that match declared values ofthe stakeholders concerned. A series of mapping sessions with di!erent stakeholders helped the inter-disciplinaryproject team to better define, what stakeholder values called for what new work patterns and what kind ofsupporting interaction design patterns these new work patterns in turn called for.

Keywords: CSCW; health information technology; interaction design; user involvement; visualisation; patterns

1. Introduction

The introduction of electronic medical analysis andrecord systems into today’s health care sector high-lights the complex interplay between new technology,users’ needs for timely and accurate information andthe highly specialised collaborative medical workroutines that shape those needs (Fussell and Benimo!1995, Cabitza et al. 2009, Swedish Ministry of Healthand Social A!airs 2010). To address this multilayeredinterplay of changing user needs, work routines andorganisational patterns, we need theoretical frame-works that can help conceptualise, compare andevaluate alternative socio-technical design solutionsin transparent ways. If the diverse set of stakeholdersa!ected by this ongoing change should have a chanceto participate as active co-designers of new ICTsystems, they also have to be involved in the ongoingprocess of conceptualising professional needs againstthe backdrop of new ICT systems, work routines andorganisational patterns (Erickson 2000, Dearden et al.2002). Thus, the theoretical framework has to providefor certain pedagogical qualities such as su!cientlyclear overviews of the processes at hand and an openand flexible vocabulary that helps stakeholders tomodel their own unique variants of known patternsand proven solutions. In this article, we reportmethodological findings from a study performed in a

gastro-medical setting, in which we have combined themethodologies of genre analysis and pattern languagesto help facilitate this kind of ICT design dialogues.

We begin by presenting the gastro-medical workand communication processes that the study is aimedat enhancing and the overall research project design inwhich the study is carried out. In the second section,we present the proposed theoretical framework interms of what earlier research we build on and ourarguments for combining the genre and the patternlanguage methodologies. In the third section, we reporton how we applied the framework to one of theproject’s ICT enhancement activities at a gastro-medical department of a Swedish University Hospitaland how the framework helped the inter-disciplinaryproject members to reflect on the proposed enhance-ment, an interactive patient case book to complementthe existing electronic patient record. In the concludingsections, we discuss the conceptual design of theproposed case book, both in terms of useful resultsfor the ongoing re-design of the department’s ICTsystems and in terms of lessons learned regardingapplying the theoretical framework. In respect tomethodological result, we conclude that the collabora-tive activity of mapping communicative genres, withtheir corresponding design patterns, against stake-holder values proved to be a promising method for

*Corresponding author. Email: [email protected]

Behaviour & Information Technology2011, 1–11, iFirst article

ISSN 0144-929X print/ISSN 1362-3001 online! 2011 Taylor & FrancisDOI: 10.1080/0144929X.2011.553749http://www.informaworld.com

Downloaded By: [The Royal Institute of Technology Kungl. Tekniska Hogskolan] At: 10:11 4 April 2011

grounding conceptual designs of ICT services forheterogeneous sets of stakeholders.

1.1. Research setting and aim of study

Beginning in 2004, the Gastro Centre at a UniversityHospital in Sweden has built up a video-mediatedconferencing system connecting six regional hospitalsto support multi-disciplinary consensus conferences,team meetings in which treatment of individualpatients are discussed, decided upon and learnt from,before and after surgery. The basic video-conferencingequipment connecting the hospitals is complementedwith two large screens at each hospital for presentingMR, CT, ultra sound and other imaging technologies.

The consensus conference system engages a widerange of direct and indirect stakeholders, includingteam members from many medical professions (about25 surgeons work at gastro), technical personnel,department management, software and hardwareproviders, public agencies, patients, relatives andcitizens as tax payers. The patient record system(Prof Doc, Take Care) is an important infrastructurefor accessing medical data all along the clinicalpathway in which the consensus conference is themain point of decision making.

The study reported here is done as a methodolo-gical development e!ort within the Funk-IS project(functionality enhancing technologies for built-insystems with application within advanced medicalprocesses), a project that focus on how haptic andinformation management support can enhance theclinical and medical e!ectiveness of the decisionconference. The project was initiated by the GastroCentre in close dialogue with the Human ComputerInteraction (HCI) researchers and the project team iscomprised of seven physicians (two of them onmanagement level) and nine HCI researchers (threePhD students, five PhDs and one research engineer).One of the key aims of the Funk-IS project, stipulatedin the call for applications by the research sponsors, isthat its results benefit commercial system providers bythe year 2016. This means quite a radical challenge fora user-centric perspective on human–computer inter-action which, not least influenced by the Scandinavianschool of Cooperative IT design, has focused IT andwork practice design on a strong participation of end-users of the new support system to be developed (Ehn1993, Bødker et al. 2000). Since changes in workforceand work practices are expected to be quite substantialin the 6 years to come (sta! turnover, new insights androutines for diagnosis and treatment, overall technicalinfrastructure), some new steps have to be taken tocomplement traditional here-and-now user participa-tion with a pedagogical method for supporting a long

term stakeholder engagement in participatory designpractices. All empirical studies performed within theproject build on established HCI methods, such asparticipatory observation, interviews and design work-shops with use scenarios for testing of mock-ups andinteractive prototypes. The aim of this methodologicalstudy is to explore a complementary method, designpattern mapping, as a feasible method for building andsustaining stakeholder engagement in participatorydesign of ICT support for the consensus conference.

2. A theoretical framework for supporting sustainableuser participation

What steps have to be taken in order to ensure thatuser engagement in analysis of new work and interac-tion patterns is sustained, beyond the users’ occasionalencounters with usability professionals during pre-studies and/or follow-ups of ICT projects? First of all,in order to transform the often ad hoc initiated userinvolvement measures into a habitual quality assess-ment practice, the usability issues addressed have to beimportant enough to be recognised by management asbeing within their area of responsibility (Brynjolfssonand Hitt 2000, Brodner 2005, Brynjolfsson andSaunders 2010). The usability e!ects of alternativedesign solutions have to be expressed in relation to thecore values of the organisation. This means that theusability e!ects on quality of treatment, patientsatisfaction, employee engagement and other corevalues have to be discernible in a way that allows forcollaborative interpretation, negotiation, decisionmaking and follow-up of usability metrics. This, inturn, means that certain pedagogical demands have tobe made on the way usability and interaction designissues are communicated between designers, develo-pers, users, management, system providers and otherstakeholders.

Based on our earlier studies on video-conferencingand quality assurance practices (Lantz et al. 2004,Walldius et al. 2005, 2009) a literature review ofcurrent studies on video-conference applications withinmedical settings (Frykholm and Groth 2009) and aseries of workshops within the interdisciplinary projectteam of HCI researchers, surgeons and radiologists,we identified two partly overlapping, complementaryapproaches that together seemed to meet our require-ments on pedagogical a!ordances. In our earlierstudies, we had built on communication genre analysisfor understanding user expectations in complex com-munication settings. For this study, we chose tocombine genre analysis with the use of patternlanguages for supporting user participation in complexICT design e!orts. In the following sections, wepresent these two approaches and how we see that

2 C.A. Walldius and A. Lantz


they interlock in ways that can underpin long-termdesign dialogues with diverse sets of stakeholders.

2.1. Genre analysis for matching interaction designsolutions against desired communication patterns

Since the early 1990s, there has been a sustainedinterest in applying the concept of genre for analysingorganisational communication in the computer scienceand HCI literature. Based on analysis of genres inliterature (Bakhtin 1986) and on speech act theory(Searle 1969, Austin 1975), Yates and Orlikowski(1992) began a series of empirical studies of organisa-tional communications in which business letters,memos, face-to-face meetings and di!erent kinds ofreports were analysed in terms of communicativepurpose and regularities in form and content. Buildingon these studies, Yoshioka and Herman (2000)analysed coordination mechanisms of a conferencewebsite from the genre perspective and, together withYates and Orlikowski (Yoshioka et al. 2001), modelleda Process Handbook developed at the MIT Center forCoordination Science from the same perspective.Here they use the formula ‘5W1H’ to summarise thesix core questions posed in genre analysis of sets oftypified activities: why (purpose), what (content), who(participants), when (timing), where (location) andhow (form of communication). Since the notion ofgenre is extended to genre systems and since arepertoire of coordination mechanisms is presentedalong this ‘genre taxonomy’, we found that both theselatter contributions o!er particularly useful insights.Not only about how social expectation and interac-tion are shaped by communication artefacts but alsoabout how the genre perspective can support delib-erate re-design. In other words, how collaborativelyanalysis and design e!orts can shape – by collabora-tively enacting – new genre expectations manifested innew artefacts and activities.

The possibilities of genre analysis as a tool fordesign e!orts is even more pronounced in Antunes andCosta (2003), a contribution that explicitly builds onthe work referred to above and which reports theenhancement, through genre system analysis, of aseries of meetingware applications. What specificallycaught the attention of our project team in the paper ofAntunes and Costa (2003) was the clear, pedagogicalmapping of the full meeting genre system, a systemwhich incidentally coincides to a large degree with theconsensus conference we were studying at the uni-versity hospital. The straightforward approach forusing the concept of genre as a tool for design was alsoinspirational for our study. As we have interpretedtheir overall process for genre-based design, it buildson the three main steps of:

(1) identifying the habitual activity patterns per-formed in the existing genre system,

(2) o!ering enhanced functionality by designingdigital support for critical activity patterns inthe genres, and

(3) providing an integrated view of the, thus,enhanced genres in order to gain acceptance,enact the new enhanced genres and buildparticipant expectations for how things aredone in the new way.

Although the literature on genre analysis oftenrefers to regularities of content and form as ‘patterns’,it is striking that the connection to another strand ofHCI literature, that of pattern languages, is seldommade. As the genre approach is used in this study as anoverarching methodology for integrating aspects ofparticipants, purposes and regularities of content andform, we see the pattern language approach as adesign-oriented method that may inform the genreperspective with an open formalism for highlightingand matching social and technical aspects in atransparent and practical way.

2.2. Pattern languages for collaborative assessment ofdesign solutions

The pattern language (PL) method is a method forrepresenting design experiences that originally ema-nated from the field of architecture in the 1970s andthat today has been successfully applied to the fields ofobject-oriented programming and interaction design(Alexander et al. 1977, Borchers 2001, Gamma et al.1995, Tidwell 2005). By emphasising the use of visualsketches and providing a formalism for accounting forcontext, the method helps designers and researchersdescribe practical solutions to recurrent design pro-blems in its di!erent fields of application (Guy 2005,Schummer and Lukosch 2007).

As the PL method is a well-known approach inthe computer science community with its ownconference tracks for di!erent application domains,it will not be summarised here (see e.g. Erickson(2009) for an overview of its application within HCItoday). What we will do, in the following sections, isto exemplify how design patterns can express designexperiences from our empirical work in terms of itsmain formalism (context – problem – counteractingforces – solution –supportive patterns). What isparticularly relevant here is in what way patternlanguages can inform genre analysis with a systematicway of matching relevant interaction design solutionsagainst desired work patterns. This amounts to thechallenge of finding a pedagogical way of collabora-tively assessing trade-o!s between the content (what),

Behaviour & Information Technology 3


form (how), timing (when) and location (where) ofalternative design solutions based on how therespective solution fulfils the purposes (why) of theengaged participants (who).

One major goal of PL in all its guises, regardless ofif it is applied to architecture, object-oriented pro-gramming, organisational or interaction design, is tofacilitate the emergence of a common ground for thediverse set of stakeholders involved in complex designventures (Tidwell 2005). This pedagogical aspect of thePL approach makes it ideal for our project, which aimsfor involving a diverse set of stakeholders in long-termcollaboration. What we specifically wanted to test wasthe potential of pattern maps (Schummer and Lukosch2007) to explicate genre patterns visually in a way thatwould give all stakeholders concerned a sense ofoverview and understanding of the overall context.

2.3. Design pattern mapping for aligning designproposals to stakeholder values

We propose the use of design pattern mapping as a keyactivity for supporting ongoing conceptual investiga-tions within the research team and, more importantly,between the team and the di!erent stakeholdersinvolved in the project. With pattern mapping wemean the activity to iteratively map out, visually, theset of design patterns selected for making the solutionmeet declared user needs and values. To map out therelationships between design pattern visually has been acommon exercise in the pattern language literature.Christopher Alexander, the inventor of the method,did it in his influential doctoral thesis on architecture(Alexander 1964), the so-called ‘Gang of four’ did it intheir application of the method to object-orientedprogramming (Gamma et al. 1995) and recently(Schummer and Lukosch 2007) have used mapssystematically in their application to computer-mediated communication. After having discussed thegeneral relationships between patterns thoroughly inthe introduction through the use of maps, Schummerand Lukosch (2007) lay out each of the three levels ofpatterns (Community support, Group support, Basictechnology) as well as each sub-level (e.g. on theCommunity level Welcome me, Guide me, Save me)through maps that depict the relationships between thepatterns on the respective level (e.g. on the Welcomeme sub-level Quick registration, Login, Welcome area,Mentor, Virtual me, User gallery, Buddy list). InFigure 1 we exemplify how a digitally supportedwork pattern in a consensus conference can be mappedout for collaborative evaluation in the strategy mapformat, a format that will be introduced shortly.

It is important to note that this mapping exerciseof work enabling patterns, specifically in the early

stages of a project, may be local in kind and deal onlywith minor aspects of the work and technology underconsideration. They may be sketches rather thanthoroughly worked out analysis of all aspects. Thepoint is that the map allows all the team members, andthen all the stakeholders, to collaboratively anditeratively see and reflect upon a set of complexrelationships between new work routines (refer toscientific evidence) and new digital artefacts (sharedannotation device) in order to discuss which e!ects(multi-disciplinary participation, decision quality) andvalues (patient health, citizen health) these routines andartefacts should enhance during an agreed upon testperiod.

What an ongoing pattern mapping activity mayamount to in the month-to-month conceptual reassess-ment of work and interaction design aspects is acollaborative reflection on how the work and interac-tion patterns proposed a!ord the values of allstakeholders. In other words, design pattern mapsmay not only cast new light on the hands-on methodsfor design of new work routines and technology butalso help build a long-term involvement of physicians,management, procurers and systems providers byo!ering a key activity in which they can negotiatetheir respective expectations. To make this integrativepotential of pattern mapping more clear, we willexemplify with a brief account of how it was exploredin one of the strands of work-and-technology redesignof the Funk-IS project, the introduction of a new‘clinical case book’ (Frykholm and Groth 2009).

3. Mapping the case book proposal

The case book is an idea that has been developed in theearly phase of the Funk-IS project to conceptualise

Figure 1. Example of a partial strategy map detailing linksbetween values for owners and patients, process e!ects, workand interaction patterns.



how the gathering, organisation and presentation ofcritical patient data, and relationships between patienttreatments, can be recorded and accessed along thewhole care chain. The case book will support thevideo-mediated team meetings where surgeons, radi-ologists, pathologists, oncologists and many otherprofessions discuss individual patients and decidewhether and how surgery should be performed. Themulti-disciplinary teams have access to a patient recordsystem but there is a wealth of information discussedduring the meetings that is not saved. The case book,or the shared clinical workspace as it was initiallycalled, is based on an overall need to record thatinformation, to visualise available patient informationin e!ective ways, thereby supporting e"cient decisionmaking and planning (Frykholm and Groth 2009). Thepatient record system, combined with dedicatedsystems, for example image management, is designedto cover all relevant data about the individual patient.But there is no readily accessible record system, forexample linked to special purpose quality records, thatcovers lessons learned from treatment of di!erentkinds of patient cases. Therefore, the video-mediatedteam meetings, as well as the whole care chain, wouldbenefit substantially from a shared interactive casebook that organises the hitherto quite unstructuredactivity of comparing outcomes of treatments betweenpatient cases. In the early discussion between theresearch team and the medical personnel, it alsobecame evident that such a system would enhancethe information management between the di"erentphases of the clinical pathway for each patient case.The comprehensiveness, relevance and accessibility ofpatient and clinical information could be substantiallyenhanced during and between the preparation, co-ordination, decision, pre-op and post-op phases of theconference through a case book that provided supportfor recording, analysis and access to the textual, verbaland image annotation data that got accumulated alongthe clinical pathway.

3.1. Initiating the conceptual exploration

How did our pattern mapping exercise help the projectteam develop its vision of a shared case book? Theexercise was based on the Funk-IS project’s extensivefield work, workshops and prototyping which has beenreported elsewhere (Frykholm et al. 2010, Olwal et al.2010). To date, in year two of the three year project,the methodological e!ort to try to summarise andmake sense of successive empirical results in patternmaps has been carried out in discussions and mappingactivities during research team meetings with succes-sive short presentations to individual surgeons, sur-geons with management responsibilities within the

project, with system vendors that have provided someof the technology tested in the project, and withfacilitators at the university commissioned to helptechnical innovators reach out to existing systemvendors. Since Swedish hospitals have used BalancedScore Cards (BSC) extensively to report their perfor-mance in relation to patients and personnel alongsidefinancial data, we first introduced the idea of usingdesign pattern maps in the format of Kaplan/Nortonscore card strategy maps for gaining overview of thedesign challenges (Kaplan and Norton 1992, Olve et al.2003). This was done in an early seminar with ourmedical partners who were all familiar with BSC,although neither with strategy maps nor with designpatterns. From our point of view, the advantage ofadhering to the strategy map layout was not only thatit would comply with the hospital’s BSC reportingformat but also that it fitted well with the case bookgenre perspective in which the values (why) of di!erentstakeholders (who) are highlighted in relation todesired e!ects (what) of proposed design solutions(where, when, how). For investigating the prospectivegenre of case books, comparing it to an instance of thegenre patient record systems seemed to be a goodstarting point. The strategy map shown to the projectpartners was part of the 2009 User Certified protocolof Take Care, the patient record system used at thehospital (Figure 2).

The rational for strategy maps is to map out,collaboratively, the interplay between planned devel-opment e"orts, their anticipated e"ects in the workprocess and the value fulfilment these e!ects representfor di!erent stakeholders/participants. In the caseof the Take Care map, the underlying analysis wasdone in an investigation of user and management

Figure 2. A summary of values for owners and patients,and process e!ects gained from the deployment of a patientrecord system, as witnessed by users and management in asoftware certification assessment.



satisfaction with the software at three hospitals, aninquiry carried out according to the Users Certified2006 certification programme (Walldius et al. 2007). Asthe medical partners were familiar with both theunderlying BSC method and the Take Care systemdepicted in the map they found this format useful as away to gain an overview to strategic concepts and theirinterplay. What was particularly interesting was that,although BSC had been used for some years at thehospital, the relationships between development ef-forts, desired e!ects and stakeholder values – that isthe logic behind yearly development commitments –had never been displayed in a strategy map and,possibly related to this, the engagement by di!erentstakeholders had been rather limited. Since BSCroutines were already in place at the hospital forfinancial, social and process related follow-ups, theydeemed that the strategy map format might well be agood way to get an overview of the development of thenew case book for consensus conferences.

In genre terms, the Who of a communication genrecan be displayed as ‘Owners’ (of di!erent kinds) on thetopmost layer of the strategy map and, as in the healthcare domain, ‘Patients’ on the level below that. TheWhy of a genre is displayed as expressions for corevalues of the identified stakeholders. We started byidentifying the values of the two main stakeholder ofthe case book genre, which in our case coincided withthe values identified in the Take Care map – Citizenhealth on the ‘Owner’ level, and Patient health on the‘Patient’ level.

When the Who and the Why in the prospectivegenre were agreed upon, the next step was to reach anagreement within the project on what detailed e!ectson the work process we deemed the case book shouldideally have e!ects that would be in line with theidentified stakeholder values. In genre terminology thisamounted to identifying What the case book woulda!ord. Based on four key themes that had beenrepeatedly referred to by the physicians and in theresearch team we identified the following e!ects on the‘Process’ level of the map: decision quality, accessibilityof relevant info, information expressiveness and multi-disciplinary participation (we chose lower case lettersfor e!ect and pattern identifiers to indicate theirpreliminary status). The next step was to identify theWhen, Where and How aspects of the case book’sa!ordances on the ‘Development’ level of the map. Inthe Take Care map, as in many other applications ofthe format, these technical aspects are summarised inthe very wide concepts of New software and New workroutines. In line with our genre pattern perspective, wedecided to identify, first a set of work patterns thatwould provide the case book’s desired e!ects, thena set of interaction design patterns that in turn

would enable those new work patterns. As displayedin Figure 3, the candidate work patterns identifiedfrom our participatory observation, interviews anddiscussions with the physicians were condense relevantinformation (e.g. from patient records, verbal com-ments of imagery analysed), annotate imagery (dis-played at the conference meetings), refer to scientificevidence (quoted by participants in conference meet-ings) and refer to personal experience (as communi-cated during the conference meetings).

We were still in too early a situation to speculateabout strengths of the relationships between tentativelyidentified work patterns and their e!ects to find itmeaningful to draw distinct cause and e!ect linesbetween them. The extension of the ‘Development’layer of the map into a ‘Work pattern’ and an‘Interaction pattern’ layer was a novel trait of thestrategy map format. Therefore, we decided that weneeded a su"ciently rich exemplification of possiblepatterns for our medical partners to recognise, choosebetween and evaluate – if these were indeed the newwork patterns they had talked about, or if still otherways of working could be more rewarding. In our owninter-disciplinary research team, we articulated a smallset of candidate interaction patterns that would sensitisethe physicians to what kind of technological enablersthat could support the new work patterns. The tentativenames we gave those candidate patterns were: a sharedwork surface (to present and visualise case data), a set ofshared selection techniques (for all participants to selectand highlight items, areas, objects etc.), a set of sharedannotation techniques (to mark items and add data andcomments to the selected items) and a set of expressivedocument formats (for representing a wide range ofmultimodal data types, see Figure 3).

Figure 3. Work in progress – a preliminary strategy mapoutline for the proposed case book.



3.2. Involving the users in the conceptual design

The exercise to map the socio-technical patterns of ourprospective case book genre in the slightly revisedstrategy map format was our main methodologicalexploration for making our development work moretransparent and accessible for a wide range ofstakeholders. In parallel, an empirical study wascarried out on how conference participants engagedin discussions during the conference, what kind ofexperiences and scientific references their commentsdrew upon and with what kind of technical supportfor pointing (e.g. laser pointers) the comments weremade (Moll and Sallnas 2009). It was observed that,while participants collaboratively analysed patientdata, the meetings were a!ected by the introductionof the laser pointers in respect to deictical referencesand referring to di!erent places in an image. This wasobserved to facilitate the discussion since complexexplanations of where the current focus was could beavoided.

To transform the finding from this and upcomingstudies into empirically validated interaction pat-terns, a list of possible user stories was set upfollowing the user story formalism of the Scrumdevelopment method: As 5kind of profession4

I want, during 5phase of clinical pathway4 to beable to 5desired interaction4 with 5desired inter-action support4 in order to 5direct process e"ect4so that 5indirect process e"ect4. The list displayeda wide spectrum of alternatives for the di!erentprofessions to choose between, and to add tothemselves. Figure 4 shows (a translated version of)this list. One of the most experienced surgeons wasthen asked to mark what he deemed to be the mostuseful alternatives. In Figure 5, the surgeon’spreferences for one of the conference phases aredisplayed. At the end of the same session, thesurgeon was asked to comment on the strategy mapagreed upon in the research team. As most of thework patterns and process e!ects in the mapresonated with the patterns and e!ects he had justmarked as desired in the user story list, hiscomments were overall confirmative. However, hewas neither inclined to prioritise among patterns andresulting e!ects, nor comment on whether the fourinteraction patterns were the most fitting ones tosupport the new way of working.

A technical design session was conducted 3 weeksafter the user story session with the surgeon, where thesame surgeon provided qualitative feedback on a set ofprototypes that were implemented to provide new

Figure 4. List of user stories about how to navigate, indicate and mark on screens.



means for navigating the medical imagery, indicatingregions of interest and supporting annotations. Theprototypes were based on combinations of interactiontechniques and hardware to allow manipulation withpen and finger on displays and surfaces of varioussizes. A set of di!erent interaction techniques wereimplemented for a Tablet PC that provided a pen and amulti-touch input, a pen-sensing Wacom tablet, and amobile phone with a pressure-sensitive touch-screen.The interaction on the devices was wirelessly synchro-nised with a large display (Olwal et al. 2010) with thesame medical imagery, to simulate the setup usedduring the medical team meetings. The results werecompared with the surgeon’s hypothesised preferencesexpressed previously as user stories. It showed that thefeedback from the prototyping session supportedseveral statements in the user stories, which indicatesthe complementary nature of these two independentmethods.

The results from the initial prototyping sessionwere then used, together with the user story list as acheck-list for comprehensiveness, to prepare a userstudy of the shared interaction devices. Twelvesurgeons tested three kinds of devices (Wacom tablet,iPod and iPad touch screen interaction) in individualsessions of a simulated team meeting in which theshared devices were synchronised with a large display.Pre- and post-session questionnaires were used as wellas think-aloud combined with interview questionsduring the test. This study is reported in Frykholmet al. (2010) and Olwal et al. (2010). What is relevantfor our conceptual design exercise is the followingobservations which will help us elaborate our workand interaction patterns’ problem-forces-solution

definitions. (1) When all team members, even remoteparticipants, have access to image interaction devices,the work pattern annotate imagery (which denotes thefull spectrum of pointing selecting and annotatingactivities) will have to address the need for a socialprotocol that enhances, or replaces, the (also technol-ogy based) control of image management that todayresides with the radiologists. (2) The user study showedthat the surgeons were more excited by the possibilityof pointing than doing actual annotations, thus, theinteraction pattern shared selection device was in thisstudy found to be more enabling than shared annota-tion device. Among the a!ordances quoted in regardsto pointing/selecting were the ability to ‘resolveambiguous references’, ‘avoid misunderstanding’ and‘improve communication for local and remote users’,forces that in turn all promote the desired e!ects multi-disciplinary participation and information expressive-ness. (3) Although pointing received stronger interestthan the actual annotation possibilities (at this stageonly simple symbols like arrows, labels and numbers)some important enabling forces of annotation werealso observed: ‘clearer referencing for younger sur-geons’, ‘stored annotations would help surgeonsunderstand patient cases faster if they could not attendthe team meeting’, ’annotations would be useful forlectures and educational material’.

The results of this and upcoming user studies withsurgeons and radiologists will provide further insightsinto how the problem-forces-solution sections of ourpreliminary work and interaction pattern should bedetailed, revised and/or replaced by more fittingpattern definitions. However, our most recent feedbackon the case book pattern map does not come from

Figure 5. Surgeon’s marked user stories for the pre-op phase.



direct users, but from an interview with one of thesurgeons with management experience and from aworkshop arranged by the institution awarding theFunk-IS research grant. There is only space here forshort summaries of four more relevant points fromthese two encounters, points that, in our judgement,indicate the overall usefulness of pattern maps as avehicle for collaborative reflection.

First, the interview with the surgeon started with ascreening of the strategy map summarising essentialvalues, e!ects and development e!orts with the TakeCare system, the existing patient record system at thehospital (Figure 2). The surgeon could confirm that allbut one of the entities were still valid as directives forcurrent development. Better research was the e!ectthat he was not willing to validate. As the system hadgrown, Less lost time and Better use of colleagues noteshad been sustained, but the database design had notbeen able to meet the needs of comparisons betweenpatient cases or fast and flexible statistical inferences.To us, this showed that strategy maps, even used asstraightforward overviews (without specification ofquantitative target measures or problem-forces-solution argument for each entity as exemplified inFigure 1), could be used to articulate well-specifiedcritique of a system in use. Second, while inspectingour proposed conceptual design of the clinical casebook, the surgeon validated the urgent need for alldisplayed e!ects, while he argued that informationexpressiveness probably had the highest potential valuefor long-term patient and citizen health, exactlythrough a more systematic use of visualisation –enabled by visual interfaces and visual interactiontechniques. Third, remarking on the two work patternscondense relevant information and refer to personalexperience, the surgeon emphasised that patientparticipation also had an interesting potential. Toengage patients in tighter loops via the internet could,if done with care at critical points of theclinical pathway, provide for better access to relevantinformation and, in turn, higher decision quality.Finally, the workshop, geared at helping to facilitatethe commercialisation of the Funk-IS project results,was the first chance to test the pattern map with asystem provider (Sense Graphics, a provider ofequipment and software for haptic feedback used inthe project) and a group of commercialisation facil-itators of the university and the grant awardinginstitution found that the conceptual design of thecase book genre in the form of a pattern map seemedto be a useful complement to other instruments fordialogues with vendors (e.g. risk analysis tools androadmaps for market introduction).

While surgeons-as-users had given feedback on thepotential usefulness of case books in their daily work,

this was an example of using the map with stake-holders in the other end of the spectrum – systemvendors and market facilitators. Again, we want topoint out that the pattern map is still more of aconceptual outline than a fully featured map. Whatremains now is to negotiate indicators and quantitativegoals (for e!ects) as well as problem-forces-solutiontriplets (for work and interaction patterns) withstakeholders, designers and technologists (of theFunk-IS projects as well as of future system providersengaging in development and deployment). Here, wefollow the advice of ‘gradual sti!ening’ given byChristopher Alexander and his colleagues in theiroriginal definition of pattern languages, to get thegreater pattern interplay right before fine-tuningrelations between subsets of patterns and the defini-tions of individual design pattern (Alexander et al.1977).

4. Conclusion

On the whole, the spontaneous comments given by theinvolved medical team members in design workshopsand user studies indicate that they have found theproject’s design methods inspiring and that themethods have given them new insights on how theywork today and on weaknesses that call for new workpatterns (Frykholm et al. 2010). In regards to thepattern map as a method to involve a broad set ofstakeholders in conceptual design of new workroutines and enabling technology, we have concludedthat the feedback we have received is promisingenough to continue the mapping exercise and to goon to define e!ects and individual design patterns insu"cient detail by involving a broader group ofmanagers, physicians and technologists in the mappingprocess. We base this on the way our preliminary mapwas validated by the lead surgeon’s user storypreferences and by the user study results, the latterproviding us with four important qualifications of maprelationships and definitions. Also, the interview withthe management representative and the discussions inthe commercialisation workshop brought forth fouradditional pointers to how the maps could be used – toidentify weaknesses in systems use (the e!ect betterresearch was negated in the Take Care map), to rankimportance of system e!ects (information expressive-ness deemed most important in case book map), toelicit new needs and opportunities (patient participa-tion), to complement existing methods for collabora-tive reflection on system design (e.g. marketingroadmaps).

Since there is a growing interest in the HCI fieldfor both genre analysis and design patterns, webelieve that the combination of the methods can be



beneficial for users of both approaches – genreanalysis giving a wider and more systematic accountof why, who demands what e!ects and design patternsgiving more detailed definitions of how, when andwhere new routines and digital services can enablethose e!ects.

5. Discussion

In the theoretical section we argued that, for userparticipation in ICT review and design e!orts tobecome a regular follow-up activity, there is a need forindividual businesses and public services to institutio-nalise e!ective user involvement in which users,management, service procurers and providers canmonitor and review quality issues with the help ofpedagogical overviews of the logic and outcome ofICT development e!orts. In our view, the mainadvantage of using design pattern maps for facilitatingsuch involvement may be that the maps clarify thatpowerful ICT services is not a goal unto itself, but ameans for enhancing the values of the organisationand the processes that support them. As we have seenin the comments from the physicians during ourworkshops and seminars, they are eager to learn fromeach other and to share opinions on how to increasethe quality of their clinical and medical workpractices. The idea to use strategy maps as a vehiclefor stakeholder dialogues on ICT development pro-jects is originally proposed by Falk and Olve (1996)and has been further developed within the UsersA-ward project (Walldius et al. 2010). The strategy mapprompts us, in a very practical way, to articulate thenew work routines which would provide the desirede!ects in the overall work process and which, in theirturn, would have to be enabled by the new technicalsupport listed as interaction patterns in the presenta-tion of the case book (see e.g. Eppler and Platts (2009)for application of strategy maps in collaborativestrategic activities).

It is important to note that our exercise to date hasconcentrated on the challenge to create a sensibleoverview of the interrelation between values, desirede!ects, and the new work and software interactionpatterns that may support them. As our scope is aprospective case book genre, the patterns should detailthe ‘what’ of the case book (in terms of e!ects oninformation content and function) while the interac-tion patterns should detail the ‘how’, ‘when’ and‘where’. The original strategy map practice developedby Kaplan and Norton requires that each e!ect bedetailed in terms of, for example, focus area, keyperformance indicators, targets and initiatives. Simi-larly, what is outlined here as tentative work andinteraction pattern names have to be detailed in terms

of problem, forces and solution (note that the twoother main definition subsections, context and suppor-tive patterns, are given by the respective pattern’sposition and links in the map). This detailing of e!ectsand patterns would then provide the basis for drawingconnecting lines between entities, as in the Take Caremap (Figure 2).

The Scrum user story format proved itself adap-table to the generic questions of: ‘who, when (andwhere)’ wants to be able to ‘what, how and why’. In ourconceptual investigation, geared to inform, and beinformed by, the parallel empirical and technicalinvestigations, the user stories have pointed out a setof gaps in our current strategy map, gaps that promptnew conceptual questions. In workshops and seminars,the physicians have emphasised the need for securefeedback loops that guarantee the integrity of theinformation in two respects: (1) that no important datagets lost or misinterpreted along the clinical pathwayand (2) that important findings from all patients andinvestigation get learnt from and applied to all futurepatients with the same kind of diagnosis. How shouldthis concept of ‘Secure feedback loops’ become visiblein the pattern map? Should it be defined as an aspect ofthe process e!ect decision quality? If so, should it beexplicated under the heading of focus area, andperhaps quantified in some way as a key performanceindicator for decision quality?

Another example of ‘missing links’ in the emer-ging pattern map regards the patient in her role as anactive member in the health care team. Fieldobservations, interviews and workshops revealedthat that there were possibilities for an enhancedinformation exchange with patients on several pointsalong the clinical pathway. Not only in respect topatient input of timely and relevant data but alsopossibilities of providing patients and relatives withupdates on where she was in the pathway and, ifapplicable, on results from investigations and treat-ments. So, should ‘Patient information exchange’ alsobe understood as an element in the focus area ofdecision quality? Which prompts a further question –should the case book have two-way communicationlinks to the patient record system (which hasregulated communication channels to patients andrelatives) to such an extent that, for example thevalue patient involvement should be declared as astrategic value on the Patient level of the map? Inwhich case, lines from the two desired e!ectsaccessibility of relevant data and information expres-siveness (in turn prompting suitable patient version-ing) would go to this new declared value. As thevalue-e!ect-pattern linkages become collaborativelyreflected upon, such questions will be discussed andnegotiated among the medical expert and the



management of the clinic. In parallel prototypingsessions, the supportive, underlying interaction pat-terns can be iteratively implemented, tested and, inturn, feed the ongoing conceptual investigation.

As of this writing, the pattern mapping exercise hasinvolved single representatives of most stakeholders(technology and design colleagues in the research team,surgeons, surgeons with management experience andsystem vendors and commercialisation facilitators).Surgeons have participated directly, through validatinguser stories and strategic patterns during workshops,and indirectly, through their use of new or revisedconcepts in workshops and interviews. In view of thegrowing popularity of the Scrum development method,we can summarise our ambition for the third year ofthe project in terms of how design pattern maps mayhelp bridge some of the all too common language gapsbetween systems procurers and Scrum developerteams. While interactive strategy maps may emergeas a channel for dialogue between procurers and Scrummasters responsible for interpreting the procurers’intentions (Eppler and Platts 2009), interactive designpattern maps, derived from and even drawn asextensions of strategy maps, may emerge as a similardialogue channel for Scrum masters as usabilityexperts and the programmers that implement theinteraction design solutions in successively new ver-sions of quality software.

AcknowledgementsThis work was supported by VINNOVA (Swedish Govern-mental Agency for Innovation Systems). We thank all themembers of the Gastro Centre team who have cooperatedand participated in this study.

References

Alexander, C., 1964. Notes on the synthesis of form.Cambridge, MA: Harvard University Press.

Alexander, C., et al., 1977. A pattern language: towns,buildings, construction. New York: Oxford UniversityPress.

Antunes, P. and Costa, C.J., 2003. From genre analysis todesign of meetingware. In: Proceedings of Group’03, 9–12,November 2003. Sanibel Island, FL: ACM Press, 302–310.

Austin, J.L., 1975. How to do things with things with words.In: J.O. Urmson and M. Sbisa, eds. Cambridge, MA:Harvard University Press, 169.

Bakhtin, M.M., 1986. Speech genres and other late essays. In:V.W. McGee, Trans. C. Emerson and M. Holquist, eds.Austin, TX: University of Texas Press.

Borchers, J., 2001. A pattern approach to interaction design.Chichester: John Wiley.

Brynjolfsson, E. and Hitt, L.M., 2000. Beyond computation:information technology, organizational transformationand business performance. Journal of Economic Perspec-tives, 14 (4), 23–48.

Brynjolfsson, E. and Saunders, A., 2010. Wired forinnovation – how information technology is reshapingthe economy. Cambridge, MA: The MIT Press.

Brodner, P., 2005. Software is ogware – a semioticperspective on computer artifacts. In: Y. Sundbladand A. Walldius, eds. Proceedings of the UITQ 2005workshop, 55–62. Available from: ftp://ftp.nada.kth.se/IPLab/TechReports/HCI-15.pdf [Accessed 25 May2010].

Bødker, S., et al., 2000. Co-operative design perspectives on20 years with ‘The Scandinavian IT Design Model’. In:Proceedings of NordiCHI 2000, 1–9 October 2000.Stockholm.

Cabitza, F., Simone, C., and Zorzat, G., 2009. PRODOC: anelectronic patient record to foster process-orientedpractices. In: Proceedings of the 2009 Eleventh EuropeanConference on Computer-Supported Cooperative Work.London: Springer, 85–104.

Dearden, A., et al., 2002. Using pattern languages inparticipatory design. In: T. Binder, J. Gregory, and I.Wagner, eds. PDC 02 Proceedings of the ParticipatoryDesign Conference, 23–25 June 2002.

Ehn, P., 1993. Scandinavian design: on participationand skill. In: D. Schuler and A. Namioka, eds.Participatory design, 41–77. Hillsdale, NJ: LawrenceErlbaum.

Eppler, M.J. and Platts, K.W., 2009. Visual strategizing: thesystematic use of visualization in the strategic-planningprocess. Long Range Planning, 42, 42–74.

Erickson, T., 2000. Lingua Francas for design: sacred placesand pattern languages. In: The Proceedings of DIS 2000,17–19 August 2000. Brooklyn, NY: ACM Press, 357–368.

Erickson, T., 2009. The interaction design patterns page,information about pattern language resources [online].Available from: http://www.visi.com/*snowfall/InteractionPatterns.html [Accessed 15 May 2010].

Falk, T. and Olve, N.-G., 1996. IT som strategisk resursForetagsekonomiska perspektiv och ledningens ansvar.Liber-Hermods AB, Malmo.

Frykholm, O. and Groth, K., 2009. The ‘‘clinical informa-tion workspace’’: Supporting collaborative medicalteam work, position paper presented at theINTERACT 2009 workshop ‘‘Medical team meetings’’,Uppsala 2009.

Frykholm, O., Lantz, A., Groth K. & Walldius, A., 2010.Medicine Meets Engineering in Cooperative Design ofCollaborative Decision-supportive System. In Proceed-ings of CBMS, The 23rd IEEE International Symposiumon Computer-Based Medical Systems. Perth, Australia,October 12–15 2010.

Fussell, S.R. and Benimo!, N.I., 1995. Social and cognitiveprocesses in interpersonal communication: implicationsfor advanced telecommunications technologies. HumanFactors, 37 (2), 228–250.

Gamma, E.R., Johnson, H.R., and Vilissides, J., 1995.Design patterns: elements of reusable object-orientedsoftware. Reading, MA: Addison-Wesley.

Guy, E.S., 2005. ‘‘. . .real, concrete facts about whatworks . . .’’: integrating evaluation and design throughpatterns. In: Proceedings of the International ACMConference on Supporting Group Work. Group ‘05, 6–9November 2005, Florida. New York: ACM Press.

Kaplan, R.S. and Norton, D.P., 1992. The balancedscorecard – measures that drive performance. HarvardBusiness Review, Jan-Feb, 71–79.



Lantz, A., Cerratto Pargman, T., and Walldius, A., 2004.Can the Genre Perspective Help Designers Anticipatehow Videoconference Systems are Actually Used? In:H.M. Khalid, M.G. Helander and A.W. Yeo, eds. Workwith Computing Systems. Damai Sciences, Kuala Lum-pur, 2004, pp. 888–893.

Moll, J. and Sallnas, E-L., 2009. User laser pointing devicesto support collaboration in multi-disciplinary medicalmeetings. Position paper presented at the INTERACT2009 workshop ‘‘Medical team meetings’’, Uppsala 2009.

Olve, N.G., et al., 2003. Making scorecards actionable,balancing strategy and control. Chichester, UK: JohnWiley.

Olwal, A., Frykholm, O., Groth, K. & Moll, J., 2010.Exploring the Potential for Collaborative Interaction inMedical Team Meetings (manuscript submitted toINTERACT 2011).

Schummer, T. and Lukosch, S., 2007. Patterns for computer-mediated interaction. Chichester, UK: John Wiley.

Searle, J.R., 1969. Speech acts: an essay in the philosophyof language. Cambridge, UK: Cambridge UniversityPress.

Swedish Ministry of Health and Social A!airs, 2010.National Strategy for eHealth. Available from: www.sweden.gov.se/ehealth [Accessed 25 May 2010].

Tidwell, J., 2005. Designing interfaces, patterns fore!ective interaction design. O’Reilly 2005. Onlineversion of book accessed June, 2009. Available from:at http://designing interfaces.com/ [Accessed 25 May2010].

Walldius, A., Sundblad, Y., Borning, A., 2005. A FirstAnalysis of the UsersAward Programme from a ValueSensitive Design Perspective, In: O. Bertelsen, N.O.Bouvin, P.G. Krogh and M. Kyng, eds. Proceedings ofthe Fourth Decennial Aarhus Conference August 20–24,2005, pp. 199–202.

Walldius, A., Sundblad, Y., Bengtsson, L., Sandblad, B., and&, Lind, T., 2007. The User Certified 2006 Programmefor Workplace IT Quality Assurance, Proceedings ofWork With Computing Systems, May 2007, KTH,Stockholm.

Walldius, A., et al., 2009. User Certification of workplacesoftware – assessing both artefact and usage. Behaviour &Information Technology, 28(2), 101–120.

Walldius, A., Olve, N.G. and Lind, T., 2010. Intervening forICT usability, position paper presented at the Secondnational workshop om development and use of Enter-prise Resource Planning systems in the private and publicsectors, available at http://www.iei.liu.se/eis/wsafas09/1.132459/ASAF-nils-goran-091016.doc (accessed at2010-05-15).

Yates, J. and Orlikowski, W.J., 1992. Genres of organiza-tional communication: a structurational approach tostudying communication and media. Academy of Man-agement Review, 17, 299–326.

Yoshioka, T. and Herman, G., 2000. Coordinating informa-tion using genres, Massachusetts Institute of Technology,Sloan School of Management, Center for CoordinationScience, CCS WP #214. Available from: http://dspace.mit.edu/bitstream/handle/1721.1/2774/SWP-4127–45322547-CCS-214.pdf?sequence!1 [Accessed 15 May 2010].

Yoshioka, T., et al., 2001. Genre taxonomy: a knowledgerepository of communicative actions. ACM Transactionson Information Systems, 19 (4), 431–456.



Documents

([SORULQJ WKH XVH RI GHVLJQ SDWWHUQ PDSV IRU … · The study reported here is done as a methodolo- ... (Bakhtin 1986) and on speech act theory (Searle 1969, ... Here they use the