Leveraging Coordinative Conventions to Promote Collaboration Awareness. The case of Clinical Records

The paper discusses the conventions used by medical practitioners to improve their collaboration mediated by Clinical Records. The case study focuses on the coordinative conventions identified in two wards of an Italian hospital and highlights their role and importance in the definition of the requirements of any system supportive of collaborative work practices. These requirements are expressed in terms of the provision of artifact-mediated information that promotes collaboration awareness. The study identified several kinds of Awareness Promoting Information (API): the paper discusses how they can be conveyed both in the web of documental artifacts constituting a Clinical Record and in its computer-based counterpart, the Electronic Patient Record (EPR). The paper ends with the implications for the design of EPRs and for their integration with Hospital Information Systems in light of the findings. 1 Background and Motivations The paper illustrates the outcomes of a field study conducted in the hospital documental domain in order to uncover how physicians and nurses coordinate each other through their official documentation, the patient-centered clinical record (CR) and to identify supportive functionalities in view of its digitalization. As documented in the literature (e.g., [6, 26, 36, 37, 47]), the hospital domain and its document systems are characterized by a number of aspects that pose significative challenges to the CSCW research: the Corresponding Author Details: Cabitza F. Present Address: Edificio U14, Viale Sarca 336, 20126 Milano, Italy Email: [email protected] Phone:+39 02 6448 7815 2 Federico Cabitza, Carla Simone, Marcello Sarini fragmented and event-driven nature of hospital work; the substantial unpredictability of the illness trajectories managed within recurrent patterns of interventions; the challenging requirements of distributed, mobile and heterogeneous actors such as doctors, nurses and their assistants; along with the multidimensional nature of the CR itself. The clinical record contains at least two partly disjointed sets of documents that are characterized by different colors, stickers and tags: the medical record and the nursing record, where doctors and nurses are supposed to document their interventions and activities. Moreover, the CR reveals a twofold nature: on the one hand, the nature of comprehensive repository for all the information concerning a single patient stay; on the other hand, that of a composite and multiperspective web of documental artifacts. In fact, when the whole clinical record is seen ‘at work’ [29] during the patient’s stay, it can be viewed as a series of heterogeneous and cross-referencing artifacts, which are each very specific for a certain aspect of care and hence are possibly used by different actors at the same time [7]. Because of this twofold nature, the CR is a typical example of “coordinative artifacts” [52]; this is a term that emphasizes the close relationship between artifacts and coordinative practices in the settings where inscribed artifacts are needed to support actors in managing task interdependencies that are too complex to be articulated by ad-hoc interaction and improvisation based on mutual awareness. In these settings, coordinative artifacts can be used by competent actors according to three general use modalities: as (a) templates, when they specify the properties of the result of cooperative work; (b) maps when they specify interdependencies of tasks or resources in a cooperative arrangement; or as (c) scripts when they specify a strict protocol of task articulation and resource interaction [52]. Several examples can be given for each of these categories, drawn from several studies (e.g., [34,38,53,57]): forms, drawings, manuals, lists, spreadsheets, even MS Word documents. Differently from scripts, which offer a limited selection of valid, legal, efficient or otherwise prescribed ‘moves’ [49], both templates and maps support coordination for their ability to let competent actors take heed of what was – or is – going on in their cooperative arrangement and to put this work context in relationship with their goals and expectations, as well as with those of their colleagues. Our observations showed that the documents which make up the CR play both the roles of templates and maps: in fact, they support articulation work only in close connection with what practitioners find and recognize in them beyond the CR’s primary and most apparent function to represent traces of past work (e.g., in the notes and matrixes of the record) and flexible plans for future work (e.g., in process maps, checklists, and excerpts of clinical guidelines). In other words, the coordinative capability of these documents is fully exploited only insofar as practitioners are able to make sense of the subtle ways these artifacts afford meaningful cues and hints on how to articulate a given activity with respect to the other on-going activities. Moreover, we observed that these ways have a predominantly conventional Title Suppressed Due to Excessive Length 3 nature, i.e., they depend on a progressive – and yet selective – stratification of local agreements on what a particular inscription – or a particular way to refer to an inscription – means with respect to work activities. For these reasons, the main outcomes of our field study is twofold: on the one hand, the identification of conventions acting on and throughout the clinical record; and on the other hand, how these conventions can contribute to making the functionalities of the expected digitized CR more acceptable and profitable. After the definition of the scope of our study, we articulate the coordinative role of conventions and describe the main ones identified from the field. Then, we show how these conventions are useful to discuss with practitioners on what information they should be aware of and how technology can promote this awareness. We will refer to the literature that we used and shared with our interlocutors to substantiate our discussions and mutual understanding. The paper ends with some implications on design and how our work will proceed in the future. 2 The Scope of Our Field study We conducted our study in two hospital wards with different specialties and critical needs (an Internal Medicine ward and a NICU Neonatal Intensive Care Unit) at the Manzoni Hospital of Lecco, a large teaching hospital in Northern Italy. In these settings, we observed and studied the situated practices of making sense of CR through which practitioners articulate their actions across wards and shifts in different clinical cases. From the methodological point of view, we followed a “quick and dirty” approach [39]: we made observations in the wards as unobtrusively as possible, and made informal and semi-structured interviews with key practitioners to discuss the results of our observations and to collaboratively identify problematic situations and technological means that could play a role in alleviating the identified problems. Finally, we mocked-up these supportive means using an original computational framework [18], and we used the mock-ups as a basis for further discussions about the optimal artifact mediated functionalities for effective collaboration. In agreement with the hospital practitioners, we focused on a set of records of the CR called Single Sheets (SS). Before entering into their specific characteristics, it is important to mention that in light of the expected digitization, the observed departments had previously undertaken the redesign of several components of the CR with the direct involvement of all practitioners [15]. This situation is perhaps unique in the landscape of hospital work analysis: it created a very constructive interaction with representatives of the practitioners, since most of the conflicting points of view had already been included in the current CR. Moreover, these representatives added a point of view that goes beyond their individual perspective since they were very committed to this cooperative elaboration. Single sheets are denoted as “single” since they integrate in one single sheet sections which, for their own function, could be parts of either the 4 Federico Cabitza, Carla Simone, Marcello Sarini medical or nursing record. Single sheets are used by physicians to order drugs, prescribe treatments or referrals and establish particular therapies: in short, they are artifacts that “mediate” the process of ordering and prescribing drugs to inpatients, i.e., the so called Physician Order Entry (POE). The POE is one of the most crucial document-mediated coordinative moments in hospital work. In the POE doctors give nurses orders about either diagnostic or therapeutic interventions; while nurses give doctors clinical accounts on which doctors can take appropriate clinical decisions with a rigidly differentiated assignment of concerns and responsibilities. The artifacts used in the POE then mediate two kinds of coordinative behaviors: a more prescriptive one, in which doctors commit and delegate nurses to accomplish an intervention on the patient, and nurses make themselves accountable for the intervention to be executed as doctors expect; and a more descriptive one, where nurses give doctors feedback on the completion of the related task and corresponding clinical data, thus enabling further activities. 3 The Nature of Conventions in Cooperative Work Since the conventional nature of hospital work emerged soon in our investigation, we were forced to clarify the very nature of the common sense term ’convention’ with our interlocutors and among ourselves in order to avoid the risk of a misleading communication. In our investigation, we use the term coordinative conventions (CCs) to combine the common-sense meaning of ‘shared agreement and related practice that is either established or consolidated by usage’ with the emphasis on the modalities by which practitioners articulate their activities in their mutual cooperative effort. Specifically, we referred to artifact-mediated coordinative conventions, i.e. conventions that regard how and when documental artifacts are used to either articulate or document work activities. Moreover, we agreed to associate to the term coordinative convention the fact that they are usually formed in an ad-hoc manner with respect to the domain and work arrangement. These conventions are fairly flexible agreements that actors share on ‘what should be done if a certain condition recorded in a document occurs’ (i.e., actions), or on ‘what a certain condition related to any document means from the coordination point of view’ (i.e., interpretations). The term convention has already been used in relation to cooperative work. For example, Lewis [42] defines conventions as solutions to a recurrent coordination problem. They are not habits, but something ranging from rather prescriptive rules (at least in terms of social approval and desiderability) to informal codes of appropriate conduct [44]. Clark also reaches this conclusion, when he points out that conventions can solve coordinative problems since actors can “mutually expect each other to do [something] based on the regularity in their recent behavior” [22]. As an example of convention from the clinical work, during surgery, just indicating a certain element in the environment can make it clear to all practitioners involved what to do next, much alike a specific mark at the border of the page. Title Suppressed Due to Excessive Length 5 During the study we made explicit that CCs are regularities in the behavior, which actors of a cooperative arrangement prefer to conform to, relying on the fact that also others do, so that mutual coordination and comprehension is facilitated. The expression “prefer to” hints at two important aspects of CCs: on the one hand, conformance to CCs is a voluntary act, not imposed by an organizational entity (either role or unit) acting as a superior authority. Even when conventions are established intentionally and do not simply emerge from habitual practice, actors follow them since they want or need to, not because some organizational entity has forced them to. On the other hand, conventions are conformed to since they are worth complying with, even irrespective of the number of actors that have agreed upon them. Differently from Lewis, we relax the requirement that “everyone or almost everyone” has to conform to a behavior to make it a convention: the conventional use of documents is any meaningful habit that has been established between actors, even between two single ones. Reciprocity is hence the condition ‘sine qua non’ by which conventions can be applied, since they are built upon and are part and parcel of the common ground that is essential for any ensemble of actors to cooperate and even communicate with [43, 55]. This common ground is by nature cumulative and is developed as actors share experiences and solve coordination problems while on the job. 4 Coordinative Conventions in hospital work In the study of the practices around the use of the Single Sheets we drew evidence that the conventional ways by which practitioners make sense of their contents and become aware of conditions and events that are useful to coordinate with each other are local, temporary, open-ended and partial : they are neither self-explicating nor fully-specified, but rather become meaningful only in the precise situations that “complete” and activate them. Moreover, they slowly change according to what actors accumulate as experience by acting in conventional ways. Driven by the interests of our interlocutors, we focused on coordinative conventions that mainly regard either reading or writing data in the clinical record. Our study complements those studies considering coordinative conventions that regard the physical and spatial use of these artifacts, i.e., those conventions that regard the coordinative implications of how documental artifacts look and where they are located. The hospital domain gives several examples of these kinds of conventions by which practitioners can infer either the role, status and location of colleagues from simple cues scattered in the working environment: e.g., in [32] the authors describe the multiple ways practitioners draw inferences about patients and tasks based on the external appearance of records alone; in [5], it is reported how the specific position of a folder within the arrangement provides clinicians with important cues on the clinical status of the patient, the level of priority of the case, 6 Federico Cabitza, Carla Simone, Marcello Sarini and the progression through the different clinical activities; likewise, in [8], a case where medical records are made clearly visible at the edge of desks in order to make doctors aware that new patients have arrived is reported as an example of artifact-mediated social awareness. While this specialist literature focuses on the coordinative nature of peculiar arrangements, their conventional nature is usually left in the background, notwithstanding the fact that these arrangements work well in virtue of their ad-hocness, locality and informality. In what follows, some sketchy vignettes describe the role of coordinative conventions when clinicians use the Single Sheets in their daily work at the Manzoni Hospital of Lecco: they are drawn from our field study as the most significative and most extensively treated in the interviews with clinicians for their strong impact on their daily work practices (see Tables in Figure 1 for a tabular summary). Conventions on proper timing Documental artifacts are used to convey meaning besides what practitioners write on them according to their boilerplate and structure. For instance, when a doctor requires a laboratory test she is supposed to indicate whether the examination is urgent or the blood sample can be taken and sent to the laboratory with all the other routine examinations. Since the indication ‘routine’ conventionally refers to the next day early in the morning, for routine examinations the physician is usually exempted from recording the precise time and even the date of the request. Conversely, for requests marked as ‘urgent’ this indication is necessary because only in this way nurses can correctly prioritize due tasks and realize whether they must quickly take the blood sample. In addition, the locality of the conventional nature of urgency was made clear in both the observed Internal Medicine ward and NICU: at the former ward, checking the ‘urgent’ box on the single sheet for a request, means “please, send me back the lab results in half an hour”, while at the NICU, “urgent” means “right now” with no exception, due to the typical critical conditions of the premature newborns. Right timing on order completion is therefore a clear example in which unwritten CCs are at work, specifically on the notion of urgency that is taken for granted in a given setting with all the coordinative consequences of deeming something urgent: for instance, consider the CC by which nurses make sense of the time elapsed from a request, in order to understand whether they are late or not about an order. Or the CC by which nurses are supposed to explicitly notify doctors that lab reports have just been sent back from the lab and are ready to be reviewed (as in the case when they are urgent) instead of letting doctors look the reports up in the clinical record on their own. The point on the proper timing CC is not whether ward practitioners need to be supported in realizing what an urgent order means every time, but rather it is how a (digital) documental system could remind them of urgent orders timely. Conventions on proper redundancy In a previous analysis of cooperative work in the Internal Medicine ward [15], we pointed out the manifold Title Suppressed Due to Excessive Length 7 ways the phenomenon of data redundancy occurs in the daily documental work of nurses and doctors. We denoted with the expressions redundancy by duplicated and replicated data those cases in which the same data are reported either in two or more documents of the clinical record or in different points of the same artifact, respectively. Also at NICU, redundancy can at times play an important role in supporting both coordination among practitioners and their decision making. For instance, it is only on a conventional basis that members of a specific NICU team want to have data on the weight, age and height of newborns reported in every single sheet of drug prescription: in fact, they want them only when a newborn is in life-threatening conditions, i.e., when it could be useful to have these data immediately available in the current sheet to calculate precise drug or nutrient dosage on the fly. Conversely, the fixed and good-for-the-whole hospital organizational rule on data replication that is irrespective of patients’ condition would likely neglect this local and team-based conventional requirement. In doing so, it would also expose practitioners to the risk of both being provided with irrelevant and overloading information and losing the unobtrusive reminder on critical conditions that the presence or absence of this data could play at the very point of order entering. Conventions on proper compilation of records A similar case regards the infusional therapy sheet and the conventions we observed pertaining to whether a compiled sheet is considered complete/accurate or not within some practitioners’ community. At the NICU, nurses are conventionally used to not reporting liquid intake values -or to reporting them only by a rough estimatewhenever these values are within normal range; this is done for two main reasons. They adopt this convention for the sake of conciseness and to convey an implicit reminder that “all is well” to the colleagues of the next workshifts. We then observed how traditional dimensions of data quality like accuracy, completeness and timeliness [10], which are usually taken as intrinsic to a document or data set, assume a more conventional and context-dependent nature in a highly dynamic and frantic domain as clinical work is. We also observed that actors perceive how well work is documented depending on local conventions, which determine what fields are really mandatory or what could be the most convenient order of their compilation on the basis of the current workload and kind of work (e.g., whether critical or stable patients). This is also a case in which CCs and the business logic of a Hospital Information System (HIS) could be discordant with each other: the risk is that administrative managers and biostatistical researchers could have designers embed their own quality requirements (e.g., for accurate and complete clinical data) irreversibly into the EPR forms and workflow in terms of corresponding constraints that straightjacket the coordinative and informational needs of clinicians at the point of care [17]. 8 Federico Cabitza, Carla Simone, Marcello Sarini Conventions on content When practitioners document and make their daily work accountable, they jot down in the CR contents that can be produced and consumed in light of conventions that affect the very meaning they convey. For instance, we observed in almost any ward that the long and continuous working together of the staff generates a very complex but still yet unofficial jargon by which medical terms and habitual examinations and treatments are abbreviated in shorthand. As the novices and frequent job-hoppers that we interviewed confirmed to us, besides very ordinary ways to shorten medical expressions that are common to a certain discipline or scientific community, also other much less common naming conventions are employed, especially in spoken language. For example, in the very same hospital, practitioners referred to their ward as either ‘reparto’ or ‘divisione’, or with abbreviations such as U.O. (for Unità Operativa) or S.C. (for Struttura Complessa) according to their length of service: corresponding “ward-wide” conventions became then consolidated according to the average age of ward staffs. These and similar conventions, once introduced within a certain group of practitioners even by chance, can become more and more consolidated over time, either by sheer habit or even for the often implicit intention of fencing off outsiders or ward patrons for whom it is better that they can not catch everything said in the ward (e.g., patients or their relatives). Drop-down menus that are employed in EPR pages and forms usually disregard these local abbreviating conventions, or even worse, tend to impose their own “standard” acronyms: that notwithstanding, doctors usually fill in free-text fields with ward-wide abbreviations that make sense only on a conventional basis. Forgetting these conventions in the process of the EPR design could seriously undermine the effectiveness of any computer-based support for the mutual articulation of ward activities. Conventions on record-based practices Other times, naming conventions come from the clash between precise marketing strategies of pharmaceutical companies and regional-wide or hospital-specific drug supplying policies: practitioners make sense of what is written on clinical records from these conventions: e.g., when they prescribe name-brand drugs, while, in doing so, they mean any drug with the same active principle; or when viceversa, nurses administer specific branded drugs instead of tantamount others once that doctors have prescribed a generic drug. The point here is that doctors and nurses cooperate about pharmaceutical treatment more on the basis of wardor even doctor-specific conventions, rather than on what it is actually written on the single sheets. Again, forgetting these ordering conventions undermines the effectiveness of automatic drug dispensers [4] and can hinder their actual inclusion in clinical practice. We also observed a set of even more articulated conventions that –consolidating across, rather than within single wards– “regulates” how nurses should prepare patients for certain treatments or tests, especially when the latter are accomplished in an exterTitle Suppressed Due to Excessive Length 9 nal facility or another ward [16]. EPRs and request forms are usually intended to mediate the booking of a time slot at the external facility and they limit themselves to supporting just the “scheduling” dimension of articulation work between multiple wards: instead, the pragmatic dimension of articulation, i.e., handing over patients so that their care trajectories result in no seams or discomforts, is left to the ad-hoc externalization and combination of CCs across different communities of practice. The fact that a patient must fast a predefined number of hours before undertaking a test, or that she must be provided with either a local or systemic sedative; and even how and to which extent she should be informed about the sequence of treatments she will undergo, is a matter of more or less externalized conventions between nurses of the referring and of the accepting wards. We discussed with practitioners how it can be frustrating and unrealistic to embed these conventions into any global logic. Global organizational rules usually do not take into account practitioners’ idiosyncrasies, particular testing modalities and other local practices. Conventions on underspecification The above mentioned conventions not only involve record keeping practices but also how clinicians make sense of what is written on these records. For instance, blank fields in the liquid intake sheet are interpreted as either ‘good news’ or ‘missing values’ according to the general condition of the newborn. In this case the intended meaning of a blank structure for whoever conceived and designed the form (i.e., ‘no value here’) differs from the context-situated meaning that that same structure acquires for whoever fills in it during her work (i.e., ‘no significant value to report’). These misalignments are inevitable but they are not bad per se. Traditional paper-based artifacts, with their silent but still present structure allow for these minor misalignments to be integrated into the usual daily work. This is because the constraints associated with structure and its intended function are left purposely underspecified, and hence they do not have the capability to overwrite the doers’ actions. The opposite would happen in the case of an EPR page that forced practitioners to fill in its form completely in order to give them access to other forms to use in the next activities. Another example of conventions thriving between the folds of underspecification concerns another sheet from the CR, the ‘First Care Planning’ (FCP) sheet. At the Lecco hospital the physician admitting a new inpatient is supposed to formulate at least three, but not more than five, diagnostic hypotheses since this is considered to be the optimal ‘medical framing’ of a given symptomatology irrespectively of its clarity or uncertainty. The hypotheses must be filled in in order of reliability and likelihood and must all be pertinent to the signs collected in the preliminary objective examination. Correspondingly, the FCP sheet has five blank rows and no space for further comments, which the physician should report on a specific section of either the Problem List or the Clinical Notes sheets, if necessary. Notwithstanding these hospital10 Federico Cabitza, Carla Simone, Marcello Sarini Figure 1 Synoptic table summarizing the vignettes of the field case. wide rules, the convention among admitting physicians is to only report the hypothesis of the most probable diagnosis, possibly adding related comments if other hypothesis should be considered. We conclude this section by emphasizing that CCs like the ones illustrated above are seldom supported by document and record systems in a native way: rather, coordinative conventions involved in paper-based practices usually only survive in the grey zones of systems that have the potential to almost completely thwart and block them (see Figure 2). This fact is mainly due to the little attention to conventional practices that analysts and designers pay when they are committed to the elicitation of functional requirements, since their attention is more directed toward activities and operations that generate value from a management perspective. Our point is that full-fledged functional requirements can not come (nor should they come) from the externalization of coordinative conventions. The latter are local, informal and can pertain to extemporaneous groups of actors that develop across institutional communities of users: in our case, these groups encompassed doctors and nurses, clinical and administrative practitioners, and nurses of different facilities. For this reason, after discussing with practitioners about how CCs should be supported in their EPR, we agreed that freezing these conventions in ad-hoc and narrow-scoped functionalities would have resulted in giving these informal habits more “visibility” and formality than necessary; and that this could also backfire on the effectiveness of their practices. Therefore, we agreed that CCs should inform mechanisms which are able to evoke these conventions without constraining or imposing them within their daily practice. This agreement stimulated the active participation of practitioners in the further elaboration that we describe in the next sections. Title Suppressed Due to Excessive Length 11 Figure 2 Shortcomings of EPRs related to the case vignettes. 5 Local Conventions as Basis for Awareness Promotion From the cases illustrated above it is clear that any (computer-based) support for cooperative practices should recognize the value of coordinative conventions and preserve them by avoiding undue constraints on work practices. This conservative view of coordinative conventions, though not trivial, neglects the possibility of leveraging conventions to provide practitioners with a support which is more aligned with how they really work. The fact that a) conventions are more informal than the rules and policies of an organization; b) their scope is more local, i.e. specific to small groups of users rather than to the whole hospital; and c) their life-cycle is bound to the contingent and transitory needs of communities that are much more fuzzydelimited than institutional units and departments of the organization; all these facts suggest that conventions can not be the basis of functionalities that prescribe well-cut flows of work, establish preconstituted orders in the work arrangement or mediate formal communication. Rather, conventions can be leveraged to indicate opportunities to practitioners for informal but correct interpretations, suggest for them behaviors which could meet the local and informal expectations of their colleagues and help them make sense of highly context dependent situations. The idea of considering the convention-based provision of suitable information and indications as a way of supporting cooperative work was first seminally introduced by Mark [43]. Mark suggested conceiving the requirement of establishing and maintaining appropriate conventions within a distributed group of cooperating actors akin to the requirement of building and maintaining “an active learning device” aimed at improving what in the CSCW literature is called collaboration awareness. This regards the capability of the actors involved in a collaborative setting of being constantly aware of the conditions that are relevant for performing their work and coordinating each other. We were influenced by Mark’s suggestion of relating conventions and collaboration awareness with some differences. First, while Mark’s investigations focused on loosely distributed groups, our field studies focused on relatively closed communities of practitioners that share strong common interest (the care of patients), hold and develop a common ground of knowledge and experience (the clinical cases), cooperate by means of offi12 Federico Cabitza, Carla Simone, Marcello Sarini cial inscriptions on a web of documental artifacts (records and reports), and develop and maintain local conventions and agreements for making sense of these inscriptions. Second, during our study we progressively interpreted such a learning device as a proactive way of improving coordination and promoting collaboration awareness by making actors aware of contextual conditions only when they are relevant to a specific convention. To support memory, judgment, communication and coordination clinicians agreed that records and reports should be augmented by functionalities that go beyond mere annotating capabilities [12]: e.g., adding informal notes, extemporaneous links, marks, colored strips and the like to the “data layer” [13] of records and official documents. In addition, the computer-based system should provide practitioners with contextual indications conveyed as graphical clues through the artifacts used in a seamless and unobtrusive way with respect to clinical practices of data accumulation and task coordination [11]. The contextual indications would be meaningful, and hence pragmatically useful, only in virtue of specific conventions that the clues evoke, leaving to practitioners the task of recognizing the relevance of the conventions evoked and exploiting them if needed. The study was then oriented towards the identification of suitable information to promote awareness – called Awareness-Promoting Information (API) – according to the precise requirements that clinicians expressed for a suitable support of their convention-based articulation work. Novices advocated API provision as a support for their ‘practice learning’ and inclusion in the ward habits. On the contrary, seniors and experts appreciated the possibility to be reminded about relevant information when hectic action and frequent interruptions could hamper the proper compliance to conventions, local agreements and evidence-based recommendations. Within the CSCW community, recent surveys have ended up by listing and describing up to nineteen different types of awareness information (e.g., [33,40]). Generalizing the situated phenomenon of collaboration awareness can be useful to detect common features and recurrent patterns of provision of this kind of information, and hence to extract similar requirements for a supportive technology. Nevertheless, one should never overlook the domain specificity of awareness promotion [51]. Much of what an actor needs to know about others and of the work context heavily depends on the application domain and on the means actors use to get this information: in our case, through the direct reading and writing interactions with documental artifacts and records, like SSs. 1 In a private communication with one of the authors, K. Schmidt pointed to the misleading nature of this expression, since awareness relates more to a state of mind, rather than to the provision of data in a specific context and for a specific