Seite - 176 - in Applied Interdisciplinary Theory in Health Informatics - Knowledge Base for Practitioners

to sense making and organizational work in which key people may seek to anticipate these outcomes, it is often the case that complex interventions will require additional adaptive work as implementation progresses. For example, a key problem of telemedicine systems historically has involved additional work required in communication and interpretation of complex clinical information, when compared with co-present consultations [7]. What we are interested in with respect to interactional workability, is the work that people have to do with objects (i.e. the physical implements that accompany an intervention, such as a new interface for patient record retrieval), new practices (e.g. a new way of performing diagnostic assessments), and each other to accommodate and adapt to new ways of working. How are confidence in, and accountability for the intervention built? (Relational Integration) Relational integration refers to forms of knowledge work that participants do to build accountability and maintain confidence in a set of practices and the people involved with them. Accountability can here be thought of as processes that give participants access to information (e.g. formal reports, or informal observations) about the outcomes of a given practice. Through such processes, confidence in an intervention and its associated practices and objects can be built and/or undermined. For example, confidence in a new teledermatology intervention was undermined when clinicians began to doubt the integrity of the images transmitted by the system, and began to examine patients in person alongside digitized images (resulting in greatly increased workload and increased pressure on their clinical department) [7]. Clinicians in this case undertook knowledge work that resulted in a loss of confidence in what was being transmitted, indicating not only why confidence was undermined, but how, and thereby identifying a point of failure at which such issues might be addressed (e.g. through development of image verification procedures that help clinicians to build accountability and confidence in the system). Who does what? (Skill set workability) Who should perform a given task? What are the processes for allocating responsibilities as the intervention progresses? Are they formal (for example, allocation by rota, or contractual changes to responsibilities), or informal through voluntary agreements between participants. Implementation of complex interventions often requires adaptation and renegotiation of roles and responsibilities, which can involve trade-offs between resource allocation (i.e. the time that specific people can contribute) and degree of need for specialist knowledge within a given part of the process. For example, a research group investigating the effectiveness of a decision aid for medication choice after a serious illness event had to decide whether the decision aid should be administered by trial managers with no clinical responsibility for the patient, or nurse practitioners actively involved in their care [3,8]. The trade-off here was between those with greater familiarity with and attachment to the intervention, compared with those closer to the field in which the decision aid intervened (i.e. the care pathway of patients recovering from serious illness events). M.BracherandC.R.May / ImplementingandEmbeddingHealth InformaticsSystems176