Seite - 36 - in Applied Interdisciplinary Theory in Health Informatics - Knowledge Base for Practitioners

Neither of these goals is easy to achieve in health informatics, for a number of reasons. Firstly there are many ways to measure success and not every success measure tells us the same thing. Should we pay more attention to surveys of user satisfaction, rates of adoption, or should we only focus on improvements in clinical outcomes? Further, given the complex organizational space within which any informatics intervention must co-exist, it is not surprising the same system, when tested in different clinical settings, usually achieves different outcomes[1]. The effects of the way a technology is implemented, the specific context in which it is used, and how it is used, all shape the outcomes of system use. With so many confounding factors, getting design right or demonstrating system success is non-trivial. Information value chain theory provides a straightforward approach to both system design and evaluation[1 2]. With its foundations in classical decision science, it provides a mechanism to tease apart the different benefits that might be associated with the use of a given information technology, and also helps identify why expected benefits may not be detectable. Using classical decision theoretic measures such as utilities, and by sequencing the different types of information system functions and their associated outcome measures, value chain analysis helps focus on the reasons for system implementation success or failure. It also assists in making comparative assessments amongst different solutions, to understand which might be best suited for different clinical contexts. 1.1. A value chain extends from system use to health outcome To undertake a value chain analysis, we begin by describing a value chain that connects use of an information system to final outcomes (Figure 1). Figure 1. The information value chain starts with a user interacting with an information system, but must go through many steps before changing clinical outcomes (from Coiera, 2015). The chain typically begins with a user interacting with a system (interaction). Some but not all of these interactions potentially providing information that is then received by a user (information received). Some of this information may lead to a decision being changed (decision changed), sometimes leading to a change in the process of care (care process altered). Finally, some of these process changes may then impact the outcome for a patient (outcome changed). For example, a clinician may interact with an electronic health record, and examine a patient’s laboratory test results. Amongst all the information in the available results, one specific test provides important new information. All the other information received adds nothing new and has no impact on the clinician’s thinking. Based on that specific new information, the clinician ceases a medication and prescribes a new one (a change in care process). In some circumstances this change is beneficial or harmful to the patient – both leading to a change in outcome for the patient. It is also quite possible that this clinical process change, one amongst many that happen daily, leads to no change in the patient’s outcome. E.Coiera /AssessingTechnologySuccessandFailureUsing InformationValueChainTheory36