Seite - 41 - in Applied Interdisciplinary Theory in Health Informatics - Knowledge Base for Practitioners

and process improvement goals are reserved for other functions such as CPOE. We can thus imagine different systems having quite different profiles for their expected utility at different stages of the information value chain. In Figure 3 hypothetical utility profiles are presented for four different classes of informatics intervention. They illustrate that an intervention: 1. May be designed to provide value by improving the quality of interactions in a health service but may provide little additional information compared to current practice (teleconsultation); 2. May optimize the quality of information capture (EHR); 3. May be designed to improve the quality and efficiency of clinical processes (care pathways) or 4. May be intended to intervene in the decision-making process to improve clinical outcomes. Some downstream benefits may even incur an upstream cost (e.g. interacting with some EHRs requires more time than normal practice). The actual benefits for these intervention classes may be very different, depending on the specific bundle of services offered. For example, it is likely a system that bundles together EHR and decision support will have higher utility than each system alone. Figure 3: The profile of expected utility for an intervention will vary across the steps of the information value chain, depending on the primary purpose of the system (from Coiera, 2015). 2. Use of information value chain theory in health informatics Given its relative simplicity, and its foundation in standard decision theoretic concepts such as utility and value of information, value chain analysis has broad application in healthcare. In particular, it can be used to assess the specific benefits of a given technology, or make comparative assessments between competing technologies. Such evaluations might happen post-hoc, for example trying to explain why outcomes for a particular technology implementation did not meet expectations. They can also be used much earlier on, in system design, when the likely impact of different technology bundles is compared and decisions made about system design. E.Coiera /AssessingTechnologySuccessandFailureUsing InformationValueChainTheory 41