Page - 43 - in Applied Interdisciplinary Theory in Health Informatics - Knowledge Base for Practitioners

2.2. Case Study 2: Clinical Audit and Feedback Audit and feedback (A&F) interventions have had mixed success in ensuring patients receive improved care [13 14]. Unlike clinical decision support tools, which provide clinicians with patient-specific advice at the point of care, A&F tools provide data about quality indicators at a population level over a period of time. Reasons for their variable effectiveness are unclear because the mechanisms behind intervention success or failure are poorly understood [15]. Value chain analysis can assist in identifying where potential barriers to effective use of A&F reside2. For example, in a situation in which A&F is focussed on improving prescribing, does the type and number of feedback alerts a clinician receives influence the probability that clinicians actually notice them, or subsequently influence their decision making, or which medications are dispensed by pharmacists or finally how many unscheduled hospital admissions are prevented? In a study by Gude et al., the number of events at each stage of the A&F value chain for medication prescription were measured [16]. System designers were faced with a situation in which A&F was not having any perceptible impact on clinical outcomes, and wanted to understand why this was the case. Analysis of the A&F value chain (Figure 5) reveals a major disconnect between events. Firstly there is a steep reduction between the number of indictors demonstrating poor performance, and the number of indicators flagged for action. An even more dramatic reduction occurs between the problems identified by these indicators, and any action to change clinical process. Of 379 indicators targeted for an action, only 31 were addressed. The study noted “feedback did not lead to teams focusing their quality improvement decisions on low performance areas, and that planned improvement actions were often not completed”. Figure 5. The information value chain for a computerized Audit and Feedback (A&F) intervention in cardiac rehabilitation. Clinical teams received feedback multiple times on a set of eighteen quality indicators (adapted from Gude et al. 2016 [16]). Focusing just on the probabilities of events in the value chain, as shown in Figure 5, can tell us where a problem is occurring. The next stage of analysis requires measuring the utility of events at each step, to provide more focused information on why events do or do not occur. In this case study, measuring utility can help identify the source of the problem more precisely. Was the lack of outcome change because the alerts about abnormal indicators (information received), were of low perceived value (perhaps 2 See Chapter 14 “Control Theory to design and evaluate audit and feedback interventions” for an analysis of the same case using Control Theory. E.Coiera /AssessingTechnologySuccessandFailureUsing InformationValueChainTheory 43