Systematic Design of Meaningful Presentations of Medical Test Data for Patients

点击翻译按钮获取中文摘要

Both patient-centered care approaches and health information technology advances (e.g. patient portals to electronic health records) are increasing how often patients are directly presented with medical test results that identify health concerns, monitor health status, or predict future health risk. In principle, such data enable patients to actively mange health conditions and participate in care decisions. In practice, availability of data may not result in understanding, as test results are often presented in confusing formats with little context. Many patients, especially those with lower numeracy skills (i.e., poor ability to draw meaning from numbers), may be unable to interpret test outcome data and use it in decision making. For these patients, "knowing" test results or risk estimates does not ensure that they understand what those numbers imply or what actions they need to consider. Such data can be, quite literally, "meaning-less," and patients are likely ignore such information in decision making even when they are fully "informed." We propose to draw on research methodologies from design science, decision psychology, human- computer interaction, and health communication and integrate them into a single, highly innovative research process that will tackle the problem of how best to present Hemoglobin A1c values and similar test results to patients with diabetes as an exemplar of the larger problem of meaningless medical test data. We will (a) define the problem space from multiple perspectives, (b) clarify what we can hope to achieve when we present diabetic patients with their test results, and (c) and identify possible approaches for improving data meaningfulness. Our iterative research approach involves three phases. In Phase 1, we will use intensive "deep dive" design sessions (a methodology borrowed from design science) with a multidisciplinary team combining experts in health communication and human-computer interaction with both practicing clinicians and expert patients. These sessions will identify discrepancies between diabetic patient needs for test result data and the formats in which such data are provided to patients, identify when low numeracy skills will be a barrier to patient interpretation and use of such data, and brainstorm potential solution concepts. In Phase 2, we will conduct rigorous comparative evaluations of proposed designs using (a) user-experience design sessions, and (b) an iterative sequence of large-sample, multi-factorial, randomized-controlled experiments in order to identify what formats make Hemoglobin A1c values and other types of test data maximally meaningful and useful for facilitating informed patient decisions about medical care. In Phase 3, we will take our identified test results communication "best practices" and develop, program, and disseminate a test results "display generator" application that will be able to be integrated with existing electronic health record systems and other applications and will be made available to patients via a freely available website.