Characterizing Patient-Friendly “Micro-Explanations” of Medical Events

Patients’ basic understanding of clinical events has been shown to dramatically improve patient care. We propose that the automatic generation of very short microexplanations, suitable for real-time delivery in clinical settings, can transform patient care by giving patients greater awareness of key events in their electronic medical record. We present results of a survey study indicating that it may be possible to automatically generate such explanations by extracting individual sentences from consumer-facing Web pages. We further inform future work by characterizing physician and non-physician responses to a variety of Web-extracted explanations of medical lab tests.