Benefit of Report Card Feedback After Point-of-Care Assessment of Communication Quality Indicators

Background: Communication in health care is crucial for patient experience and biomedical outcomes, but problems with communication are often seen in health care. Training can improve communication, but skills must be reinforced after graduation to remain improved. Since educational methods are too resource intensive for sustained use throughout the Aurora Health Care system, it is necessary to develop affordable, quantitative methods. The first author has developed necessary techniques, including behavior-specific measures called communication quality indicators. Purpose: To demonstrate secure audio recording in an outpatient visit and to use communication quality indicators with a heterogenous set of patient-clinician conversations. Methods: Thirty primary care physicians were audio-recorded with one or more patients via a secure Internet application running on exam-room computers. Transcripts were abstracted quantitatively using explicit-criteria definitions for two groups of communication quality indicators: assessments of understanding (AU) and jargon explanations (JE). There are four separate behaviors within the AU group: open-ended, close-ended, the highly effective “request for teachback,” and the least effective “OK?” question. Quality indicator data were returned using a previously described report card. After feedback, one or more follow-up recordings were done for comparison. Results: Baseline transcripts included a mean of 15.5 unique jargon words, but words were often used more than once so the mean total jargon count was 25.1. JEs were rare at baseline, with a median of 1 per transcript. The JE ratio (fraction of jargon words that follow a JE for that word) averaged 0.26 out of a best-possible 1.0. AUs were found in 61.1% of transcripts, but most were “OK?” (median 2.13/transcript) or close-ended questions (median 0.52/transcript). After the report card, the median number of JEs improved to 4 per transcript (P \u3c 0.01 by Wilcoxon), and the JE ratio improved to 0.36 (P \u3c 0.01 by matched t-test). AUs improved to 81.3% of transcripts (P \u3c 0.04 by chi-squared). Most of the increase was found in close-ended AUs (median 0.97/transcript by, P \u3c 0.04 by Wilcoxon). Conclusion: This project demonstrated that it is feasible to record at the point of care, abstract transcripts at a central office and improve communication quality via a report card. The small sample size was acceptable for a demonstration project, but a larger, multifaceted program could improve patient experience and biomedical outcomes across Aurora