3017 — Understanding the Language Used by Clinicians in Describing Patients with PTSD
Tuepker A, Portland VA Medical Center; Zickmund SL, VA Pittsburgh Healthcare System, CHERP, and University of Pittsburgh; Nikolajski C, CHERP; Post LA, Yale University; Hahm BM, James A. Haley Veterans Hospital, HSR&D/RR&D Center of Excellence, Tampa, FL; Butler J, Consortium for Healthcare Informatics Research, CHIR, VA Salt Lake City Healthcare System; Weir C, CHIR; Hickam DH, Portland VA Medical Center, HSR&D;
The VA’s electronic medical record (EMR) is a potentially valuable source of research data, but its capacity to capture accurate information depends on the practices of its users. The purpose of this research, conducted as part of the Consortium for Health Informatics Research (CHIR), was to understand the thought processes that guide clinicians’ word and topic choices when writing EMR notes for their patients with post-traumatic stress disorder (PTSD).
We conducted semi-structured interviews with 28 VA psychiatrists, psychologists, and social workers at five geographically dispersed VA facilities. Respondents reviewed 1-2 of their own progress notes for patients with PTSD. The interview probed about the actual language used to indicate changes in patient status (such as symptoms), as well as thematic questions about what information was considered important or difficult to include. The interviews were transcribed and coded by two qualitative researchers using a combination of directed and inductive techniques, with a subset of interviews double-coded to enhance reliability.
Clinicians generally considered change in patient status, including the effects of treatment, as the most important information in the note. The language used to record change in status was influenced by provider type, by treatment pursued, and by VA facility or program practices. Symptoms and behaviors of interest to clinicians included, but went well beyond the DSM criteria for PTSD diagnosis. Much of the language that clinicians identified as most important related to functional status or engagement with the treatment itself. Clinicians often chose to describe status changes in non-medical language or the patients’ own words. Discussion of sensitive topics (non-combat trauma, sexual orientation) was frequently consciously omitted from the note.
Clinicians stressed the importance of capturing clinically relevant changes in status. These changes are frequently described using language drawn from the patients’ unique statements and circumstances. Providers also were cautious when dealing with sensitive topics, leading to a systematic absence of data that PTSD researchers may seek in the EMR.
Extracting data from the EMR about PTSD symptoms, status change, and treatment effectiveness will require methods that capture the diverse approaches used by clinicians for recording this information.