SUICIDE RISKS - NEW WAYS TO IDENTIFY PATIENTS

Words Matter: Clinicians' Notes May Hold Key to Suicide Risk

It may be possible to identify patients at risk for suicide by analyzing clinicians' notes, a novel study finding suggests.

Investigators from White River Junction Veterans Affairs Medical Center, in Vermont, found a link between clinicians' use of distancing language and increased suicide risk.

Individuals who commit suicide often have had recent contact with a healthcare provider, yet suicide remains tough to predict. Linguistic analysis has been applied to diaries, suicide notes, and social media postings of people who later died by suicide, but this is the first study to apply it to clinicians' notes as a way of identifying suicidal patients, the investigators, led by Christine Leonard Westgate, note.

The study was published online October 1 in Psychiatric Services.

Linguistic Signaling

The researchers studied the content of clinicians' notes regarding a random sample of 63 US military veterans who died by suicide during 2009 and who had received outpatient VA healthcare services in the year preceding suicide.

Roughly half had received mental health services. The investigators matched them regarding mental health service use with living VA outpatients. The researchers used linguistics software to construct quantitative theme-based categories related to distancing language and examined temporal trends using keyword analysis.

The analysis showed more frequent use of distancing language in clinicians' notes of patients who later died by suicide.

Emotional distancing can be thought of as a "largely unconscious effort that clinicians use to spare themselves from the emotional trauma associated with the suicide of one of their patients. Frequent use of first-person and second-person pronouns signifies interpersonal closeness, whereas frequent use of the third-person pronouns characterizes interpersonal distance," the investigators write.

In the primary analyses, they confirmed their hypothesis that there would be more third-person pronoun use in the suicide group than in the nonsuicide group. In an exploratory analysis, they confirmed their hypothesis that key words relating to distancing language would emerge as the dates of suicides by users of mental health care services neared.

If replicated in additional studies, the findings could have important implications in the identification of suicide risk, the researchers note. For example, routine linguistic analysis of clinical notes could be shared with clinicians, perhaps via a clinician "dashboard," to help guide psychotherapy practice. And linguistic "signals" could be combined with other demographic information and risk factors to create a more comprehensive risk profile to guide decision making and to help individual clinicians tailor their interpersonal style to their patients' needs, they suggest.

Innovative Prevention Approach

Caitlin Thompson, PhD, national mental health director of suicide prevention and community engagement for the US Department of Veterans Affairs, welcomes this type of research.

"We need to find tools in order to predict suicidal behavior, so it's absolutely essential that articles and work like this be done," she told Medscape Medical News. "We are continuing to think through innovative and novel ideas about predicting suicide and suicidal behavior, and this is just one of them. I really thought this article was very interesting," said Dr Thompson, who was not involved in the research.

"The beauty of some of these qualitative software programs is the ability to really look at language and pull themes. There is a lot of work being done in terms of qualitative looks at social media, Facebook pages, but use of clinicians' notes is novel. It's super interesting, and I look forward to what [the researchers] decide to do next."

The study had no commercial funding. The authors and Dr Thompson have disclosed no relevant financial relationships.

Report By Megan Brooks November 04, 2015.

Psychiatr Serv. Published online October 1, 2015. Abstract

Comments