This article is more than 5 years old. Information may no longer be current.

Discharge notes may help identify suicide risk following hospital release

Multiple clinical features available at hospital discharge identified individuals with significant risk for suicide, suggesting the feasibility of this data for stratifying risk for death by suicide following hospital discharge.

“With 41,149 completed suicides reported by the CDC in 2013, suicide is the 10th leading cause of death in the United States (12.6 cases per 100,000) and the second leading cause among individuals aged 15 to 24 years (10.9 cases per 100,000),” Thomas H. McCoy Jr, MD, of Massachusetts General Hospital, Boston, and colleagues wrote.

Thomas H. McCoy

To determine if including natural language processing of narrative discharge notes improves stratification of risk for suicide death after hospital discharge, researchers evaluated clinical data for 845,417 hospital discharges among 458,053 individuals from two large academic medical centers between 2005 and 2013. Median follow-up was 5.2 years.

All-cause mortality was 19% during the 9-year study period, including 0.1% of the cohort who died by suicide during 2.4 million patient-years of follow-up.

Positive valence in narrative notes was associated with a 30% reduced risk for suicide, according to models adjusted for coded sociodemographic and clinical features (HR = 0.7; 95% CI, 0.58-0.85; P < .001).

“The present study demonstrates the feasibility of characterizing suicide risk based on data available as part of routine clinical care as a possible step toward clinical risk stratification. Even limited to coded data, our prediction substantially improves on chance or on the current standard of no systematic assessment. Furthermore, it illustrates the application of simple machine-learning techniques to extract additional data from clinician notes as a means of capturing more detail than is available in coded data sets and crucially shows that even a coarse measure may substantially improve risk stratification,” the researchers wrote. “While the value of large data sets in health care has undoubtedly been the subject of substantial hyperbole, our results add to a growing body of work indicating the feasibility of leveraging such data sets with standard computational tools to make predictions that may be applied to stratify risk.” – by Amanda Oldt

Disclosure: McCoy reports no relevant financial disclosures. Please see the full study for a list of all authors’ relevant financial disclosures.