PLoS ONE (Jan 2022)

Characteristics and predictive factors of severe or fatal suicide outcome in patients hospitalized due to deliberate self-poisoning.

  • Stefanie Geith,
  • Maja Lumpe,
  • Johannes Schurr,
  • Christian Rabe,
  • Armin Ott,
  • Tobias Zellner,
  • Michael Rentrop,
  • Florian Eyer

DOI
https://doi.org/10.1371/journal.pone.0276000
Journal volume & issue
Vol. 17, no. 11
p. e0276000

Abstract

Read online

Prediction of clinical course of intoxication is essential for timely initiation of appropriate medical treatment in patients hospitalized due to suicidal self-poisoning. In this retrospective single-centre study in patients hospitalized due to suicidal poisoning in a specialized clinical toxicology unit, we aimed to identify predictive factors associated with severe or fatal course of self-poisoning. All patients underwent at least one psychiatric exploration during their inpatient stay. Severity of poisoning was assessed on admission and after 24 hours according to the Poison Severity Score index (PSS). Spearman's rank correlation coefficient was used to test the association of PSS with sociodemographic, anamnestic and (pre-)clinical parameters. Multivariable binomial logistic regression analysis was performed to determine predictive factors for severe and/or fatal self-poisoning. 1090 patients were included in the study. Median age was 39 years (range 13-91), 66.7% of patients were female. PSS was classified in the majority as "minor" (n = 558, 51.2%) or "moderate" (n = 264, 24.2%). 61 patients (5.6%) had PSS "severe"; 14 patients (1.3%) died. A higher severity of poisoning positively correlated with duration of inpatient therapy (p<0.001, Spearman's rho = 0.454) and duration of ventilation (p<0.001, rho = 0.474), and it inversely correlated with initial Glasgow Coma Scale (GCS) score (p<0.001, rho = -0.437). Multivariable analysis identified no alcohol co-ingestion (OR 3.23; 95%CI 1.3, 8.07; p = 0.012) and self-poisoning with non-medicinal substances (OR 5.4; 95%CI 1.78, 16.34; p = 0.003) as factors predictive for "severe" or "fatal" suicide outcome. In contrast, female gender (OR 0.4; 95%CI 0.2, 0.81; p = 0.011), not using an antidepressant as the method for self-poisoning (OR 0.27; 95%CI 0.12, 0.59; p = 0.001) and a higher initial GCS score (OR 0.79; 95%CI 0.73, 0.85; p<0.001) reduced the risk of a severe or fatal course of self-poisoning. The conclusion for clinical practice is that male patients hospitalized due to self-poisoning, with a low initial GCS score, who did not co-ingest alcohol, attempted suicide with non-pharmaceutical substances or antidepressants are at a higher risk of severe/fatal outcome of suicide. Determination of these risk factors at admission could be potentially used to guide treatment intensification in patients hospitalized due to deliberate self-poisoning.