JMIR Publications

ABSTRACT

Background:

The very highest risk of suicide is seen after discharge in patients hospitalized with suicidal thoughts and behaviors. Suicidal ideation (SI) is a well-known precursor of suicide attempts (SA) and suicide and is a central indicator of subjective suffering.

Objective:

To examine how the presence, intensity and trajectory of SI in newly discharged patients relate to patient characteristics and symptom severity levels before discharge.

Methods:

Before discharge, patients hospitalized due to high suicide risk were screened for diagnosis (MINI.7.0.2 and SCID-5-PD), general symptom severity (Outcome Questionnaire-45, OQ-45), depression (Patient Health Questionnaire- 9, PHQ-9), suicidal ideation (Suicide Status Form-IV, SSF-IV) and suicidal behaviors (Suicide Attempt Self-Injury Count). Twenty male and 16 female adult participants had a mobile application installed on their smart-phones and reported level of SI on a 1–5-point Likert scale (1= not at all, 5= very much) five times a day for 10 days after discharge. Associations between baseline characteristics and post-discharge SI were analyzed with logistic binary regression, median regression and mixed linear regression.

Results:

SI was present in 78% of participants and in 51% (n= 528) of all delivered surveys (N=1078). Women more frequently reported SI than men (x2= 63.39, P< .001), while participants hospitalized with SA (n=24, 67%) reported SI less frequently than those hospitalized with high suicide risk (n=12, 33%) (x2= 22.64, P< .001). The intensity of SI did not change over the 10-day period (B = 0.29, 95%CI -.01 - .06, P = .09), and varied largely between participants (B = .80, SE .20). Participants with SI after discharge, compared to those without, had higher scores on PHQ-9 (OR 1.21, 95% CI 1.03, 1.41) and SSF-IV (OR 1.28, 95% CI 1.05, 1.57) and evaluated their subjective risk of suicide as higher (OR 17.62, 95% CI 1.85, 168,05) before discharge. Participants hospitalized without a SA had a higher frequency of peak scores of SI after discharge (B=16.00, 95%CI 8.05, 23.95, P< .001). This association was moderated by the presence of Personality disorder (B=18.00, 95%CI 4.85, 32.15, P< .009). Previous suicide attempts (B = .02, 95% CI 0.01, .04, P < .008) and baseline scores on OQ-45 (B = .02, 95% CI.01, .03, P < .006) and SSF-IV (B = .09, 95% CI .03, .14, P < .004) predicted intensity of SI, while a subscale of SSF-IV (“SSF-IV-chronic”) predicted the intensity (B= .20, 95%CI .15, .26, P< .001) and the trajectory (B= .01 95%CI .00, .02, P< .005) of SI after discharge.

Conclusions:

SI after discharge was common, varied substantially between participants, and may be more prevalent than previously reported, particularly in groups with high morbidity and prior SAs. Identified baseline predictors of SI may aid clinicians in tailoring treatment and follow-up services. Clinical Trial: ClinicalTrials.gov ID NCT05955807