Suicidal ideation in medical inpatients: psychosocial and clinical correlates.

OBJECTIVE: To identify psychosocial and clinical correlates of suicidal ideation in medical inpatients.

METHOD: In a cross-sectional study, all adults consecutively admitted to the medical wards of a University Hospital had their names recorded and were randomized and evaluated during the first week of admission. Suicidal ideation was assessed using Item 9 of Patient Health Questionnaire-9. The Beck Depression Inventory, the Beck Anxiety Inventory, the WHO Subjective well-being scale, the Charlson Comorbidity Index and other numerical rating scales (pain and self-reported physical illness severity) were used. Patients with less than four confidants were considered with poor social support. The Student's t test, Mann-Whitney U test, chi-square test and stepwise logistic regression analysis were used.

RESULTS: Of the 1092 patients who composed the sample, 7.2% reported having suicidal ideation. After adjusting for psychosocial and clinical confounders, prior suicide attempts (OR: 4.41; 95% CI: 2.12-9.15; P<.001), depressive symptoms (OR: 1.11; 95% CI: 1.06-1.17; P<.001), severe anxiety symptoms (OR: 3.04; 95% CI: 1.47-6.26; P=.003) and poor social support (OR: 2.02; 95% CI:1.03-3.96; P=.04) were independently associated with suicidal ideation.

CONCLUSIONS: Three out of the four correlates of suicidal ideation in medical inpatients are potentially modifiable factors: severe anxiety, depressive symptoms and poor social support. The fourth variable, prior suicide attempts, is not modifiable but should serve as a red flag to suspect and investigate current suicide risk. These findings highlight the importance of suicidal ideation as a proxy for the distress that is incumbent upon physicians to manage if they wish to provide excellent and comprehensive inpatient care.