Suicidal ideation in patients with cancer: Its prevalence and results of structural equation modelling

Objective Patients with cancer have a higher risk of suicidal ideation (SI) and suicidality than the general population. This study was designed to investigate the prevalence of SI and its association with psychosocial and sociodemographic factors and tumour entity. Methods In this observational cross-sectional study, 4372 adult patients with different cancer entities were enrolled. We assessed the outcome variables (i.e. SI, depressive and anxiety symptoms, mental and physical fatigue and sociodemographic data) using self-report questionnaires. Data were analysed via descriptive statistics, binomial logistic regression and structural equation modelling (SEM). Results Among all patients, 627 (14.3%) reported SI, of whom 12.8% reported SI on several days, 0.9% on half of the days and 0.6% nearly every day. Age, anxiety, mental fatigue and the Patient Health Questionnaire-9 items 'feeling down, depressed and hopeless', 'feeling bad about oneself' and 'slowing or agitation' were significant predictors of SI. SEM, including all significant predictors with a latent depressiveness–demoralisation variable, explained 30.3% variance of SI, showing a good fit. Conclusions Our results showed that a significant number of patients with cancer show SI. Future long-term studies are needed to address the differential contribution of depression and demoralisation on SI in patients with cancer.