The risk factors associated with suicidal ideation among patients with head and neck cancer (HNC) have rarely been investigated. Thus, this study proposed an effective screening model to facilitate the identification of patients with HNC who are at risk of committing suicide.
Retrospective longitudinal study.
We recruited 286 inpatients with HNC who were referred to the psychiatry department for an emotional distress assessment during cancer treatment. Sociodemographic and clinical factors, as well as Distress Thermometer (DT) scoring, were gathered and analyzed using multivariate logistic regression.
Among the patients, 27 (9.4%) were recognized as having suicidal ideation. A logistic regression analysis revealed five significant indicators associated with suicidal ideation, namely depression (odds ratio [OR]: 42.83), dysphoria (OR: 8.06), hypopharyngeal cancer (OR: 4.61), cancer history (OR: 4.53), and average DT scores (OR: 1.84). Based on the receiver operating characteristic and logistic regression, the model can predict suicidal ideation with 93.4% accuracy (area under the curve value).
Our findings provide strong evidence that patients with depression, hypopharyngeal cancer, higher DT scores, and previous cancer history can be quickly screened for suicidal ideation, which may help to prevent suicide events in patients with HNC.