Objective To assess the sensitivity and specificity of emergency physicians in detecting anxiety and depression in patients requiring admission to the emergency department (ED) observation care unit for complementary investigations/treatment.

Methods 339 consecutive patients admitted to the emergency observation care unit of 14 EDs were interviewed with standardised questionnaires. The characteristics of the patients, EDs and attending ED physicians were collected. Patients' anxiety and depression were identified using the Hospital Anxiety and Depression Scale (HADS), a self-administered questionnaire. ED physicians were blind to the HADS score and were asked to declare whether they perceived anxiety and depression in each patient. The judgement of ED physicians and the HADS score were compared using sensitivity, specificity, positive and negative likelihood ratios.

Results The HADS questionnaire was correctly completed by 310 patients who comprised the study population. HADS detected symptoms of anxiety in 148 patients (47%) and symptoms of depression in 70 patients (23%). ED physicians determined the presence or absence of anxiety with a sensitivity of 48% (95% CI 40% to 56%) and a specificity of 69% (95% CI 61% to 75%). Positive and negative likelihood ratios were 1.54 (95% CI 1.16 to 2.06) and 0.75 (95% CI 1.28 to 3.28) for anxiety. They detected the presence or absence of depression with a sensitivity of 39% (95% CI 28% to 51%) and a specificity of 78% (95% CI 72% to 83%). Positive and negative likelihood ratios were 1.75 (95% CI 1.20 to 2.56) and 0.78 (95% CI 1.26 to 3.87) for depression.

Conclusion Although patients presenting to the ED often experience anxiety and depression, these symptoms are poorly detected by ED physicians.