Psychometric properties of the Hospital Anxiety and Depression Scale in an inpatient video-monitoring epilepsy cohort

Lara McCartney, Benjamin Johnstone, Terence O'Brien, Patrick Kwan, Tomas Kalincik, Dennis Velakoulis, Charles Malpas

Research output: Contribution to journalArticleResearchpeer-review

4 Citations (Scopus)


Background: People with epilepsy (PWE) have high rates of comorbid anxiety disorders and depressive disorders, from 25% in general population cohorts to rates of 55% in people with treatment resistant epilepsy. High rates are also seen in patients with psychogenic nonepileptic seizures (PNES). Depressive disorders and anxiety disorders in PWE are associated with decreased quality of life measures and are the strongest risk factors for increased suicidality, rates of which are markedly elevated in PWE, at 12%, compared to the general Australian population (1.8%). The Hospital Anxiety and Depression Scale (HADS) is one of the more commonly used screening tools in medical populations. Past studies of the HADS in general outpatient populations with epilepsy have demonstrated promising validity for detecting depression. Objectives: The following were the objectives of the study: 1. To examine the validity of HADS in detecting depressive disorders and anxiety disorders in an inpatient population of patients admitted for video monitoring. 2. To investigate the measurement structure of the HADS across the diagnosis groups of epilepsy subtypes and PNES. Methods: This was a retrospective cohort study of 485 patients admitted to a tertiary epilepsy video electroencephalography (EEG) monitoring unit. All patients received clinical neurological, neuropsychiatric, and neuroimaging assessments to arrive at consensus epilepsy and psychiatric diagnoses. Clinical psychiatric diagnosis of depressive disorders and anxiety disorders, based on the assessment of a neuropsychiatrist, were compared to accepted HADS cutoff scores for these conditions. Findings: Of the 485 patients, 229 patients were with epilepsy, 28 had both epilepsy and PNES, and 121 had PNES. In 107 cases, no definite diagnosis could be made. At a cutoff score of 7 HADS was able to significantly classify patients with depression (area under the curve [AUC] = 0.79, 95% confidence interval [CI] = 0.72–0.82) with a sensitivity of 70% and a specificity of 83%. A similar result was observed for anxiety disorders; a cutoff score of 7 (AUC = 0.75, 95% CI = 0.72–0.81) was able to significantly classify anxiety disorders in patients with a sensitivity of 88% and specificity of 54%. Conclusions: This study has found that HADS measures two separate, yet correlated, constructs of anxiety disorders and depressive disorders. Our results indicate that while the HADS is sensitive to distress in this population, relatively low cutoff scores would be required to achieve highly sensitive screening. This sample includes patients with a diagnosis of epilepsy and/or PNES, and thus, the findings have clinical applicability to screening in tertiary epilepsy video-EEG monitoring units where both these conditions frequently co-occur.

Original languageEnglish
Article number106631
Number of pages8
JournalEpilepsy & Behavior
Publication statusPublished - Feb 2020


  • Epilepsy
  • Prevalence
  • Psychiatric disorder
  • Screening instruments
  • Validity

Cite this