Original article
Performance of the Hospital Anxiety and Depression Scale as a screening tool for major depressive disorder in cancer patients

https://doi.org/10.1016/j.jpsychores.2007.01.009Get rights and content

Abstract

Objectives

(1) To assess the overall performance of the Hospital Anxiety and Depression Scale (HADS) as a screening instrument to identify cases of major depressive disorder (MDD) in mixed cancer outpatients. (2) To determine the sensitivity and specificity of various HADS cut-off scores. (3) To recommend an optimal HADS cut-off score for use in this population.

Methods

A cross-sectional questionnaire and interview survey of consecutive patients attending outpatient clinics in a regional cancer centre in Edinburgh, Scotland, United Kingdom.

Results

Thirty (8.3%) of 361 patients met criteria for MDD at interview. With the use of the total HADS score to identify these cases of MDD, the area under the receiver operating characteristic curve was 0.94 (95% CI, 0.91–0.97). A cut-off of 14/15 gave a sensitivity of 0.87 (95% CI, 0.70–0.95), a specificity of 0.85 (95% CI, 0.81–0.89), and a positive predictive value of 0.35, and was considered optimal.

Conclusion

The HADS can be used effectively as an initial screening tool for the detection of MDD in outpatients attending mixed cancer clinics.

Introduction

Major depressive disorder (MDD) is common in outpatients attending cancer services [1]. It is associated with increased symptom burden, greater disability, and reduced quality of life [2], as well as reduced compliance with medical treatment and poorer outcome [3].

Despite its importance MDD is frequently undetected and untreated [4]. One way of improving the detection rate is to screen all outpatients for MDD. However, if all patients were given a diagnostic interview this would place an unrealistic burden both on patients and on clinical resources. Instead, a two-stage screening process is frequently used; the first stage is a self-report scale and the second stage a diagnostic interview [5]. Only patients scoring above a specified cut-off on the self-report scale are asked to participate in the diagnostic interview.

The question then arises of which self-report scale to use in the first stage for screening large numbers of outpatients with a variety of cancer types for MDD. The Hospital Anxiety and Depression Scale (HADS) is the most popular. The HADS is a 14-item scale which was originally designed to detect symptoms of depression and anxiety in patients attending medical outpatient clinics [6] and excludes somatic symptoms, which may be attributable to cancer and its treatment. But how good is the HADS for this purpose and which cut-off score should be used to identify patients for interview?

Literature reviews of the performance of the HADS as a screening tool indicate that a variety of cut-off scores have been recommended [7], [8]. Table 1 summarizes the studies, published to date, which have investigated the optimal cut-off score of the HADS to screen cancer patients for psychiatric disorders. Despite the large number of publications, our specific question remains unanswered. There are several reasons for this: Firstly, most studies have been of specific cancer subgroups with only two of mixed cancer outpatients [9], [10]. Secondly, studies have evaluated the HADS's ability to detect a variety of psychiatric diagnoses, but only three [11], [12], [13] have specifically addressed the detection of MDD. Thirdly, most of the studies have been based on small samples, and, consequently, confidence intervals around the reported sensitivity and specificity, although often unpublished, are likely to have been wide. We therefore aimed (1) to assess the overall performance of the HADS when used to identify cases of MDD in mixed cancer outpatients; (2) to determine the sensitivity and specificity of various HADS cut-off scores; and (3) to recommend an optimal HADS cut-off score for use in this population.

Section snippets

Design

This study was a cross-sectional survey with completion of the HADS followed by a diagnostic interview. It was conducted as part of a larger research programme investigating the detection and management of depression in cancer patients.

Sample

A consecutive series of 635 patients attending a range of outpatient clinics (prostate and bladder, testicular, colorectal, gynaecological, and breast cancer clinics) in a regional cancer centre in Edinburgh, Scotland, United Kingdom, who completed routine

Characteristics of the sample

Fig. 1 shows the recruitment process and number of patients at each stage of the study.

Six hundred thirty-five patients attended the selected clinics, of whom 445 (70%) completed the HADS. SCID interviews were carried out with 361 (81%) of these patients. Reasons for failure to conduct an interview included not having a contact telephone number, being unable to contact the patient within 2 weeks, the patient being too ill or admitted to hospital, the patient being too distressed or refusing to

Performance of the HADS in oncology outpatients

This study found that the HADS performed well when used as a screening tool for MDD in mixed cancer outpatients. The areas under the ROC curves were close to 1.00, indicating good overall performance and the scale offered good levels of sensitivity and specificity.

We have focused on the performance of the HADS total score, as this is the most widely used form of the HADS in clinical practice. However, we also tested the subscales as a shorter screening tool would impose less burden on patients.

Conclusion

The HADS is a reasonable choice for the first stage of a two-stage screening programme for MDD in mixed cancer outpatients. We suggest that a total score cut-off of ≥15 offers a good trade between sensitivity and specificity. The HADS is only one of a range of available screening instruments [37], but one of the most widely used. The development of more efficient screening tools is required and may be assisted in the future by using computerized adaptive testing [38]. Shorter tools which

Acknowledgments

The authors would like to thank Dr Ann Smyth and the staff and patients of the Edinburgh Cancer Centre who were involved in this study. This study was partly funded by Cancer Research UK (KP and RR).

References (41)

  • D Razavi et al.

    Screening for psychiatric disorders in a lymphoma out-patient population

    Eur J Cancer

    (1992)
  • MJ Massie

    Prevalence of depression in patients with cancer

    J Natl Cancer Inst Monogr

    (2004)
  • JS McDaniel et al.

    Depression in patients with cancer. Diagnosis, biology, and treatment

    Arch Gen Psychiatry

    (1995)
  • M Sharpe et al.

    Major depression in outpatients attending a regional cancer centre: screening and unmet treatment needs

    Br J Cancer

    (2004)
  • MP Pignone et al.

    Screening for depression in adults: a summary of the evidence for the U.S. Preventive Services Task Force

    Ann Intern Med

    (2002)
  • AS Zigmond et al.

    The Hospital Anxiety and Depression Scale

    Acta Psychiatr Scand

    (1983)
  • RM Berard et al.

    Depressive disorders in an out-patient oncology setting: prevalence, assessment, and management

    Psychooncology

    (1998)
  • A Kugaya et al.

    Screening for psychological distress in Japanese cancer patients

    Jpn J Clin Oncol

    (1998)
  • AW Love et al.

    Diagnostic efficiency of the Hospital Anxiety and Depression Scale in women with early stage breast cancer

    Aust N Z J Psychiatry

    (2002)
  • AW Love et al.

    Screening for depression in women with metastatic breast cancer: a comparison of the Beck Depression Inventory Short Form and the Hospital Anxiety and Depression Scale

    Aust N Z J Psychiatry

    (2004)
  • Cited by (118)

    • Effect of Stepped Care on Sexual Interest and Enjoyment in Distressed Patients with Head and Neck Cancer: A Randomized Controlled Trial

      2021, Sexual Medicine
      Citation Excerpt :

      The HADS is a 14-item psychometrically sound, self-assessment scale for measuring distress (total HADS score) with 2 subscales, anxiety (HADS-A) and depression (HADS-D), developed for non-psychiatric patients. The total HADS score ranges from 0 to 42, the subscales from 0 to 21, where higher scores represent more distress.21,26 HADS at baseline was assessed by telephone or by means of OncoQuest.22,23

    View all citing articles on Scopus
    View full text