One-off screen predicts future symptoms of depression and anxiety in Coronary Heart Disease

J. E. Palacios, M. Khondoker, A. Tylee, M. Hotopf

Research output: Contribution to journalAbstractpeer-review


Background: There is a lack of evidence among longitudinal analysis than can accurately assess just how persistent symptoms of depression and anxiety are in the course of CHD, and whether an initial screen for these symptoms can predict this persistence. Currently, there is controversy surrounding the alleged benefits of screening for depression. We report on a longitudinal, multi-wave analysis of a primary care cohort of CHD patients, with the aim of determining the differences of reporting positive symptoms of depression and anxiety amongst people with a baseline positive and negative screen.

Method: The population used for this study is the cohort conducted by the UPBEAT UK research programme, consisting of 803 patients on CHD registers in general practices throughout South London. Baseline measures using the Hospital Anxiety and Depression Scale (HADS) for depression and anxiety were then repeated at 6, 12, 18, 24, 30, and 36 months. Using a multi-level, mixed effects model, we determined the positive and negative predictive values (PPV and NPV) for baseline screen of depression and anxiety symptoms using a cutoff of 13 on the HADS-total, and also determined the PPV and NPV for two consecutive screens (baseline + 6 months).

Results: Patients who had increased anxiety and depression at baseline (cutoff of 8 on HADS-D, HADS-A) continued to have high scores on the HADS in 61.2 and 59.7% of time points, respectively. Among patients who had low scores at baseline the values were 9.7 and 9.3%. The PPV of a baseline screen for symptoms of distress (HADS-total) was 67.8%, and the NPV was 95.7%. When measuring both baseline and 6-month positive screens for HADS-total, the PPV rose to 86.7%, whilst the NPV dropped to 79.5%.

Conclusion: There is a large difference of the continuing reporting of symptoms between patients with scores above and below the cut-off on the HADS at baseline. Results suggest that a single, one-off screen can be used both to identify and follow-up at-risk patients who screen positive, and to remove from follow-up those patients who screen negative. Two consecutive positive screens further suggest the strong predictive value of this measure. We believe these results show that screening for depression and anxiety symptoms in CHD population is a valid and necessary measure to identify patients at risk for the detrimental outcomes associated with this comorbidity.
Original languageEnglish
Pages (from-to)617
Number of pages1
JournalJournal of Psychosomatic Research
Issue number6
Early online date2 May 2015
Publication statusPublished - Jun 2015

Cite this