The Beck Depression Inventory requires modification in scoring before use in a haemodialysis population in the UK

Abstract

Background: There are high rates of mental illness, especially depression in people with end-stage renal disease (ESRD) on haemodialysis (HD). Depression can be difficult to diagnose, as depression and the medical problem share many symptoms. It is essential that the diagnosis of a depressive illness be accurately made in order that subsequent psychiatric management can occur. This is the first study done in the UK which aims to validate a simple method for detecting depression, the Beck Depression Inventory (BDI), in people on HD against a structured psychiatric interview with a trained psychiatrist using the ICD-10 classification of depression. The BDI is validated in a normal population, but in patients with physical ill health there has been a difficulty in defining appropriate cut-off scores.

Methods: Patients dialysing for over 3 months from a single HD unit were eligible for participation. Patients were excluded if a known psychiatric illness existed or if they were receiving medication for a psychiatric illness. Patients who had a recent major illness requiring hospitalisation were excluded. 57 consenting participants completed the self-reported BDI. Within the next week a psychiatrist carried out a clinical interview to diagnose depression, based on the ICD-10 classification of a depressive disorder. Patients were classified as not being depressed or having mild, moderate or severe depression according to their BDI score or the ICD-10 classification.

Results: The BDI gave more scores for depression and more severe scores than ICD-10. 30 cases had scores which agreed, 27 cases resulted in a higher BDI score than the ICD-10 classification, in no case was the ICD-10 classification more severe (Wilcoxon signed rank test p < 0.001). All 7 diagnosed as being depressed by clinical interview were also depressed by BDI, although in 2 cases the BDI worsened the depression score.

Conclusions: Using the general population cut-off score, the BDI significantly overdiagnosed depression in this HD population. This study defines a BDI cut-off score which more reliably detects major depression in a HD population. Unrecognised depression remains high in those with no known psychiatric illness within the HD populations.