Comparison of self-report and structured clinical interview in the identification of depression

Abstract

Background

A self-report method seeking a binary response for assessing depression is a cost-effective and time-efficient way to obtain a psychiatric history, yet the reliability of this method is largely unknown. The aim of the study was to compare and assess the validity of two methods for identifying a past history of depression in a population-based study.

Methods

This study examined data collected from 891 men and 1086 women participating in the Geelong Osteoporosis Study. Self-reports of depression were compared with results obtained using the Structured Clinical Interview for DSM-IV-TR Research Version, Non-patient edition (SCID-I/NP).

Results

Using the SCID-I/NP, 146 (16.4%) men and 285 (26.2%) women met criteria for a lifetime depression. Of those participants, 61.0% (n = 263) self-reported a history of depression. The level of agreement between self-reporting depression and the SCID-I/NP depression module was reasonably high; 61% sensitivity, 89.5% specificity and the overall level of agreement (kappa) was 0.5.

Limitations

Results may not be generalizable to other self-report instruments or be suitable for use in clinical samples.

Conclusion

The SCID-I/NP remains the gold standard for identifying depression; however, given the moderate level of agreement between the self-report questionnaire and SCID-I/NP in our current study, we conclude that simple self-report methods can be used to identify depression with some degree of confidence.

Introduction

Self-report methods and symptom scales are often utilised to obtain diagnostic information in large-scale epidemiological studies, as they are cost-effective and time-efficient. There are many simple scales available to measure symptoms of depression that are either self-administered or administered by a lay person. The number of items on a depression symptom scale can vary, however, scales with as little as two-items have been successfully utilised to detect depression [1]. Research comparing depressive symptoms measured using such convenient survey instruments and results from ‘gold standard’ psychiatric examinations, such as the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) (SCID-I), report varying levels of agreement among clinical and survey settings [2].

Simple self-report methods that seek a binary (yes/no) response have been used to ascertain the presence of medical conditions in a medical history exam. Past research has shown good sensitivity and specificity for self-reported medical conditions against physician reports for well-defined chronic conditions, such as hypertension, whereas over-reporting has been identified with conditions such as arthritis [3].

A self-report yes/no questionnaire assessing psychiatric disorders is an effective way to obtain information regarding psychiatric history, yet the reliability of this method remains in question. One study, the Segiumiento Universidad de Navarra (SUN) project [4], compared self-reported physician diagnosis of depression with the SCID-I in a group of university graduates. In that study, the clinical assessment confirmed depression in 74% of those with a self-reported diagnosis [4].

This current analysis is a head-to-head comparison of a simple, self-reported identification method for lifetime depression compared against a gold-standard clinical psychiatric assessment, using data from a large, randomly-selected, population-based sample of men and women.