Open Access Research article

A population study comparing screening performance of prototypes for depression and anxiety with standard scales

Helen Christensen1*, Philip J Batterham1, Janie Busby Grant2, Kathleen M Griffiths1 and Andrew J Mackinnon3

Author Affiliations

1 Centre for Mental Health Research, The Australian National University, Canberra, Australia

2 Centre for Applied Psychology, University of Canberra, Canberra, Australia

3 Orygen Research Centre, The University of Melbourne, Melbourne, Australia

For all author emails, please log on.

BMC Medical Research Methodology 2011, 11:154  doi:10.1186/1471-2288-11-154

Published: 22 November 2011



Screening instruments for mental disorders need to be short, engaging, and valid. Current screening instruments are usually questionnaire-based and may be opaque to the user. A prototype approach where individuals identify with a description of an individual with typical symptoms of depression, anxiety, social phobia or panic may be a shorter, faster and more acceptable method for screening. The aim of the study was to evaluate the accuracy of four new prototype screeners for predicting depression and anxiety disorders and to compare their performance with existing scales.


Short and ultra-short prototypes were developed for Major Depressive Disorder (MDD), Generalised Anxiety Disorder (GAD), Panic Disorder (PD) and Social Phobia (SP). Prototypes were compared to typical short and ultra-short self-report screening scales, such as the Centre for Epidemiology Scale, CES-D and the GAD-7, and their short forms. The Mini International Neuropsychiatric Interview (MINI) version 6 [1] was used as the gold standard for obtaining clinical criteria through a telephone interview. From a population sample, 225 individuals who endorsed a prototype and 101 who did not were administered the MINI. Receiver operating characteristic (ROC) curves were plotted for the short and ultra short prototypes and for the short and ultra short screening scales.


The study found that the rates of endorsement of the prototypes were commensurate with prevalence estimates. The short-form and ultra short scales outperformed the short and ultra short prototypes for every disorder except GAD, where the GAD prototype outperformed the GAD 7.


The findings suggest that people may be able to self-identify generalised anxiety more accurately than depression based on a description of a prototypical case. However, levels of identification were lower than expected. Considerable benefits from this method of screening may ensue if our prototypes can be improved for Major Depressive Disorder, Social Phobia and Panic Disorder.