Can general practitioners identify people with distress and mild depression? A meta-analysis of clinical accuracy

Abstract

Background: There is increasing emphasis on distress and mild depression but uncertainty regarding how well general practitioners (GPs) identify these conditions. Further, the proportion of attendees suffering distress is also unclear.

Aim: To quantify the rate of distress in primary care and to clarify the ability of GPs to identify distressed and/or mildly depressed individuals using their clinical skills.

Methods: Meta-analysis of clinical recognition of distress and mild depression defined on a continuum (severity scale) or categorically (semi-structured interview).

Results: From 157 studies that examined the ability of GPs to diagnose any emotional or mental disorder, we identified 23 that focused on defined distress and 9 that reported on mild depression. The prevalence of broadly defined distress was 37.4% (n=23, 95% CI=29.5% to 45.5) although it was 47.3% (n=14, 95% CI=38.0% to 56.7%) using self-report methods. GPs correctly identified distressed individuals in 48.4% (n=21, 95% CI=42.6% to 54.2%) of presentations and identified non-distressed people in 79.4% (n=21, 95% CI=74.3% to 84.1%) of presentations without distress. GPs correctly identified 33.8% (95% CI=27.3% to 40.7%) of people with mild depression and had a detection specificity of 80.6% (95% CI=66.4% to 91.6%) for the non-depressed. Clinicians' ability to recognize mild depression was significantly lower than their ability to recognize moderate-severe depression. Out of 100 consecutive presentations, a typical GP making a single assessment would correctly identify 19 out of 39 people with distress, missing 20. He or she would correctly re-assure 48 out of 61 people without distress, falsely label 13 people as distressed. For mild depression, out of 100 consecutive presentations, a typical GP would correctly identify 4 out of 11 people with mild depression, missing 7. GPs would correctly re-assure 72 out of 89 people without distress, falsely diagnosing 19.

Conclusions: Clinicians have considerable difficulty accurately identifying distress and mild depression in primary care with only one in three people correctly diagnosed. Clinicians are better able to identify distress than mild depression but success remains limited. However not all such individuals want professional help, and some people who are overlooked get help elsewhere, or improve spontaneously, therefore the implications of these detection problems are not yet clear.