Research reportCan general practitioners identify people with distress and mild depression? A meta-analysis of clinical accuracy
Introduction
Mild mental disorders are increasingly recognized as clinically significant (Backenstrass et al., 2007). People with major depression experience cumulatively more subthreshold symptoms than syndromal episodes (Furukawa et al., 2009). Further, depressive disorders that do not meet diagnostic criteria for major depression as well as those only just meeting criteria are associated with significant morbidity, functional impairment and reduced quality of life (Pincus et al., 1999, Wagner et al., 2000, Preisig et al., 2001, Cuijpers et al., 2004). Recognition and treatment of mild cases might prevent a substantial proportion of future serious cases (Kessler et al., 2003a, Cuijpers & Smit, 2004, Lyness et al., 2006). That said minor depression and distress can be resolved without professional help (Hermens et al., 2004, Forsell, 2007). Beyond depression, distress has similar clinical and prognostic implications (Bultz et al., 2007). Distress refers to significant emotional upset that is common to a range of psychological and psychiatric conditions. In fact a greater proportion of primary care attendees suffer distress than suffer depression. In one study, 28.4% of attendees reported emotional distress but less than half of these reported discussing emotional distress with their primary care physician in the past year (Gross et al., 2007a). Using the HADS total score, de Waal et al. (2005) found distress in 27% of 1500 primary care attendees (de Waal et al., 2005). There is a close relationship between distress and poor mental health (Ustun and Von Korff, 1995). High rates of distress are a risk factor for functional impairment (Klapow et al., 2002). In recent years General Practitioners (GPs) have been observed to have limited success in identifying clear psychiatric disorders such as depression and anxiety (Hirschfeld et al., 1997, Wittchen et al., 2002, Weisberg et al., 2007, Mitchell et al., 2009). However, an alternative approach is to view symptoms of depression on a continuum with the majority suffering mild disorder or subsyndromal conditions (symptoms insufficient to meet operational criteria) (Backenstrass et al., 2006, Prisciandaro & Roberts, 2005). Many suggest that the focus of detection should not be a psychiatric disorder per se but rather the attempt to identify people suffering broadly defined distress who require professional help (Saraceno et al., 1994, Mulder, 2008). Against this some argue that screening has not proven successful and further that detection of mild disorders has equivocal benefits (Palmer & Coyne, 2003, Magruder & Yeager, 2009).
Individuals meeting criteria of a depressive episode but with 4 or 5 symptoms are defined as mild depression in ICD10. In DSMIV mild major depression is defined as those with symptoms barely meet criteria for major depression and result in little distress or interference with the patient's ability to work, study or socialize. Those with core symptoms not fulfilling full criteria are labelled with minor depression. One difficulty, however, is that the definition of distress has not been robustly operationalized in ICD10 or DSMIV. Distress is often used as marker of clinical significance in relation to generalized anxiety disorder, PTSD, dysthymia and depression. Indeed broadly defined distress may not be compatible with an operational definition but may be better graded on a severity scale (Goldberg, 1992). To date most methods that have been developed to evaluate distress have been evaluated in general hospital settings (Forsell, 2007, Derogatis, 1994, Hausam et al., 1998, Roth et al., 1998, Seelert et al., 1999, Ritsner et al., 2002). For example, the US National Comprehensive Cancer Network (NCCN) developed a definition of distress1 along with a single item visual-analogue scale (NCCN, 2007). In primary care, the most common scale to elicit distress is the General Health Questionnaire (GHQ). The GHQ is a self report questionnaire designed to identify distress associated with common mental disorders. The main formats are the GHQ-60, GHQ-30, GHQ-28 and the GHQ-12. Its sensitivity and specificity appear to be close to 80% when measured against a structured interview (Goldberg & Williams, 1988, Makowska et al., 2002, Schmitz et al., 1999a, Schmitz et al., 1999b, Bell et al., 2005). Several alternative tools have also been validated and these include the Hopkins Symptom Check-List (SCL-90-R) (Derogatis, 1977) and the Kessler Psychological Distress Scale K10 (Kessler et al., 2002). The SCL-90-R is a 90 item instrument that aims to evaluate a broad range of psychological problems including somatization, obsessive–compulsive, interpersonal sensitivity, depression, anxiety, anger–hostility, phobic anxiety, psychotism, and paranoid ideation. The Kessler (K10) is a simple measure of psychological distress. The Kessler K10 involves 10 questions about emotional states each with a five-level response scale. The measure is often used as a brief screen for distress or any mental disorder (Kessler et al., 2003b). In a primary care sample of 422 attendees both the SCL90R and the GHQ12 showed similar validity against the SCID (SCL-90-R: mean AUC = 0.75; GHQ: mean AUC = 0.73) (Schmitz et al., 1999c).
Given the uncertainty about distress in primary care, the aim of this quantitative data synthesis was 1. To clarify the typical rate of distress in primary care (prevalence) 2. To summarize the ability of primary care clinicians to detect distressed patients 3. To clarify the ability of clinicians to detect mild depression in primary care settings (either alone or in comparison with non-mild cases).
Section snippets
Inclusion/exclusion criteria
The principle inclusion criterion was the examination of the clinical ability of GPs to detect defined mild mental disorder. We focussed on distress and mild depression as a preliminary search revealed insufficient studies for a meta-analysis anxiety disorders. We separated results for outcomes defined by severity scale from those using research interview (see Tables 1 and 2 for details of criterion standard).
We were interested in unassisted clinical diagnoses without systematic help from
Study description and methods
We identified 157 potential studies but 86 did not adequately define a group with distress or mild depression and 39 studies did not have data on unassisted clinical accuracy of GPs but concerned screening evaluations. We excluded one study measuring distress via CES-D (Chung et al., 2003). We excluded one study where GPs were responsible for generating the gold standard (Jencks, 1985). One large scale comparison of the GHQ12 with the CIDI-PC was excluded as extractable data was not presented
Comment
There has been no previous attempt to summarize how well clinicians in primary care are able to identify people with mild mental disorders, specifically mild depression and distress. Indeed the proportion of attendees suffering distress has itself been unclear although the largest studies suggest at least a quarter have significant distress (de Waal et al., 2005, Ustun & Von Korff, 1995). This meta-analysis of 17,783 individuals found an even higher rate of distress between 37.4% and 44.1% with
Role of funding source
None.
Conflicts of interest
None.
Acknowledgements
Many thanks to the staff of the medical library, Leicester General Hospital.
References (107)
- et al.
A comparative study of nonspecific depressive symptoms and minor depression regarding functional impairment and associated characteristics in primary care
Compr Psychiatry
(2006) - et al.
Cost-effectiveness of usual general practitioner care with or without antidepressant medication for patients with minor or mild-major depression
Journal of Affective Disorders
(2008) - et al.
Effectiveness of an intervention to reduce sickness absence in patients with emotional distress or minor mental disorders: A randomized controlled effectiveness trial
General Hospital Psychiatry
(2006) - et al.
Minor depression: Risk profiles, functional disability, health care use and risk of developing major depression
J Affect Disord
(2004) - et al.
The reporting of specific physical symptoms for mental distress in general practice
Journal of Psychosomatic Research
(2005) A classification of psychological distress for use in primary care settings
Social Science & Medicine
(1992)- et al.
Global versus specific symptom attributions: Predicting the recognition and treatment of psychological distress in primary care
Journal of Psychosomatic Research
(2004) Counselling in primary care: A randomised controlled trial
Patient Education and Counseling
(1997)- et al.
The prognosis of minor depression in the general population: A systematic review
Gen Hosp Psychiatry
(2004) - et al.
Meta-analytic methods for diagnostic test accuracy
J Clin Epi
(1995)