Determinants of (non-)recognition of depression by general practitioners: Results of the Netherlands study of depression and anxiety

Abstract

Background

Although most depressed patients are treated in primary care, not all are recognized as such. This study explores the determinants of (non-)recognition of depression by general practitioners (GPs), with a focus on specific depression symptoms as possible determinants.

Methods

Recognition of depression by GPs was investigated in 484 primary care participants of the Netherlands Study of Depression and Anxiety, with a DSM-IV diagnosis of depression in the past year. Recognition (yes/no) by GPs was based on medical file extractions (GP diagnosis of depressive symptoms/depressive disorder and/or use of antidepressants/referral to mental health care). Potential determinants of (non-)recognition (patient, depression, patient-GP interaction, and GP characteristics) were bivariately tested and variables with a p-value ≤ 0.2 entered into a multilevel multivariate model. Subgroup analysis was performed on 361 respondents with more reliable GP diagnosis data.

Results

60.5% of patients were recognized by their GP. Patients who did not consult their GP for mental problems, and without comorbid anxiety disorder(s) were less often recognized. In the subgroup, where 68.7% was recognized, in addition to these, decreasing number of symptoms of depression and increased appetite were associated with decreased recognition. No GP characteristics were retained in the final model.

Limitations

Some data on recognition were collected retrospectively.

Conclusions

In addition to patients without a comorbid anxiety disorder or who did not consult their GP for mental problems, GPs less often recognized patients with fewer depression symptoms or with increased appetite. Recognition may be improved by informing/teaching GPs that also increased appetite can be a symptom of depression.

Introduction

Depression is a common condition, associated with a large burden for patients and society due to its chronic or recurrent nature (Murray and Lopez, 1997). Most patients with depression are treated in primary care, although often in a non-specific way (Kessler et al., 2003, Wilson et al., 2003). Adequate recognition and treatment can decrease the burden of disease (Claxton et al., 2000, Hirschfeld, 2001, Melfi et al., 1998). It is reported that general practitioners (GPs) recognize depression poorly, perhaps due to their more physical and demand-led orientation (Berardi et al., 2005, Simon and VonKorff, 1995, Wittchen et al., 2001). However, various definitions of ‘recognition’ were used in these studies. Those that applied a cross-sectional design and relied solely on GP diagnosis at time of consultation found lower recognition rates compared to studies that used medical file extraction over extended time periods (Kessler et al., 2002, Mitchell et al., 2009).

Although recognition alone does not necessarily imply appropriate treatment (Dowrick and Buchan, 1995). It seems obvious that recognition of a patient as having depression or as ‘a psychological case’, or at least a discussion of the symptoms, is essential for adequate treatment. Documentation of an International Classification of Primary Care (ICPC) code of depression in the GP's records might not be required to ensure appropriate treatment, as GPs might decide not to diagnose depression because they (or the patient) might consider a diagnosis of depression as stigmatizing (Barley et al., 2011). Also, not all GPs code every consultation with an ICPC code. Finally, not every patient with depression needs (immediate) treatment. With a reasonable chance of spontaneous recovery within three months, several guidelines recommend ‘watchful waiting’ or a minimal intervention as an option during the first months, especially for patients with a first and mild depression (Meeuwissen et al., 2009, National Collaborating Centre for Mental Health, 2009, Spijker et al., 2002). On the other hand, many patients with depression do need treatment, and recognition alone might not be sufficient to ensure adequate follow-up and treatment in these patients (Claxton et al., 2000, Hirschfeld, 2001, Melfi et al., 1998). Therefore a definition of recognition measuring ‘active recognition’ i.e. receiving treatment such as antidepressants or a referral to mental health care might be more suitable.

When it is established which patients remain unrecognized, GPs can be advised to focus on these groups which, in turn, might improve recognition. Although some studies examined determinants of recognition of depression, the results were ambivalent and the sample sizes small. As possible determinants, mostly depression severity and demographics were investigated.

Some studies reported that depression severity predicts recognition (Klinkman et al., 1998, Simon and VonKorff, 1995, Tylee and Walters, 2007, Wittchen et al., 2001), or that patients presenting with mental problems were better recognized (Furedi et al., 2003, Menchetti et al., 2009, Wittchen et al., 2001). Patient characteristics such as age, gender, ethnicity and marital status have also been investigated, but with mixed results. Some found that women and older persons were identified more easily, whereas others found no differences (Fernandez et al., 2010, Gater et al., 1998, Rifel et al., 2008, Wittchen et al., 2001). An elderly primary care sample showed that clinical clues to better identify depression were female gender, the presentation of vague symptoms, and gastrointestinal symptoms (van Marwijk et al., 1996). Another study performed in the Netherlands found that not only patients with low severity of depression, but also those without chronic somatic comorbidity, with lower educational level and with fewer visits to the GP, were at higher risk for non-recognition (Nuyen et al., 2005).

Physician factors such as gender, experience, depression interest and courses on depression were also investigated, again with mixed results. Wittchen et al. found that physician experience of more than 5 years increased recognition. Tylee and Walters found that interest in psychiatry and empathy increased recognition, while pre-occupation with organic disease decreased recognition (Tylee and Walters, 2007, Wittchen et al., 2001).

Only one study investigated the different symptoms of depression as possible determinants, and found that only ‘loss of self-confidence’ was associated with recognition (Wittchen et al., 2001).

It is unclear which determinants predict GP's recognition of depression when using a broader, longitudinal measured definition of recognition and examining a wide spectrum of potential predictors.

The main aim of the present study was to identify determinants of (non-)recognition of depression by GPs (longitudinally measured) in patients with DSM-IV diagnosed depression. Characteristics of the patient, depression, patient–GP interaction and GP were investigated. Of the depression characteristics, we focused on the influence of specific depression symptoms on recognition rate. We hypothesized that GPs would less often recognize less severe cases (including patients without suicidal tendency), those who did not present with mental problems, and/or patients with few(er) visits to their GP.