Increasing attention has been drawn to the burden of disease associated with depression, and the level of unmet need of people with depression in the community. Much of this attention has focused on the role of primary care and its capacity to care adequately for these people. While it is true that much depression is treated in primary care, it also appears that general practitioners fail to recognise depression in about 50% of cases.1-3 Although the many barriers to the recognition of depression make this degree of non-recognition understandable,4 GPs could reasonably argue that the standards against which they are tested are not valid or appropriate to their setting;5 after all, primary care psychiatry is not specialist psychiatry.6
One of the impediments to progress in this area is the lack of a satisfactory taxonomy that has validity and meaning for patients and clinicians in the primary care arena. There have been attempts to adapt standard classifications (eg, the Diagnostic and statistical manual of mental disorders [DSM]7 and the International classification of diseases [ICD]8) for use in primary care, but evidence suggests that these classifications are no more “user friendly” than the specialist psychiatric classifications from which they were derived.9 The fundamental question of whether these classifications have face validity in general practice remains.
In an important study, 27 GPs observed and reported on three videotaped interviews with patients with mental health problems.10 Not only was there poor agreement about diagnostic labels among the GPs, but, contrary to expectation, there was no common, discernable GP language. “There is evidence and growing need to draw up a new classification of psychiatric morbidity in general practice”, concluded the study authors.10 Since that study, there has been no substantial work published to rectify this problem.5,6
Here, we assumed that experienced GPs have a concept of what depression is. Using observed consultations, we examined the process of making a psychiatric diagnosis in general practice, in order to reveal the rules of classification used by GPs. From this, we aimed to create a taxonomy of depression and distress that mirrors the thinking and practice of experienced GPs.
Data were collected in 2005 through observation of routine clinical consultations conducted over the course of 1 day by a convenience sample of 14 GPs in metropolitan Melbourne. With patients’ informed consent, consultations were videotaped, while GPs also completed Patient Encounter Forms that summarised the reasons for and focus of the consultations, and the assessments or diagnoses they made. Consultations that had depression as a focus of attention were analysed. The analysis involved three steps, outlined below.
Videotaped consultations were studied and analysed using a qualitative method described as ethnomethodology. By studying both the spoken and unspoken word, a “commonsense” process of making a psychiatric diagnosis was revealed.11,12 This process identified the language of depression used, any concepts considered important by GPs, and any assumptions that were applied to the various categories of depression. It enabled the development of a draft taxonomy, which was then refined using data from interviews with GPs.
After analysis of the videotaped consultations, nine GPs agreed to participate in an in-depth interview. Following a method described elsewhere,13 three types of ethnographic questions (descriptive, structural and contrast) were used to further elicit the unspoken process of decision making used by GPs. A loose interview guide was used during the interviews, in which we asked GPs to describe how they recognised and diagnosed depression in their patients. To avoid a-priori knowledge, an interviewer (K C) naïve to the subject matter used follow-up questions to explore the topics raised by the GPs using the same terms and definitions GPs themselves employed.
These interviews were conducted and analysed sequentially (using content analysis14) at the same time as the taxonomy was being developed, so that each interview informed the process in increasingly deeper ways. Later interviews asked specifically about the emerging taxonomy.
A total of 139 consultations were filmed and evaluated. Of these, 36 (26%) had depression as a focus of attention and were analysed.
A diagrammatic representation of the process through which GPs thought about and categorised “depression” is shown in the Box. Essentially, GPs worked in three diagnostic steps. The first step was perception of a change in the patient’s emotional state, evidenced by certain symptoms and signs; the second was a determination of cause, linked with beliefs about depression; and the third was a differentiation based on the most prominent symptoms.
It is difficult to describe what first triggered GPs to think about depression in a patient. Nevertheless, once the “penny dropped”, there was a common pattern of probing that GPs undertook.
GPs indicated that they began to pick up on signals of depression, lowered mood, or distress even before the patient had entered the consultation room. These preliminary signals centred on a facial expression; a certain “look”, sometimes described as “dullness in the eyes”:
I go a lot on the way they look, the way they act ... their affect. (GP; interview)
Their eyes, their facial expression, just changes altogether. (GP; interview)
Dull eyes, avoiding eye contact, no enthusiasm ... (GP; interview)
I think you can tell by someone’s eyes when they get better. It’s got something to do with the way they look at you. This sort of blackness has gone. (GP; interview)
They just get that sparkle back in their eyes. (GP; interview)
In most cases, GPs did not ask patients directly about depression or feeling depressed. Rather, signs and symptoms of distress were used as pointers to possible underlying depression.
Then I ask them how they’re going in themselves and I suppose it’s not coming straight out and saying “Are you depressed?” or “Have you got anxiety?”, but it’s “What do you feel?”, “How are you going at the moment?”, “Do you feel good in yourself, are you happy?” It’s a bit like asking someone about their sex life. You can’t just sort of jump straight in there. (GP; interview)
Once GPs had concluded that something was wrong, they often followed up by asking a direct question about depression.
After identifying a patient’s distress, the second step GPs undertook was to determine if there was an identifiable environmental cause. If there was not, patients were generally considered to have an “endogenous” form of depression. One GP described this distinction:
And it [the depression] might be a long-term thing that’s been investigated for biological causes, or it can be a new presentation in somebody that’s had several stressors recently. (GP; interview)
The following excerpt from a consultation with a previously depressed patient shows the GP considering with the patient the possible contribution of both biological (medication) and psychosocial factors.
The identifying feature of the reactive type of depression was the presence of an identifiable cause.
Psychological distress is much more common in my patients when they are going through difficult times, for example, job hassles, family or marital disharmony, money worries, unemployment, chronic illness, drug and/or alcohol use; and the list goes on and on. (GP; interview)
The most common form of reactive distress was identified with “worry” about things, or “stress”.
Additionally, patients often express these anxieties as difficulties in coping.
GPs identified a spectrum of reactive depression, from one form with predominant anxiety to another characterised by helplessness and hopelessness.
Patients at the helpless–hopeless end of the spectrum were not frequently observed during the videotaped consultations. However, one patient and one GP described their previous experiences with this sort of depression:
In this situation, GPs are commonly slow to prescribe antidepressants:
GPs identified a category of “endogenous depression”, characterised by a lack of interest or motivation with no identifiable cause, although this was rarely observed in the videotaped encounters. Nevertheless, it was a concept in the minds of the GPs, linked with a chronicity of course and a biological theory of depression and its treatment.
The endogenous ones are the ones who, right from childhood, for as long as they can remember, have episodes when things are black. I mean they may not even tell anybody about it, even their closest and nearest don’t pick it up. (GP; interview)
If it is purely chemical or endogenous, they often still blame other reasons for it. But once they get treatment, they’re better off. (GP; interview)
This study revealed a definite pattern of thinking among GPs about depression, and the clinical features that are important to them in identifying depression.
The symptoms and signs that either trigger their thinking about depression, or are enquired of to confirm a diagnosis of depression, include affect; crying; loss of interest, motivation or drive; sleep disturbance; and impairment of functioning. Even in a brief consultation, these stand out, particularly if there is a change from usual, or if it is brought directly to the GP’s attention in some way (eg, by obvious crying).
Once distress is recognised, GPs naturally look for an environmental cause or stressor, and if one is found, the distress or depression is considered “reactive”. This is the most common form of distress, although it is expressed along a continuum with “worry” prominent at one end (sometimes referred to as “stress”) and helplessness–hopelessness at the other. The latter, accompanied by feelings of “not coping”, is very similar in its severest form to the experience of demoralisation observed in the severely medically ill.15,16 The most common forms of sleep disturbance — difficulty getting to sleep or staying asleep — were associated with anxiety and worry.
Although infrequently observed in this study, GPs had a clearly defined concept of endogenous depression as being characterised by lack of interest and motivation without a psychosocial trigger, having a biological if not characterological basis, and requiring or being responsive to pharmacological treatment. Early morning waking was associated with this form of depression.
These results are interesting in that they converge with contemporary psychiatric practice in some respects, but diverge in others. The symptoms that GPs recognise as indicating depression are all key depression criteria in standard classifications (DSM and ICD). Depressive affect is often hard to describe, but experienced GPs recognise something in facial expression. This “look” is reminiscent of the description used by Robert Burton with respect to the sorrow of melancholia in The anatomy of melancholy (first published in 1621): “It dries up the bones ... makes them hollow-eyed, pale, and lean, furrow-faced, to have dead looks, wrinkled brows, rivelled cheeks”.17 Although these signs of distress and depression can be subtle and are subjective, they are clearly important in this first step of recognition.
There was divergence between GPs and contemporary psychiatric practice in the use of the terms “endogenous” and “reactive” and their theoretical underpinnings. The third edition of the DSM18 ushered in an era of taxonomy in which diagnostic categories were divorced from aetiological theories,19 and the “atheoretical” term “major depression” was coined. However, for the GPs we interviewed, it made clinical sense to consider aetiology and to name the depression accordingly. It also linked with their ideas about treatment (eg, endogenous depression has a biological basis and responds to pharmacotherapy).
The most commonly observed forms of distress and depression were the reactive forms, particularly “anxious depression”. For these patients, GPs tended to use a “watchful waiting” approach to pharmacotherapy, consistent with current clinical guidelines.20
The particular strength of this taxonomy is the inherent validity that arises from it being derived from GPs, their patients, and a study of the clinical decision making that goes on between them. Its other major strength is its simplicity and parsimony.
This study worked on the assumption that experienced GPs have an intuitive sense based on experience and reflection. It is interesting that they used concepts, like reactive and endogenous, that have not been used by specialists for 20 years. Is this their intuition, or because that is what they have been taught? It is of course impossible to study experienced yet naïve practitioners. All GPs have participated in some educational activities in mental health. Subjective and qualitative methodologies are useful, but they need to be supported by more objective research methods.
Two limitations of this study reflect its primary care focus. The first is the limited range of psychopathological disorders observed. GPs had a concept of endogenous (or severe) depression, though there was only one example of this in the consultations studied. Data to clarify concepts of severe, endogenous depression cannot be derived from general practice.21
Second, this study was only able to shed light on the types of depression that were recognised by the GPs. Although appropriate for this study’s aims, it will be important for future research to examine in more detail the nature of the depressions that are not commonly recognised by GPs.
A study such as this will not be the last word on the matter of a preferred taxonomy for depression. There is a need for continued research that listens carefully to patient or consumer descriptions and links these subjective accounts with clinical models. Nevertheless, the results of this and other similar work22 should be taken seriously, as they are derived from practice in the field. Current psychiatric classifications are not applicable to the primary care setting. Future research needs to establish the level of understanding that GPs and patients have of this taxonomy, and the reliability and practicality of its use. Although not comprehensive, this taxonomy offers a simple and parsimonious system of diagnosis for distress syndromes that are common in general practice.
Abstract
Objective: To create a taxonomy of distress and depression for use in primary care, that mirrors the thinking and practice of experienced general practitioners.
Design: Qualitative study, using an ethnomethodological approach, with observation of videotaped routine GP–patient consultations and in-depth interviews with GPs.
Setting and participants: The study was conducted in metropolitan Melbourne in 2005. Fourteen GPs conducted 36 patient consultations where depression was a focus; nine GPs participated in in-depth interviews to elicit details of how they recognised and diagnosed depression in their patients.
Results: GPs consider distress and depression in three steps. In the first step, a change in a group of symptoms and signs is observed (eg, facial expression, loss of drive). The second step categorises the syndrome according to whether or not there is an identifiable environmental cause (reactive or “endogenous”), with the final step categorising the reactive syndromes according to their most prominent symptoms: either anxiety and worry, or helplessness and hopelessness. The resulting taxonomy includes: endogenous depression (a chronic and perhaps characterological depression characterised by a lack of interest and motivation); anxious depressive reaction (stress or worry); and hopeless depressive reaction (demoralisation).
Conclusion: This simple and parsimonious taxonomy has validity based on its derivation from within the primary care setting.