Making fewer depression diagnoses: beneficial for patients?

Abstract

Currently, general practitioners actively search for depressive disorders in their patients. When they diagnose ‘depressive disorder’, they tell their patients that they have a disease and can be treated accordingly. This is probably an important reason for the huge prescription rates of anti-depressants. In doing so, general practitioners implement specialised, psychiatric diagnostic methods in a setting characterised by patients with symptoms that superficially may resemble those of depressive disorder but in reality mainly arise from normal problems in everyday life due to losses of valued relations or failure to achieve desired goals. We argue that it might be beneficial for patients if general practitioners, in a stepped care approach, hold back on specialised methods of psychiatry and instead use a more generalist approach as first step, in which patients' problems are formulated in their own words, and efforts are directed in helping patients regain their self-confidence to solve them. Our arguments for directing attention away from diagnosing depressive disorder are: depressive disorder is a diagnosis by agreement and therefore relative, so there are other ways to look at problems than though psychiatric glasses; depression has unclear boundaries with other mental disorders and with normality; depression is often not an adequate summary of the real problems of the patient; the patient often has a very different conception about what is wrong and often does not agree with the proposed presence of a mental disorder; to diagnose depressive disorder may have more disadvantages than advantages for the patient;. the efficacy of anti-depressants is very modest.

Introduction

Currently general practitioners (GPs) diagnose depression and tell their patients that they have a disease.¹ As such, depression resembles appendicitis. A matter of making a diagnosis and starting treatment. According to some, GPs underdiagnose depression.² However, Parker recently maintained that depression is overdiagnosed.³ He argued that depression has reached a status that goes far beyond its true status and that the concept has been widened gradually to a concept containing many heterogeneous categories. This makes it more understandable that GPs currently prescribe antidepressants too often, mostly in the first consultation.⁴

So, in the recent past, GPs underdiagnosed and undertreated depression. Now, they diagnose and treat depression, frequently with medication, and they are criticised as well!

To get out of the trouble between GP and depression, we propose that the GP makes fewer depression diagnoses and instead treats patients more as the generalist (s)he is supposed to be. Physicians should have an indication before they apply the medical model to the patient's story, and this indication should be more than the mere presence of a number of symptoms for a certain time.

We base our proposition on the following arguments: the concept of depression is a relative one because depression is a diagnosis by agreement and not by essence; the concept of depression suffers from problems with validity and reliability; the concept of depression does not distinguish between normal and disordered depressive feelings; patients reject being stigmatised as having a mental disorder; the effectiveness of antidepressant medication is limited; and, finally, patients are reluctant to take antidepressants.

The model

There are two kinds of diseases: diseases characterised by a biological essence and diseases characterised by agreement.⁵ Examples of the first kind are appendicitis and myocardial infarction. These diseases are unambiguously demonstrable in all cultures and all times, on the condition that the necessary technical possibilities are in stock. They are universal and are discovered as soon as the technical possibilities become available. They are based on anatomical or physiological abnormalities. This basic abnormality is a thing and has a biological essence: an infected appendix or an obstructed coronary vessel. Examples of diseases of the second kind are: depression and attention-deficit hyperactivity disorder (ADHD). These diseases are not patiently waiting for their discovery. They are constructed when their time is there. Based on the available facts, a number of influential experts determine that, for example, depression is a syndrome consisting of a number of symptoms – at least five out of a list of nine with at least one of two obligatory symptoms. Those symptoms have to be present over a specified period of time and have to be sufficiently serious. The agreement gives a name to a certain combination of symptoms. One speaks of BOGSAT diagnoses: diagnoses that develop from a ‘bunch of guys sitting around a table’. One could argue that depression does have a biological essence, because there are different patterns of colouring in functional magnetic resonance pictures. Although these patterns have been demonstrated unambiguously, there remain questions about the specificity of these findings and about whether these findings are causes or consequences of depression. After all, every thought or emotion will sooner or later have a visible counterpart in the brain.

A general problem with agreement diagnoses is that the longer they are used, the more people are inclined to see them as a thing in the patient, tangible and objective, seemingly an essence (reification).⁶ This is comfortable for physicians: fatigue in combination with depressed feelings in the patient's story guide the doctor to the diagnosis of depression, just as pain in the right lower abdomen in combination with pain during movement will lead to the diagnosis of appendicitis. The physician is willing to recognise the depression as this is controllable and treatable with evidence-based treatments. A disadvantage of reification is that the physician loses possibilities to view presented problems differently.

The disorder

There is evidence that depression is not a well-circumscribed syndrome with clear boundaries with anxiety disorders and somatoform disorders.⁷ More than half of the patients with anxiety or depressive disorder satisfy criteria for somatoform disorder, and more than a quarter of patients with somatoform disorder have anxiety or depressive disorder.⁸ Alongside unclear boundaries with other disorders, there is a gradual transition from depressive disorder to minor disorder to normal mental health.⁹ Further, the combination of symptoms in recurrent depressions is not similar to the combination of symptoms in previous episodes: the correlation for each symptom between episodes is rather low. This is strange, as for a diagnosis of depression it is required that at least five out of nine symptoms are present.¹⁰ Moreover, the combination of symptoms appears to depend on the type of life event that precedes the episode: death of close relatives or friends is more often associated with a depression characterised by depressed feelings and loss of pleasure, while chronic stress is more often associated with a depression characterised by fatigue and sleep problems.¹¹ Finally, the definition of depressive disorder does not deal with the distinction between normal sadness and disordered sadness, except in bereavement.¹² These issues raise questions about the validity of the concept of depressive disorder.

The patient

Does the patient accept depression as an adequate description of his or her problems? Or has the patient a very different idea of what is wrong?

The patient's opinion about depression

The attitude of patients regarding the diagnosis ‘depression’ is ambivalent. Our experience as practising physicians reveals that some patients object to depression as a description of their problem. Patients are often not inclined to see themselves as ‘mentally ill’ and have good explanations for their complaints. Their explanation is strongly determined by their context and is formulated in a personal story. This is also apparent from research findings.¹³ Patients more often see the cause of depression in relational problems, domestic violence, financial problems, unemployment and somatic problems. Patients are worried about the stigmatising effect of the diagnosis ‘depression’. Patients report being bothered more by the feeling of loss of control than by the symptoms. However, for some patients the diagnosis is a liberation: finally, there is something that the doctors are able to recognise and treat. It has been given a name and therefore is controllable. On the other hand, the diagnosis places the patient in the category of the ‘mentally ill’.¹⁴ Patients with opinions resembling doctors' opinions consider the diagnosis as the confirmation that they are not responsible for the condition themselves, because the essence is a biochemical abnormality in the brain. Unfortunately, this has a negative influence on the feeling of being in control.

The medicine

Suppose the patient agrees with the diagnosis ‘depression’ and suppose he is willing to take medication, then what should we tell him? Should the patient still be willing to take an antidepressant after having been given the details of a limited effectiveness?

Effectiveness of antidepressants

GPs prefer to treat their depressive patients with antidepressants. In trials, on average about 50% of the patients benefit from an antidepressant whereas 30% of placebo users improve. This means that of five patients treated with an antidepressant, only one recovers as a consequence of antidepressant use.¹ Trials from primary care yield about the same success rates: the number needed to treat for selective serotonin reuptake inhibitors (SSRIs) is six, for tricyclic antidepressants (TCAs) four. Reformulated, it means that 80–85 from every 100 patients with depression treated with SSRIs will have no benefit from taking an antidepressant. Probably, these numbers are an overestimation of the real effect. Firstly, the circumstances for patients in randomised controlled trials are ideal. Secondly, in antidepressant trials there is a problem with blinding. Patients will find out whether they take the antidepressant or the placebo because of the adverse effects. They thus will have the extra beneficial effect of knowing that they use a real medicine. Trials dealing with this problem (using active placebos) generally show smaller effects.¹⁵ Thirdly, most trials of antidepressants are sponsored by the pharmaceutical industry, which is well known to have more favourable results than non-sponsored research.¹⁶ Fourthly, there is publication bias in antidepressant trials.¹⁷ Apart from their limited effectiveness, antidepressants have high dropout rates due to adverse effects.¹⁸

What patients think of antidepressants

A systematic review of research in primary care shows that patients prefer psychotherapy when they are given the choice between psychotherapy or medication, because they suppose psychotherapy is more directed at the cause of the depression.¹⁹ Reasons for not preferring antidepressants are the conviction that antidepressants are addictive, the refusal to see depression as a disease, worry that antidepressants suppress ‘normal’ grief, and prior negative experiences with antidepressants.²⁰ Medication use diminishes the feeling of being in control, while regaining control is considered essential for recovery. Antidepressant users are ambivalent about stopping their medication. On the one hand they feel that they have to stop some time, on the other they fear a relapse.²¹

How to continue?

Diagnosing patients' problems as depression is not obligatory, as depression is a diagnosis by agreement and therefore relative. Diagnosing patients' problems as depression is often not a good reflection of the real problems of patients – patients frequently have a very different viewpoint about what's wrong. Diagnosing patients' problems as depression leaves out the question of normality and omits other mental disorders. And finally: the diagnosis leads to a treatment that is not very effective or acceptable.

Therefore, we propose not to use the map of the medical model before the patient's story has been fully explored²² and has been discussed in the patient's own language. We propose not to start with a specific diagnosis and therapy, and to use the medical model only when indicated. We propose a stepped care approach on a generalist basis. The first step is characterised by an emphasis on illness instead of on disease, by attending primarily to the patient's story, and by not focusing on the search for symptoms. Listening is the main activity in the first stage. Stimulating people to tell their story is in itself therapeutic.²³ Next, the physician and the patient try to reach consensus about the definition of the problem, and when many problems seem to co-exist – as is often the case – try to agree about the relative importance of each problem. Following the shared definition of the problems and the relative importance of each problem, the physician firstly pays attention to the patient's own questions. The physician tries to restore the patient's perspective and self-efficacy and thereby relies on the patient's natural healing capacities.²⁴ The attention on the restoration of the patient's perspective is important as, in longstanding problems, loss of perspective is frequently the trigger to consult a GP. All activities presuppose an alliance of doctor and patient in which mutual trust and shared goals have a central role. Here, the GP has a benefit over other healthcare workers because of personal continuity in a longstanding relation with the patient. A central feature in this stage is that not only is treatment stepped, but so is diagnosis.²⁴ If the foregoing activities do not result in better health, we continue with the second step: the use of a medical diagnosis and treatment.

The thoughtful use of the medical model has to be, or has to become again, a specific competence of the GP. We think that our approach does more justice to the situation in primary care, where the problems are seldom simple and often benefit from a broad approach.