Why screening for depression in primary care is impractical

In its latest set of guidelines, the Canadian Task Force on Preventive Health Care presents clear and comprehensive conclusions on screening for depression in adults.¹ The document replaces the task force’s 2005 guidelines, which recommended screening adults in the general population for depression in primary care settings that have integrated systems to manage treatment.² That approach is no longer recommended.

The current guidelines focus on screening in primary care settings, because it already seems to be well accepted that screening the general population outside the treatment setting is not recommended. This focus is most appropriate, because most people with depression will receive treatment in primary care, many solely in primary care.

Unfortunately, the task force does not state how it defines “screening.” In practice, this may mean asking as few as 2 questions, as is recommended in the National Institute for Health and Clinical Excellence (NICE) guidelines (the so-called “Whooley questions”).^3,4 Clearly, responses to these questions would not be diagnostic and would serve only as clues to investigate further.

Alternatively, clinicians can screen patients for depression using a more comprehensive assessment tool, such as the 9-item Patient Health Questionnaire (PHQ-9).⁵ However, this questionnaire has more of the characteristics of a diagnostic instrument rather than a screening tool. Further, as a diagnostic instrument for major depression, the PHQ-9 or similar tools may miss patients with minor depression, dysthymia, recurrent brief depression or bereavement, or depression associated with a major medical condition, substance use or an organic mental state, unless there is further inquiry.

In addition, because comorbidity is common in depression, there may be considerable distress and morbidity from contemporaneous subthreshold disorders that may not be detected by a screening instrument. Comorbidity is also an important driver of treatment seeking.

Depression is a disorder defined by its symptoms. If the DSM-IV-TR (Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision)⁶ is followed, the diagnosis is categorical; that is, one either has it or one does not. But what about the person who is only a single symptom short of meeting the criteria for such a categorical diagnosis? In this situation, comorbidity is likely to be an important determinant of the actual diagnosis and the treatment plan, which supports the need for assessing distress rather than a specific diagnosis.

In the current set of guidelines, the task force defines major depressive episode according to the DSM-IV-TR criteria. However, only the “A” criteria are used, not the “B, C, D or E” criteria, and the categories of depressive disorders mentioned earlier are not included.¹ By screening for a single disorder such as depression, one may miss other diagnoses such as anxiety, which is more frequent than depression and commonly associated with it. If screening is to be done, perhaps a better case could be made for the use of screening instruments such as the Kessler Psychological Distress Scale (K10)⁷ or the General Health Questionnaire,⁸ which are designed to detect levels of mental distress that should lead to further inquiry to establish a more definitive diagnosis.

The task force mentions in passing some of the problems caused by false-positive diagnoses, but it does not address the magnitude of erroneous diagnoses. If we use the 12-month prevalence of 5% for depression reported in the guideline¹ and the K10 tool’s sensitivity of 71% and specificity of 90%,⁷ the false-positive rate will be nearly 73%. Given that family physicians and mental health workers have difficulty dealing with their existing caseloads, it is not feasible to ask them to follow up everyone identified by a screening instrument, when only about one-quarter will actually have depression.

The K10 tool is used to screen for a variety of disorders, not just depression. Therefore, even if we use the average of the prevalence rates for those disorders (including depression) among men and women (17.3%),¹ then the false-positive rate drops to 40% — much better, but still too high to justify use of the tool for screening in a primary care setting. Use of a simpler screening method, such as the Whooley questions,³ is likely to lead to even higher false-positive rates. These points argue strongly against any form of routine screening.

There is no question, as the task force amply illustrates, that depression constitutes a major public health problem. Although milder cases may require only watchful waiting rather than treatment, about 15% of people with major depression go on to a chronic course, with much residual disability.⁹ Family physicians have been criticized for failing to recognize depression. However, studies have shown that many missed cases are those of milder depression, which often remits spontaneously, and that patients with milder forms of depression may experience adverse effects and other complications if the depression is treated.¹⁰ Family physicians have also been criticized for not treating depression even when it is diagnosed. In certain situations, however, a physician may decide not to treat after an assessment of the patient’s social circumstances and situation.

The task force is correct in drawing attention to the lack of evidence in some areas and suggesting that there is little evidence to guide practitioners about what type of screening, if any, to use. In this type of report, which is based on synthesizing and grading the available information, recommendations can be made only when there is sufficient evidence. This leaves areas where recommendations for screening cannot be made, at least for the time being. Clinicians and program managers thus do not have scientific evidence to support routine screening for depression and will need to make decisions based on their experience and practical knowledge.

Key points

• There is little evidence of sufficient quality to guide practitioners about what type of screening, if any, to use to detect depression in adults in primary care settings.

• The number of false-positive screens with current assessment tools is too high, and the follow-up required to rule them out too time-consuming, to justify routine screening for depression in primary care practices.

• If false-positive screens are not ruled out, patients are at increased risk of receiving the wrong diagnosis and inappropriate treatment.

See related guidelines from the Canadian Task Force on Preventive Health Care on page 775 and at www.cmaj.ca/lookup/doi/10.1503/cmaj.130403