Canadian Task Force on Preventive Health Care

Early Detection of Depression

Prepared by John W. Feightner, MD, MSc, FCFP, Department of Family Medicine, The University of Western Ontario

Objective
To make recommendations regarding routine screening for depression in asymptomatic adults, updating a report published in May, 1990.

Burden of Suffering
The lifetime prevalence of clinically significant depression is 15% to 30%; it is about twice as common among women as among men, and is thought to have increased among children and adolescents. People who are single, divorced, separated, seriously ill, recently bereaved or those with a family history of depression have a greater incidence than others. Depressed individuals frequently present with physical symptoms, which may make diagnosis more difficult, resulting in diagnostic testing or treatment for other illnesses. Affected people are more likely than others to be suicidal: 30% to 70% of people who have committed suicide were previously identified as having major depression. In Canada, in 1986, suicide accounted for an estimated 97,600 potential years of life lost among men and 25,300 among women.

Options
To screen or not to screen for depression using the General Health Questionnaire (GHQ) or the Zung Self-Rating Depression Scale.

Outcomes
Validity, sensitivity and specificity of questionnaires; physician detection and treatment of depression (p451).

Evidence
MEDLINE was searched up to May, 1993 using the MeSH headings "depression" and "mass screening". Study results were synthesized in table or graphic format only.

Values
The 13-member Task Force of experts in family medicine, geriatric medicine, pediatrics, psychiatry and epidemiology used an evidence-based method for evaluating the effectiveness of preventive health care interventions. Recommendations were not based on cost-effectiveness of options. Patient preferences were not discussed.

Background papers providing critical appraisal of the evidence and tentative recommendations prepared by the chapter author were pre-circulated to the members. Evidence for this topic was presented and deliberated upon in 1- to 2-day meetings, 2 to 3 times per year from January 1993 to June 1993. Consensus was reached on final recommendations.

Benefits, Harms, and Costs
The validity of many self-administered questionnaires has not been fully established. Using clinical criteria as the diagnostic standard, questionnaires have shown sensitivities of 64% to 91% and specificities of 56% to 82%. Correlations with other tests or with clinical assessments ranged from 0.40 to 0.89 (p451).

4 RCTs evaluated effects of the routine use of a screening questionnaire on the detection and management of depression. In 3 of these trials, patients completed the GHQ, and the results were presented to the physician before seeing the patient. A study of 1242 patients attending an inner-city primary care clinic found that GHQ results had an effect on physician detection of psychological problems only for those ł 65 years, and ultimately had no effect on patient management for any age group [3]. Similarly, a study of 1469 patients in a primary care office found no effect of physician knowledge of a positive GHQ result on detection of psychological distress [4]. A methodologically weaker study of 1093 patients found that 16% of patients were given a diagnosis of psychiatric illness when the physician had no GHQ results; an additional 11% of patients were diagnosed after GHQ results were given to the physician [5]. Another study found that when informed of a positive patient score on the Zung-Self-Rating Depression Scale, physicians detected depression in more patients (68% of 102) than when they were uninformed (15% of 41). This study was flawed in design and execution [6]. (p451)

Potential harms and costs were not discussed.

Recommendations
Recommendation grade [A, B, C, D, E] and level of evidence [I, II-1, II-2, II-3, III] are indicated after each recommendation. Citations in support of individual recommendations are identified in the guideline text.

Fair evidence exists to exclude (from the PHE) routine testing of asymptomatic persons for depression using the GHQ or the Zung Self-Rating Depression Scale [D, I]. It is suggested, however, that clinicians maintain a high degree of clinical suspicion for depression among their patients.

Validation
Recommendations were peer-reviewed and are similar to those of the 1989 U.S. Preventive Services Task Force.

Sponsors
The Canadian Task Force on Preventive Health Care developed this guideline with funding from Health Canada.

Selected References

Source Document
Feightner JW. Early detection of depression. In: Canadian Task Force on the Periodic Health Examination. Canadian Guide to Clinical Preventive Health Care. Ottawa: Health Canada, 1994;450-4.

Other
Canadian Task Force on the Periodic Health Examination. The periodic health examination, 1990 update: 2. Early detection of depression and prevention of suicide. Can Med Assoc J. 1990;142:1233-8.