Systematic Evidence Review

Number 6

Submitted to:Agency for Healthcare Research and Quality 6010 Executive Boulevard, Suite 300 Rockville, Maryland 20852

Submitted by:Research Triangle Institute3040 Cornwallis RoadP.O. Box 12194Research Triangle Park, North Carolina 27709

Contract No. 290-97-0011Task No. 3

RTI Project No. 6919-003

April 2002

Mike Pignone, MD MPH
Bradley N. Gaynes, MD MPH
Jerry L. Rushton, MD MPH
Cynthia D. Mulrow, MD, MSc
C. Tracy Orleans, PhD
B. Lynn Whitener, DrPH
Catherine Mills, MA
Kathleen N. Lohr, PhD
for the
Research Triangle Institute - University of North Carolina
Evidence-based Practice Center

Depressive disorders are an important cause of morbidity and are common in primary care settings. Previous research suggests that depression is underrecognized and undertreated. Screening for depression in primary care settings may improve recognition, treatment, and outcomes of depressive disorders.

To review systematically the literature regarding the effectiveness of screening for depressive disorders in primary care settings.

We systematically searched MEDLINE from 1994 through 1999 using 2 Medical Subject Headings (MeSH) terms, depression and depressive disorders, and combined them with predefined strategies to identify diagnostic accuracy studies and randomized controlled trials of screening and treatment. We used the second edition of the Guide to Clinical Preventive Services, recent systematic reviews, and focused searches of MEDLINE from 1966 to 1994 to identify older articles of interest. We also used hand checking of bibliographies; a search of the Cochrane depression, anxiety, and neurosis database; and extensive peer review to identify articles not captured through our main search strategy.

Diagnostic accuracy studies were included if they reported sensitivity and specificity results based on evaluation against a criterion standard. Treatment studies were included if they were randomized trials in primary care populations. Screening outcome studies were included if they were randomized trials that reported outcomes of change in recognition or treatment of depression or change in health outcomes.

Two reviewers initially examined titles and abstracts of articles and excluded those that clearly did not meet inclusion criteria. Two reviewers then examined the full articles of the remaining studies and determined final eligibility by consensus.

A single reviewer abstracted the relevant data from the included articles and entered them into evidence tables. A second reviewer checked the accuracy of the tables against the original articles.

Studies examining the diagnostic accuracy of depression screening instruments generally have found sensitivity results of 80% to 90% and specificity results from 70% to 85% in adults and 60% to 100% and 60% to 85% in children.

For adult primary care patients with major depression, treatment with pharmacotherapy or psychotherapy reduces symptom duration and severity. Pharmacotherapy and psychotherapy appear to produce a similar magnitude of effect. Approximately 4 patients must be treated to produce 1 additional clinical remission. Cognitive-behavioral therapy appears to reduce depression scores in children and adolescents. Data on pharmacotherapy for children and adolescents are mixed: tricyclic agents appear ineffective, and data for selective serotonin reuptake inhibitor drugs are inconclusive but promising.

Feedback of screening results to providers appears to increase recognition of depression in adults compared with usual care but its effect on treatment and clinical outcomes are mixed. Screening appears to be more effective when coupled with systematic efforts to ensure adequate treatment and follow-up. The effect of screening has not been evaluated in children.

Other than medication side effects, little evidence is available about the potential harms of screening and treatment of depression.

Accurate screening tests and effective therapies for depression are available. Screening for depression can improve outcomes compared with usual care in adults, particularly when coupled with efforts to ensure adequate treatment and follow-up. The effect of screening in children and adolescents is unknown.