Context:

Depression during pregnancy or the first year postpartum is impressively common and can have devastating consequences for the woman, her children, and other family members.

Objectives:

We systematically review the evidence on (1) the prevalence and incidence of perinatal depression, (2) the accuracy of different screening instruments, and (3) the effectiveness of interventions for women screened as high risk for perinatal depression

Data Sources:

MEDLINE, CINAHL, PsycINFO, Sociofile, and the Cochrane Library (1980 through March 2004); bibliographic hand searches; and experts.

Study Selection:

The English-language studies assessed women for major depression alone or for major or minor depression. Studies of the prevalence and incidence of depression and the accuracy of screening tools had to include diagnostic confirmation by a reference standard. Studies involving interventions required a comparison group. Two reviewers independently evaluated each abstract to determine inclusion by consensus.

Data Extraction:

A primary reviewer abstracted data on key variables from the articles directly into detailed evidence tables; a second reviewer confirmed accuracy.

Data Synthesis:

We conducted a meta-analysis of the prevalence and incidence estimates to compute combined estimates for particular periods and points in time. We also conducted meta-analyses of the sensitivity and specificity of different screening instruments. For screening outcome studies, we were only able to synthesize qualitatively.

Results:

We identified 30 studies of prevalence. For major depression alone, point prevalence estimates ranged from 3.1 percent to 4.9 percent at different times during pregnancy and 1.0 percent to 5.9 percent at different times during the first postpartum year. For major and minor depression, estimates of the point prevalence ranged from 8.5 percent to 11.0 percent during pregnancy and 6.5 percent to 12.9 percent during the first year postpartum. However, these prevalence estimates were not significantly different from those of similarly aged nonchildbearing women. Data on incidence were more limited.

We identified 10 studies of screening accuracy. One small study reported on accuracy during pregnancy. For postpartum depression, screeners appeared feasible, but the small number of depressed patients involved precluded identifying an optimal screener or threshold for screening. Screening instruments studied are generally good at identifying major depression alone, with accuracy consistent with reports from primary care settings, but they performed poorer for the major or minor depression category.

We found no studies directly testing whether screening improved outcomes. However, we identified 15 studies that used some sort of screening to identify women at risk of depression and for whom a subsequent intervention was provided. The results of four small studies of various psychosocial interventions during pregnancy did not demonstrate consistently superior outcomes. Results were also mixed for postpartum interventions. Six of nine studies of various psychosocial interventions reported significant improvement in depression for the experimental group. Two studies with pharmacologic interventions provided conflicting results.

Conclusions:

Although limited, the available research suggests that depression is one of the most common perinatal complications and that fairly accurate and feasible screening measures are available. Studies with larger sample sizes and a greater racial and ethnic mix are needed. Researchers also need to determine whether screening itself leads to better access to proven treatment and improved outcome relative to usual care.