Validity of the Edinburgh Postnatal Depression Scale for screening pregnant and postpartum adolescents: a systematic review

Main Article Content

Fernanda Barassi
Annmarie Grealish

Keywords

Adolescent mothers, antenatal depression, postpartum depression, psychometric properties, rating scales

Abstract

Objective: To examine the validity of the Edinburgh Postnatal Depression Scale (EPDS) for screening depression in pregnant and postpartum adolescents. Background: The incidence of postpartum depression (PPD) in 15 to 19-year olds is double the rate reported among mothers older than 25 years. EPDS threshold scores that indicate possible depression among adolescents have not been established and may differ from those validated for adults. Study design and methods: A systematic review of the literature between 1987 and 2020 was conducted using five databases: CINAHL, EMBASE, MEDLINE, MIDIRS and PsycInfo. Studies that sampled adolescent mothers in the perinatal period, that screened for depressive symptoms using the EPDS, and which assessed the validity of the tool were included. The studies were grouped according to their methodology with results presented as a narrative synthesis. Two researchers independently reviewed search results, study selection and data extraction, and undertook quality appraisals using the Quality Assessment of Diagnostic Accuracy Studies checklist. Results: Five studies that sampled a total of 1,241 participants were included in the review: four validated the EPDS against diagnostic reference standards and one against other depression screening tools. The EPDS demonstrated high levels of sensitivity and specificity although optimal cut-off scores for possible depression were between 2 and 7 points lower than that recommended for adult samples. Overall performance of the EPDS was equivalent or better when compared to other screening tools. Discussion: The standard EPDS cut-off score (≥12) does not identify all adolescents at high risk of depression during the perinatal period. Scores of ≥9 may be more appropriate. However, different reference standards and sampling methods used in the studies compromise the review’s strength of evidence. Conclusion: Despite development of the EPDS more than 30 years ago, research into its validity among adolescents is still in its infancy. This review makes an important contribution to that body of evidence by revealing its limitations and highlighting trends upon which further research can build. Implications for research, policy and practice: Although Australian guidelines for perinatal depression screening are some of the most detailed in the world, limited guidance is offered for EPDS use among adolescents. The findings of this review raise important points for that guidance and for practitioners. Concerns for possible depression in adolescents should be triggered at an advisory EPDS cut-off score of ≥9. Further research is needed to confirm or refute these findings.


What is already known about the topic?



  • Postpartum depression (PPD) among adolescents is prevalent and a recognised public health concern due to its significant association with morbidity.

  • Perinatal screening for depression using validated instruments, including the Edinburgh Postnatal Depression Scale (EPDS), is recommended in national and international guidance.

  • There is no validated version of the EPDS for adolescents although there is evidence to suggest that cut-off scores indicative of possible depression may need to be lower than those recommended for adults.


What this paper adds:



  • The paper synthesises evidence that demonstrates the EPDS is a valid scale for perinatal screening of depression in adolescents.

  • It identifies lower cut-off scores for suspected depression among adolescents compared to adults, and also highlights score differences in the pre and postnatal periods.

  • The paper provides advisory screening guidance for practitioners and emphasises the importance of monitoring for depressive symptoms in perinatal adolescents at every healthcare encounter.

Abstract 244 | View PDF Downloads 110 Supplementary Data Downloads 85