In what clinical contexts is the PHQ-9 preferable to the EPDS, despite the EPDS having stronger perinatal-specific evidence?

The PHQ-9 is preferable when: (1) the practice uses it across all adult populations and EHR integration and staff familiarity are already in place; (2) the clinician needs to assess functional impairment domains (PHQ-9 item 9 criterion set maps directly to DSM-5 MDD criteria, which can be useful for documentation and collaborative care communication); (3) the patient population is more diverse in the perinatal versus general mental health mix and a single standardized tool reduces training burden. The PHQ-9 also has a validated 2-item screener (PHQ-2) that is useful for ultra-brief initial screening at busy OB visits, with a full PHQ-9 administered for any positive PHQ-2 result. The EPDS does not have an equivalent validated abbreviated version.

Can the GAD-7 replace the EPDS anxiety subscale for perinatal anxiety screening, or should both be used?

For practices specifically concerned about perinatal anxiety as a primary presentation, the GAD-7 provides more granular anxiety symptom assessment than the EPDS anxiety subscale (3 items). However, the GAD-7 does not screen for depression and is not validated as a standalone perinatal tool to the same degree as the EPDS. Practices seeing high rates of perinatal anxiety without comorbid depression can add the GAD-7 as a companion screen to the EPDS, using the EPDS anxiety subscale as a pre-screen and the full GAD-7 for patients who score 4 or above on the subscale. This approach adds minimal time while improving sensitivity for anxiety disorder presentations. Running both tools at every visit is logistically unnecessary for most practice volumes.

EPDS vs PHQ-9 vs GAD-7: Perinatal Screening Tool Comparison

Perinatal mood and anxiety disorders affect approximately 20 percent of pregnant and postpartum women. Three validated instruments dominate clinical practice: the Edinburgh Postnatal Depression Scale (EPDS), the Patient Health Questionnaire-9 (PHQ-9), and the Generalized Anxiety Disorder-7 (GAD-7). Each has distinct strengths, limitations, and practice contexts where it performs best.

Choosing the wrong tool does not necessarily produce invalid results -- but it can produce misleading ones, particularly in postpartum populations where somatic items overlap with normal new-parent physiology.

---

Side-by-Side Comparison

| Feature | EPDS | PHQ-9 | GAD-7 | |---|---|---|---| | Items | 10 | 9 | 7 | | Constructs screened | Depression, anxiety (partial) | Depression | Generalized anxiety | | Time to complete | 3 to 5 min | 2 to 3 min | 2 to 3 min | | Positive screen threshold | ≥10 | ≥10 | ≥10 | | Probable major depression threshold | ≥13 | ≥15 | N/A | | Somatic item confound | Minimal (designed for peripartum) | Moderate (sleep, appetite) | Low | | Validated for pregnancy | Yes | Yes (less robust) | Yes | | Validated for postpartum | Yes (primary validation) | Yes | Yes | | Validated for fathers | Yes | Yes | Yes | | Screens for suicidal ideation | Yes (item 10) | Yes (item 9) | No | | Anxiety subscale | Partial (items 3, 4, 5) | No | Primary purpose | | OB guideline endorsement (ACOG) | Yes | Yes (alternate) | Supplemental | | AAP endorsement | Yes (preferred) | Yes (alternate) | Supplemental | | Public domain | Yes | Yes | Yes |

---

Edinburgh Postnatal Depression Scale (EPDS)

Why the EPDS performs better in postpartum settings

The EPDS was designed in 1987 specifically for postpartum screening. Its key advantage over general depression instruments is the deliberate exclusion of somatic items that are normal in the peripartum period. It does not ask about sleep, appetite, weight, or fatigue -- all of which are confounded in the postpartum period by infant feeding and care demands.

The 10 items focus on mood, anhedonia, anxiety, self-harm ideation, and cognitive symptoms. This specificity produces better sensitivity and specificity in postpartum samples compared with general instruments.

EPDS psychometric properties

Across multiple validation studies in perinatal populations:

Sensitivity at threshold ≥10: approximately 86 percent
Specificity at threshold ≥10: approximately 78 percent
Sensitivity at threshold ≥13: approximately 78 percent
Specificity at threshold ≥13: approximately 90 percent

(Levis et al., British Journal of Psychiatry, 2020 -- meta-analysis of 58 studies)

The EPDS performs best as a screening tool, not a diagnostic instrument. A score of 13 or above suggests probable major depression; it does not confirm it. Clinical interview remains the diagnostic standard.

Item 10 and suicidal ideation

Item 10 asks: "The thought of harming myself has occurred to me." Response options are:

0: Never
1: Hardly ever
2: Sometimes
3: Yes, quite often

Any response above 0 requires direct clinical assessment, regardless of total score. Do not interpret item 10 in the context of total score alone.

Limitations

The EPDS does not reliably screen for postpartum OCD, PTSD, or bipolar disorder. A patient with primarily intrusive thoughts, checking behaviors, and fear of harming the baby may score below 10 on the EPDS while meeting full criteria for OCD. The EPDS anxiety subscale (items 3 through 5 scored 6 or above) has been studied as an adjunct anxiety screen, but this application has less validation than the full-scale depression application.

---

PHQ-9

When the PHQ-9 is appropriate in perinatal settings

The PHQ-9 performs well in perinatal screening with one important caveat: item 3 ("Trouble falling or staying asleep, or sleeping too much") consistently produces false elevation in postpartum samples because sleep disruption is nearly universal in the early postpartum period.

This item contributes up to 3 points to the total score. A patient with a PHQ-9 score of 10 driven largely by sleep and fatigue items (3, 4, 7) may not be depressed -- she may be caring for a newborn. Clinical interpretation must account for this.

Adjustments:

Some clinicians score the PHQ-9 with item 3 excluded and use a modified threshold.
Alternatively, directly assess whether sleep disruption is infant-related (can return to sleep after infant's needs are met) vs. neurovegetative (lying awake despite infant sleeping).

When to prefer PHQ-9 over EPDS

Your practice already uses PHQ-9 for all adult patients and EHR workflows are built around it
Screening a mixed population (perinatal + non-perinatal) and need a single tool
Later postpartum period (6 months and beyond), where newborn sleep disruption is less severe
Research or registry contexts that require PHQ-9 for comparability with non-perinatal datasets

PHQ-9 clinical thresholds

| Score | Category | Perinatal clinical action | |---|---|---| | 1 to 4 | Minimal | Monitor | | 5 to 9 | Mild | Psychoeducation, rescreening in 4 weeks | | 10 to 14 | Moderate | Referral for therapy | | 15 to 19 | Moderately severe | Referral + medication consultation | | 20 to 27 | Severe | Urgent referral; assess for higher level of care |

---

GAD-7

The case for routine anxiety screening

Anxiety disorders are the most common psychiatric presentation in the perinatal period. GAD, panic disorder, health anxiety, OCD, and specific phobias all have elevated prevalence in pregnancy and postpartum. Many clinicians screen primarily for depression and miss the anxiety presentations that significantly impair functioning and carry their own risks, including preterm birth, low birth weight, and impaired maternal-infant bonding.

A 2019 systematic review and meta-analysis in JAMA Psychiatry (Fawcett et al.) found perinatal anxiety prevalence of approximately 15 percent in pregnancy and 10 percent postpartum -- comparable to or exceeding depression rates.

The GAD-7 is not perinatal-specific, but it has been validated in perinatal populations and performs adequately for generalized anxiety screening.

Limitations in perinatal anxiety identification

The GAD-7 screens for generalized anxiety. It captures diffuse worry and tension well, but does not distinguish GAD from:

Postpartum OCD (ego-dystonic intrusive thoughts + compulsions)
Health anxiety focused on the infant or pregnancy complications
Birth trauma and PTSD
Panic disorder

Patients presenting with intrusive thoughts, rituals, or hypervigilance focused on infant safety should receive targeted clinical interview regardless of GAD-7 score.

Practical use: pairing GAD-7 with EPDS

The EPDS and GAD-7 take approximately 5 to 8 minutes combined and together provide coverage of depression, partial anxiety, and suicidal ideation (EPDS) plus a fuller anxiety picture (GAD-7). This combination is particularly useful in:

Prenatal intake assessments
NICU settings where anxiety presentations are common
Fertility clinic contexts where patients have high pre-existing anxiety
Any patient with a history of anxiety disorder

---

Special Contexts: Which Tool to Use

First trimester prenatal visit

EPDS (sensitivity for depression in pregnancy is well-validated) plus GAD-7 if anxiety history or current presentation suggests it. Establish a baseline score early.

Third trimester

EPDS again. Scores in the third trimester predict postpartum PMAD with moderate sensitivity. Patients with third-trimester EPDS of 10 or above should be flagged for close postpartum follow-up regardless of symptomatic presentation at delivery.

Postpartum OB visit (4 to 6 weeks)

EPDS. Consider supplementing with a brief inquiry about intrusive thoughts if not covered by EPDS score. "Some new parents have bothersome or frightening thoughts they don't want to have. Have you experienced anything like that?"

Pediatric well visits (1, 2, 4, 6 months)

EPDS per AAP recommendation. The EPDS is the preferred tool for maternal screening in pediatric settings and is referenced specifically in the AAP Policy Statement on Incorporating Recognition and Management of Perinatal Depression into Pediatric Practice.

NICU

EPDS plus GAD-7 given the dual burden of depression and anxiety. Begin screening within 2 weeks of admission. Edinburgh scores in NICU populations are often in the moderate-to-severe range -- thresholds and referral workflows should be calibrated accordingly.

---

Screening Is Not Diagnosis

Regardless of tool, a positive screen requires clinical follow-up. The score informs urgency and disposition; it does not replace assessment. The diagnostic conversation should include:

Direct inquiry about the nature and onset of symptoms
Functional impairment assessment
Safety assessment (including thoughts of self-harm or harm to the infant)
Prior psychiatric history
Current support and stressors
Breastfeeding status (relevant to pharmacological options if medications are being considered)

For patients with a positive screen who need referral, the next step is a formal diagnostic assessment and treatment planning with a perinatal mental health specialist. Understanding what happens in that referral process helps providers communicate clearly to patients about what to expect.

For postpartum depression, the EPDS at cutoff 10 demonstrates sensitivity of approximately 86% and specificity of 78% in the original Cox et al. validation and comparable values in subsequent perinatal-population studies. The PHQ-9 at cutoff 10 shows sensitivity of 75 to 89% and specificity of 79 to 91% in perinatal populations, depending on the study population and comparator. The GAD-7 at cutoff 10 for GAD shows sensitivity of 89% and specificity of 82% in the general adult population, with limited perinatal-specific validation data. The EPDS includes an anxiety subscale (items 3, 4, 5) that provides some GAD screening utility. For practices choosing a single tool, the EPDS has the most robust perinatal-specific validation and covers both depression and anxiety dimensions.

EPDS vs. PHQ-9 vs. GAD-7: Which Screening Tool for Which Perinatal Setting?