In a 2010 study of 198 mothers attending well-child visits at an urban pediatric clinic, Linda Chaudron and colleagues at the University of Rochester ran a quiet experiment: give every mother three depression screeners and a structured psychiatric interview, then check whether the screens matched the diagnosis. Fifty-six percent of the mothers met criteria for major or minor depressive disorder. The Edinburgh Postnatal Depression Scale, the most widely used perinatal depression tool in American obstetrics, missed a substantial share of them at the standard cutoff of 10. The optimal cutoff in this sample, which was predominantly Black and low-income, was 9 for major depression and 7 for any depressive disorder (Chaudron et al., Pediatrics, 2010, PMID 20156899).
Read that again. The score your obstetrician or your baby's pediatrician circles on the EPDS, the score the front-desk nurse hands you on a clipboard at the six-week postpartum visit, is calibrated to a population that probably did not include you. The EPDS was developed in 1987 by John Cox and colleagues using a sample of white women in Edinburgh and Livingston, Scotland. The cutoff of 10, sometimes 13 for probable major depression, traces back to that original sample. It has been validated in dozens of populations since then, but validated does not mean equivalent. The same number on the same form can mean different things in different bodies.
S. Darius Tandon's team took this seriously. In a comparison of three screeners administered to low-income African American women enrolled in home visiting programs, the EPDS at the standard cutoff of 13 caught 81.5 percent of women who met DSM criteria for major depression on structured interview. Drop the cutoff to 11 and sensitivity climbed to 88.9 percent without crashing specificity (Tandon et al., Journal of Affective Disorders, 2012, PMC3789596). Almost one in five depressed Black mothers in that sample would have walked out of a screening encounter at the standard cutoff with a normal score and no follow-up.
Patricia Lee King, in a 2012 factor-analytic study of the EPDS in 169 postpartum African American women with low socioeconomic status, found that the instrument's underlying symptom structure did not map cleanly onto the depression-only frame the EPDS was built around. The three-factor model that fit best separated depression, anxiety, and anhedonia as distinct symptom clusters, and the author closed with an explicit warning: cautious use of the EPDS among perinatal women of racially, ethnically, and socioeconomically diverse backgrounds until its factorial invariance is better understood (Lee King, Archives of Women's Mental Health, 2012, PMID 22297555). That sentence, buried in a methods journal, is the entire problem in one line. The tool does not behave the same way across groups, and we are using it as if it does.
The PHQ-9, the other workhorse, has its own measurement-bias literature in Black adults. The factor structure shifts. Somatic items (sleep, appetite, energy) get endorsed at rates that reflect everything from shift work to chronic pain to caring for other people's kids, and they pull total scores up or down in ways that have nothing to do with mood. The instruments were not designed to be wrong. They were designed on convenience samples that did not include enough of us to surface the drift.
None of this is hypothetical anymore. In June 2023 the American College of Obstetricians and Gynecologists released its first Clinical Practice Guideline on perinatal mental health, recommending that every pregnant and postpartum patient be screened for depression and anxiety using validated tools at the initial prenatal visit, later in pregnancy, and at postpartum visits, with the EPDS and PHQ-9 listed as recommended instruments (ACOG Clinical Practice Guideline No. 4, Obstetrics & Gynecology, June 2023, PMID 37486660). The American Academy of Pediatrics, through Bright Futures, recommends pediatricians screen the birth parent for depression at the 1-, 2-, 4-, and 6-month well-child visits. The U.S. Preventive Services Task Force gave perinatal depression preventive counseling a B recommendation in 2019.
Universal screening is the standard of care. That is real progress. It is also where the conversation usually ends, and where the harm hides. "We screened her" is not the same sentence as "we caught it." When the screen is calibrated for someone else's symptom profile and yours comes in under the line, you are documented as negative. The chart says screened. The follow-up never happens. The eight months you spend rage-crying in the pantry at 3 a.m., the months you cannot feel your baby's face when she smiles at you, those months unfold without anyone in the medical record believing anything is wrong. The data on Black maternal mental health bears this out at every step of the pipeline: Black birthing people are less likely than white birthing people to be diagnosed with postpartum depression at similar symptom levels, and less likely to receive treatment when diagnosed.
Some of that gap is provider bias, plain. Some of it is shorter visits, less continuity, more turnover, and clinics where the doctor does not know your name. Some of it is the well-documented Strong Black Woman script that teaches you not to answer question 10 honestly. (Question 10 is the suicide-ideation item. The thing you do not say out loud at a six-week check because your mother did not say it and her mother did not say it and you have a baby on your hip and you do not want anyone calling DCF.) And some of it, the part this piece is about, is upstream of all of that: the instrument itself.
You can be honest on the EPDS, score a 7, and still be sick. You can be a chronically high-functioning person who has learned to perform okayness in clinical settings, score a 4, and still be dissociating in the shower every morning. The cutoff was set in Scotland in 1987. The cutoff is not your symptoms. The cutoff is a number somebody picked.
Here is the practical part.
If your screen score came back "normal" and you are not okay, that result is not the final word. Tell the clinician, plainly: I scored low but I am not well. Ask for a referral to a perinatal mental health specialist regardless of the score. The ACOG 2023 guideline directs clinicians to use clinical judgment and patient-reported concerns alongside the screening number, not as a tiebreaker against it. Bring that up by name if you need to: ACOG Clinical Practice Guideline Number 4, June 2023.
Ask your OB and your baby's pediatrician these exact questions. What screening tool did you use, what was my score, and what is the cutoff this clinic uses? Do you ever refer based on patient-reported symptoms when the screen is below cutoff? If I want to see a perinatal psychiatrist or therapist, can you make that referral today rather than at the next visit? These are not rude questions. They are the questions an engaged patient asks, and a competent clinician will answer them without flinching.
Use the Postpartum Support International provider directory to find a perinatal mental health specialist. PSI is the professional society that maintains the Perinatal Mental Health Certification (PMH-C) credential, so a clinician listed there has at minimum committed to specialty training. The searchable directory is at postpartum.net/get-help/provider-directory. PSI also lists Black maternal mental health support groups and its Alliance for People of Color at postpartum.net/get-help.
Know the symptoms that are not on the EPDS and that require urgent attention. The EPDS asks ten questions and was built to detect depression and anxiety. It was not built to detect postpartum psychosis, which is a psychiatric emergency that affects roughly 1 to 2 of every 1,000 deliveries and presents with sudden onset (usually within the first two weeks postpartum) of confusion, paranoia, hallucinations, rapid mood swings, inability to sleep even when the baby is sleeping, or thoughts of harming yourself or the baby. Postpartum psychosis does not wait for the six-week visit. It is an emergency-department-tonight situation. If you or someone you love is in that state, do not wait for a screen.
The National Maternal Mental Health Hotline is 1-833-TLC-MAMA (1-833-852-6262), free, confidential, 24/7, in English and Spanish, by call or text. It is operated by the federal Health Resources and Services Administration and staffed by counselors trained specifically in perinatal mental health. Partners and family members can call too. The 988 Suicide and Crisis Lifeline (call or text 988, or chat at 988lifeline.org) is for any mental health crisis including suicidal ideation, and it works during pregnancy and postpartum the same way it works any other time. Both numbers are free and neither creates a record that follows you. Calling does not trigger a child welfare report by default; counselors are trained to assess safety without escalating reflexively.
If your insurance is the barrier, ask the clinic for a list of sliding-scale or Medicaid-accepting perinatal mental health providers in your area, and ask whether postpartum care is covered under your state's Medicaid postpartum extension. The majority of states have now extended Medicaid postpartum coverage from sixty days to twelve months. This is the single biggest policy lever for Black maternal mental health access in the last decade and a lot of people who qualify do not know they qualify. Your state Medicaid agency website will tell you what is covered.
The screening instruments will get better. The factor-structure work King flagged in 2012 is slowly being done. The cutoff numbers will eventually move. None of that helps the mother sitting in the postpartum clinic this week, who knows something is wrong, and whose chart says she screened negative. Until the tools catch up, the only person who knows for sure is you.
