Secondary Logo

Share this article on:

Universal Perinatal Depression Screening in an Academic Medical Center

Gordon, Trent E. J. MS1; Cardone, Ida A. PhD1,3; Kim, Jennifer J. PhD1; Gordon, Scott M. MD2; Silver, Richard K. MD1

doi: 10.1097/01.AOG.0000194080.18261.92
Original Research

OBJECTIVE: To develop a department-based program to identify and treat women at risk for perinatal depression.

METHODS: Private and employed physician groups were engaged to conduct antepartum maternal depression screening using the Edinburgh Postnatal Depression Scale. A comprehensive program was established to ensure that patients identified as being at risk would receive appropriate care. The program 1) developed a network of existing community mental health providers to accommodate screen-positive referrals, 2) created a 24/7 hotline staffed by mental health workers to respond to urgent/emergent patient needs, 3) provided nursing and physician education via a comprehensive curriculum on perinatal depression, and 4) facilitated outpatient depression screening that included a centralized scoring and referral system.

RESULTS: A total of 4,322 women completed 4,558 screens during the initial 24 months (June 2003–May 2005). Although initial uptake of the screening program was gradual, all 20 departmental obstetric practices were screening their patients at the end of the first year. Depression screening was accomplished between 28–32 weeks of gestation, and postpartum screening (during the 6-week postpartum visit) was subsequently added. Overall, 11.1% of women screened positive in the antenatal period, and 7.3% screened positive in the postnatal period. Three hundred three women were referred for evaluation and care.

CONCLUSION: Department-based, perinatal depression screening was feasible when individual physician practices were not required to develop the infrastructure necessary to respond to at-risk patients. We believe that the provision of clinical safety nets (mental health provider network and the hotline) were essential to the universal acceptance of this program by practitioners.


Department-based universal screening for perinatal depression was introduced in the context of a comprehensive program that provides for triage and referral of at-risk patients.

From Evanston Northwestern Healthcare, the Jennifer Mudd Houghtaling Postpartum Depression Program, Departments of 1Obstetrics & Gynecology, 2Psychiatry, and 3Pediatrics, Northwestern University’s Feinberg School of Medicine, Evanston, Illinois.

The authors thank Dr. Charles Mudd and Joan Mudd for their financial support of the program.

Corresponding author: Trent E. J. Gordon, MS, Department of Obstetrics & Gynecology, 2650 Ridge Avenue, Evanston, IL 60201; e-mail:

Pregnancy and the postpartum period include multiple interactions between a patient and her caregiver. These contact points represent excellent opportunities to engage women in a myriad of health issues related to their pregnancy, including perinatal depression, which affects 13–18% of all pregnant women.1–3 However, discussion surrounding a patient’s mental well-being rarely occurs because of several barriers that exist. For example, obstetricians, family medicine physicians, and certified nurse midwives may not feel comfortable discussing a patient’s mental health. They may lack formal training in this area or might fear what the patient may disclose. They may also worry that their practice is not positioned to easily respond with a prompt mental health referral if a need is identified. Alternatively, a patient may not initiate such discussions because of the stigma that surrounds mental illness. In addition, the litany of routine screening tests and office activities required in the course of prenatal care is daunting and leaves little room for additional activities. Thus, in the usual clinical setting, a patient’s emotional state may never be discussed, leaving depressed patients in an undiscovered and, perhaps, intensified state, at risk of hurting either themselves or their children if they remain unidentified.

In response to these challenges, we convened a multidisciplinary steering committee that included obstetricians, psychiatrists, psychologists, nurses, administrators, and technical communication specialists to develop a comprehensive program that would require limited direct initiative from individual department faculty. The principal goal was to identify perinatal depression early in its clinical course and meet this problem head-on with an integrated system of assessment and intervention. This was accomplished first through education, providing heightened awareness of this illness for patients and caregivers alike. Next, a network of mental health providers was organized and a functioning hotline was established, 2 initiatives that set the stage for universal patient screening to be instituted. Screening would itself help to destigmatize this condition and send a clear message that health professionals care about their patients’ psychological wellness. Incorporating screening into physician practices gave providers some ownership of the process and emphasized their stake in the emotional well-being of their patients.

Back to Top | Article Outline


This study was considered exempt from review by the Institutional Review Board of Evanston Northwestern Healthcare. Patients included in the study were those who delivered at an Evanston Northwestern Healthcare facility during the first 2 years (June 2003–May 2005) after screening was implemented. The order of programmatic implementation was intentional and included the following steps accomplished in a serial, albeit overlapping, fashion: 1) physician and nursing education, 2) development of a mental health provider network, 3) establishment of the 24/7 depression hotline, and finally, 4) initiation of antepartum depression screening. Providers knew that before their first patient completed the Edinburgh Postnatal Depression Scale (EPDS), an integrated system was in place to meet the needs of screen-positive women.

Because of the department’s professional staff composition (predominantly private practice groups) members could not be mandated to administer the screens. Therefore, the program had to earn the trust of obstetric caregivers and motivate them to voluntarily screen their patients. First, a structured primer on perinatal mood disorders was given to each office (physicians, nurses, and allied staff with direct patient care contact) by the program’s medical director, a psychiatrist. Pertinent research on the history of screening was shared to establish its validity, and education was provided about the various pharmacological and psychotherapeutic treatment alternatives, including a focus on safety data in pregnancy and during lactation.

The medical director also met with each practice individually to discuss the virtues of screening. In going to the practitioners’ offices, he was able to establish personal contact with those nursing and allied staff members who would be directly involved with administering the EPDS. Important in this process was obtaining a commitment from a designated point person, an office liaison, whom the program could rely on for troubleshooting going forward. These visits also allowed for potential barriers to compliance to be explored, such as limitations in office technology or predominance of non-English–speaking patients. In addition, clearly delineated contingency plans were established for hypothetical patients who staff felt were unsafe because of suicidal ideation.

A select group of mental health workers was recruited to staff a 24/7 crisis hotline. A written manual was created and reviewed with this group to standardize the phone-based assessments and intervention options to be used. Training sessions for the hotline workers were provided by both the program and medical directors, including mock scripting for a variety of caller situations that might be encountered. Available technology allowed for automatic forwarding of each call to personal cell phones used by hotline workers so that they were able to immediately answer calls without automated triage, thus avoiding potential gaps in service at change of shift. The hotline number was well publicized and made available repeatedly to the patient and her family at multiple points in her care.

An inventory of all available outpatient mental health services was constructed within the ZIP code–based catchment area of patients from our department. Direct contact was made with each provider or group to ascertain their expertise in the area of perinatal depression and their interest in the program. In return for their willingness to participate, continuing education on perinatal depression was provided through biannual network meetings. To be included in the network, providers (social workers, psychologists, and psychiatrists) made a commitment to see a referred patient urgently (within 48 hours) on either a sliding scale of charges or at no cost to the patient for at least the initial visit. The membership of this network was continuously updated and expanded to cover the appropriate counties of patient residence, and the hotline workers received monthly electronic updates of this roster.

Choosing an instrument for perinatal depression screening was not straightforward to the extent that a gold standard for this purpose did not exist. An initial literature review identified multiple alternatives, including the Edinburgh Postnatal Depression Scale (EPDS),2 the HANDS Depression Screening Tool,15 the Center for Epidemiologic Studies Depression Scale (CES-D),5 and the Postpartum Depression Screening Scale.6 Data on all scales were examined relative to appropriateness for prenatal use as that was our primary screening time frame. Other selection factors included ease of administration and scoring, predictive value (specificity and sensitivity), utility for partners as well as patients, associated proprietary costs, content validity, and any preexisting published experience. Ultimately the EPDS was chosen, followed coincidentally by its endorsement by our state’s public health department (Table 1).

Table 1

Table 1

We intentionally began with prenatal screening to allow for preemptive interventions to be instituted before delivery in at-risk patients. The notion of prenatal screening is supported by ample data suggesting the potential onset of depression during pregnancy.3,7–11 The increased frequency of office visits during the third trimester provided multiple opportunities to both conduct EPDS and provide follow-up for positive screens in a familiar setting. To standardize the timing of administration, offices were asked to link the prenatal EPDS to diabetic screening, which usually takes place by 28 weeks of gestation. When the second screen was added to the program, the 6-week postpartum appointment time was chosen for reasons of convenience combined with the anticipated peak incidence of postpartum depression occurring in this time frame.

Although copies of the EPDS are readily available online via a public web site (,12 we chose to provided all practices with electronic copies of English and Spanish versions to facilitate their ease of duplication and administration. Practices could request screens be translated into other languages, which the program accomplished on their behalf. Completed screens were submitted same-day via the program’s centralized e-fax number and entered into the department’s obstetric database, which would ultimately include delivery data and then be supplemented with the postpartum EPDS entry.

The range of scores possible on the EPDS is 0–30, with a score of 0 representing no evaluation of depression and a score of 30 consistent with a major risk for depression. Although a cutoff score of 10 or greater is thought to be optimal for use in the research settings and is likely to minimize false-negative assessments,1 our initial experience with this breakpoint resulted in an inordinately large number of patients in our population who presented with no suggestion of depressive symptoms determined at the time of follow-up assessment, such that referral was not needed. We therefore adopted a more conservative cutoff (≥ 12), consistent with EPDS use as a screening tool in primary care settings.2 Published data suggested that this breakpoint produces a sensitivity of 67–95% and a specificity of 78–96%.2,13,14 Screens with a score of 12 or greater or surveys endorsing any response other than “never” to the question describing thoughts of harming oneself were given to the program’s primary director (psychologist), medical director (psychiatrist), or coordinator (psychologist) for immediate patient phone contact, evaluation, and triage. At the conclusion of the call, the mental health professional obtained the patient’s verbal consent to contact the clinician who was providing her obstetric care. Feedback to the provider consisted of a faxed communication outlining the patient’s EPDS score, the date of phone contact, and the interventions/referrals made. For this study, categorical and continuous variables were compared using χ2 and t tests, respectively, with P < .05 considered as significant.

Back to Top | Article Outline


Within the first year of the program, universal participation was achieved from all 20 clinical practices associated with the department. Uptake and uniform application were gradual, and not all practices initially agreed to screening. However, the available safety nets, combined with positive program experiences publicized among their peers, resulted in eventual commitment from the more reluctant practices. During these first 2 years (June 2003–May 2005), 4,558 EPDS screens were completed by 4,322 women. During this same time period there were 10,108 deliveries at the department’s 2 obstetric campuses. Nine hundred thirty-three of these mothers were either transfers from other institutions or did not receive prenatal care at a department-affiliated office. An uncertain number of additional women were beyond 28 weeks at the program’s onset, then delivered during the study interval or were screened antenatally but had not yet delivered. Thus, the 4,322 women who completed screening represented about half of the eligible patients, leaving an opportunity for more complete ascertainment in the future. A more accurate measure of compliance during the program was the rate of screening as a percentage of delivered patients per month (Fig. 1), which was 63.7% in the final month of the period of study and 67.0% at the time of this article’s submission.



A total of 4,038 antenatal and 520 postpartum screens were processed, with 236 of the 520 representing repeat screens for patients screened antenatally as well. Postpartum screening was initiated only during the last 3 months of the study time frame. Of the 4,038 antenatal screens, 354 (8.8%) scored 12 or higher. There were an additional 95 women who endorsed an answer other than “never” in response to suicidal ideation, making the total screen-positive rate of 449/4,038 or 11.1%. In contrast, only 38 of the 520 postpartum screens (7.3%) were positive. All screen-positive women received a call from either the program director, the coordinator, or a hotline worker, during which evaluation and triage were accomplished. Three hundred three patients accepted referrals for evaluation and care.

The subset of women who were screened both antenatally and in the postpartum period allowed for a preliminary analysis of screening results before and after delivery (Table 2). A majority of women (88.6%) screened negative both times, consistent with the overall rate of negative screening in the study population. Of potential interest were the 14 women (5.9%) who screened positive in the antenatal period but were negative in the postpartum period, perhaps in relationship to their evaluation and/or an intervention they received before delivery (Table 3).

Table 2

Table 2

Table 3

Table 3

The crisis hotline received 514 calls from patients, their friends, or their family members. Triage for urgent/emergent care was accomplished 328 times, while 82 calls (16%) were from family members or professionals and referrals were not warranted. Fifty-seven patients (11%) were already seeing a mental health professional and did not require a referral. These callers often were using the hotline because their counselor was not readily available. Twenty-six callers (5%) declined referrals; many of these women were approaching this phone contact as a one-time support encounter and were not seeking additional attention beyond the call. Twenty-one callers (4%) were from out of state and could not be given specific local referrals.

Back to Top | Article Outline


It has been suggested that 50% of women with symptoms of perinatal depression remain undetected by medical providers, despite the availability of easily administered and clinically validated screening tools.4 The program described herein was designed to solve that problem and our data suggest that we have made headway toward that goal but that more work is required. Successes to date include universal physician practice participation, a steadily growing percentage of screened women, appropriate follow-up and triage for those who screen positive, significant use of the crisis hotline, and appropriate referral of at-risk patients to community mental health providers. Creating a "chaperoned program" that places the screening activity in provider offices with all other functions centralized appears to strike the appropriate balance between practice autonomy and achievement of programmatic goals. Physician practices are inclined to administer the screen with the knowledge that they have organized resources and a team of mental health experts behind them.

However, the degree of participation and associated favorable statistics should not be the only measure of programmatic success. To truly determine efficacy, additional qualitative research will be required. Examples of questions to be answered include an exploration of patient perceptions of screening, the degree to which call-backs for positive screens are truly helpful, the frequency with which patients follow through with their referrals, the effectiveness of those referrals, and the ultimate impact on the incidence and severity of perinatal depression among our patients.

Further modifications and improvements to the current program are also being considered. We wish to involve partners in screening so as not to ignore the impact they may have on their family’s mental health and vice versa. Recent data suggest that paternal depression can negatively influence infant development, such that its discovery and treatment is also warranted.18 To this end, we are considering hospital-based EPDS screening of partners immediately after delivery, because they are most likely to be accessible in this time frame. Adding a third maternal screen at 12 weeks postpartum, (via pediatric practices) is another initiative under consideration because it would extend the safety net for detection to better cover the known range of onset of postpartum maternal depressive illness.19 This modification will require repeating the educational offerings tailored to the offices of our pediatric colleagues and, more importantly, will necessitate a mind-set shift on the part of pediatricians who do not necessarily view their clinical role as including direct maternal care.

To the extent that additional research on the program confirms its efficacy, expansion to other departments is a future goal. Because our hotline has sufficient capacity for greater call volume, it may be appropriate to extend its apparatus to other departments in our geography, thereby reducing their start-up time by making available this critical prerequisite. In a consulting role, the program’s leadership could certainly provide the guidance and problem-solving skills necessary for an efficient implementation of universal screening in other departments.

Finally, funding of this program can be problematic because charging patients or their insurers for screening services, while permissible, creates a de facto barrier to access that we wish to avoid if at all possible. Our solution has been to rely on philanthropy combined with intramural and external grant support to cover start-up costs. Once the initial impact of the program was quantified and presented to our hospital system, its perceived value to the community prompted our administration to provide an extended commitment to cover personnel costs associated with program and maintenance of the hotline.

Back to Top | Article Outline


1. Cox J, Holden J. Perinatal mental health: a guide to the Edinburgh Postnatal Depression Scale. London (UK): The Royal College of Psychiatrists; 2003.
2. Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression: development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry 1987;150:782–6.
3. Evans J, Heron J, Francomb H, Oke S, Golding J. Cohort study of depressed mood during pregnancy and after childbirth. BMJ 2001;323:257–60.
4. Chaudron LH. Postpartum depression: what pediatricians need to know. Pediatr Rev 2003;24:154–61.
5. Radloff LS. The CES-D Scale: a self-report depression scale for research in the general population. Appl Psychol Meas 1977;1:385–401.
6. Beck CT, Gable RK. Postpartum Depression Screening Scale: development and psychometric testing. Nurs Res 2000;49:272–82.
7. Buesching DP, Glasser ML, Frate DA. Progression of depression in the prenatal and postpartum periods. Women Health 1986;11:61–78.
8. Gotlib IH, Whiffen VE, Mount JH, Milne K, Cordy NI. Prevalence rates and demographic characteristics associated with depression in pregnancy and the postpartum. J Consult Clin Psych 1989;57:269–74.
9. O’Hara MW. Social support, life events, and depression during pregnancy and the puerperium. Arch Gen Psychiatry 1986;43:569–73.
10. Stowe ZN, Hostetter AL, Newport DJ. The onset of postpartum depression: implications for clinical screening in obstetrical and primary care. Am J Obstet Gynecol 2005;192:522–6.
11. Watson JP, Elliott SA, Rugg AJ, Brough DI. Psychiatric disorder in pregnancy and the first postnatal year. Br J Psychiatry 1984;144:453–62.
12. The National Health Service. Edinburgh Postpartum Depression Scale. Available at: Retrieved October 27, 2005.
13. Harris B, Huckle P, Thomas R, Johns S, Fung H. The use of rating scales to identify post-natal depression. Br J Psychiatry 1989;154:813–7.
14. Murray L, Carothers AD. The validation of the Edinburgh Postnatal Depression Scale in a community sample. Br J Psychiatry 1990;157:288–90.
15. Baer L, Jacobs DG, Meszler-Reizes J, Blais M, Fava M, Kessler R, et al. Development of a brief screening instrument: the HANDS. Psychother Psychosom 2000;69:35–41.
16. Beck CT, Gable RK. Further validation of the Postpartum Depression Screening Scale. Nurs Res 2001;50:155–64.
17. Boyd JK, Weissman MM, Thompson WD, Myers JK. Screening for depression in a community sample: understanding the discrepancies between depression symptom and diagnostic scales. Arch Gen Psychiatry 1982;39:1195–200.
18. Ramchandari P, Stein A, Evans J, O’Connor TG, ALSPAC study team. Paternal depression in the postnatal period and child development: a prospective population study. Lancet 2005;365:2201–5.
19. Chaudron LH, Szilagyi PG, Kitzman HJ, Wadkins HIM, Conwell Y. Detection of postpartum depressive symptoms by screening at well-child visits. Pediatrics 2004;113:551–8.
© 2006 The American College of Obstetricians and Gynecologists