Psychiatry Research 226 (2015) 113–119

Contents lists available at ScienceDirect

Psychiatry Research journal homepage: www.elsevier.com/locate/psychres

Combined use of the postpartum depression screening scale (PDSS) and Edinburgh postnatal depression scale (EPDS) to identify antenatal depression among Chinese pregnant women with obstetric complications Ying Zhao a,b, Irene Kane c, Jing Wang d,n, Beibei Shen d, Jianfeng Luo e, Shenxun Shi b,f,n a

School of Nursing, Fudan University, No. 305 Fenglin Road, Shanghai 200032, PR China Psychiatry Department, Fudan University Affiliated Huashan Hospital, No. 12 Wulumuqi Zhong Road, Shanghai 200040, PR China School of Nursing, University of Pittsburgh, 3500 Victoria St. VB 420, Pittsburgh, PA 15261, United States d Fudan University Affiliated Women's Hospital, No. 128 Shenyang Road, Shanghai 200090, PR China e Department of Biostatistics, School of Public Health, Fudan University, No. 130 Dongan Road, Shanghai 200032, PR China f Shanghai Jiao Tong University School of Medicine Affiliated Shanghai Mental Health Centre, No. 600 Wan Ping Nan Road, Shanghai 200030, PR China b c

art ic l e i nf o

a b s t r a c t

Article history: Received 25 June 2014 Received in revised form 4 December 2014 Accepted 12 December 2014 Available online 20 December 2014

The purpose of the present study was to evaluate antenatal depression screening employing two scales: the Postpartum Depression Screening Scale (PDSS) and Edinburgh Postnatal Depression Scale (EPDS) for the population of Chinese pregnant women with obstetric complications. A convenience sample of 842 Chinese pregnant women with complications participated in this study. The PDSS total score correlated strongly with the EPDS total score (r¼ 0.652, p¼0.000). Each tool performed extremely well for detecting major and major/minor depressions with PDSS resulting in a better psychometric performance than EPDS (po 0.01). If combined use, the recommended EPDS cut-off score was 8/9 for major depression, at which the sensitivity (71.6%) and specificity (87.6%) were the best, and the recommended PDSS cut-off score was 79/80 for major depression, along with its best sensitivity (86.4%) and specificity (100%). The study concluded that EPDS and PDSS appear to be reliable assessments for major and minor depression among the Chinese pregnant women with obstetric complications. Combined use of these tools should consider lower cutoff scores to reduce the misdiagnosis and improve the screening validity. & 2015 Elsevier Ireland Ltd. All rights reserved.

Keywords: Antenatal depression High risk Obstetric complications Screening Postpartum Depression Screening Scale Edinburgh Postnatal Depression Scale

1. Introduction Obstetric complications are common and highly distressing events (Ariadna et al., 2009). Women with obstetric complications reported significantly more negative experiences during their recent childbirth (Thornton et al., 2010; Blom et al., 2010;Gausia et al., 2012). Early recognition of depression in pregnant women can eliminate the length of time that these women have to suffer with debilitating perinatal depression and can decrease the potentially harmful effects on the infants involved (Neiman et al., 2010). Only a minority of pregnant women suffering from depression are identified by health care providers despite its importance. A major impediment to depression detection is the difficulty in the administration of depression screening tests in busy clinical settings. If there were simplified and appropriate screening instruments, the

n

Corresponding authors. Tel.: þ 86 13816203171. E-mail addresses: [email protected] (J. Wang), [email protected] (S. Shi). http://dx.doi.org/10.1016/j.psychres.2014.12.016 0165-1781/& 2015 Elsevier Ireland Ltd. All rights reserved.

obstetricians and obstetrical nurses would be able to identify women with obstetric complications who have depressive symptoms more efficiently and effectively (Choi et al., 2012). In both research and clinical practice, a positive screen for depression in the obstetrical patient should be followed by a confirmed clinical diagnosis through a structured diagnostic survey or semi-structured interview based on DSM disorders (SCID) criteria for major depression (Daniels, 2013; Tandon et al., 2012). However, Gjerdingen et al. (2011) found the SCID might be less convenient or comfortable for mothers because some individuals cannot be reached for an interview, resulting in missed opportunities for diagnosis, selection bias, and possible treatment disparities. In contrast, using a depression survey during a convenient perinatal visit to the obstetrician's office, though perhaps less accurate, could be easily administered, is more cost-effective, and more inclusive of screening all women. Fortunately, there are many different, easily administered depression screening tools to help identify those who are at greatest risk, according to SCID criteria, are widely accepted and have been developed to screen for postnatal and antenatal

114

Y. Zhao et al. / Psychiatry Research 226 (2015) 113–119

depression in the primary care setting (Boyd et al., 2005; Breedlove and Fryzelka, 2011), two widely utilized screening scales are the Edinburgh Postnatal Depression Scale (EPDS) (Cox et al., 1987) and the Postpartum Depression Screening Scale (PDSS) (Beck and Gable, 2000). Some studies have compared the psychometric properties of the EPDS and other screening scales. For example, Tandon et al., (2012) enrolled 32 pregnant women and 63 women with a child o6 months in home visitation programs. Each woman completed a structured clinical interview and three depression screening tools: the EPDS, Center for Epidemiologic Studies Depression Scale (CES-D), and Beck Depression Inventory II (BDI-II). The results indicated that each screening tool appear to be reliable and brief assessments of major and minor depression among low-income African American perinatal women. However, given that no women during pregnancy met criteria for minor depression, it was not possible to determine optimal prenatal cutoff scores. In addition, Zhong et al. (2014) recruited 1517 women receiving prenatal care to evaluate the psychometric properties of the Patient Health Questionnaire (PHQ-9) and EPDS, and found both of them are reliable and valid scales for antepartum depression assessment. But the study results could not determine the extent to which scores from each scales are predictive of adverse maternal and perinatal outcomes. There are few studies evaluating the comparability of the EPDS and PDSS, to date, in China, only one study (Li et al., 2011) screened 387 normal pregnant mothers within 12 weeks postpartum and compared the performance of the Chinese version of the EPDS and PDSS. Given the high prevalence of depressive symptoms among high-risk pregnant women and the lack of studies combined the EPDS and the PDSS during pregnancy, we conducted the present study to expand the body of research for perinatal depression screening by employing and analyzing the validity of the combined use of them, for antenatal depression screen in a population of Chinese pregnant women with obstetric complications. More specifically, if combined use, we wanted to determine the cutoff scores of the PDSS and the EPDS to screen for antenatal depression among this population. Previous research indicates that screening pregnant women for depression is clinically indicated; and, combining the EPDS and the PDSS might build upon the strengths of each scale to increase the validity of antenatal depression screening. Such information could encourage other researchers and clinicians to test the validity of the combined use of the different screening instruments to identify depression during the perinatal period in multiple obstetric settings. 2. Methods 2.1. Sample The convenience sample included 842 high-risk pregnant women in various gestational weeks during November 2013 to January 2014 at the antenatal department of the Fudan University affiliated Women's Hospital. 2.2. Data collection Convenience sampling included all of the women with obstetric complications who attended antenatal clinics within the data collection period noted above. The researcher approached the eligible women and invited them to participate in the study. All of the women who were approached were given a full explanation of the study and informed of their right to refuse to participate. All study participants who agreed to participate signed a study consent form prior to completing two depression screening scales (EPDS and PDSS) and while waiting for their routine antenatal check-up at the Fudan University affiliated Women's Hospital. The socio-demographic and obstetrical risk factors questionnaire included questions on age, parity, gestational weeks, highest level of education completed, occupation, past and/or current health problems, and past history of adverse

obstetrical outcomes such as abortion or abnormal delivery were extracted from the interviews and the hospital medical records.

2.3. Follow-up procedures During screening, whenever a mother's depression score was in the moderate to severe symptom range (i.e., EPDS Z 13 or PDSSZ 80), the researcher suggested the mother receive the further mental health evaluation such as SCID or MINI and subsequently encouraged the mother to contact other mental health clinicians. In addition, the researcher contacted all mothers whose scales scores placed them at a high risk for antenatal depression (i.e. EPDS score is 9–12 or PDSS score is 60–79) by: 1) a brief telephone, 2) reminded to adjust the mood during pregnancy; and, 3) follow up the antenatal depression screening actively at the next obstetric examination or with a mental health clinician referral.

2.4. Instruments 2.4.1. The Edinburgh Postnatal Depression Scale (EPDS) The EPDS, developed by Cox et al. (1987), is the most widely used measure of PPD symptoms and is commonly used as a screening tool for prenatal depression symptoms, as well (Gaynes et al., 2005). The EPDS does not directly correspond to DSM criteria. It excludes somatic depressive symptoms (appetite change and fatigue), as well as psychomotor agitation/retardation and diminished concentration (Cox et al., 1987). Participants base their answers on their experiences and feelings over the previous week. Each item is scored on a 4-point Likert scale from 0 to 3 with possible total scores ranging from 0–30. A higher score indicates higher reported frequency or severity of symptoms (Cox et al., 1987). Previous clinical research suggests that a cut-off for probable depression has been suggested at 12/13, and for possible depression at 9/10. However the cut-off points of 9/10 are used as markers of possible minor depression (Gibson et al., 2009) and scores4 12 are associated with a diagnosis of major depressive disorder (Cox et al., 1987). The Chinese vision of EPDS has good validity. The area under the curve was 0.91, suggesting excellent psychometric properties in screening for depressive illness (major and minor depression) at 6-week post-partum. At the conventional 12/13 cut-off, the sensitivity of the scale was only 0.41, with specificity of 0.95 (Lee et al.,1998). If being used to screen the antenatal depression in women of the third trimester of pregnancy, the optimal critical value was 9.5, and the sensitivity and specificity was 0.786 and 0.834, respectively (Guo et al., 2009). In the population, scoresZ 9 suggest risk for major or minor depression. A cutoff score of 13 or more indicates a positive screen for major depression. Cronbach's alpha coefficient of the EPDS in this study was 0.78.

2.4.2. The Postpartum Depression Screen Scale (PDSS) The PDSS is a 35-item self-report instrument that can be completed by the respondent in 5–10 min. All 35 items were derived from Beck's series of qualitative studies on this mood disorder. Although developed for postpartum depression, PDSS is accurate to screen for antenatal depression (Beck and Gable, 2002; Pereira et al., 2010). The total score indicates whether the woman needs to be referred for additional diagnostic evaluation (Pereira et al., 2011). The items represent dimensions include sleeping/eating disturbances, anxiety/ insecurity, emotional lability, mental confusion, loss of self, guilt/shame, and suicidal thoughts. For each item, the woman is asked to rate the feelings that she has experienced the last 2 weeks in a Likert scale from 1 (strongly disagree) to 5 (strongly agree). The total possible scores for the instrument range from 35 to 175 points (Beck and Gable, 2000). Higher scores on the PDSS indicate higher levels of PPD symptomatology. The total score indicates whether the woman needs to be referred for additional diagnostic evaluation (Beck and Gable, 2001). ScoresZ 60 suggest risk for major or minor depression. A cut-off score of 80 or more indicates a positive screen for major postpartum depression (Sit and Wisner, 2009; Beck and Tatano, 2008). Li et al., (2011) translated the PDSS from English into Chinese, reporting a Cronbach alpha of 0.96 and intraclass correlation coefficient of 0.79. Compared with other screening scales, PDSS is reported to have satisfactory reliability and validity (Zhang and Li, 2007; Yawn et al., 2009). In the present study, scoresZ 60 suggest risk for major or minor depression. A cut-off score of 80 or more indicates a positive screen for major postpartum depression, and, the Cronbach's alpha was 0.95, indicating that the test has good instrument internal reliability.

2.4.3. Diagnosis interview The Mini-International Neuropsychiatric Interview (M.I.N.I) is a short structured diagnostic interview, developed jointly by psychiatrists and clinicians in the United States and Europe, for DSM-IV and ICD-10 psychiatric disorders (Sheehan et al. 1998). The participants whose EPDSZ 9 or PDSS Z60 were interviewed using the Chinese version of MINI following informed consent at a scheduled appointment. The MINI was administered by the trained research assistant.

Y. Zhao et al. / Psychiatry Research 226 (2015) 113–119

2.5. Data analyses SPSS 19.0 for Windows was used for all statistical analyses. Descriptive statistics were used to describe demographics and depressive disorders prevalence. Internal consistency was measured with Cronbach's coefficient alpha. Receiver Operative Characteristics (ROC) analyses were performed to obtain PDSS and EPDS Areas Under the ROC Curve (AUCs) and to select cutoff scores, based on the best sensitivity and specificity combinations. Interpretation of AUCs as a summary statistic of the overall accuracy of a screening scale followed the guidelines proposed by Swets (1988): AUC ¼0.5, non-informative; 0.5 oAUC o 0.7, less accurate; 0.7 o AUCo 0.9, moderately accurate; 0.9 oAUC o 1, highly accurate. The closer the AUC is to 1.0, the more accurate the test. Chi-square tests were used to compare the AUCs of each screening tool; all study participants completed both of the screening tools and each participants' results were correlated.

2.6. Ethical considerations The study was approved by the Ethical Committee of the Fudan University affiliated Women's Hospital. Informed consent was obtained from the participants before screening was initiated. Appropriate medical (obstetric) and psychological (higher EPDS OR PDSS scores) services were offered and provided to the participants if indicated during the study duration.

3. Results

115

Table 1 Baseline characteristics of participants (N ¼ 842). Variables Age (mean 7S.D.) 20–26 years (n¼ 105) 27–34 years (n ¼605) 35 years or more (n¼ 132) Education Elementary school and lower (n¼ 24) Middle school (n ¼84) College and higher (n ¼734) Occupation Yes (n¼ 704) No (n¼ 138) Parity Primi (n¼ 720) Multi (n¼ 122) Gestational weeks r 12weeks (n¼ 57) 13–27weeks (n¼ 413) 28–34weeks (n¼ 334) Z 35 weeks (n¼ 38) Number of complications One (n¼ 573) Two (n ¼219) Three or more (n¼ 50)

30.5 7 3.7 105 (12.5%) 605 (71.8%) 132 (15.7%) 24 (2.8%) 84 (10.0%) 734 (87.2%) 704 (83.6%) 138 (16.4%) 720 (85.5%) 122 (14.5%) 57 (6.8%) 413 (49.0%) 334 (39.7%) 38 (4.5%) 573 (68.1%) 219 (26.0%) 50 (5.9%)

3.1. Sample characteristics The demographic characteristics of the participants are presented in Table 1. Eight hundred and forty two Chinese pregnant women with obstetric complications in their different gestational weeks (M ¼ 24.86, S.D. ¼7.75, range¼10–39 weeks of gestation) participated in the study. Among study participants, 57 (6.8%) were 12 weeks or less gestational age, 413 (49.0%) were between 13 and 27 weeks, 334 (39.7%) were between 28 and 36 weeks, and 38 (4.5%) were 37 weeks or more. The mean participant age was 30.5 years (S.D. ¼3.7; range¼21–43 years). Most were primiparas (85.8%), and the remaining were multiparas. Around 2.85% had primary and lower education, 9.98% had middle school education and 87.2% had college and higher education. The majority (83.6%) was still working. The majority of the participants (n ¼573, 68.1%) had just one complication, 219 (26.0%) had two complications, whereas only 50 (5.9%) experienced three or more complications. The obstetric complications conclude body mass index (BMI) abnormal, ageZ35years old, hypertension, diabetes mellitus, heart disease, thyroid disease, and so on. Table 2 shows all of the obstetric complications in this sample. Of these 842 Chinese pregnant women with obstetric complications, 81(9.6%) had EPDS Z13 or PDSS Z 80 in an abnormal psychological state (depression), 257 (30.5%) had EPDS score between 9 and 12 or PDSS score between 60 and 79 in borderline psychological state, 504 (59.9%) had EPDS score o 9 or PDSS o60 in normal psychological state. 3.2. Concurrent validity Convergent validity was assessed to determine the extent to which the PDSS score correlated with the EPDS score. The total PDSS score average was 56.177 16.55, while the total EPDS score average was 4.90 73.84. By calculating the Pearson's correlation coefficient, a significant positive correlation (r ¼0.652, p ¼0.000) was found between the PDSS and EPDS total scores. The findings are presented in Table 3. 3.3. ROC curves for screening tools The results of the ROC analysis are presented graphically in Figs. 1 and 2. For the full sample, each tool performed extremely

well for both major depression (Fig. 1) and major/minor depression (Fig. 2), with AUCs 40.80. For the PDSS and EPDS, the AUCs for major depression were of 0.983 (SE¼ 0.006; p ¼0.000, 95% CI 0.972–0.995) and of 0.898 (SE¼ 0.019; p ¼0.000, 95% CI 0.861–0.936), respectively, while the AUCs for major/minor depression were of 0.979 (SE¼ 0.005; p¼ 0.000, 95% CI 0.969–0.989) and of 0.822 (SE ¼0.016; p¼ 0.000, 95% CI 0.791–0.853), respectively. There were statistically significant differences in the AUCs for detecting antenatal depression (p o0.01), which indicates that the PDSS has a better psychometric performance for both major depression and major/minor depression than the EPDS.

3.4. The optimal overall cutoff scores for major depression (MDD) and major /minor depression (MDD/MnDD) Sensitivity and specificity were calculated for both the PDSS and the EPDS according to standard cutoff scores. For both of the screening tools, the cutoff scores which corresponded to the optional sensitivities and specificities for MDD/MnDD are presented in Tables 4 and 5, and the cutoff scores which corresponded to the optional sensitivities and specificities for MDD are presented in Tables 6 and 7. Beck and Gable (2001) recommend a higher cutoff score of 80 for MDD and a lower cutoff score of 60 for MDD/MnDD. In the present sample, based on the ROC curve for the PDSS, a cutoff score of 59.5 for MDD/MnDD produced sensitivity of 92.3%, specificity of 99.8%, a nearly similar cutoff score of 79.5 produced higher sensitivity (i.e., true positive rate) value of 86.4% and higher specificity value of 100%. Thus, for the PDSS, the optimal overall cutoff score was a little slightly lower for MDD and MDD/MnDD than the standard cutoff scores. Based on the ROC curve for the EPDS, a cutoff score of 8.5 produced sensitivity of 71.6% and specificity of 87.6%. Cox et al. (1987) report an optimal cutoff score of 12 or 13. In the present sample, however, measures of validity produced when using a cutoff score of 12.5 included lower sensitivity of 39.5% and higher specificity of 99.6%. Thus, for the EPDS, the optimal overall cutoff scores for MDD and MDD/MnDD were much lower than the commonly recommended cutoff scores.

116

Y. Zhao et al. / Psychiatry Research 226 (2015) 113–119

Table 2 Summary of reported prior pregnancy complications. (n¼842). Types

Number (%)

Body mass index (BMI ) abnormal AgeZ 35years old Pregnancy induced hypertension Diabetes mellitus Heart disease Thyroid disease Hepatic disease Blood disease Placenta previa Multiple pregnancy Scarred uterus Siphilis Previous History of anxiety Previous History of abnormal Gestation Previous History of infertility Genital tumors Inflammation In-vitro-fertilization ABO incompatibility Abnormal pelvic Physical discomfortable Amniotic fluid anomaly Myopia Abnormal fetal position

178(15.3%) 109(9.4%) 9(0.8%) 105(9.0%) 9(0.8%) 142(12.2%) 80(6.9%) 9(0.8%) 16(1.4%) 31(2.7%) 135(11.6%) 7(0.6%) 1(0.1%) 55(4.7%) 33(2.8%) 79(6.8%) 20(1.7%) 36(3.1%) 24(2.1%) 11(0.9%) 59(5.1%) 1(0.1%) 11(0.9%) 3(0.3%)

Table 3 Correlation between the EPDS and PDSS (N ¼ 842).

EPDS PDSS

M

S.D.

Pearson

P value

4.90 56.17

3.84 16.55

0.652

0.000**

The Edinburgh Postnatal Depression Scale ¼EPDS. The Postpartum Depression Screen Scale¼ PDSS. nn

Fig. 2. Major/minor depression.

Table 4 Performance of PDSS over a range of cutoff scores for major and minor depressions. PDSS cutoff score

Sensitivity

Specificity

56.50 57.50 58.50 59.50 60.50 61.50 62.50

0.935 0.935 0.926 0.923 0.893 0.867 0.843

0.919 0.937 0.966 0.998 0.998 0.998 0.998

PDSS¼ Postpartum Depression Screening Scale.

Correlation is significant at the 0.01 level (two-tailed).

depressive disorder” and “increased risk for major depressive disorder”. Concordance of the EPDS and PDSS were present in 591 women (70.19%): 504 (59.86%) had “normal” score on both, 66 (7.84%) had increased risk for minor depressive disorder and 21 (2.49%) had increased risk for major depressive disorder for both. Discordant scores in 251 women included 186 with elevated PDSS scores but normal EPDS scores and 25 with elevated EPDS scores but PDSS scoreso60. The Index of Inconsistency for answer normal minor depressive disorder and major depressive disorder are 59.9%,77.9% and 63.5%, and the Aggregate Index of Inconsistency is 67.5%.

4. Discussion

Fig. 1. Major depression.

3.5. The concordance and discordance in the EPDS and PDSS Table 8 presents the number and proportion of antenatal women who met cutoff for major and minor depressions based on EPDS scores and PDSS scores. The primary outcome of this study was to find rates of concordance and discordance in the EPDS and PDSS categories of “normal”, “increased risk for minor

Consistent with previous research (Blom et al., 2010; Tan et al., 2011), major and minor depressions were highly prevalent in the present study sample of Chinese pregnant women with obstetric complications. The prevalence of antenatal major and minor depressions was 9.6% and 30.5%, respectively, which reflects a higher than normal depression rate level reported in other studies done by Qiao et al. (2009), Bunevicius et al. (2009) and Gaynes et al. (2005). Our study indicated that pregnancy with complications can be experienced as a traumatic event and as such can lead to partial or full depression symptoms during antenatal period, which was almost consistent with the results of Li (2006), who investigated 160 highrisk pregnancies and found that tension (78.1%) was common in the adverse psychological reaction, and the pregnant women with different education levels had different proportions of the adverse emotional reaction. The depression not only affect high-risk pregnant

Y. Zhao et al. / Psychiatry Research 226 (2015) 113–119

Table 5 Performance of EPDS over a range of cutoff scores for major and minor depressions.

Table 8 Concordance and discordance in the EPDS and PDSS.

EPDS cutoff score

Sensitivity

Specificity

PDSS

3.50 4.50 5.50 6.50 7.50 8.50 9.50 10.50

0.820 0.760 0.704 0.607 0.533 0.447 0.311 0.207

0.581 0.724 0.819 0.909 0.962 0.998 0.998 0.998

o 60 (%)

EPDS ¼Edinburgh Postnatal Depression Scale.

Table 6 Performance of PDSS over a range of cutoff scores for major depression. PDSS cutoff score

Sensitivity

Specificity

76.50 77.50 78.50 79.50 80.50 81.50 82.50

0.889 0.877 0.864 0.864 0.765 0.679 0.654

0.974 0.986 0.992 1.000 1.000 1.000 1.000

PDSS ¼Postpartum Depression Screening Scale.

Table 7 Performance of EPDS over a range of cutoff scores for major depression. EPDS cutoff score

Sensitivity

Specificity

4.50 5.50 6.50 7.50 8.50 9.50 10.50 11.50 12.50 13.50

0.951 0.951 0.889 0.790 0.716 0.630 0.531 0.407 0.395 0.259

0.581 0.669 0.765 0.823 0.876 0.928 0.963 0.987 0.996 0.999

EPDS ¼Edinburgh Postnatal Depression Scale.

women's own health and marriage, but also cause serious and long-term cognitive and behavior problems in their children (Wang, 2011; Wood et al., 2010). Because of the high prevalence of depression, active screening, early identification and management of pregnant women with antenatal depression are of imminent importance for their health and, importantly the health of their families (Siu et al., 2010). Relative to other studies that reported the sensitivities and specificities of the screening measures for postpartum depression in comparison with diagnostic instruments (e.g., SCID or DIS) (Melville et al., 2010; Tandon et al., 2012; Davis et al., 2013; Pitanupong et al., 2007) the present study combined two depression screening tools to administer an antenatal sample of Chinese pregnant women with obstetric complications to determine the efficacy without comparison with any diagnostic (e.g., SCID or DIS) instruments. The results of the present study indicated that the convergent validity of the PDSS was satisfactory, confirming the positive relationship between the scores of the PDSS and the EPDS. This means that the two tools measure similar concepts, and those who had higher scores in PDSS showed higher level in EPDS, consistent with the findings of Beck and Gable (2001b), Karacam and Kitiz (2008), Hanna et al. (2004) and Li et al. (2011).

117

EPDS o9(%) 9–12(%) Z13(%)

60–79 (%)

Pearson chisquare

P value

264.58

0.000

Z 80 (%)

504(59.86) 163(19.36) 23(2.73) 25(2.97) 66(7.84) 26(3.09) 0(0.0) 14(1.66) 21(2.49)

EPDS: Edinburgh Postnatal Depression Scale. PDSS: Postpartum Depression Screening Scale.

For the EPDS, the AUCs for major or minor depression suggested good psychometric properties in screening for major or minor depression during antepartum. At the conventional 12/13 cut-off, the sensitivity of the scale was only 0.395, with specificity of 0.996. Using the receiver operating characteristic curve, if combined use, we identified 8/9 as the optimal cut-off for this study population, at which the sensitivity was 71.6% and specificity was 87.6%, which is lower than the result of the Lau et al. ( 2010), who found the EPDS threshold score was 10.5 (sensitivity, 81.25%; specificity, 80.67%) for clinical depression in Chengdu mothers. When a low false negative rate is desired in our study, a cut-off score of 6/7 can be used. For the PDSS, the AUCs for major or minor depression suggested excellent psychometric properties in screening for major or minor depression during antepartum. Just like the conventional cut-off, if combined use, using the receiver operating characteristic curve, we identified 79/80 as the optimal cut-off for major depression, at which the sensitivity was 86.4% and specificity was 100%, and 59/60 as the optimal cut-off for minor depression, at which the sensitivity and specificity were 92.3% and 99.8%, respectively. The ROC analysis confirmed the effectiveness and accuracy of the PDSS and the EPDS in the detection of antenatal depression and indicated the proposed threshold scores. Both screening tools were highly accurate when used in the antenatal period. Comparison of AUCs revealed statistically significant differences between the screening tools for detecting MDD and MDD/MnDD ( p o0.01), which indicates that the PDSS has a better psychometric performance for both major depression and major/monor depressions than the EPDS, which is consistent with the results of White (2008), who compared PDSS with EPDS and a structured clinical interview on a sample of 60 New Zealand women of European origin, and found that the PDSS was favored over the EPDS because the women considered that the statements more comprehensively captured their feelings. Although the PDSS is more accurate to screen depression than the EPDS, it still can lead to an inaccurate screen for depression. According to the results (Table 8), 25(2.97%) women with normal PDSS score but an EPDS score between 9 and 12 suggested increased risk for minor depressive disorder requiring clinical follow up for the women. Furthermore, 14(1.66%) women with the borderline PDSS score (60–79) but an EPDS score (elevated 4 13) indicated an increased risk for major depressive disorder with the further professional psychological assessment indicated for these women. The discordant results between PDSS and EPDS indicated that each screening scale had both a false positive result and a false negative result. Combined use of both scales, potentially reduces the incidence of depression screening errors and improves screening validity. Based on the present study's sample, in terms of PDSS, a cutoff of 59/60 was recommended for major and minor antenatal depressions, and a cutoff of 79/80 was proposed for major antenatal depression. This finding was comparatively and clearly lower than the cutoff value of 387 in a previous study of mothers

118

Y. Zhao et al. / Psychiatry Research 226 (2015) 113–119

within 12 weeks postpartum among Chinese postpartum women (Li et al., 2011), but consistent with the cutoff value of 60 for major and minor depression and 80 for major PPD in another study of American mothers (Beck and Gable, 2001b). Regarding EPDS recommended cutoff scores to detect minor or major antepartum depression, the present sample cutoff is 8/9 with the sensitivity of 71.6% and specificity of 87.6%. As mentioned earlier, Cox et al. (1987) suggested that a cutoff score of 12/13 be used to identify minor and major depressions, the cutoff in this population of Chinese pregnant women with obstetric complications is much lower than the traditional cutoff scores in other populations (Cox et al., 1987; Rowan et al., 2012) because the EPDS had decreased sensitivity when using standard cutoffs, which may lead to missing women who are in the midst of a depressive episode. One possible explanation for this variability of score cutoffs might be a discrepancy in the prevalence of antenatal depression due to the different cultures, social contexts and screening times. Another explanation might be related to the bias caused by specific target populations. A recent review of EPDS classification functioning (Gibson et al., 2009) found wide variability across studies, attributed to differences in the selection of reference “gold standard” diagnostic criteria employed, and in a given study population. The present study was focused on the population of Chinese pregnant women with obstetric complications who not only experienced greater physical stresses, but also generated many more complex psychological problems (Li and Li, 2009). Their depression varied with the obstetric complications at the different gestational weeks. Therefore, the present study mirrored the findings of Chaudron et al. (2010) suggesting that the optimal cutoffs for detecting antenatal depression among each screening tool may be lower than generally recommended (i.e., standard) cutoffs in the population of Chinese pregnant women with obstetric complications. The screening tools had decreased sensitivity when using standard cutoffs, which may lead to missing Chinese pregnant women who are in the midst of a depressive episode. Importantly, using both the PDSS and the EPDS does not substitute for a full psychiatric assessment as indicated. These scales do not predict postpartum depression, nor do they provide a measure of severity; but simply provide a score which indicates the likelihood of depression for action of anticipatory guidance or counseling by the obstetricians. Routine screening for perinatal depression is feasible, but general screening guidelines and recommendations remain controversial. For example, the American College of Obstetricians and Gynecologists. Committee on Obstetric Practice (2010) noted that there was insufficient evidence to support firm recommendations on when and how often perinatal depression screening should be conducted. Based upon the present study, we advocate that the optimal time to screen for antenatal depression in pregnant Chinese women is at the first antenatal obstetrical visit. We suggest that a combination of the EPDS and the PDSS be used for screening for antenatal depression in pregnant women with obstetric complications at the first pregnancy visit. Both of the screening scales should be given at appointment check-in, completed by the pregnant women, and the completed scales returned to the obstetricians, for scoring. If the EPDS or the PDSS score is more than the cutoff of the minor depression, the women should be asked to follow up at the next obstetrics visit, once a month until the delivery. If the EPDS or the PDSS score is more than the cutoff for major depression, referral for further psychological assessment should be recommended and discussed with the patient. 4.1. Study limitations Limitations in this study include the use of a modest sized convenience sample; use of one hospital may inadvertently skew the study sample thereby limiting generalization of results.

5. Conclusion In conclusion, depression screening plus “high-risk” feedback to providers improves the recognition of depression (Gjerdingen and Yawn, 2007). However, for screening to positively impact clinical outcomes, it needs to be combined with the complete screening system that provides valid screening tools, strong collaborative relationships between primary care and mental health providers, and longitudinal case management, to assure appropriate treatment and follow-up. In the present study, employing the EPDS and PDSS together appear to be reliable screening tools for major and minor depressions among Chinese pregnant women with obstetric complications. When combined using these tools, obstetricians and obstetrical staff should consider using lower cutoff scores to most effectively identify women in need of depression treatment, improve the screening validity toward successful identification of depression in these women. Future studies are needed to confirm these findings, examine and compare the effects of combined use of different depression screening scales during the perinatal period in different populations. Further longitudinal studies with longer follow-up periods are also encouraged.

6. Role of funding source This study (Protocol no. 10-020-201206) was funded by the China Medical Board (CMB).

7. Conflict of interest There are no personal, organizational or financial conflicts of interest.

Acknowledgment We are deeply grateful to all women who participated in the present study. To doctors, nurses and administrative personnel, we would equally like to thank their collaboration. References American College of Obstetricians and Gynecologists. Committee on Obstetric Practice, 2010. Committee opinion no. 453: screening for depression during and after pregnancy. Obstetrics & Gynecology 115, 394. Pereira, Ana Telma, Bos, Sandra C., Marques, Mariana, Maia, Berta R., João Soares, Maria, Valente, José, Allen Gomes, Ana, António Macedo, Maria, Helena Pinto, de Azevedo, 2011. The postpartum depression screening scale: is it valid to screen for antenatal depression? Archives of Women's Mental Health 14, 227–238. Ariadna, F., Linda, C., Urania, M., et al., 2009. Prevalence of post-traumatic stress disorder in pregnant women with prior pregnancy complications. The Journal of Maternal-Fetal and Neonatal Medicine 22 (6), 522–527. Beck, C., Tatano, D., 2008. State of the science on postpartum depression: what nurse researchers have contributed-part2. The American Journal of Maternal/ Child Nursing 33 (3), 151–156. Beck, C.T., Gable, R.K., 2000. Postpartum Depression Screening Scale: development and psychometric testing. Nursing Research 49 (5), 272–282. Beck, C.T., Gable, R.K., 2001b. Comparative analysis of the performance of the Postpartum Depression Screening Scale with two other depression instruments. Nursing Research 50, 242–250. Beck, C.T., Gable, R.K., 2001b. Further validation of the Postpartum Depression Screening Scale. Nursing Research 50 (3), 155–164. Beck, C.T., Gable, R.K., 2001. Comparative analysis of the performance of the Postpartum Depression Screening Scale with two other depression instruments. Nursing Research 50 (4), 242–250. Beck, C.T., Gable, R.K., 2002. Postpartum depression screening scale manual. Western Psychological Services, Los Angeles. Blom, E, Jansen, P, Verhulst, F, Hofman, A, Raat, H, Jaddoe, V, Coolman, M, Steegers, E, Tiemeier, H, 2010. Perinatal complications increase the risk of postpartum

Y. Zhao et al. / Psychiatry Research 226 (2015) 113–119

depression. The generation R study. British Journal of Obstetrics and Gynaecology 117, 1390–1398. Boyd, R.C., Le, H.N., Somberg, R., 2005. Review of screening instruments for postpartum depression. Archives of Women's Mental Health 8, 141–153. Breedlove, G., Fryzelka, D., 2011. Depression screening during pregnancy. Journal of Midwifery and Women's Health 56, 18–25. Bunevicius, R., Kusminskas, L., Bunevicius, A., Nadisauskiene, R.J., Jureniene, K., Pop, V.J., 2009. Psychosocial risk factors for depression during pregnancy. Acta Obstetricia et Gynecologica Scandinavica 88 (5), 599–605. Chaudron, L.H., Szilagyi, P.G., Tang, W., Anson, E., Talbot, N.L., Wadkins, H.I., Tu, X., Wisner, K.L., 2010. Accuracy of depression screening tools for identifying postpartum depression among urban mothers. Pediatrics 125, 609–617. Cox, J.L, Holden, J.M, Sagovsky, R., 1987. Detection of postnatal depression. Development of the 10-item Edinburgh postnatal depression scale. British Journal of Psychiatry 150, 782–786. Daniels, V., 2013. Antepartum depression–screening and treatment. International Journal of Childbirth Education 28 (3), 67–71. Davis, K., Pearlstein, T., Stuart, Scott, O’Hara, Michael, Zlotnick, Caron, 2013. Analysis of brief screening tools for the detection of postpartum depression: comparisons of the PRAMS 6-item instrument, PHQ-9, and structured interviews. Archives of Women's Mental Health 16, 271–277. Gausia, K., Ryder, D., Ali, M., Fisher, C., Moran, A., Koblinsky, M., 2012. Obstetric complications and psychological well-being: experiences of Bangladeshi women during pregnancy and childbirth. Journal of Health, Population and Nutrition 30 (2), 172–180. Gaynes, B.N., Gavin, N., Meltzer-Brody, S., Lohr, K.N., Swinson, T., Gartlehner, G., et al., 2005. Perinatal depression: prevalence, screening accuracy, and screening outcomes. Evidence Report/Technology Assessment 119, 1–8 (Summary). Gibson, J., McKenzie-McHarg, K., Shakespeare, J., Price, J., Gray, R., 2009. A systematic review of studies validating the Edinburgh Postnatal Depression Scale in antepartum and postpartum women. Acta Psychiatrica Scandinavica 119 (5), 350–364. Gjerdingen, D.K., Yawn, B.P., 2007. Postpartum Depression Screening: importance, methods, barriers, and recommendations for practice. Journal of the American Board of Family Medicine 20, 280–288. Gjerdingen, D., McGovern, P., Center, B., 2011. Problems with a diagnostic depression interview in a postpartum depression trial. Journal of the American Board of Family Medicine 24 (2), 187–193. Guo, X.J., Wang, Y.Q., Liu, Y., Chen, J., Pu, X.F., 2009. Study on the optimal critical value of the Edinburgh Postnatal Depression Scale in the screening of antenatal depression. Chinese Journal of Nursing 44, 808–810. Zhang, H.L., Li, L.Z., 2007. Analysis and comparison of the application of three foreign postpartum depression scales. Chinese Journal of Nursing 42 (2), 186–188. Karacam, Z., Kitiz, Y., 2008. The Postpartum Depression Screening Scale: its reliability and validity for the Turkish population. Turk Psikiyatri Dergisi 19 (2), 187–196. Lee, DTS, Yip, S.K., Chiu, HFK, Leung, T.Y., Chan, K.P., Chau, I.O., Leung, H.C., Chung, T. K., et al., 1998. ,Detecting postnatal depression in Chinese women. Validation of the Chinese version of the Edinburgh postnatal depression scale. British Journal of Psychiatry 172, 433–437. Li, L.Z., Liu, F., Zhang, H.L., Wang, L., Chen, X.F., 2011. Chinese version of the Postpartum Depression Screening Scale translation and validation. Nursing Research 60 (4), 231–239. Li, Y.Y., Li, Q.Y., 2009. Psychology and health education survey on pregnant women. Nursing Practice and Research 6 (4), 109–110. Li, Y.P., 2006. Analysis of 160 high-risk pregnant women's psychological state and strategies. Today Nurse 8, 56–58. Melville, J., Gavin, A., Guo, Yuqing, Fan, Ming-Yu, Katon, W.J., 2010. Depressive disorders during pregnancy: prevalence and risk factors in a large urban sample. Obstetrics & Gynecology, 116; , pp. 1064–1070. Neiman, S., Carter, S., Sharon, V.S., Kindred, C., 2010. Best practice guidelines for the nurse practitioner regarding screening, prevention, and management of postpartum depression. Critical Care Nursing Quarterly 33 (3), 212–218.

119

Pereira, A.T., Bos, S., Marques, M., Maia, B.R., Soares, M.J., Valente, J., Gomes, A., Macedo, A., Azevedo, M.H., 2010. The Portuguese version of the postpartum depression screening scale. Journal of Psychosomatic Obstetrics and Gynecology 31, 90–100. Pitanupong, J., Liabsuetrakul, T., Vittayanont, A., 2007. Validation of the Thai Edinburgh Postnatal Depression Scale for screening postpartum depression. Psychiatry Research 149, 253–259. Rowan, Paul, Greisinger, A., Brehm, B., Smith, F., McReynolds, E., 2012. Outcomes from implementing systematic antepartum depression screening in obstetrics. Archives of Women's Mental Health 15, 115–120. Choi, Sae Kyung, Kim, Jung Jin, Park, Yong Gyu, Ko, Hyun Sun, Park, In Yang, Shin, Jong Chul, 2012. The simplified Edinburgh Postnatal Depression Scale (EPDS) for antenatal depression: is it a valid measure for pre-screening? International Journal of Medical Sciences 9 (1), 40–46. Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J., Weiller, E., Hergueta, T., Baker, R., Dunbar, G.C., 1998. The Development and Validation of a Structured Diagnostic Psychiatric Interview for DSM-IV and ICD-10 59, 22–57. Sit, D., Wisner, K., 2009. Identification of postpartum depression. Clinical Obstetrics and Gynecology 52 (3), 456–468. Siu, Bonnie W.M., Patrick, Ip, Chow, Hesione M.T., Kwok, Shirjolina S.P., Li, Oi-Lin, Koo, Mei-Ling, et al., 2010. Impairment of mother-infant relationship: validation of the Chinese version of Postpartum Bonding Questionnaire. Journal of Nervous and Mental Disease 198, 174–179. Swets, J.A., 1988. Measuring the accuracy of diagnostic systems. American Association for the Advancement of Science 240 (4857), 1285–1293. Tandon, S., Cluxton-Keller, Darius, Leis, Fallon, Le, J., Perry, Huynh-Nhu, Deborah, F., 2012. A comparison of three screening tools to identify perinatal depression among low-income African American women. Journal of Affective Disorders 136, 155–162. Tan, Xiaoyan, Guo, Weihang, Wang, Aimei, 2011. Analysis of psychological state and influencing factors of high risk pregnancies before cesarean section. Journal of Nursing 18 (4B), 74–76. Thornton, D., Guendelman, S., Hosang, N., 2010. Obstetric complications in women with diagnosed mental illness: the relative success of California's county mental health system. Health Services Research 45 (1), 246–264. Wang, C.X., 2011. Analysis of maternal postpartum depression related factors and cognitive function nursing intervention. Nursing Practice and Research 8 (18), 40–41. White, G., 2008. A comparison of the Postpartum Depression Screening Scale (PDSS) with the Edinburgh Postnatal Depression Scale (EPDS). New Zealand College of Midwives Journal 39, 28–32. Wood, A., Middleton, S., Leonard, D., 2010. “When it's more than the blues”: a collaborative response to postpartum depression. Public Health Nursing 27 (3), 248–254. Yawn, B., Pace, W., Wollan, P., et al., 2009. Concordance of Edinburgh Postnatal Depression Scale (EPDS) and Patient Health Questionnaire (PHQ-9) to assess increased risk of depression among postpartum women. Journal of the American Board of Family Medicine 22 (5), 483–491. Lau, Y., Wang, Y.Q., Yin, L., Chan, K.S., Guo, X.J., 2010. Validation of the Mainland Chinese version of the Edinburgh Postnatal Depression Scale in Chengdu mothers. International Journal of Nursing Studies 47, 1139–1151. Qiao, Y.X., Wang, J., Li, J., Ablat, A., 2009. The prevalence and related risk factors of anxiety and depression symptoms among Chinese pregnant women in Shanghai. Australian and New Zealand Journal of Obstetrics and Gynaecology 49, 185–190. Zhong, Q.Y., Gelaye, Bizu, Rondon, Marta, Sánchez, Sixto E., García, Pedro J., Sánchez, Elena, Barrios, Y.V., Simon, G.E., Henderson, D.C., Cripe, Swee May, Williams, M. A., 2014. Comparative performance of Patient Health Questionnaire-9 and Edinburgh Postnatal Depression Scale for screening antepartum depression. Journal of Affective Disorders 162, 1–7.