Dr CE Hewitt, Department of Health Sciences, University of York, York YO10 5DD, UK. Email email@example.com
Background Postnatal depression (PND) is a common mental health problem, which is associated with adverse consequences beyond the individual with depression. It is not known whether using formal methods to identify PND are clinically and cost effective in improving maternal and infant outcomes.
Objectives To evaluate the clinical and cost effectiveness of antenatal and postnatal identification of depressive symptoms.
Search strategy Twenty electronic databases were searched to retrieve English and non-English language articles published until February 2007.
Selection criteria Randomised controlled trials or controlled trials comparing the use of formal methods to identify PND, with or without enhancement of care, or feedback of scores with not using formal methods to identify PND or usual care.
Data collection and analysis Two reviewers independently assessed studies for inclusion and extracted data. Results from the trials were combined to calculate odds ratios and 95% confidence intervals for dichotomous outcomes.
Main results Five studies were identified that compared formal use of a method to identify PND, with or without enhancement of care, or feedback of scores with not using a formal method or usual care. All of the studies used the Edinburgh Postnatal Depression Scale (EPDS) to identify women with PND. The results of the studies indicated beneficial effects of using the EPDS in reducing EPDS scores (OR = 0.61; 95% CI 0.48–0.76).
Author’s conclusions Despite some apparent beneficial effects of using formal methods to identify PND, it is difficult to disentangle the effects of the screening component alone from interventions linked to a positive screen as some of the studies included enhancements of care and/or an intervention.
Postnatal depression (PND) is a common mental health problem. A meta-analysis of 59 studies (including 12 810 women, mainly from first world countries) found the prevalence of depression within the first few postnatal months was 13% (95% CI 12.3–13.4%).1 Most cases develop within the first three postnatal months,2 with a peak incidence at around 4 to 6 weeks.3 Although one study showed that most cases last around 3 months and resolve spontaneously without treatment,2 another study demonstrated the presence of depression, with over 50% lasting over 6 months and some being still present at 4 years.4
Depression in all populations is associated with profound decrements in quality of life, social functioning and economic productivity.5 However, in the case of PND, the adverse consequences are felt beyond the individual with depression. There is now considerable evidence to show that postnatal depression has a substantial impact on the mother and her partner;6 the family;7 mother–baby interaction8 and on the longer-term emotional and cognitive development of the baby,9 especially when depression occurs in the first year of life.10 Unfortunately, <50% of cases of PND are detected by healthcare professionals in routine clinical practice.11
Given the adverse consequences of PND and the general underidentification of this problem, routine use of methods to identify PND has been proposed. However, this screening has attracted substantial controversy.12 No systematic review or meta-analysis has critically appraised currently available data to establish whether the routine use of methods to identify PND are clinically and cost-effective. Previously published reviews in this area have focused on prevention and treatment of antenatal and postnatal depression,13–19 rather than evaluating methods to identify PND. The aims of the review were to explore whether the routine use of antenatal and postnatal identification of depressive symptoms (‘screening or case finding’) or the integration of these methods with enhancements of care are clinically and cost effective.
Searches were undertaken across 16 databases to identify relevant clinical and cost-effectiveness literature: MEDLINE; CINAHL; PsychINFO; EMBASE; Maternity and Infant Care; CENTRAL; DARE; CDSR; Science citation index web of knowledge; NRR; ReFeR; metaRegister of Controlled Trials; Health Services Research Projects in Progress; LILACS; Inside Conferences; Dissertation Abstracts. In addition, four databases were searched specifically for cost effectiveness studies: NHS EED; HEED; IDEAS; EconLit. All databases were searched from their inception until February 2007. No language or other restrictions were applied.
We used three sets of key terms: pregnancy (e.g. Pregnancy/, Postnatal Care/, new mother$, postpartum, postpartum), depression (e.g. depress$, Depression, Postpartum/, PND) and screening and diagnosis terms (instrument specific: e.g. Edinburgh adj5 (Inventory or Questionnaire or scale or index or checklist or interview and general terms (e.g. screen$, diagnos$, detect$, identif$). An example of our search strategy is provided in Appendix S1. Reference lists of all studies were inspected to ensure that all potentially relevant studies had been identified.
All records were imported into a bibliographic referencing software programme (endnote version 9 Thomson, Philadelphia, PA, USA). Duplicate references were deleted. Two reviewers screened titles and abstracts generated by the searches to identify potentially eligible studies. Any disagreements were resolved by consensus or deferred to a third party if necessary. Full papers of potentially eligible studies were obtained and assessed for inclusion independently by two reviewers. Articles were eligible for inclusion if they fulfilled the criteria outlined below.
For the clinical effectiveness review, randomised controlled trials or controlled trials comparing the implementation of any method to identify PND, with or without enhancement of postnatal care, in any setting were included. There is some suggestive evidence from a review by the US Preventative Services Task Force20 that using formal methods to identify depression can become effective when they are accompanied by organisational enhancements of care, involving clinician education, support from case managers or a collaborative care approach between specialists and primary care physicians.21 Hence, in this systematic review, we examined the impact of using formal methods by themselves, as well as studies where the formal method had been incorporated into some from of educational or organisational enhancement of care.
Any type of intervention was included as long as it was compared with usual care for expectant or postnatal mothers (up to one year postnatally). Trials assessing interventions aimed at healthcare professionals and/or at women with PND were included in the review. Data on maternal and infant outcomes were included, however defined. Short- (<6 months), medium- (6–12 months) and longer-term (>12 months) outcomes were all considered. Trials not reporting maternal or infant outcomes were excluded. Head to head trials comparing interventions without the option of participants being randomised to a usual care group (for example, a trial comparing antidepressants with placebo) were excluded. Trials that included a method to identify PND as a strategy to gain a baseline measure of depression and did not use the results from this for any other purpose were also excluded.
For the cost-effectiveness review, articles were eligible for inclusion if they were full economic evaluations (cost benefit analyses, cost effectiveness analyses and cost-utility analyses) of methods to identify PND. We followed explicit guidelines laid down by the Centre for Reviews and Dissemination (CRD) in the preparation of the NHS Economic Evaluations Database (NHS EED).22
Data extraction and quality assessment
Data were extracted by one reviewer and checked by a second reviewer, using a predefined data extraction form. The following data were extracted: author, year of publication, study design, setting, patient population, details of the intervention and usual care, sample size and results. Where there were multiple publications for the same study, data were extracted primarily from the most recent and complete publication. In cases where the duplicate publications reported additional relevant data, these data were also extracted.
The quality of any available studies for the clinical effectiveness review was assessed using the following criteria: randomisation, concealment and loss to follow up. The quality and relevance of any available economic data were judged from the perspective of the UK NHS according to criteria laid down by Drummond and colleagues.23
For each dichotomous outcome, the numbers of patients experiencing the outcome were extracted for each group. The Odds Ratio (OR) and 95% confidence interval (CI) were calculated for each study outcome. Where there was more than one study for a comparison, the ORs were pooled using a fixed-effect model (the Mantel–Haenszel (M–H) method) and the corresponding 95% confidence intervals (CIs) were calculated. Statistical heterogeneity was assessed using the I2 statistic. It was intended that continuous data would be analysed by calculating the difference in means and corresponding 95% CI for each study. To perform a meta-analysis of continuous data, we needed to extract mean values of the outcome, the standard deviation and the number of participants included in the analysis of the outcome data. In the two studies reporting continuous outcome data, it was impossible to obtain standard deviations for each group from the data summaries presented in each publication. Hence, formal pooling was not undertaken for the continuous outcome data.
Overview of included studies
The study selection process is outlined in Figure 1. In total, 15 612 potentially relevant studies were identified from the searches, of which 227 were selected for full assessment.
Four studies were identified that met the inclusion criteria for the clinical effectiveness review (Table 1).24–27 One of the studies compared the rate of identification of PND using the Edinburgh Postnatal Depression Scale (EPDS) compared with spontaneous detection from routine clinical evaluations by physicians and mid-level healthcare providers in a residency training programme.24 Women were recruited from a single hospital in the US and allocated to groups based on the date of delivery; those women who delivered on even days were assigned to the intervention group. Women assigned to the intervention group were posted the EPDS to complete at 6 weeks postnatally. Data for the control group were gathered by review of the postnatal record only; hence, the EPDS was not completed by the control group. Only 37% (n = 79) of women in the intervention group returned the EPDS and only 54% (n = 96) of the control group women attended the 6 week postnatal visit. Women scoring 10 or more on the EPDS were considered to be at higher risk of postnatal depression. The number of women considered to be at higher risk of postnatal depression in the intervention group was 35.4% (n = 28) and 6.3% (n = 6) in the routine clinical evaluation group (P < 0.0001).
Table 1. Characteristics of included studies
Setting, patient population
Intervention, usual care
Evins, 2000, US Individual RCT (odd/even dates of delivery)
Hospital for recruitment. EPDS by mail Consenting women who delivered at the hospital (n = 391 women)
Intervention: screening for postnatal depression with the EPDS (postal). Women at high risk (EPDS ≥ 10) for postnatal depression were referred to the social worker for assessment. Usual care: no screening with the EPDS only spontaneous detection during routine clinical evaluation
Threshold EPDS scores at 6 weeks
Kung, 2002, China Individual RCT
Hospital for recruitment. Home for the intervention. Chinese women, who were Hong Kong residents, spoke Cantonese and who gave birth to a baby in a single hospital were included (n = 405 women)
Intervention: early intervention including in depth assessment by midwives (women were asked to describe any sleep disturbance, poor appetite, self-blame or if they had ideas of hurting the baby), subsequent phone follow-up service by midwives and volunteers to the women with high EPDS score. Therapeutic listening, information and appropriate care were provided according to need. If women were in severe distress they were referred to an obstetrician. Information on community resources were also offered if required. A contact telephone number was given to anyone who refused weekly telephone follow up so they could contact the midwife if they needed help. Enhancement: training on basic counselling and interviewing skills, emotional problems of postnatal women were provided to midwives, social workers and volunteers before the study Usual care: women were not assessed with the EPDS until 6 weeks after delivery. Their emotion was assessed by routine clinical evaluation and counselling provided if needed. A contact telephone number was also given to the women if they needed help or wanted to talk with the midwife.
Threshold EPDS scores at 6 weeks; mean EPDS scores at 6 weeks; use of contact number
MacArthur, 2002, UK Cluster RCT
Women who received postnatal care in the recruited practices (n = 36 GP practices)
Intervention: care was led by midwives, with contact with GPs based on referral, including home visits and the final discharge consultation. A symptoms checklist was used at day 10 and 28 and at the discharge consultation at 10–12 weeks. The EPDS was also used at day 28 and discharge consultation. Enhancement: midwives trained in postnatal care and health and trial design. Training to implement new model of care. Usual care: standard community postnatal care consisting of 7 home midwife visits to 10–14 days (can continue to day 28) after birth and care from health visitors thereafter. GPs did routine home visits and final 6–8 week check. Some health visitors use the EPDS routinely so some controls screened. Enhancement: midwives trained in postnatal care and health and trial design.
Mean mental health score at 16 weeks; mean EPDS score at 16 weeks; mean physical health score at 16 weeks; threshold EPDS scores at 16 weeks postnatally; overall satisfaction; better than expected; planning care score; continuity of care; maternity discharge score; talk to midwife about most/all health symptoms; no difficulty talking to midwife about health symptoms; midwife visits (from midwife records); Midwife visits (from women’s diaries); GP visits (from 4 months questionnaire); GP visits (from women’s diaries)
Webster, 2003, Australia Individual RCT
Obstetric (prenatal clinic) hospital Risk factors for PND: (1) low social or partner support, measured by Maternity Social Support Scale ≤ 24, (2) a past history of mental illness, (3) family psychiatric history, (4) past PND or (5) having a mother who had PND. Only positively screened women were randomly allocated (n = 600 women)
Intervention: educational intervention. Providing women with a booklet about PND and a list of phone contacts of PND resources prenatally. Letters were also sent to referring GP and to their Child Health Nurse with details of their risk status. Usual care: standard care, including case management and referral to a hospital social worker or psychiatrist if required
Threshold EPDS scores 16 weeks
The remaining studies focused on detecting and subsequently reducing the number of women at higher risk of PND or the severity of PND (indicted by a reduction in identification scores), rather than detecting women at higher risk of PND alone as in the previous study. All of the studies used the EPDS (alone or in combination with other strategies) as a formal method to identify PND, although the EPDS was administered at various time points: 4 weeks postnatally;25 1 day before discharge, 1 week postnatally and 6 weeks postnatally;26 or it was unclear.27 The EPDS was used as an outcome measure across all of the studies, either as a continuous variable comparing mean EPDS scores between the intervention and control group or as a dichotomous variable comparing the number of women scoring above or below a cut point on the EPDS. Outcomes were assessed at 6 weeks postnatally26 or 4 months postnatally.25,27 Additional outcomes to be considered were: number of visits to midwife and GP (n = 1), psychical and mental components of the SF36 (n = 1), number of referrals (n = 1), women’s views about care (n = 1), overall satisfaction with care (n = 1) and the use of support contact number (n = 1). The results of each individual study for the threshold EPDS scores and mean EPDS scores across are displayed in Figures 2 and 3 respectively. For the studies providing dichotomous outcomes, the I2 value was 2%. The methodological quality of included studies was variable (Table 2).
Table 2. Methodological quality of included studies
Loss to follow up
The aim of the first study was to evaluate the effectiveness of a prenatal intervention in reducing the incidence of PND.27 Women attending their first prenatal appointment at a single hospital in Australia were screened for risk factors of PND (MSSS < 25; past personal or family history of mental illness or PND). Those women identified with elevated risk of PND were randomised to the intervention (n = 299) or control group (n = 301). The intervention consisted of a booklet about PND with contact numbers, use of the EPDS, a discussion with the women about her risk of developing PND and a letter to the woman’s GP and local child health nurse alerting them to the woman’s risk of PND (i.e. feedback of the risk). Women in the control group received standard care, which included case management and referral to a hospital social worker or psychiatrist if required. Sixteen weeks postnatally women were asked to complete a second EPDS. Of the 509 women sent, the second EPDS, 73% (n = 371) responded. The number of women with EPDS scores of 12 or above at 16 weeks postnatally in the intervention group was 24% (n = 46) and 28% (n = 50) in the control group (OR = 0.80, 95% CI 0.50 to 1.28).
The second study investigated whether the provision of an early intervention by midwives decreased the incidence of depression at 6 weeks postnatally.26 The sample of women was recruited from a single teaching hospital in Hong Kong. Women in the intervention group were assessed with the EPDS three times during the postnatal period (1 day before discharge, 1 week postnatally and 6 weeks postnatally). The intervention included in depth assessment by midwives (women were asked to describe any sleep disturbance, poor appetite, self-blame or if they had ideas of hurting the baby), with subsequent phone follow-up service by midwives and/or volunteers to the women with high EPDS scores (cut point 10). Training on basic counselling and interviewing skills and emotional problems of postnatal women, were provided to midwives, social workers and volunteers before the study commenced. Women in the control group were not assessed with the EPDS until 6 weeks postnatally. In the intervention group, 12 (5.9%) women had EPDS scores above 9 at 6 weeks postnatally versus 24 (11.8%) women in the control group (P = 0.03). The mean EPDS score of the intervention group was 4.32 whereas that of the control group was 4.97; an exact P-value was not reported in the article it was only reported that the P-value was not statistically significant.
The final study explored the effects of redesigning community postnatal care on women’s health 4 months postnatally.25 Thirty-six general practices from the West Midlands were cluster-randomised using minimisation to the intervention or control group. General practices were randomised rather than women to prevent contamination. In the intervention group, care was led by midwives and included extended care to 28 days, use of a symptoms checklist and the EPDS at day 28, a referral to a general practitioner as necessary and a 10–12 week discharge visit. The control group received routine care that included seven midwifery home visits until 10–14 days postnatally (although this could be extended to 28 days) and care by health visitors thereafter. General practitioners completed routine home visits and a final check-up at 6 to 8 weeks postnatally. Additional training in postnatal care and health and trial design was provided to all midwives. Seventeen practices (1087 women) were randomised to the intervention group and 19 practices (977 women) were randomised to the control group. The mental health score was significantly better in the intervention group than controls (OR = 3.03, 95% CI 1.53 to 4.52, ORadj = 4.31, 95% CI 2.50 to 6.12), as were mean EPDS (OR = −1.92, 95% CI −2.55 to −1.29, ORadj = −2.68, 95% CI −3.46 to −1.89) and the proportion of women with an EPDS score of 13 or above (OR = 0.57, 95% CI 0.43 to 0.76, ORadj = 0.47, 95% CI 0.31 to 0.76). The physical health score did not differ between study groups (OR = −0.79, 95% CI −1.91 to 0.34, ORadj = −0.80, 95% CI −2.32 to 0.72). Overall satisfaction with care from the community midwives did not differ statistically between groups (OR = 1.09, 95% CI 0.72 to 1.63).
The EPDS was the most frequently used method to indicate women at higher risk of postnatal depression. All of the studies indicated beneficial effects of using the EPDS in reducing EPDS scores, although some of the individual studies did not show statistically significant differences. Studies reporting dichotomous outcomes were combined and the pooled estimate was OR = 0.61 (95% CI 0.48 to 0.77). Thus, the odds of scoring above the threshold for depression in a population where a formal method was used to identify women at higher risk of PND (with an intervention for those identified) were 0.61 times the odds of scoring above the threshold for depression in a population where there was no formal method used (with an intervention for those identified).
On the basis of their title and abstract, two studies from the original searches and two studies from the economic searches appeared to be potentially eligible for inclusion in the cost-effectiveness review. Nevertheless, on closer inspection, none of the studies were full economic evaluations of methods to identify PND (Table 3).
Preventative intervention versus routine primary care
Not a full economic analysis of a method to identify PND; cost effectiveness analysis of the intervention under study
Discussion and conclusion
For the clinical effectiveness review, four studies were identified that compared using a formal method to identify women with PND, with or without enhancement of care, or feedback of scores with not using a formal method or usual care. All of the studies used the EPDS as a method to identify women at higher risk of PND and the outcome of choice within these studies were depression scores, both as continuous score (i.e. mean score) or as a binary measure (i.e. number of women scoring above and below a threshold).
It would have been interesting to explore the impact of using the EPDS on additional maternal and infant outcomes; however, an insufficient number of studies reported such outcomes to undertake any analyses. Another limitation of the studies included was that measures of dispersion were often not reported; thus, analyses were unable to be undertaken for studies reporting continuous outcomes.
Difficulties arose when interpreting the results as two out of the four studies also included additional enhancements of care (e.g. training on basic counselling and interviewing skills). Furthermore, in some studies, those women identified as being at higher risk of PND were offered an intervention. As screening was incorporated with other changes, which were beyond those needed to implement the screening, it was difficult to highlight the impact of the screening component. For example, it could be that the training on basic counselling is the important factor and use of the formal methods to identify PND has little impact. Hence, overall, it was difficult to draw any firm conclusions about clinical effectiveness. The research evidence, in this respect mirrors the evidence that has been used to establish the effectiveness for screening for depression in medically ill populations and in primary care.28
Additional differences between the procedures used in the studies were also apparent. For example, in one of the studies, the women were sent a postal EPDS to complete, whereas, in the other studies, women were asked to complete the EPDS during a routine visit with a health professional. Postal questionnaires are unlikely to be a method that would be used in routine clinical practice.
A number of strategies have been proposed to improve PND identification, which broadly fall into five categories:
• Postnatal identification using specially developed standardised postnatal questionnaires (such as the Edinburgh Postnatal Depression Scale);
• Postnatal identification using standardised generic questionnaires for depression (such as the Beck Depression Inventory);
• Prenatal screening using standardised depression questionnaires to identify those with pre-existing depression and those at risk of developing significant depression in the postnatal period;
• Prenatal screening using known risk factors for PND (such as previous history of depression and lack of social support) to identify those who are likely to subsequently develop depression in the postnatal period;
• The use of training packages targeted at healthcare professionals designed to enhance awareness and recognition of clinical signs of postnatal depression and to ensure that a thorough psychosocial assessment is provided on a routine basis.
We considered all five categories of identification within this review; however, this could be seen as a potential limitation of the review.
There were a number of methodological limitations identified among the studies included in this review. We included controlled trials as well as randomised trials, for example, in some of the studies, odd or even expected dates of delivery were used to allocate participants to treatment groups. Obviously, such methods of allocation are not truly random and thus, in practice, this often results in selection bias being introduced. In addition, in some of the studies described as randomised controlled trials, it was often unclear how the randomisation sequence was generated and what methods were used to conceal the sequence. Hence, it was difficult to judge whether the methods used to randomise participants to treatment groups were subject to bias. Another frequently occurring problem was attrition, with most studies reporting that some women were lost to follow up. Noncompliance or nonattendance was also reported across studies. It was often unclear whether analyses were undertaken using intention-to-treat or not. An intention to treat approach should be used as the primary analysis when noncompliance occurs as this ensures that the initial randomisation sequence is preserved.
Another issue that arose within this research was the cut point used to distinguish between those women who were at higher risk of having PND and those who were not. The choice of cut point to indicate a positive test creates a unique issue that arises within diagnostic studies. Variations in cut points lead to variations in measures of the accuracy of the test (i.e. sensitivity and specificity) and thus created difficulties within this review as the studies used different cut points on the EPDS.
Despite an extensive systematic search of the literature none of the studies identified presented full economic evaluations of methods to identify PND. Future research needs to address this gap in the literature. We are aware of one large, but yet unpublished, randomised controlled trial with an economic evaluation (PoNDER trial, http://www.controlled-trials.com/ISRCTN92195776). The PoNDER trial aimed to assess the costs and effectiveness and broad impact of two health visitor psychological interventions for PND.
Screening is only one way in which recognition and management of PND might be improved and there are clear criteria laid down, which should guide the adoption of a screening strategy as part of national health policy. Screening tests can only be justified if the instrument is: accurate; results in a more effective treatment than would otherwise be the case and; does so with a favourable ratio of costs to benefits.29 These criteria have been codified in the UK by the establishment for the National Screening Committee (NSC) and the publication of clear criteria, which must be satisfied prior to adoption of a screening strategy.30 When these criteria were previously applied to the case of screening for PND,31 insufficient evidence was found to support the implementation of this strategy. This NSC recommendation has not been without controversy, but genuine concerns remain regarding the acceptability, validly, clinical effectiveness and cost effectiveness of identification methods for PND.
One of the criteria of the NSC states that ‘There should be evidence from high quality randomised controlled trials that the screening programme is effective in reducing mortality or morbidity’. The clinical effectiveness review highlights that insufficient evidence is available to conclude that the EPDS is effective in improving maternal and infant outcomes. Some suggestive evidence indicates that the EPDS, maybe with some enhancement of care, may lead to reductions in the number of women with EPDS scores above a certain threshold or reduction in EPDS scores. Hence, this criterion would not be met. A second criteria from the NSC states that ‘The opportunity cost of the screening programme (including testing, diagnosis and treatment, administration, training and quality assurance) should be economically balanced in relation to expenditure on medical care as a whole (i.e. value for money)’. As no full economic evaluations of methods to identify PND, this criterion would also not be met.
In conclusion, evidence surrounding the clinical and cost-effectiveness of PND screening is lacking and further research is required in this area to address this gap.
Disclosure of interest
None to declare.
Contribution to authorship
SG had the original idea for the research. CH selected studies and extracted data. CH analysed and interpreted the data. Both authors drafted and revised the manuscript and approved the final version for publication.
Details of ethics approval
NHS R&D Health Technology Assessment Programme.
We are grateful to the individuals who translated the non-English language publications. We also thank authors who responded to requests for further data, copies of articles or points of clarification. Finally, we thank Stephen Brealey for assistance with selecting studies and data checking.