Postnatal depression and puerperal psychosis
Section 2: Diagnosis, screening and prevention

2.1 Diagnosis

Depression is a common condition, affecting a large proportion of women of childbearing age. Studies are evenly divided in reporting postnatal depression as either more or less severe than depression at other times16,17,18,19 and there is little evidence that the nature of symptoms differs between postnatal and non-postnatal depression.20,21 In diagnosing depression in the postnatal period, there is a risk that normal emotional changes may be mistaken for depression or may mask depressive symptoms.7 Evidence level 2+

A large number of studies have assessed the prevalence of postnatal depression.6,19,23,24,25,26,27,28,29,30,31,32,33,34,35 In those where robust methodology was used, prevalence (whether point or period) ranges from 4.5% to 28% of women in the postnatal period. The majority cluster around 10% to 15% with one meta-analysis giving a prevalence of 13%.2 Variance between studies can be accounted for in part by the period under evaluation and the method of assessment used. There is some evidence that, while the overall prevalence of postnatal depression is not significantly different from that of depression at other times, there is an increased risk of depression occurring in the early postnatal period (threefold in the first five postnatal weeks).35,36 Evidence level 2++

Although research exists on the prevalence of postnatal depression in other cultures, little work has been published on ethnic minority groups within Scotland. It is important to remember that there are widely varying cultural traditions and rituals surrounding pregnancy and childbirth and a lack of cross-cultural equivalence in concepts of depression. Effective detection and management requires an understanding of these differences.

Puerperal psychosis is a much less common condition, affecting one to two per thousand women.3,37 Most studies agree that this rate represents a significantly increased risk for psychotic illness when compared with other times in a woman's life. Puerperal psychosis is largely affective in nature, although several studies comment on atypical features in the presentation such as mixed affective state, confusion and disturbed behaviour.38 It typically presents in the early postpartum period, usually within the first month.

2.2 Risk factors



If risk factors predicting postnatal depression can be identified by screening this would allow optimum targeting of effective interventions.

The evidence suggests that risk factors for postnatal depression are no different to the risk factors for non-postnatal depression. Three systematic reviews identified the following risk factors as having moderate to strong associations with postnatal depression:2, 39,40


Weak associations have been found with obstetric complications, a history of abuse, low family income and lower occupational status.2,39,40,41 An American review found no evidence regarding the effect of early postpartum discharge.42 Evidence level 3

In addition to the above factors, cohort and case control studies have identified the following as risk factors:19,43,44,45,46,47,48,49,50


There is no conclusive evidence on hormonal changes as a risk factor for postnatal depression. In a small experimental study, simulating hormone changes after delivery led to a significant change in mood in five of eight women with a previous history of postnatal depression compared with none of eight comparison women, suggesting differential sensitivity to hormone change.51 Evidence level 1-

Mothers' mental health may also be affected by the health of the baby. In cohort studies depression has been associated with neonatal risk,52 stillbirth, neonatal death or Sudden Infant Death Syndrome (SIDS),53 and very low birth weight (less than 1500 g).54,55 Evidence level 2+


Factors that increase the risk of puerperal psychosis include a past history of puerperal psychosis, pre-existing psychotic illness (especially affective psychosis) of severity requiring admission to hospital, and family history of affective psychosis in first or second degree relatives.56,57,58,59 Women who have had a previous puerperal psychosis are at significant risk of future puerperal and non-puerperal episodes.56,57,60,61,62,63 The risk of a future puerperal episode lies between 25% and 57% and the risk of non-puerperal relapse is even higher. Evidence level 2+

2.3 Screening


Screening for postnatal depression has gained in popularity since the original studies on the effectiveness of screening by health visitors in primary care were published.64 Screening can have negative consequences however, particularly so in the field of mental health. It is therefore important that the health professionals administering any aspect of a screening programme are adequately trained to do so.

Many areas throughout Scotland have already instituted screening programmes, often in the context of integrated care pathways for the detection and management of postnatal depression.65 To be effective, screening programmes should meet certain criteria, the most important of which include:

The issue of screening for postnatal depression is currently being assessed by the Department of Health National Screening Committee. In practice, screening is already taking place, albeit in varying styles and with varying levels of resources. While evidence may not yet be available to meet the strictest criteria for recommending screening programmes, if programmes are instigated they should conform to best available research evidence on effectiveness, be adequately resourced, and include ongoing evaluation as an integral part of the programme. The SIGN guideline development group's recommendations are based on these premises.



Screening tools have been devised to predict postnatal depression in the antenatal period. These have been based around known risk factors for postnatal depression (see section 2.2.1), but many have not been properly evaluated to determine sensitivity, specificity and predictive value. The Edinburgh Postnatal Depression Scale (EPDS) has also been examined as an antenatal screening tool.67 As yet, no antenatal tool has been devised which will accurately predict those who go on to develop postnatal depression. Evidence level 2+

There is no evidence to support routine screening in the antenatal period to predict the development of postnatal depression.


While no specific screening tools have been devised to identify women at high risk of puerperal psychosis, there is ample evidence that risk factors can be easily identified and are highly predictive (see section 2.2.2). Based on this evidence, enquiry about such risk factors has been recommended by several expert reports.5,68 Evidence level 4



The most commonly used screening tool in the postnatal period is the EPDS. There is good evidence for its effectiveness, although its sensitivity, specificity and predictive value are dependent on the cut-off scores chosen. A cut-off of greater than 9 has been suggested for 'possible depression' and greater than 12 for 'probable depression'.67,69,70 The lower cut-off will ensure that very few women are missed, but at the expense of a high false positive rate. Taking this into account, the positive predictive value of the EPDS varies from 44% to 73%. 71,72,73 Evidence level 2+

Concerns have been expressed that the EPDS may perform less well in cases where there are psychomotor symptoms (often suggestive of severe depression).74 More work is required on the timing and number of administrations, and on appropriate cut-offs to use. Evidence level 2+

There is some evidence that, in research settings, combining two screening tools (the EPDS and the General Health Questionnaire, GHQ) may be more effective than either tool alone.75 Evidence level 4




An effective intervention to prevent postnatal depression in high risk groups would benefit women in reducing depression and its impact on the child and marital relationships. However, the evidence for the effectiveness of interventions to prevent postnatal depression is conflicting. Few good quality randomised controlled trials have been published and those that have give inconclusive findings.

The provision of home support workers,76 midwife managed care,77 and postnatal check up at six weeks78 have all been found in randomised controlled trials to have no significant effect in women with no complications in pregnancy or delivery. No long term effect of a doula (female birthing companion) was found in a study of pre-term babies in South Africa, although the follow-up rate in this study was low (50%).79 An antenatal preparation for parenthood intervention (six one-hour antenatal classes) was found to have no effect on depression or psychosocial risk factors at three months, although only 45% of women attended more than two sessions.80 A small study found that individualised, family-based interventions resulted in improved maternal psychological well-being.81 Evidence level 1+

Weekly visits by a child health nurse significantly reduced the scores on the EPDS at six weeks in a group of mothers screened for high risk factors, although only 53% returned questionnaires.82 Reduced anxiety and depression has also been found after debriefing by midwives, although the morbidity in the control group in this study was high (55% above threshold for depression on the Hospital Anxiety and Depression (HAD) scale). A large Australian study found no effect of debriefing in women following operative childbirth.83,84 In another small study, women with at least one predictor of postnatal depression who were given interpersonal therapy had reduced Beck depression scores compared with the control group, but 33% of the control group had major postnatal depression.85 Evidence level 1+

A cohort study in mothers identified antenatally as vulnerable to depression found that first time mothers who had taken part in a parenthood educational programme had significantly lower EPDS scores compared with routine care.86 Evidence level 2+

A placebo controlled randomised clinical trial examined the use of nortriptyline to prevent recurrence of depression in 51 postnatal women who had a previous history of postpartum depression. No difference in the rate of recurrence was found between the two groups.87 Evidence level 2+

In a small study of eleven women with a history of puerperal psychosis or puerperal major depression but no history of non-puerperal depression, prophylactic oestrogen prevented relapse in all but one.88 Evidence level 2+

The current research base for preventive interventions in low risk women is extremely limited.


Two cohort studies have examined the use of prophylactic lithium given either in late pregnancy or immediately after delivery for the purpose of preventing the development of puerperal psychosis in high risk women. The first found that fewer than 10% of treated patients developed illness, substantially lower than their estimate of 20%.89 The second study examined 27 women with bipolar mood disorder.90 Only one of 14 who received mood stabilisers developed puerperal psychosis compared with eight of 13 untreated women.90 Both these studies were limited by their open design, but their findings are reinforced by one study showing a high risk of puerperal relapse in bipolar women who discontinue lithium during pregnancy.127Evidence level 2-

The evidence suggests that lithium is an effective treatment when used to prevent puerperal psychosis in high risk groups but it is not of sufficient quality to support a recommendation.

[Contents][Back][Top] [Next]