Correspondence: Ian Jones. Email: email@example.com
- Postpartum psychosis is a severe mental illness with a dramatic onset shortly after childbirth.
- All women should be screened antenatally for the known risk factors.
- Women with bipolar disorder have at least a 1 in 4 risk and need close contact and review during the perinatal period even if they are well.
- Prompt recognition of the illness and rapid institution of treatment are of vital importance.
- To recognise women at high risk for severe postpartum mental illness.
- To recognise and appreciate the severity of postpartum psychosis and the need for prompt assessment and treatment.
- Who should ultimately make decisions about taking medications in pregnancy – the clinician or the woman and her family?
- What advice should a woman at high risk of postpartum psychosis be given if she is considering pregnancy?
Postpartum mood disorders are of great clinical and public health importance, with suicide a leading cause of maternal death in the UK. They are commonly divided into three categories: the ‘baby blues’, postnatal depression, and postpartum (or puerperal) psychosis. This review focuses on postpartum psychosis (PP), emphasising the importance of recognising women at high risk, and the early recognition and prompt treatment of women who develop the illness. According to the Confidential Enquiries into Maternal Deaths in the United Kingdom report, a diagnosis of severe affective illness was present in about 60% of the women who committed suicide. Although suicide after childbirth is rare, affecting about 1 pregnancy in 100 000, suicide remains a leading cause of maternal death. Suicide in a new mother is a great tragedy, and for each woman who takes her life there are many near misses.
The concept of postpartum psychosis
The term ‘postpartum psychosis’ refers to a severe mental illness with a dramatic onset shortly after childbirth. It most commonly takes the form of mania, severe depression, or a mixed picture with features of both high and low mood. In addition, women show evidence of psychosis with delusions and hallucinations common, and also will often demonstrate marked confusion or perplexity.
There is little consensus regarding the classification of postpartum psychosis. Modern classification systems do not recognise PP as a separate nosological entity. In the International Classification of Diseases (ICD-10), the category ‘mental and behavioural disorders associated with the puerperium, not elsewhere classified’ should only be used when unavoidable and includes only mental disorders commencing within 6 weeks of delivery that do not meet the criteria for other diagnoses. In the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) postpartum affective episodes are treated as mood disorders with a postpartum trigger: patients must meet the criteria for a mood episode and the criteria for the postnatal-onset specifier. In DSM-IV a postnatal onset is considered to be within 4 weeks of delivery. Despite these issues, however, the term postpartum (puerperal) psychosis has remained in common clinical use.
Record-linkage studies estimate the admission rates to psychiatric hospital in the postpartum as about 1–2 per 1000 births in the general population. However, the true incidence rates for severe postpartum affective disorder may be higher, as at least some women with PP are likely to be treated at home – particularly if facilities for admission with the baby are not available.
There is consistent evidence of a specific relationship between PP and bipolar disorder (BD). Women with BD have at least a 1 in 4 risk of suffering a severe recurrence following delivery. Bipolar women with a personal or family history of PP are at particularly high risk with greater than 1 in 2 deliveries affected by PP.
Aetiology and pathophysiology
Most episodes of PP can be considered as affecting women with a bipolar disorder diathesis acted on by a specific puerperal trigger. A number of factors have been suggested that increase vulnerability to the puerperal triggering of episodes of BD.
There is robust evidence that the vulnerability to the triggering of affective psychosis by childbirth aggregates in families and may define a genetically relevant subtype of bipolar disorder. Evidence from family studies suggests that episodes of PP are a marker for a more familial form of BD and that a specific vulnerability to the puerperal triggering of BD is familial. Evidence from a linkage study indicated the possible location of a susceptibility gene on chromosome 16. Particular candidate genes, such as those involved in the serotoninergic[11, 12] hormonal[13, 14] and inflammatory pathways, have also been investigated. It is hoped that identifying the genetic factors that increase risk will lead to more individualised risk assessment, earlier identification of women at risk, and improved treatments for women who become ill.
Obstetric risk factors
An increased risk of PP has been reported with a number of obstetric factors including: primiparity; pregnancy and delivery complications; delivery by caesarean section; having a female baby; and a shorter gestation period. However, findings are consistent only for primiparity. The bias that women with a severe postpartum episode may be less likely to go on to have further children is unlikely to be the sole, or even the main, explanation. Given that there is little evidence of an association between PP and psychosocial factors, the possibility remains that the effect of primiparity is, at least in part, due to biological differences between first and subsequent pregnancies. In this regard, the overlap with other pregnancy related disorders that also occur more frequently in first pregnancies, such as pre-eclampsia, is of interest. Hormonal, immunological and other biological differences between first and subsequent pregnancies are interesting targets for further investigation into the aetiology of PP.
Changes in medications
Women with BD often come off mood stabilisers, such as lithium, preconception or in early pregnancy because of concerns over toxicity to the fetus. A survival analysis comparing women with BD who stopped taking lithium because of pregnancy compared with age-matched non-pregnant women who discontinued lithium for other reasons, reported similar rates of recurrence during the first 40 weeks after lithium discontinuation for both groups. However, among subjects who remained stable over the first 40 weeks after lithium discontinuation, postpartum recurrences were 2.9 times more frequent than recurrences in non-pregnant women during weeks 41–64 (70% versus 24%). Thus, the increased risk of recurrence following childbirth for women with BD does not appear to be merely a result of stopping mood-stabilising medication.
The lack of evidence implicating psychosocial factors and consideration of the abrupt onset during a time of major physiological change suggest that biological, possibly hormonal, factors are important. The role of several hormones (including estrogen, progesterone, prolactin, follicular stimulating hormone and luteinising hormone) has been considered, but the evidence pointing to hormones in the aetiology of PP remains predominantly circumstantial.
A plausible hypothesis is that the sleep deprivation of delivery and the immediate postpartum period is responsible for puerperal triggering of illness. Sleep loss can effectively trigger the onset of mania in people with BD and sleep loss is, of course, common for new mothers.
Screening for risk factors
In addition to a history of BD or PP, other risk factors for PP include having a first-degree relative who has experienced PP and having a first degree relative with BD. For women who themselves have a history of mood disorder, particularly on the bipolar spectrum, a family history of severe postpartum episodes is very important and may indicate a risk in excess of 50%. For women who have not suffered episodes of psychiatric illness themselves, it is not so clear that family history is relevant. While a risk of PP of 1–3% in such women represents a considerable increase on the population rate of around 1–2 in 1000 deliveries, it is unclear whether extensive efforts to identify women who are well but with a family history is a worthwhile strategy. These women may benefit from the risks being discussed with them if it is something they have identified as a concern, but it also is possible that it may cause unnecessary worry in women who will not go on to develop illness.
Because of the relapsing and remitting nature of BD, women at high risk are often currently well and not in contact with mental health services and may fail to recognise the serious risk of their situation. Thus, all women should be screened for known important risk factors at their antenatal booking visit. Protocols should be put in place to ensure that women at potential risk receive a formal risk assessment and management plan.[19, 20]
Women with schizophrenia also have an increased risk of hospitalisation after childbirth. However, the specificity of the childbirth trigger in schizophrenia is still controversial. Women with schizophrenia have a four-fold lower relative risk of admission in the postpartum period compared with those with BD. In many cases, schizophrenia is a severe chronic illness and women are commonly admitted for assessment of parenting or to help them to cope with the newborn rather than for the acute onset of a new episode.
Management of women at high risk
Women at high risk of PP need very careful care before conception, throughout pregnancy and during the postpartum period. The high risk of illness in the weeks following delivery in a woman with a history of BD must be recognised both by healthcare professionals and by the woman herself.[19, 20]
The management of women at risk raises ethical questions on the role of the patient and her family in the decision-making process. Ethical principles of patient-centred care provide the foundation for the doctor–patient relationship: autonomy, justice, beneficence and nonmaleficence. The clinician should avoid paternalistic attitude, exploring and considering the values and the expectations of the patient and leaving the final informed decision to the woman. If the woman wishes, family members can be involved in the decision making process. The NICE guidelines suggest to discuss the teratogenic risk explaining the background risk of fetal malformations in the general population (around 2–4%) and to describe the risk using natural frequencies rather than percentages (for example, 1 in 10 rather than 10%). A written plan covering pregnancy, delivery and the postnatal period should also be developed and discussed with the woman and her family.
The possibility of future pregnancy should be considered in all women with BD who are of childbearing age. Ideally the pre-conception counselling should be conducted by a perinatal psychiatrist, however, depending on availability, other psychiatric teams or GPs can provide the necessary information to woman. The risks of illness following childbirth should be discussed with women and the importance of seeking help emphasised. Decisions about continuing or stopping medications before, or during, pregnancy are difficult and should be the result of a detailed and individualised risk analysis. Although there are significant concerns about the teratogenic effects of the medications used to treat BD, the risks of stopping medication must also be considered. Data suggest that women with BD who stopped medication during pregnancy were more than twice as likely to experience a recurrence than those who remained on medication. Data on the teratogenicity of psychotropic agents are often controversial and an exhaustive discussion is beyond the scope of this review. Consistent evidence suggest that valproate should be avoided, because of the high risk of malformations, and because of negative effects on neurodevelopment. On the contrary, a recent meta-analysis found no significant association between any major congenital abnormality and lithium. However, due to the wide confidence limits, considerable uncertainty about the risk of lithium remains.
During pregnancy and after childbirth
Perhaps the most important aspect of care is to maintain close contact and review during the perinatal period. Women at high risk, even if they are well, should be referred in pregnancy for psychiatric assessment and monitored regularly for at least 3 months following delivery. The good practice guidelines developed by Royal College of Obstetricians and Gynecologists suggest that the following scenarios are indications for referral to specialised perinatal mental health services where available, otherwise general psychiatry services:
- current severe psychiatric symptom
- a history of serious postpartum illness or bipolar disorder or schizophrenia
- on complex psychotropic medications schemes.
Moreover, the guidelines suggest that referral should be considered for those with moderate symptoms developed in late pregnancy or early postpartum or mild symptoms and a family history of bipolar disorder or puerperal psychosis (COG). Psychiatric services should have priority care pathways for pregnant and postpartum women and care by multiple psychiatric teams should be avoided. Ideally, women should be managed by multiprofessional/multidisciplinary teams, with a named obstetrician (possibly with special interest in perinatal mental health), midwives, perinatal psychiatrist, community psychiatric nurse, health visitors and GP. All communication between maternity and mental health services should include primary care. A written care plan should be developed in collaboration with all relevant healthcare professionals and recorded in all versions of the woman's notes.[19, 20]
It may also be important to address other avoidable factors that may increase risk – such as decreasing general levels of stress and paying attention to sleep in late pregnancy and the early postpartum weeks. Liaison with maternity services should involve discussion about how to manage the labour to reduce the sleep deprivation that can occur if labour is prolonged. For women with a history of BD who have been off medication in pregnancy the introduction of prophylactic medication in the immediate postpartum period should be considered. Some evidence exists for the use of lithium in this context, but the few studies have been open and retrospective and there are practical problems with reaching therapeutic levels quickly to cover the period of risk. These issues have led some perinatal psychiatrists to use typical or atypical antipsychotics as prophylaxis.
Diagnosis of postpartum psychosis (PP)
History and examination
Although all women should be assessed antenatally for known risk factors, such as personal or family history of BD and psychosis, it is important to bear in mind that 50% or more of women who develop PP have no history that puts them in a high-risk group.
The distinctive clinical features include sudden onset and rapid deterioration. The vast majority of episodes have their onset within 2 weeks of delivery, with over 50% of symptom onsets occurring on days 1–3. The clinical picture often changes rapidly, with wide fluctuations in the intensity of symptoms and severe swings of mood. Common symptoms and signs include:
- A wide variety of psychotic phenomena such as delusions and hallucinations, the content of which is often related to the new child.
- Affective (mood) symptoms, both elation and depression.
- Disturbance of consciousness marked by an apparent confusion, bewilderment or perplexity.
- Primary cerebral or systemic disease (such as eclampsia or infection) should be excluded. The misattribution to psychiatric disorder has led to a number of deaths in new mothers.
- Exogenous toxic substances or hormones: History of therapeutic use and/or abuse of known causative substances or hormones, other symptoms and signs specific to the substance or substances involved should be investigated. Urine drug screen may be positive in substance abuse and identifies the substance taken, although it is not definitive for drug misuse.
- Other psychiatric disorders of the puerperium: Baby blues affects 30–80% of births and causes transient emotional lability during the first postpartum week. The mother typically presents mood swings ranging from elation to sadness, insomnia, tearfulness, crying spells, irritability, anxiety, and decreased concentration. Care of the baby is not impaired, hopelessness and worthlessness are not prominent, and women do not feel suicidal. It is self-limiting, but assessment should ensure that the woman is not and does not become more severely depressed.
- Postnatal depression (for a review see Musters et al.). The tendency for all postpartum episodes to be labelled as postnatal depression can lead to suboptimal care and, in some cases, have dramatic consequences on mothers and babies.
Any psychotic symptoms, particularly delusions or hallucinations, substantially increase risk for both mother and child. The woman should be referred for a same day emergency appointment so that a detailed risk assessment can be carried out.
Management of women with postpartum psychosis (PP)
PP is a psychiatric emergency. The clinical picture may mislead, quickly become extremely severe and vary significantly from hour to hour. Admission is usually necessary, even for women with the most supportive of families. The NICE guidelines recommend that women within a year of childbirth should be offered admission to a specialist mother and baby unit, however, the provision of services across the UK is patchy and for the majority of women there is no option of admission with her baby.
A range of psychotropic medication may need to be employed. The treatment used depends on a number of factors, including the symptoms that the woman experiences, her level of disturbance and her previous response to medication. For many women the severity of the illness does not allow breastfeeding. If breastfeeding is being considered, factors in the baby such as prematurity and systemic illness should be considered in addition to the particular properties of the medication itself. Limited data suggest that the use of lithium during breastfeeding is not as problematic as once thought but is usually avoided because of the risk of toxicity in the baby.
The short-term prognosis for PP is generally good. However, women need to be counselled about the risks they run of a further puerperal or non-puerperal episode. This will include discussing the need for longer-term mood stabilising medication and other measures that can reduce the risk of recurrence. Despite the high risk of recurrence following further deliveries, many women make the decision to become pregnant again and it is our view that women with PP, or indeed women with bipolar disorder more generally, should not be told that they should not have children.
Neglect of the baby
Referral to safeguarding teams should not be routine, but should take place as the result of a risk assessment. Extra vigilance and care are required in these cases, as it may increase the risk of deterioration in the mother's mental health, and even lead to suicide.
Suicide or infanticide
When discussing symptoms of low mood, sensitively asking whether the woman feels that life is not worth living and whether she has ever thought of harming her baby enables the clinician to assess the most serious aspects of risk. It is important to note that asking about suicidal thoughts does not increase, but rather is likely to reduce, the risk of a woman taking her life.
Although there is a paucity of information on rates of non-puerperal recurrences, Robertson et al. found that following the index episode of PP, 62% of women experienced at least one non-puerperal affective episode during a median 9 years of follow-up.
Postpartum psychosis is a severe condition complicating childbirth following approximately 1 in 1000 deliveries. In 50% of cases there is no prior history of psychiatric disorder. In other women, however, there are clear factors – a history of bipolar disorder or previous episode of PP – that identify women who are at very high risk (50% or more). Women at high risk need to be identified in pregnancy and referred to psychiatric services for further assessment. PP is a true psychiatric emergency and it is vital that it is recognised early and treated aggressively.
Disclosure of interests
ADF, IRJ and SS report no conflict of interest.