The prevalence and range of psychiatric conditions in early pregnancy is the same as in the female population of comparable age. Mental illness during pregnancy can cause management problems and effect the pregnancy outcome and fetal and infant development. Women with current mental illness may be taking medication. Stopping medication in pregnancy can lead to relapse that may compromise management. Continuing medication may affect the developing fetus. The obstetrician will be asked to give advice on medication and undertake a risk benefit analysis.

A substantial proportion of women seen by maternity services will have mental health problems. A significant, although smaller number, will be well but have risk factors for serious postpartum illness that will need to be proactively managed.

There are effective treatments for serious mental illness and the prognosis for postpartum onset affective disorder is good. However, if untreated or with delays in detection, morbidity may be prolonged with adverse effects on the family life and infant.

Perinatal psychiatry

Perinatal psychiatry is the internationally recognized term for psychiatric disorders that complicate pregnancy, childbirth and the postpartum period.

Perinatal psychiatric disorders include:

• New onset conditions in previously well women.

• The recurrence of conditions in women who have been well for some time but have a history of a previous illness.

• Relapses or deteriorations in women who are currently ill or who have not fully recovered from a previous illness.

Perinatal psychiatry is also concerned with the effects of maternal illness and treatment on the developing fetus and infant.

Specialized perinatal psychiatric services are organized to meet the special needs of women and work in close relationship with maternity services.

Antenatal psychiatric disorders

Antenatal psychiatric disorders are common, affecting between 15 and 20% of pregnancies. In early pregnancy the rate and range of psychiatric disorder will be the same as in the non-pregnant population of the same age. Maternity services will therefore see women with learning disability, substance misuse, schizophrenia, bipolar disorder, depressive illness, obsessional compulsive disorder and anxiety states; these women are likely to be receiving medication. Enquiries should be made in early pregnancy about previous psychiatric history and current mental health and medication.

The incidence of psychiatric disorder during pregnancy is slightly increased, due to an increase in the rates of anxiety and depression. The incidence of serious mental illness is much reduced in pregnancy, in marked contrast to the postpartum period.

Psychiatric medication in pregnancy

At least 10% of women will be taking psychiatric medication at the beginning of pregnancy, usually antidepressants but sometimes an antipsychotic or mood stabilizer.

With the exception of mood stabilizers, the evidence for adverse effects of antidepressants and antipsychotics is emerging and conflicting. It is likely that the absolute risks are small and need to be balanced against the established risks of stopping maternal medication and the effects of relapse on maternal and fetal health.

It is not possible to give a definitive list of psychiatric drugs that are ‘safe’ and ‘unsafe’. The reader should always check the latest systematic reviews before giving advice or initiating a prescription.

Antidepressant medication

Monoamine oxidase inhibitors (MAOIs)

These are now rarely used. They should not be used in pregnancy because of the risks of interaction with certain food stuffs and analgesics.

Tricyclic antidepressants (TCAs)

These have been in use for 40 years. They include amitriptyline, imipramine, clomipramine and doxepin (dothiepin). The usual therapeutic dose is 150 mg daily, and tablet size allows for adjusting dosage by 25 mg increments. With the exception of clomipramine, there is no evidence that TCAs are associated with an increased risk of structural or functional fetal abnormalities, early pregnancy loss, restricted interuterine growth or early delivery.

Clomipramine may be associated with the same increased risk of cardiac abnormalities as SSRIs and should be used with caution in pregnancy.

TCAs, if taken at full dose prior to delivery, can cause neonatal withdrawal effects. These include jitteriness, convulsions (rarely), hypoglycaemia, hypothermia and problems feeding. These effects, however, are short-lived.

Many practitioners would consider reducing the dose of TCAs prior to delivery. However, for women who are seriously depressed a reduction may result in a relapse. In such circumstances a managed delivery at or shortly before term should be considered.

Current guidelines support the use of TCAs in pregnancy

If antidepressants are indicated in pregnancy because of a profound depressive illness with the characteristic ‘biological’ symptoms, then a suggested regime would be 50 mg of imipramine or amitriptyline increasing by 25 mg every few days until a therapeutic dose of 150 mg is reached.

Side-effects of TCAs include dry mouth, blurred vision and sometimes difficulty passing urine. TCAs, particularly dothiepin, are cardiotoxic in overdose. TCAs should not be prescribed if there is a risk of suicide.

Selective serotonin reuptake inhibitors (SSRIs)

These are the most frequently prescribed antidepressants. They include fluoxetine, paroxetine, sertraline and citalopram.

SSRIs, particularly paroxetine, have been linked with cardiac abnormalities, specifically ventricular septal defect. The evidence is emerging and conflicting, but specific concerns have led the drug administrations in the UK and US to advise against the use of paroxetine in pregnancy. SSRIs as a class may be associated with increased rates of early pregnancy loss, growth restriction, early delivery and pulmonary hypertension in the newborn.

The evidence of neonatal compromise when taking an SSRI prior to delivery is more robust. Preterm babies are particularly vulnerable. Neonatal compromise includes difficulties in feeding, hypoglycaemia and hypothermia. However, these are temporary.

Despite the likely increase in the relative risk of these effects, the absolute risk is low (with the exception of neonatal compromise).

Current guidelines support the use of fluoxetine and sertraline in pregnancy

Some clinicians would advise the withdrawal of SSRIs prior to delivery. However, this may compromise the management of severely depressed women when a managed delivery may be considered and a watchful eye kept on the neonate.

Serotonin–norepinephrine reuptake inhibitors (SNRIs)

The lack of evidence of their safety and concerns about pulmonary hypertension suggest that these antidepressants should not be used in pregnancy.

Antipsychotic medication

There are two broad groups, the older ‘typical’ antipsychotics such as chlorpromazine, trifluoperazine and haloperidol, and depot long-lasting antipsychotics such as Modecate® and the newer ‘atypical’ antipsychotics such as olanzapine, quetiapine and risperidone. These drugs are used to treat and maintain schizophrenia, episodic psychoses and bipolar illness. Stopping them leads to a marked elevation in the risk of relapse in the next 6 months. Both groups of antipsychotics are equally effective but differ in their side-effects and acceptability.

Older ‘typical’ antipsychotics

These have been in use for 40 years. There is no evidence that they are linked with fetal abnormality or early pregnancy loss. However, schizophrenia, an indication for their use, is associated with poor obstetric outcome, including early delivery, increased rates of caesarean section, maternal and neonatal mortality, and neurodevelopmental problems.

Anticholinergic drugs used to counteract the extra pyramidal side-effects of typical antipsychotics cannot be regularly used in pregnancy.

Postural hypotension, particularly with chlorpromazine can cause falls and theoretically compromise placental function.

If well maintained on typical antipsychotic agents, they should not be changed during pregnancy. However, a watchful eye should be kept on the neonate for withdrawal symptoms to the sedative effective of maternal medication.

Current advice is that typical antipsychotic drugs may be used during pregnancy. However, depot injections such as Modecate should be avoided because of difficulties adjusting the dose.

Newer ‘atypical’ antipsychotics

There is less evidence about their effects in pregnancy than the older typical antipsychotics. There is no evidence to suggest an association with structural or functional fetal abnormality. The risk of maternal relapse if they are stopped is substantial.

There is consistent evidence linking atypical antipsychotic agents, particularly olanzapine, with gestational diabetes and rapid, substantial weight gain. There are also concerns of increased risk of venous embolism. Women on atypical antipsychotics should be closely monitored by an obstetrician. A watchful eye needs to be kept on the neonate for potential withdrawal symptoms from the sedative effect of maternal medication.

Current guidelines advise that atypical antipsychotics may be used during pregnancy. Clozapine is the exception. It is used to treat refractory schizophrenia and should not be used in pregnancy.

The adverse effects of all antipsychotics are dose related. The lowest effective dose should therefore be given. Maintaining medication will lower the risk of relapse and avoid having to use much higher doses in an acute crisis.

Mood stabilizers

Lithium carbonate and anti-epileptic drugs (AEDs) mainly sodium valproate and valproate semisodium (Depakote®) are used to treat and maintain women with bipolar disorder.

Lithium is associated with an increased risk of cardiac abnormalities of all types; the risk is 1 in 10 in exposed pregnancies. It is specifically linked to a marked increased risk of Ebstein’s anomaly. However, the absolute risk of this rare abnormality is low. Ideally women taking lithium should receive advice about the risks to their mental health due to pregnancy and the effects of lithium withdrawal that should occur before conception. If a woman becomes pregnant whilst taking lithium the following guidelines should be followed.

If she has been well for over 2 years the lithium should be slowly withdrawn under psychiatric supervision by 200 mg every 2 weeks. She will require close monitoring for any evidence of a relapse when an antipsychotic agent should be used.

If she has been recently ill or has relapsed after previous withdrawal of lithium, then continuation of the lithium may be the least detrimental alternative at the lowest dose to maintain an effective serum level.

A level three ultrasound scan should be arranged at 22–24 weeks to look for fetal cardiac abnormalities.

Theoretically, lithium can be used in the second trimester of pregnancy. However, its use in later pregnancy is problematic. Maternal serum levels will fall but the fetal level will equilibrate and fetal clearance of lithium is less than the mother’s. Lithium in later pregnancy is associated with fetal hypothyroidism and polyhydramnios. During delivery, with physiological diuresis, maternal serum lithium levels may suddenly rise to toxic levels and there may be neonatal lithium toxicity. Using lithium in pregnancy is a high-risk strategy requiring close collaboration between the obstetrician and psychiatrist. Serum lithium levels need to be checked weekly in the last trimester. The woman should be induced before term to allow for the lithium to be stopped 10 days prior to delivery. If a woman starts labour whilst taking lithium it should be immediately stopped, hydration and diuresis maintained and intravenous access obtained. The neonatal paediatrician should be alerted.

Anti-epileptic drugs

All AEDs (the possible exception of lamotrigine) are associated with an elevation in the risk of fetal malformations in general and neural tube defects in particular. The risk with sodium valproate is higher than for other AEDs, particularly for neural tube defects, neurodevelopmental delay and impaired cognitive functioning in school age children. Between 8% and 15% of all exposed pregnancies are affected, depending on dosage and polypharmacy.

Sodium valproate and valproate semisodium (Depakote) are widely used as mood stabilizers. It is no longer thought that epilepsy itself is responsible for these increased risks, and it is now accepted that anticonvulsants are responsible. Valproate used for psychiatric reasons will have the same risks.

Current guidelines on the management of epilepsy, bipolar illness and antenatal mental health advise that valproate should not be used in women of reproductive age unless there are no effective alternatives.

If a woman becomes pregnant whilst taking valproate, the following guidelines should be observed. She should be urgently reviewed by both a consultant obstetrician and psychiatrist and collaboratively managed.

The dose should be reduced to 800 mg daily or less. Daily dosage of long-acting preparations should be converted to a twice or thrice daily regime to minimize ‘pulsing’ of the fetus. The earliest anomaly scan should be arranged and the ultrasonographer warned of the fetal exposure. Valproate should be slowly withdrawn, e.g. by 200 mg every 2 weeks, and if necessary replaced by an antipsychotic drug.

Careful monitoring is required because of the possibility of a relapse.

Antenatal screening

The only reliable risk factor with a high positive predictive value for postnatal mental illness is a history of episodes following childbirth and at other times.

Women should be asked at the early pregnancy assessment about their psychiatric history, and serious illness distinguished from other conditions.

At least 50% of women with a previous history of bipolar illness, severe depressive illness, severe postnatal depression or postpartum (puerperal) psychosis will become ill.

Women with a past history of serious mental illness should be referred for psychiatric assessment during pregnancy and have a peripartum plan. The postnatal illness is likely to be severe, arise suddenly in the early days following delivery and deteriorate quickly; they may benefit from preventative treatment. Close monitoring and supervision is required. A management plan allows for early detection, prompt treatment and the avoidance of delay in providing appropriate care.

Women should be asked about a family history of bipolar disorder or severe depressive illness and if a maternal relative was ill following childbirth.

A family history of bipolar illness is a risk factor for postpartum psychosis. The risk is approximately 3% (compared with 0.2% in the general pregnant population). If the family history is of postpartum onset conditions then the risk is higher; however, between 94% and 97% of women will remain well. Unless the woman is concerned, a positive family history is not necessarily an indication for referral to psychiatric services. However, all should remain vigilant and have a lowered threshold for concern if symptoms develop following delivery.

Women should be asked in early pregnancy about substance misuse.

Substance misusers should be referred to specialized midwives and drug addiction services. They are not appropriately managed by specialized perinatal mental health services or general adult services nor by general practitioners.

Women should be asked about their current mental health in early pregnancy and on at least two occasions thereafter.

It is recommended that the Whooley questions are used:

1. During the past month have you often been bothered by feeling down, hopeless or depressed?

2. During the past month have you often been bothered by having little interest or pleasure in doing things?

3. Do you feel you need or want help with this?

Between day 3 and day 10 most experience for 48 hours emotional lability, tearfulness, exhaustion, anxiety, irritability, a tendency to catastrophize and worry about coping. The commonest day of occurrence is day 5. Similar feelings can recur periodically over the next 6 to 8 weeks, particularly if the baby has been ‘difficult’ and there has been lack of sleep.

The ‘blues’ are normal but can be distressing to a woman who has not been forewarned. No ‘treatment’ is required but understanding, explanation and reassurance is needed. If there is a risk of depressive illness, the midwife needs to be vigilant that the condition is resolving.

Postnatal psychiatric disorders

The full range of psychiatric disorders can complicate the postnatal period. Postnatal mood disorders have an increased risk of occurrence following delivery. Distinctive clinical features, and outcome are described in this chapter.

Aetiology

All women are vulnerable to postpartum mood disorders. Childbirth results in major changes to role, expectations and relationships. There are physiological, physical and neuroendocrine changes and normal increases in anxiety and instability of mood and sleep deprivation.

More vulnerable are those with relationship difficulties, socioeconomic problems, domestic violence, having been in care or sexually abused and those with a sick child or bereaved.

The aetiology of postpartum psychosis and severe postnatal depression is thought to be genetic and neuroendocrine. There is a heritable genetic vulnerability to serious affective disorder and a specific postpartum trigger. This is currently thought to be an abnormal sensitivity of the dopamine and serotonin receptors to the sudden fall in estradiol postpartum.

Psychosocial factors are most important in the aetiology of mild to moderate conditions and genetic and neuroendocrine factors for serious mental illness.

Postpartum (puerperal) psychosis

This is the least common and most serious of the postpartum conditions occurring in 2/1000 deliveries in women of all ages, backgrounds and in all cultures and countries in the world. It is commoner in first-time mothers and in older mothers and those who have had emergency caesarean section after a first baby. Risk factors include a family history of bipolar illness, a maternal family history of postpartum psychosis and previous episodes of bipolar illness, schizo-affective disorder or postpartum psychosis (Box 14.2).

Approximately 50% of women with a previous bipolar illness or postpartum psychosis will become ill. This risk justifies assessment and monitoring during pregnancy and with the woman’s consent prophylactic intervention following delivery.

The illness is characterized thus:

• Sudden onset in the early days following delivery, deteriorating on a daily basis.

• Half will present within first postpartum week, the majority within 2 weeks and almost all within 3 months of delivery.

• Psychosis, delusions, fear and perplexity, confusion and agitation and sometimes hallucinations.

• Agitation and severe disturbance.

In the early days of the illness, the picture changes frequently and is often called ‘an acute undifferentiated psychosis’. Later it is more clearly a bipolar illness. A third will be manic and the rest usually mixed with some symptoms of mania but a predominantly depressive content.

Management

Urgent admission to an inpatient mother and baby unit is usually necessary. These women should not be admitted to a general adult psychiatric unit. Specialized medical and nursing care is required. The admission of the baby with the mother is not only humane but will facilitate the mother’s treatment and ensure a good relationship with her infant.

These illnesses respond rapidly to treatment. Antipsychotics, antidepressants and mood stabilizers may be used and, on occasion, electroconvulsive therapy (ECT). The prognosis for a full recovery is good.

Postpartum psychosis can be a life-threatening condition with an elevated risk of suicide and accidental harm from disturbed behaviour. There is risk to physical health from not eating and drinking, not accessing medical care and the woman may be temporarily unable to care for the baby.

Treatment needs to be continued for some time after recovery because the risk of a relapse in the early weeks is high, particularly if she has been manic and may relapse into a depressive state.

The risk of recurrence following all future pregnancies is at least 1 in 2. She should therefore be referred early in her next pregnancy and a management plan put into place.

Although often known as ‘PND’, this is not one illness but a range of subtypes of depressive illness of different severity. The clinical symptoms of depressive illness following childbirth are similar to those at other times, but there will be additional features relating to the maternity context. Overall the incidence is 10% of all new mothers.

Severe postnatal depression

This affects about 3% of all deliveries. Important risk factors are a previous severe depressive illness and a family history particularly of postpartum onset. Psychosocial risk factors may be present. However, severe PND can occur in women from all backgrounds and favourable circumstances. A previous stillbirth or neonatal death increases the risk as does infertility, in vitro fertilization (IVF) and serious obstetric concerns during pregnancy.

Onset is gradual in the first 2 weeks following delivery, but becomes more severe and presents within 3 months of delivery. Presentation is often associated with the withdrawal of practical support (husbands going back to work, grandmothers returning home, etc.).

The classical symptoms of early morning wakening, mood worse in the morning, slowing up, impaired concentration and difficulty coping may be masked by the tasks of new motherhood.

Women with severe postnatal depressive illness feel guilty, have ideas of worthlessness, lack of enjoyment and lack of spontaneity. They are often preoccupied with ruminative worry and are very anxious (Box 14.3). Intrusive obsessional thoughts of harm coming to their babies and panic attacks are common. Sometimes they are preoccupied by the birth experience and may have some features of obstetric post traumatic stress disorder.

Management

Women with severe postnatal depression can deteriorate quickly. Early detection, prompt assessment and treatment are imperative.

Treatment is antidepressants and psychological support. Admission to a mother and baby unit may be necessary if severely ill or suicidal. The illness has a good prognosis and recovers within 8 weeks. Treatment will need to be continued for at least 6 months. However, although recovery is usually complete, there is a 1 in 2 risk of it occurring again after the next baby.

Mild to moderate postnatal depression

This is the commonest postpartum condition. The most important risk factors are psychosocial.

The symptoms will be less severe without the tendency to rapidly deteriorate and worsen and may be very variable with good days and bad days. They will often feel better in company and worse when alone. They may not have the classical sleep disturbance and loss of vitality and pleasure but nonetheless may be distressed by their lack of pleasure and enjoyment in their babies (Box 14.4). Anxiety is a prominent feature.

Management

The more severe end of the spectrum will benefit from referral to specialized community perinatal psychiatric services. The less severe will often benefit from listening visits from health visitors, social support and self-help groups.

Left untreated, two-thirds of women affected will improve by 6 months postpartum although returning to work may be preceded by an increase in symptoms. With treatment the majority should improve sooner.

Effects on the child

Chronic, untreated postnatal depression associated with social adversity can affect the infant’s social, emotional and intellectual development for many years. Prompt effective treatment of the mother is essential for the child. Additional help with the mother–infant relationship may be necessary.

There is no evidence that antenatal screening for risk factors for postnatal depression or psychosocial antenatal interventions prevents PND, however regular postnatal visiting by a health professional in an at risk population has been shown to reduce rates of PND. There is no evidence that prophylactic antidepressants prevent PND depression.

Hormones

There is some evidence that transdermal estradiol (100–200 mg twice weekly) can prevent and treat non-psychotic postnatal depression. It is likely that estradiol acts as an antidepressant. Other concerns about using estradiol at this time mean that it is not recommended as a first-line treatment nor licensed for this purpose. There is no evidence that progesterone prevents or treats postnatal depression. There is some evidence that it may make depressive symptoms worse. Its use is not recommended.

Psychiatric medication in breastfeeding

All psychotropic medications will be present in breast milk. In general, the greatest concern is breastfeeding a newborn baby. As the baby becomes older and heavier and particularly when solids are introduced, the amount of any maternal drug in the breast milk in relation to the baby’s body weight is reduced.

Antidepressant medication

TCAs

Amitryptiline, imipramine and dothiepin can be used in breastfeeding. The amount of maternal drug in the breast milk is small. However, the sedative side-effects may compromise the mother’s ability to care for her infant. When feeding a newborn, a thrice daily regime is advisable, e.g. amitryptiline 50 mg tds.

SSRIs

The half-life and quantities in the breast milk vary. Fluoxetine has a long half-life and is best avoided when feeding a newborn. Sertraline is a better choice. If the baby is over 3 months old other SSRIs can be used. Mothers can be advised to take their daily dose after the infant’s last feed and before its longest sleep.

Antipsychotic medication

There is little evidence of the quantities of atypical antipsychotics in breast milk. They are best avoided when feeding the newborn.

Small doses of typical antipsychotics (up to 5 mg daily of haloperidol or trifluoperazine) can be used but breastfeeding should be suspended if anticholinergic drugs are necessary for treating extrapyramidal side effects as they may cause bradycardia in the newborn.

Most had a previous psychiatric history. Rarely was this risk identified in pregnancy and a management plan put in place. This failure to acknowledge the risk of recurrence was not only in the maternity services, but also in the general practice and psychiatric services. Had the risk been identified and managed then at the least these women would have been closely monitored following delivery and at best some preventative action might have been taken.

The women who died

The majority were older, married or cohabiting, employed and educated, particularly those suffering from serious illness. The majority were well in pregnancy and for many years before. They would not have been recognized as ‘vulnerable’. It should be remembered that serious illness can also affect those in comfortable circumstances with no obvious personal or social problems.

Communication

Often midwives and obstetricians did not obtain or were not given vital information about a previous history by the general practitioner. In many cases the general practitioner was not aware of the pregnancy. There was poor communication between maternity and psychiatric services.

The care received

The majority were cared for by general adult psychiatric services. In the last 15 years only 4 maternal suicides have been cared for by specialized perinatal services. In all but one, care had been taken over by general adult services at the time of death. General adult (non-specialized) services had not adapted to the maternity context and the distinctive features of postpartum illness. They often responded too slowly, did not consider admission to a specialized mother and baby unit and under-estimated the risk and severity of the condition. In the last enquiry, the recent changes in psychiatric service delivery were noted with women treated by multiple teams during the short course of their illness.

Controlled drug taking during pregnancy was followed by a return to previous levels of consumption postpartum, particularly if the baby was removed by social services. Tolerance had been lost and the final recreational consumption, perhaps triggered by despair and distress, proved fatal.

Deaths from physical consequences of substance misuse

A small number died from alcoholic liver disease or from accidents when drunk. The other causes were varied including overwhelming infections, cardiomyopathies and sudden cardiac deaths induced by stimulant drugs. Often the woman and her friends failed to obtain medical care for treatable serious illness.

Avoidance of care

Most that died from the consequences of substance misuse were not in contact with specialized drug teams were late bookers and had poor uptake of antenatal care. Most were known to child care social services and death appeared to follow child protection case conferences or the removal of the child. This suggested that they avoided both maternity and psychiatric care for fear that their child might be removed.

The women who died

In contrast to suicides they were younger, living in adverse circumstances, some socially excluded and experienced domestic violence. Their partners were often substance misusers

Other maternal deaths from medical conditions associated with psychiatric disorder

Over the last 15 years many women died because the symptoms of their medical conditions were mistaken for psychiatric disorder. Other women did have a psychiatric disorder but distress and agitation related to physical illness was misattributed to their psychiatric disorder. In some cases it was a psychiatric disorder that led to avoidance of treatment, but the most concerning group were women whose symptoms of a medical condition were misdiagnosed as a psychiatric illness.

Maternal deaths from medical conditions associated with psychiatric disorder

Case study 1

A woman with schizophrenia became agitated and complained that she could not breathe. She died from asphyxiation a few days after delivery. A post mortem revealed bleeding into a thyroid adenoma.

Case study 2

A woman with a needle phobia who was unable to take anticoagulants, died from a venous embolism.

Case study 3

A woman who did not speak English was unable to eat and drink, losing weight and very distressed. She was misdiagnosed as suffering from anorexia nervosa. At post mortem she was found to have milliary tuberculosis.

Case study 4

A woman with no psychiatric history stopped her medication for systemic lupus on becoming pregnant. In mid pregnancy she developed malaise, loss of appetite and weight loss. She was diagnosed with depression and not investigated further. She died shortly after delivery. Post mortem revealed a lupus encephalopathy.

Back to basics

The Enquiries reveal a problem with recognizing the severity of both medical and psychiatric disorders in pregnancy and in distinguishing serious illness from commonplace symptoms. These include headache, pyrexia, diarrhoea and vomiting, abdominal pain as well as emotional symptoms.

Anxiety and distress in pregnancy and following delivery

Episodes of tearfulness, anxiety and depressive symptoms are commonplace, particularly in first-time mothers. Mostly these will be mild and self-limiting. However, in some these symptoms can be the early signs of a more serious illness (Box 14.5).

Unexplained physical symptoms

In a number of maternal deaths, symptoms of the medical illness were attributed to psychiatric disorder. In many this was because of non-specific symptoms such as distress, agitation and loss of appetite. In others, the symptoms of an acute confusional state condition were misinterpreted as depression.

Clinicians should be aware of the clinical features and causes of confusional states. It should be remembered that physical illness can present as, or co-exist with, psychiatric disorder (Box 14.6).

Box 14.6 Unexplained physical symptoms

Should not be attributed to psychiatric disorder:

• Unless there is a clear pathway to symptom production

• Unless there is a known previous psychiatric history

• When there is a marked change from normal functioning

• When the only psychological symptoms are non-specific for example, distress and agitation

• When the woman does not speak English or is from an ethnic minority group

Midwives and obstetricians must ensure that women are asked at early pregnancy assessment about previous psychiatric history. Those with a history of serious illness should be regarded as high risk and management plans put in place (Table 14.1). Postnatal care should be extended to include the period of maximum risk.

Table 14.1

Psychiatry and obstetrics

Time	Action	Risk factors
Booking clinic	Take family and personal history	Severe postnatal depression
	Psychiatric disorder	Puerperal psychosis
	Refer to past history	Previous serious psychiatric disorder
Antenatal care	Vigilance	Previous baby, previous loss, infertility
		Multiple antenatal admissions
		High anxiety
Delivery	Vigilance	Emergency caesarean section
Delivery	Vigilance	Intensive Care Unit
Postnatal	Vigilance	Baby admitted to special care baby unit
		Maternal readmission
		Early maternal disturbance
Postnatal examination	6 weeks	Screen for postnatal depression