Puerperal psychosis (Reference Jones and SmithJones 2009) is a psychiatric emergency in which assessing and managing risks is paramount. First is the risk of suicide by the mother. Suicide is the leading cause of maternal mortality in the UK, accounting for 28% of the deaths. Women who take their own lives in the postpartum period tend to use violent methods such as hanging or jumping. This contrasts with the usual finding that women who are not in the postpartum period are more likely to die from an overdose of medication. Also compared with the other causes of maternal death, those who die by suicide tend to be older and socially advantaged. This reflects the severity of the illness and also calls into question the so-called ‘protective effects of maternity’ (Reference OatesOates 2003).

Second is the risk of physical harm to the baby. Infants are at risk from neonaticide, infanticide and filicide, particularly in new-onset psychosis or relapse of psychosis during the postpartum period. Mothers who commit neonaticide are mostly troubled by psychosis and social problems, whereas infanticide and filicide are commonly associated with postpartum depression (Reference Oakley, Hynes and ClarkOakley 2009). In the unfortunate event that a woman causes the death of her newborn, the Infanticide Act 1922 and the revised Infanticide Act 1938 can be used to reduce the charge from culpable homicide to manslaughter, if it can be proven in a court of law that ‘at the time of the act or omission the balance of [the mother's] mind was disturbed by reason of her not having fully recovered from the effect of giving birth to the child or by reason of the effect of lactation consequent upon the birth of the child’.

Third are risks to both mother and infant (during the gestation period and also postpartum) posed by psychotropic medication. The USA Food and Drug Administration rates drugs in five categories (A, B, C, D and X). Valproic acid is listed in category X (‘proved risk in humans (no indication for use, even in life threatening situations’); lithium is listed in category D (‘human fetal risk seen (may be used in life threatening situation)’); haloperidol and chlorpromazine are listed in category C (Reference Sadock and SadockSadock 2007: pp 865–7). The risks and benefits of treatment with psychotropics versus maternal psychiatric illness must be carefully evaluated on an individual basis. Recently in the USA, GlaxoSmithKline, manufacturer of paroxetine, was ordered to pay $2.5 m to the mother of a child with birth defects (Reference TanneTanne 2009).

Case registers in pregnancy might be helpful in several ways (Reference YadavYadav 2009). A survey revealed that fewer than half of the mental health trusts in England provide specialist perinatal psychiatric services (Reference Oluwatayo and FriedmanOluwatayo 2005). Specialist services that specifically address the needs of perinatal women have been advocated by the Royal College of Psychiatrists (Perinatal Specialist Interest Group 2003) the National Institute for Health and Clinical Excellence (National Collaborating Centre for Mental Health 2007) and the Scottish Intercollegiate Guidelines Network (2002). In the absence of such specialist services, case registers in perinatal psychiatry could be set up easily by existing psychiatric teams to help with research, planning and implementation of services. Prospective registers exist in the UK: for example, the Epilepsy and Pregnancy Register (Reference Morrow, Russell and GuthrieMorrow 2006). Such case registers have achieved prominence with the advent of electronic case records and the technological capacity to derive anonymous databases from them (Reference Perera, Soremekun and BreenPerera 2009).