21 Psychiatric mother and baby units

This chapter will describe the background to the development of psychiatric mother and baby units in the setting of perinatal and general adult psychiatric services. It will also deal with the evidence for their effectiveness in treating and improving the experience of mentally ill mothers.

Psychiatric mother and baby units were developed in Great Britain in the 1950s. This was a time of great change in psychiatric care with the development of the National Health Service in 1948 leading to a major reorganization in the way that psychiatric care was provided. Psychiatric care had previously depended mostly upon the presence of large psychiatric hospitals run by physician superintendents, often with little in the way of doctors with specialist psychiatric training other than in areas of London and other parts of the country. This time also saw a move by organizations such as the National Association for the Welfare of Children in Hospital to ensure that mothers were able to be with their sick children on paediatric wards.

It would seem in the setting of the change of ethos, both politically and medically (Main 1948), towards more patient-centred care and keeping mothers with their children, that mother and baby units opened in London, Shenley, and Banstead (Margison and Brockington 1982). The theoretical basis for the benefit of such services was also influenced by Bowlby’s work (1973) on the adverse affects of separation on the mother–infant relationship. There was a general move in the 1950s towards the keeping together of mothers and infants in all medical settings because of this and the general belief and understanding in the importance of bonding between mother and infant.

The further development of mother and baby units over subsequent years often depended upon the interests of particular specialists in developing such services. The majority of mother and baby units are located in the British Isles with some in the Antipodes (Buist et al. 2004) and a few else worldwide. The presence of this marked variation in the provision of mother and baby units raises the question as to the evidence for their necessity and benefit. One would suppose that if the impact upon health outcomes and satisfaction for the mother and baby were extensive one would expect such services to be provided worldwide in those countries with developed healthcare services. There may be factors to do with the health service organization and delivery of care that have prevented their development worldwide.

The demands for the provision of mother and baby units and the ability to keep mother and baby together during times of serious mental illness have generally been supported by the medical professional and service user groups. The number of mother and baby units has waxed and waned over the years and until recent times has depended generally upon the enthusiasm of individual specialists in perinatal psychiatry and their teams to develop such services. The rationale for provision of services and different styles of service will be discussed later in this chapter but in recent years the need to show the evidence base for such decisions and for the provision of what can be a costly service has increased the threat to some services.

There have been a number of studies of the availability of mother and baby services over the years. A study by Prettyman and Friedman in 1991 of England and Wales indicated about 20% of health districts had dedicated facilities for mothers and their babies, whilst approximately a half admitted mothers with babies to acute general psychiatric wards. A number of areas were reviewing their facilities and there appeared to be a general move towards providing more joint admission but there were still large areas of the country without the ability to jointly admit mothers with their babies.

There is sometimes a debate as to the definition of a mother and baby unit. There are clearly some units which have a separate identity with specialist staff, including special equipment and arrangements for nursery and child feeding. Sometimes single side rooms are designated as mother and baby units without the comprehensive package of care described.

A later survey by Oluwatayo and Friedman (2005) carried out a similar survey of 78 Mental Health Trusts in England. This was a larger survey of perinatal psychiatry services that indicated that just over a third of Trusts had specific mother and baby beds, with approximately two-thirds of these being dedicated units. Most psychiatric services considered their resources inadequate and in that survey less than half of the Mental Health Trusts provided any specialist perinatal services and only a few provided a comprehensive service. The survey found a reduction in the number of dedicated inpatient units for perinatal women in England since 1999 and it was felt that this might be related towards the shift to community provision as an alternative to admission. These issues were further reviewed as part of the National Institute of Health and Clinical Excellence (NICE) guidelines on clinical management in antenatal and postnatal mental health published in 2007 (NICE 2007). The NICE guidelines as part of their review carried out a survey of services for pregnant women and those in the perinatal period. A brief questionnaire was sent to all Primary Care Trust chief executives in England and chief executives of National Health Trusts in Wales. They received a response from just under half of all health authorities. There was a further study of specialist mental health services which received a good response rate and this stated that 31% of respondents were direct providers of either specialist mother and baby unit or had designated beds specifically for women in the antenatal or postnatal period. A further 40% made use of mother and baby beds outside of their Trust. At the same time 52% reported using general psychiatric beds without a facility for admitting infants. This data suggested a number of Trusts were making use of several different services and were admitting mothers to different types of care.

These surveys indicate that whilst there appears to be general consensus that such facilities should be able there has not been the development of a comprehensive service across the United Kingdom (UK) ensuring that all mentally ill mothers could be admitted with their children.

There have been a number of reports and recommendations concerning perinatal mental health services. The Royal College of Psychiatrists Council report (2000) made a series of recommendations including the necessity of a local perinatal mental health strategy to provide effective treatment on a comprehensive basis. Included in these recommendations was that mother and baby units be established to allow admission of mothers with their babies. It was understood that because of economies of scale that it may be necessary for these to serve the needs of a number of health authorities.

In that report it was felt unlikely that anything above the largest health authority would have sufficient numbers of women requiring admission following childbirth to justify the setting up of a specialist mother and baby unit. There was a need for a critical mass of births and it was suggested that a number of health authorities could jointly purchase a unit to serve a population large enough to ensure the admissions necessary to maintain the skills of the staff and to provide the resources necessary for the patients’ care.

The rates of puerperal psychosis have generally been stated as one in 500 births (Brockington 1998) but more recent experience suggests that the figure may be lower than this; a health area of a million people might have a birth rate of 12 000–15 000 births suggesting 20–30 cases of puerperal psychosis per year. There are other diagnoses leading to admission but it is this group where it might be felt to be essential to have provision for joint admission. This issue of the relative rarity of admissions following childbirth necessitating large catchment areas is a significant issue in the organization and efficiency of such services.

The NICE guidelines on clinical management and service guidance for antenatal and postnatal mental health published in 2007 reviews the range of services and treatment for pregnant women. It identified the functions of the inpatient services as including: managing mental health problems during pregnancy; the assessment of mental illness in the postnatal period, including risk assessment and the assessment of ability of the mother to care for the infant; provision of expert care of women requiring admission; the expert provision of safe care for the infants and women admitted; and support for the women in caring for and developing a relationship with her baby. The care of these women is complicated by the need to work with and liaise with other services including maternity and obstetric services, GPs, and maternity-based and community mental health services. The comprehensive maternity services in the UK, which are unique in having dedicated health visitors and midwife home visits after birth, have also probably contributed to the development of perinatal psychiatry services. Other countries, notably the United States, have many different and disparate services providing (or not providing!) care during this period, which makes the organization of psychiatry services complex compared with the UK where almost all women pass through the National Health Service maternity services.

The NICE guidelines discuss that a key factor in the decision to admit a woman with an infant is consideration of the welfare of the infant; whether the infant is better to stay with his or her mother, or whether the infant should be cared for by another family member whilst the mother is receiving inpatient treatment. In areas with specialist units the infant is normally admitted with their mother. There are often logistic issues in relation to geographical proximity which may be important in relating to visiting times and contact with family and social networks. This is clearly a factor in relation to support after discharge and linking in with local services.

The NICE guidelines state that there should be careful planning about the development of mother and baby units, involving key stakeholders taking into account population needs and the influence of related services. The NICE guidelines reiterate that there are few formal evaluations of the provision of mother and baby units and the cost effectiveness of this style of care provision. Indeed, there are no economic analyses of these specialist inpatient units or specialist perinatal teams.

In Scotland the Scottish Intercollegiate Guidelines Network (SIGN 2002) produced a national clinical guideline on the management of postnatal depression and puerperal psychosis. This involved a comprehensive review of the management of psychiatric conditions during pregnancy and puerperium in relation to mother and baby units. This review of the literature describes several studies where mother and baby admissions occur (Buist et al. 1990; Kumar et al. 1995) and whilst they describe advantages in avoiding separation of mother and infant and establishing positive attachment and providing support, they are generally descriptive papers rather than clinical trials. The recommendation of this group was that the option to admit mother and baby together to a specialist unit should be available to all mothers. It also stated that mothers and babies should not be routinely admitted to general psychiatric wards. The evidence for this was not based upon trial data but upon expert opinion and so this was at the lowest grade of evidence to support this recommendation.

The framework for mental health services in Scotland (Scottish Executive 2004) in relation to perinatal mental illness, including hospital admission, contains a detailed framework for mental health services and the assessment that should occur and this endorses the right of mothers to have their babies with them during admission if this is what they wish.

An early paper by Kumar et al. 1995 is a study of the characteristics of 160 admissions to a mother and baby unit. In this study 56% of the admissions occurred within two weeks of delivery and the average duration of admission was two months. Twenty patients had schizophrenia, 56 had affective psychoses, and 24 had non-psychotic disorders. The demographic and obstetric characteristics of these groups were similar but the affective psychosis group were more likely to have acute illnesses and an earlier onset of admission occurring within two weeks of delivery. Women with non-psychotic disorders were also more likely to become unwell within two weeks of delivery but tended to be admitted later. In this study only 7% of the affective psychotic and non-psychotic women were discharged separated from their infants and they found that the women with schizophrenia required greater input of nursing resources than mothers with other illnesses and that 50% were discharged without their infants.

In the UK there was a large collection of mother and baby admission data through the Marcé clinical database (Salmon 2004). This lead to an examination of maternal, clinical, and parenting outcome related to diagnosis and associations with poor outcomes (Salmon et al. 2003). Information was collected on over 1000 mother and baby admissions including 224 mothers with schizophrenia, 155 with bipolar disorder, 409 with non-psychotic depression. There was generally a good clinical outcome in 78% of cases but there were particular predictors of poor outcome. In particular, the factors that predicted poor outcome were a diagnosis of schizophrenia, behavioural disturbance during admission, low social class, and psychiatric illness in the woman’s partner or a poor relationship with a partner. In those with poor outcome on all these four variables, 66% suffered schizophrenia. Women with schizophrenia showed more behavioural disturbance, were more likely to experience hallucinations and delusions, and were more likely to be of low social class. They were also less likely to have a partner although more likely to have a partner with psychiatric illness.

There is a limited literature in relation to studies examining outcome and benefits of mother and baby units. A Cochrane review of mother and baby units and schizophrenia (Joy and Saylan 2007) discusses the issue of admitting mothers with schizophrenia and their babies together might be felt to be particularly important because of the difficulties for women with schizophrenia in forming attachment to their children. The children of mothers with schizophrenia may also be at more risk of losing their mother as primary carer as it is often considered that they are unable to cope (Howard 2003). Mother and baby units may also help in engaging people with serious mental illness into the services (Klompenhouwer and van Hulst 1991; Appleby and Dickens 1993; Kumar and Hipwell 1994; Barnett and Morgan 1996).

The paper reviews the concerns about admitting mothers to psychiatric units due to risk of institutionalization, exposure to multiple carers, and potential physical harm from deluded mothers. Various anecdotal evidence and findings suggest the actual incidents of harm to babies in mother and baby units as very rare (Margison and Brockington 1982; Buist 1990; Salmon 2004). This paper states that there is not a good explanation as to why there are few mother and baby units in other Western countries (Kumar 1995; Cawley 1999) outside of the UK. The paper suggests that it may be in part related to differences in family structure and ences and differences in healthcare and occasionally different cultures, as well as the lack of firm evidence to support the cost effectiveness of such units.

The Cochrane review attempted to include all relevant randomized controlled trials of admission versus non-admission to mother and baby units of mothers with schizophrenia. The review was unsuccessful in finding controlled trials addressing this issue. Forty-four papers were identified but there was only one control study which had to be excluded, from Baker et al. (1961) which compared 20 mothers treated in a mother and baby unit with 20 mothers treated on acute admission ward but without admission of a baby. In this study, which was not randomized, over half the mothers in the standard care group (13 out of 20) were able to take their children home whereas all the mothers on the mother and baby unit were reported as being able to take full care of their baby on their return home. The mother and baby unit group were reported as being more seriously ill on admission but were discharged with fewer symptoms. This comparative descriptive study suggested that the admission to mother and baby units was helpful; although this study took place over 40 years ago and diagnostic and clinical practices have changed since then. The conclusion of the review was that they did not find reliable objective evidence for the efficacy of mother and baby units in schizophrenia and these findings were similar to other reviews.

In studies that included women’s views to admission, they expressed a wish to use a mother and baby unit rather than a standard ward (Margison and Brockington 1982; Kumar 1995; Neil 2006), but these papers also state that at the moment it is very unclear if they have any beneficial affect in terms of mother recovery, risks to the infant, or the development of mother and child bond compared with other models of care. The conclusion for policy makers and managers from these studies was that units for mentally ill mothers and their babies are expensive and their effects and benefits over other packages of care are uncertain. These studies show that units are preferred by patients, which is important, but concern is raised whether this is a sufficient reason for unevaluated investment, and the reviews state that if newer units are being constructed whilst older packages of care are being phased out, then this should afford an opportunity for evaluation.

There is an interesting study looking at psychiatric morbidity and mental health treatment needs in prison mother and baby units (Birmingham et al. 2006). There are four prison mother and baby units in England and Wales, although these clearly are not specifically for the treatment of psychiatric disorders, but are used for women who have babies whilst in custody. Fifty-five participants were recruited (93% of group). It was noted that 60% of the women had mental disorders with a third having a diagnosis of personality disorders but interestingly no one had a psychotic disorder. Thus the group in mother and baby units in prisons tend to have more stable backgrounds than their prison counterparts and women serving sentences for drug offences are favoured. A selection process appears to select out women with psychiatric morbidity and other difficulties that may make them unsuitable for placement in prison mother and baby units.

The case has also been made for using mother and baby units to assess parenting capacity in mothers with mental health problems—often in collaboration with social services departments. There are a number of studies (Seniveratne 2003) looking at parenting assessment in a psychiatric mother and baby unit. A case note study of 62 referrals for an inpatient parenting assessment during a six-year period found that fewer than half the mothers were discharged together with their babies at the end of the assessment period and that at follow-up less than a third were still caring for their children. Mothers with depression were more likely to remain primary carers. The general finding was that it illustrated the need for more integrated coordination between professionals in mental health and children’s services to ensure early planning for mothers and infants at risk as this appears to benefit the outcome of keeping mother and baby together.

There continues to be widespread debate as to the organization of perinatal mental health services. It would seem from a review of the literature that the position of admitting psychiatrically sick mothers with their babies appears to have become part of the culture and psychiatric practice within the UK and one or two other countries noticeably Australia and New Zealand (Brockington 2004). At the same time there are very large areas of the world where this practice is almost unknown. It is obviously important to understand the reasons for this in trying to consider the benefits and drawbacks of such a service that have lead to such a wide differences in the organization of treatment.

It would seem to some extent that differences in services may reflect cost and funding of such services. Mother and baby units in the UK were initially started as part of local initiatives by individual clinicians and not dependant upon funding through insurance schemes which would have found such admissions more expensive than those requiring admission of the mother alone.

The use of mother and baby units also arose at the time when there was far less community care and people with significant psychiatric illness were routinely admitted to hospital. The general rates of admission to psychiatric units have fallen significantly due to changes in clinical practice and the development of more comprehensive services in the community. In particular, the use of assertive outreach teams and crisis resolution teams which are able to offer home treatment on a more regular basis has reduced the number of admissions to psychiatric units. These services will presumably also impact upon the necessity of admitting sick mothers with their babies where home treatment has been shown to be a practical alternative to admission (Oates 1996).

This is an important issue because there has always been a tension related to the rates of psychiatric morbidity amongst new mothers and the number of beds required to service a population. Mother and baby units have generally been developed in large conurbations and even then have often needed to offer admissions over larger geographical areas to maintain viability. There is a tension between the desire to have specialist units with skilled staff able to care for both mothers and babies whilst at the same time wanting to keep mothers and their babies close to home and to integrate with local services. If the rates of sick mothers requiring admission fall further this threatens the viability of larger mother and baby units and more flexible and imaginative responses will be required.

Mother and baby units depend to an extent on the cultural and social background of the populations that are being served. The decision to admit mothers to psychiatric wards with their babies will depend to an extent on the support of families and local communities. There may be cultures where extended families are more willing or able to take on sick mothers or to care for their babies whilst they are admitted to hospital. The increasing use of crisis teams and their associated early discharge services may allow periods of admission to reduce and for mothers only to be admitted during acutely disturbed periods. There is also an issue within the management of mother and baby units as to whether babies should be present when their mothers are acutely psychiatrically unwell or enormously behaviourally disturbed. At these times it is often impractical or unwise for mothers to have close contact with their babies and it may be better for them to be cared for by their families.

There is no doubt from surveys and feedback from patients that generally there is a desire amongst patient groups and staff to try and admit mothers with their babies. There is a strong desire for such services to continue and it is for decision makers to weigh up the balance between the wants and desires of patients, staff, and their supporters, and the increased financial cost associated with specialist mother and baby units. This is a dilemma in the United States where there has been insufficient evidence to support the development of mother and baby units (Wisner 1996).

There is clearly a need for better research data as to the benefits and costs of running such units compared with other models of care. It is difficult to decide upon the benefits of mother and baby units because they should exist within the wider remit of a specialized perinatal psychiatry service. In this way they form part of a delivery of care that can integrate with other parts of psychiatric and general medical services. The difficulty is that certain ways of organizing services may be beneficial in one part of the country but not in another. There is a move towards providing managed clinical networks for perinatal mental health services and this should enable a more rational method of organizing such specialist services.