Effectiveness and cost-effectiveness of psychiatric mother and baby units: quasi-experimental study, British Journal of Psychiatry, 2022.

Statement from Action on Postpartum Psychosis (APP).

APP has long campaigned for women affected by postpartum psychosis (PP) to have access to a Mother and Baby Unit (MBU). MBUs are essential places of care for new mums experiencing severe postnatal mental illness. In national charity, Action on Postpartum Psychosis, we hear daily of the importance of MBUs to mothers; and we hear the devastating consequences of such units not existing.

We urge readers of this paper to bear in mind the methodological limitations when drawing conclusions regarding outcomes and cost effectiveness of MBUs. These limitations, many of which reflect common challenges of conducting controlled trials in this area of service provision, are carefully laid out by the authors in the final section of the paper.


Because of the nature of severe postnatal illness, it is difficult to recruit large samples for controlled trials. As it would be unethical to deny a woman MBU admission where space was available, a naturalistic study design was used. Due to the small sample, women with admission to General Psychiatric Units (GPU) were grouped with those who received Home Treatment. This combined group was then compared to MBU admission. The authors explain that this grouping may have limitations, with the Home Treatment group potentially masking differences in outcomes between MBU care and GPU care. Women are admitted to hospital normally because they have been deemed unsuitable for Home Treatment: Home Treatment would be unsuitable for almost all cases of postpartum psychosis because of the risks for mother and baby. The paper shows differences between the Home Treatment group and those who become inpatients – in particular, that women with severe and relapsing mental illness are under-represented in the Home Treatment sample. For those with PP, a better comparison to determine the clinical and cost effectiveness of MBUs would be to exclude Home Treatment, and compare MBU vs GPU admission. When examined separately, rates of readmission in the current study were in the expected direction (22% MBUs, 32% GPU, 21% Home Treatment).


A 12 month relapse may not be an ideal outcome measure to determine the effectiveness of MBUs: Firstly, MBUs have a lower threshold for readmission than General Psychiatric Units and so women are more likely to be readmitted; Secondly, with episodes of postpartum psychosis and bipolar episodes triggered by pregnancy, relapses are common and the expected illness course, rather than an indicator of quality of care; Thirdly, women who are admitted to MBUs are willing to be readmitted if they are struggling with their mental health, whereas, we know all too well in APP, if a woman experiences the trauma of separation from her baby during admission to a non-specialist psychiatric unit, she will fight with all her power not to return.


Some of the outcome measures have been criticised for not being responsive to change in mental health (1,2). The authors have tried to include many of the hidden costs associated with GPU admission, at least up to the month following discharge. Attempts have been made to include services such as midwives, health visitors, social workers, services provided to the father or co-parent, general practice visits, psychological services, physical health checks and services for the baby that would be integral to MBU care, but whose costs would be born elsewhere in the health service during GPU admission. Other costs may occur later; there might be costs that are not considered; and the quality of the services provided (where no specialist perinatal links or training exists) cannot be compared.

The personal stories of mothers who have experienced GPU admission show many other hidden economic costs that may occur after the end of the study period - for example, counselling to come to terms with the trauma of the experience and mandatory separation from their baby; legal aid for women fighting to regain custody of their children after recovery; fathers’ and co-parents’ lost jobs and financial hardship when they and wider family put employment on hold to look after the baby.


Consistent with findings from our surveys in APP, this study finds women are much more satisfied when admitted to MBU versus a GPU or Home Treatment. If rates of readmission at 12 months are somewhat similar, what are postnatal women valuing about MBU care? The satisfaction scale asks questions such as: “To what extent did the service meet your needs?” and “If you were to need similar help again, would you be willing to come back to this service?”

A survey of 218 women with PP in contact with APP in 2010 found not only were women more satisfied, they felt safer, more informed about their illness, more confident in staff, more supported with recovery, more fully recovered at discharge, and more confident with their baby at discharge. More women in the MBU group reported they felt fully recovered at 12 months (70% vs 40%). There is evidence from MBRRACE Enquiries (3) of fewer suicides to women admitted to MBUs versus General Unit.

Unlike MBUs, general psychiatric wards are inexperienced in managing postnatal physical care, leading to feelings of shame and indignity for mothers. They lack facilities to express milk, do not have milk fridges, changing facilities and are without safe spaces for babies, or older siblings, to visit. Many are mixed sex wards. As admission for severe postnatal illness may last for several months, the impact on family life can be catastrophic. MBUs acknowledge the mothering role, supporting breastfeeding, attachment and the development of parenting skills. Mothers are treated alongside other mothers – enabling women to maintain their connection with baby-care, and facilitating informal peer support.

Ultimately, the value of Mother and Baby Units cannot exclusively be measured in terms of clinical recovery. The costs and outcomes are much broader than those considered in the research – including the outcomes for the infant, family dynamics, the woman’s long term psychological wellbeing, and that of her partner, her legal and human rights. When women are unable to access an MBU, this often results in a domino effect of challenges that impact families in a myriad of ways. The early months of motherhood are a precious experience for many women. As noted by the authors of the paper – women have a right to adequate maternity care and family life - that should be acknowledged and supported by mental health services.

Future work

This work is a long overdue, but challenging first attempt to investigate the value of Mother and Baby Units; the study of which is an area of national importance. MBUs are considered an essential service by women and families recovering from PP. Powerful personal stories, case series and qualitative studies outline their importance (4,5,6). MBUs contribute to much wider system and societal change: building capacity, changing attitudes, increasing knowledge and skills. The UK is leading the world in the development of such units - and should continue to do so with further investment. There are regions of the country such as Northern Ireland, the North of Scotland and North Wales, with no access to these lifesaving services. We look forward to the outcome of further research in this complex area.

1. The use of EQ-5D for measuring health-related quality of life for people with (severe) mental illness e.g. Is the EQ–5D fit for purpose in mental health? | The British Journal of Psychiatry

2. The use of QALY more generally in mental healthcare e.g. QALYs and mental health care - PubMed (nih.gov)

3. MBRACCE Enquiries: https://www.npeu.ox.ac.uk/mbrrace-uk

4. https://www.bbc.com/news/uk-wales-62141119

5. https://www.app-network.org/about-us/ambassadors-and-storytellers/

6. Heron J, Gilbert N, Dolman C, Shah S, Beare I, Dearden S, Muckelroy N, Jones I, Ives J. Information and support needs during recovery from postpartum psychosis. Arch Womens Ment Health. 2012 Jun;15(3):155-65. doi: 10.1007/s00737-012-0267-1. Epub 2012 Apr 21.PMID: 22526403.