Introduction
Women with mental health disorders are more likely to be parents than men with mental health disorders (Nicholson et al. Reference Nicholson, Nason, Calabresi and Yando1999; Nicholson et al. Reference Nicholson, Sweeney and Geller1998). Also, pregnancy increases vulnerability to mental illness (Austin and Highet Reference Austin and Highet2011). Perinatal mental illness presents a major public health challenge, given its contribution to maternal morbidity and mortality (Austin & Highet Reference Austin and Highet2011), adverse obstetric outcomes (Frayne et al. Reference Frayne, Nguyen, Allen, Hauch, Liira and Vickery2019) and impact on child development (O’Connor et al. Reference O’Connor, Heron and Glover2002; Stein et al. Reference Stein, Pearson, Goodman, Rapa, Rahman and McCallum2014). Partners’ quality of life and mental health may also be affected (Paulson et al. Reference Paulson, Sharnail and Bazemore2010) and other children in the family may experience a greater risk of mental illness and adverse social and behavioural outcomes (Halligan et al. Reference Halligan, Murray, Martins and Cooper2007; Murray & Cooper Reference Murray and Cooper1996; Stein et al. Reference Stein, Pearson, Goodman, Rapa, Rahman and McCallum2014).
For a long time, women with mental health problems were discouraged from having children. In many countries this was (and sometimes still is) (Pearson, Reference Pearson1995) sanctioned by legislation (Amy & Rowlands Reference Amy and Rowlands2018). Thankfully, the rights of individuals to have families are increasingly being protected by instruments like the United Nations Convention on the Rights of Persons with Disabilities (CRPD), in particular Article 23, which provides a number of protections (United Nations 2006). It is vital that mental health professionals champion these rights and support individuals in having families.
Women are more likely to be admitted in the 2 years following delivery than at other times of their lives, particularly during the first 6 weeks after birth. A Danish study found that women were more than seven times more likely to be admitted 10–19 days post-partum compared to women who had given birth 11–12 months prior (Munk-Olsen et al. Reference Munk-Olsen, Laursen, Pedersen, Mors and Mortensen2006). Women from lower socioeconomic backgrounds have greatly increased rates of admission in this period with 79% more likelihood of being admitted involuntarily compared to non-pregnant individuals (Langan Martin et al. Reference Langan Martin, McLean, Cantwell and Smith2016).
In many jurisdictions there are no dedicated mother and baby mental health facilities. This absence makes adherence to Article 23 of the CRPD highly challenging. It states that ‘In no case shall a child be separated from parents on the basis of a disability…’. As mental healthcare becomes more complex, policies, standards and guidelines are used to maintain quality (Institute of Medicine 2000). In their review of mental health policy in an Irish regional mental health unit, McGuire et al. (Reference McGuire, Curtis and Duffy2020) found there to be a lack of consideration given to pregnant or breastfeeding patients. They also highlighted a complete absence of consideration of this cohort of patients in The Quality Framework for Mental Health Services (Mental Health Commission 2006) document, on which national inpatient policies are based.
Pregnancy creates a temporary situation where the mother–baby dyad has to be considered. This time period is often treated as an exceptional period. Consequently, it is addressed in separate policies and in isolated sections of textbooks or included in areas of special interest (Kaplan & Sadock 2000). However, pregnancy is an inherent and essential part of the human condition with all humans having been involved in at least one pregnancy. The isolation of policy and guidelines relating to pregnant women and nursing mothers, leads to a post hoc consideration of their needs and implies that pregnancy and motherhood are the exception rather than the norm.
We aimed to conduct a literature review to identify how pregnancy and breastfeeding are considered in mental health policy, particularly in relation to hospitalisation. We attempted to identify areas that should routinely be addressed in mental health policy.
Methods
The nature of this study and the research available on the topic did not allow for a systematic review. While a systematic approach was taken, results were often highly limited. We contacted relevant experts in the field requesting suggested information on this topic and on ward policy.
A literature search was undertaken in Medline, CINAHL, APA PsycINFO and EMBASE from 1970 until 2020. Terms searched were pregnancy, reproductive age, child-bearing, perinatal, postnatal, prenatal, breastfeeding, lactation, mental health, psychiatry, inpatient, hospitalisation, policy, guidance, legislation and standards. Terms relating to the topics of pregnancy and breastfeeding were searched using the logical operation ‘OR’. The same process was used for terms relating to mental health, in patient care and policy or guidelines. Then each of the four topics were combined using the logical operation ‘AND’. Search terms were limited to title or title and abstract. The search was then limited by human, English language and date of publication 2000 until present due to the small number of papers published before this date. A hand search of references was also performed. Guidelines for inpatient policy both in general adult wards and mother baby units (MBUs) were also reviewed.
Exclusion criteria included papers that did not relate to policy or were not in the English language. Two authors reviewed each paper for relevance and referred any disagreements to a Consultant Perinatal Psychiatrist.
We tried to focus on the implications for inpatient policy for perinatal patients but in some areas it was not possible to separate out the specific considerations for admitted individuals. As a consequence, some of the discussed topics are not exclusively related to inpatient care.
Results
Of the 262 papers identified in our initial search, 44 met the criteria (Fig. 1) and are briefly summarised here. The papers largely fell into 10 distinct categories. The main findings from these papers are outlined in Table 1. The findings of these papers informed our recommendations for policy which are presented in Table 2.
Discussion
Proactive preconception psychoeducation
Fertility among individuals with mental illness has increased in parallel with de-institutionalisation (Ødegård Reference Ødegård1980). Many of these pregnancies are unplanned or unwanted (Guedes et al. Reference Guedes, Moura, Evangelista and da Conceicao2009; du Toit et al. Reference du Toit, Jordaan, Niehaus, Koen and Leppanen2018). Given the bidirectional nature of the association between unplanned pregnancies and mental illness (du Toit et al. Reference du Toit, Jordaan, Niehaus, Koen and Leppanen2018) fertility and contraception should be considered and discussed both at the initial consultation and at all other interactions with individuals of reproductive age. Solari et al. (Reference Solari, Dickson and Miller2009) suggest improving education and reducing barriers to accessing birth control. Access to abortifacients is unfortunately not ubiquitous. Policy should highlight the needs of this cohort of patients and outline strategies to ensure that these women are not discriminated against by virtue of their illness, inpatient legal status or location.
Screening
Screening can help identify mental health disorders at both the antenatal (Sharif et al. Reference Sharif, Shangaris, Jefferies, Gangopadhyay and Pann2016) and post-partum period (Turella et al. Reference Turella, Donolato, Ciulli and Riolo2016; Wisner et al. Reference Wisner, Sit, McShea, Rizzo, Zoretich and Hughes2013). However, the effectiveness of screening varies. In their robust appraisal of a well-established screening programme, Rowan et al. (Reference Rowan, Greisinger, Brehm, Smith and McReynolds2012) found that universal screening alone did not lead to increased uptake of behavioural healthcare, and suggested that this may be due to barriers such as lack of coverage, lack of proximity of an acceptable provider, difficulty obtaining time off from work, stigma and wariness of psychopharmacologic treatment (Horowitz et al. Reference Horowitz, Murphy, Gregory and Wojcik2009). Lee et al. (Reference Lee, Mihalopoulos, Chatterton, Chambers, Highet and Morgan2019) and Chambers et al. (Reference Chambers, Randall, Mihalopoulos, Reilly, Sullivan and Highet2018) identified that those in rural areas and those from ethnic minorities had less access to this specialised service. Two studies (Wisner et al. Reference Wisner, Sit, McShea, Rizzo, Zoretich and Hughes2013; Turella et al. Reference Turella, Donolato, Ciulli and Riolo2016) noted difficulty in contacting vulnerable women from lower socioeconomic groups and migrant populations for follow-up. Policy should aim to target and minimise these barriers. The screening of all mothers of young children may well be warranted (Wisner et al. Reference Wisner, Sit, McShea, Rizzo, Zoretich and Hughes2013), particularly for those presenting with subthreshold symptoms.
Sawati and Wijesiriwardena (Reference Sawati and Wijesiriwardena2013) noted a number of shortcomings in the quality of screening in a general hospital. This can lead to a sense of inadequacy on the part of providers and may contribute directly to the failure to provide universal screening (Horowitz et al. Reference Horowitz, Murphy, Gregory and Wojcik2009). Policy should focus on equipping staff to screen for and identify perinatal mental health problems. Also, there is increasing evidence to support screening for trauma symptoms in relation to pregnancy and delivery (Dikmen-Yildiz et al. Reference Dikmen-Yildiz, Ayers and Phillips2017, HSE, 2017).
Undoubtedly, an appropriate service should be available to treat women who screen positive. Policy should outline clear referral and treatment pathways.
In Australia, Chambers et al. (Reference Chambers, Randall, Mihalopoulos, Reilly, Sullivan and Highet2018) noted that changing mental health policies, with more emphasis on prevention, screening and early detection, resulted in an increased uptake of services by women and an associated increase in cost of provision of services. Lee et al. (Reference Lee, Mihalopoulos, Chatterton, Chambers, Highet and Morgan2019) identified that universal screening resulted in an increase in admissions antenatally but a reduced number of admissions postnatally, particularly for adjustment disorders. This implies that early detection and management is effective at reducing postnatal morbidity. These studies highlight the need for adequate provision of appropriate structures, policies and training in admission units to ensure high quality care for these women.
Reproductive health-informed psychotropic prescribing
The need for pre-pregnancy planning is well established for women with bipolar affective disorder (BPAD) (Viguera et al. Reference Viguera, Whitfield, Baldessarini, Newport, Stowe and Reminick2007) and schizophrenia (Solari et al. Reference Solari, Dickson and Miller2009) but is also important in depression, anxiety and other mental health disorders. Guidelines and information should be shared with women in order that they can make a fully informed decision regarding their treatment (HSE 2017). Decisions should highlight future reproductive choices where relevant and where known. With such high rates of unplanned pregnancy (du Toit et al. Reference du Toit, Jordaan, Niehaus, Koen and Leppanen2018) it is vital that this is considered at the time of changing or initiating medication.
Detailed consideration is of particular importance with sodium valproate (which is not licensed for women of child-bearing age in many countries), other mood stabilisers and medications (Ornoy et al. Reference Ornoy, Weinstein-Fudim and Ergaz2017) including venlafaxine (Anderson et al. Reference Anderson, Lind, Simeone, Bobo, Mitchell and Riehle-Colarusso2020). In New Zealand, a collaborative patient–clinician system has been developed which targets patient education and recommends electronic alerts for prescribers (Goldspink et al. Reference Goldspink, Pumipi, Dey and Menkes2020). The European Medicines Agency (2018) has published strong guidelines to prevent foetal exposure to valproate including a ban on its use for migraine and BPAD during pregnancy and conditions on its use in epilepsy. Individual EU Countries have incorporated these guidelines, many of which are legally binding, into their national prescribing policies and have produced their own information on the topic. It is important that this information is integrated into local and national guidelines.
Rybakowski et al. (Reference Rybakowski, Cubała, Gałecki, Rymaszewska, Samochowiec and Szulc2019) recommend minimising polypharmacy prior to a planned pregnancy or in some cases reducing and discontinuing medication altogether. Counselling of patients should include address the risks and benefits of drug administration and discontinuation in pregnancy. Specific medication recommendations, such as these, may be included in mental health policy.
Psychotropic use while pregnant or breastfeeding has been widely examined (Winans Reference Winans2001; Kennedy Reference Kennedy2007; Pacchiarotti et al. Reference Pacchiarotti, León-Caballero, Murru, Verdolini, Furio and Pancheri2016). Adverse outcomes such as gestational diabetes and hypertension are higher in women prescribed antipsychotics compared with unexposed women, yet diabetogenic drugs are not adequately monitored in this context (Newman & Thamban Reference Newman and Thamban2015). With the high prevalence of gestational diabetes and its high transition rate to type II diabetes (Zhu & Zhang Reference Zhu and Zhang2016) it is important that policy considers all pharmacological options. Policy should recommend the drafting of guidelines to appropriately prescribe for and monitor women for side effects during this period. In addition to the direct effect on the baby through the presence of psychotropic medication in the breastmilk, the physical health consequences for the woman and the impact and risk of maternal sedation should be considered.
Symptom monitoring
Given the heterogeneity of symptom profiles, and the dynamic nature of risk of relapse throughout the perinatal period, close monitoring of the patient throughout their perinatal period, with rapid intervention at the first sign of relapse, is paramount. Viguera et al. (Reference Viguera, Whitfield, Baldessarini, Newport, Stowe and Reminick2007) suggest the use of a streamlined but flexible treatment algorithm based on illness history and the acceptability and safety of pharmacologic or non-pharmacologic treatments. Adaptation of generic guidelines as suggested by Kelly & Sharma (Reference Kelly and Sharma2010), may provide a framework on which local and national policy makers could build an adaptable strategic policy to address this. Wards and community teams should have high levels of flexibility. Rigid policy on missed appointments or criteria for home visiting may need to be adapted to reflect the demands of this period.
Emergencies
McGuire et al. (Reference McGuire, Curtis and Duffy2020) highlighted the lack of consideration given to the pregnant and breastfeeding patient in inpatient policy, in particular during emergency situations. Seclusion and restraint policies should consider both pregnancy and breastfeeding, as involuntary treatment in the perinatal period is sadly common (Langan Martin et al. Reference Langan Martin, McLean, Cantwell and Smith2016). It is when restraint is required the simple strategy of using a wedge under the woman’s right hip may prevent aortocaval compression (Aftab & Shah. Reference Aftab and Shah2017; Solari et al. Reference Solari, Dickson and Miller2009). This could also be used when positioning a patient during ECT (electroconvulsive therapy). In the case of catatonia or severe agitation, generic guidelines recommending the use of high dose benzodiazepines may pose a threat to the pregnant woman and her foetus (Gonzales et al. Reference Gonzales, Quinn and Rayburn2014). Ladavac et al. (Reference Ladavac, Dubin, Ning and Stuckeman2007) recommended using the minimal amount of sedation in these situations. These practices may also limit someone’s ability to breastfeed or to express breastmilk. This enforced abrupt cessation of breastfeeding raises important ethical problems and can also directly worsen an individual’s mental state (Nam et al. Reference Nam, Choi, Kim, Cho and Park2017). Policy makers should aim to implement programmes to promote breastfeeding continuation during the post-partum period. Wards that cannot admit the mother with the baby should facilitate the expression of breastmilk if the woman desires and is well enough. Policy should make clear recommendations for treating the pregnant or breastfeeding female in emergency situations.
Holistic treatment
Mental health policy recognises the need for the holistic treatment of patients. Food orders for pregnant and breastfeeding women should be modified to increase the amount provided to them and they should also be provided with the necessary vitamin supplements (Solari et al. Reference Solari, Dickson and Miller2009). Grube (Reference Grube2005) found that having a supportive male partner was associated with reduced length of hospital stay for women admitted during the perinatal period and had a positive impact on the woman’s outcome in the postnatal period (Matthey et al. Reference Matthey, Barnett, Ungerer and Waters2000; Kitamura et al. Reference Kitamura, Shima, Sugawara and Toda1996; Marks & Lovestone Reference Marks and Lovestone1995). Policy should ensure that women have access to this vital support during their admission. This may require enhanced levels of flexibility and accommodation in relation to ward policy.
Substance misuse
It is well recognised that dual diagnoses with substance misuse disorders and mental health disorders are common, particularly in relation to depression and anxiety. International guidelines (Thibaut et al. Reference Thibaut, Chagraoui, Buckley, Gressier, Labad and Lamy2019) recommend brief interventions in the case of low or moderate risk of alcohol use, the use of low doses of benzodiazepines to prevent alcohol withdrawal symptoms when necessary and avoidance of the use of other pharmacological treatment for maintenance of abstinence in pregnancy. This is due to the low level of evidence and/or low benefit to risk ratio associated with these treatments. The perinatal period presents a unique opportunity to address substance misuse issues as many women have higher levels of motivation during this period. Given the positive impact that brief interventions have been shown to have on both maternal abstinence and neonatal outcomes (O’Connor & Whaley Reference O’Connor and Whaley2007), policy should make clear and strong recommendations for the management of alcohol misuse in pregnancy. Consideration should also be given to the screening for and treatment of co-occurring substance use disorders, such as benzodiazepine or hypnotic drug addiction (Gopalan et al. Reference Gopalan, Moses-Kolko, Valpey, Shenai and Smith2019). Where inpatient detoxification and stabilisation services are limited, policy and guidelines should acknowledge the unique opportunity that the perinatal period presents and prioritise pregnant women and young mothers.
Eating disorders
Women with eating disorders are at a particularly high risk for miscarriage, preterm labour, delivery by caesarean section and for post-partum depression (Mazer-Poline & Fornari Reference Mazer-Poline and Fornari2009; Bulik et al. Reference Bulik, Sullivan, Fear, Pickering, Dawn and McCullin1999; Franko et al. Reference Franko, Blais, Becker, Delinsky, Greenwood and Flores2001). The risks to the newborn include a greater likelihood of low birth weight, microcephaly and being small for gestational age. Patients with anorexia and bulimia nervosa need to have a ‘team approach’ to treatment with coordinated care from the dietitian/nutritionist, obstetrician and mental health clinician (Mazer-Poline & Fornari Reference Mazer-Poline and Fornari2009). Clearer practice guidelines are needed for the care of pregnant women with eating disorders.
Bonding
Women with major mental illness are more likely to be separated from their babies in the post-partum period with more than one quarter (27.2%) of mother–baby dyads separated at the end of their MBU stay (Ces et al. Reference Ces, Falissard, Glangeaud-Freudenthal, Sutter-Dallay and Gressier2018). While there are risks to children being cared for by a parent with a psychotic illness (Hammond & Lipsedge Reference Hammond and Lipsedge2015) there are also risks associated with care in an institution (Johnson et al. Reference Johnson, Browne and Hamilton-Giachritsis2006). A decision to separate a child from their mother should always be a last resort.
In order to facilitate the bonding process when a woman is admitted in the post-partum period, Hurt & Ray (Reference Hurt and Ray1985) recommend an interdisciplinary protocol to include family meetings, staff meetings with the father and baby’s visits with the mother. Ces et al. (Reference Ces, Falissard, Glangeaud-Freudenthal, Sutter-Dallay and Gressier2018) call for strengthening of training for psychiatrists and other health professionals involved in the care of these women. In the absence of MBUs, local policy should ensure that there is a private room available which could accommodate a mother and baby, with adequate paediatric supports if required. At the very least, a visiting area specifically designated for children should be available. Inpatient admission for women in the perinatal period may provide an opportunity to optimise their social support situation and this should be highlighted in policy. Multidisciplinary input at this time may also be invaluable with significant roles for social work and occupational therapy. A prioritisation of pregnant women and new mothers should be included in policies relating to multidisciplinary input.
Targeted national or regional policy
Specific policy targeted at women in the perinatal period has been shown to be effective. Rowan et al. (Reference Rowan, Duckett and Wang2015) described state level interventions in the USA, to enhance the treatment of post-partum depression. Thirteen states enacted policies to address perinatal mental health issues. These policies were found to fall into four broad categories – education mandates, screening mandates, post-partum depression awareness campaigns and task force mandates. Two other states (Vermont and Illinois) enacted policies that addressed psychosocial issues in the perinatal period. While these mandates had very little impact on influencing clinical outcomes overall, it was found that home visits postnatally with a mental health component may be effective (Rowan et al. Reference Rowan, Duckett and Wang2015) They suggested that state policies should mandate for insurance to cover this, along with three sessions of counselling.
The Australian National Perinatal Depression Initiative (NPDI, 2008/09) (Chambers et al. Reference Chambers, Randall, Mihalopoulos, Reilly, Sullivan and Highet2018) saw the universal routine screening of perinatal women and provided follow-up support and care. A number of free perinatal-specific counselling sessions were introduced for economically disadvantaged perinatal women with a mild to moderate psychiatric diagnosis. This resulted in an increased incidence of antenatal admissions for depression and a reduced incidence of post-partum admissions. Lee et al. (Reference Lee, Mihalopoulos, Chatterton, Chambers, Highet and Morgan2019) suggest that this may be due to earlier detection and improved patient and carer awareness which may have positive flow implications for the well-being of the child. These studies suggest that universal screening for perinatal mental health problems along with the provision of increased psychosocial supports during this period may improve outcomes for both mother and child.
Limitations
While our study highlights an important and neglected area, it has many limitations. First, pregnancy and breastfeeding represent an important but very narrow segment of reproductive health. Mental health policy should also consider inter alia sexual functioning, fertility, access to contraception and, where legal, access to termination of pregnancy. In addition, many individuals attending mental health services will be parents and policy should also consider this, in particular early parenting. These important topics were outside the scope of our current review.
Second, our study design and the majority of studies that we reviewed gave a voice to the clinician or allowed the clinician to shape how the questions relating to the individuals attending mental health services were answered. Qualitative research of the view of people of child-bearing age admitted to mental health services would be vital in further expanding this area of research. Also, narrative reviews by their nature are prone to certain biases including selection bias.
Third, policy is a blunt instrument and is ineffectual without support from and training of staff. The inclusion of many of the identified items in ward policy would not be sufficient to implement change unless accompanied by health policy leadership, appropriate training (Heiman et al. Reference Heiman, Smith, McKool, Mitchell and Roth Bayer2015) and engagement with service users.
Finally, policy does not always capture all the pertinent issues in a given topic and research into the impact of policy is less developed compared to clinical matters, hence our methodological approach may omit key areas.
Conclusion
Female reproductive rights are given sparse consideration in mental health policy. This is despite the fact that many women who attend mental health services are of reproductive age, and many psychotropic medications have serious implications for sexual and reproductive health. These topics need to be prospectively considered by all mental health professionals and governing bodies and not just considered after the fact. Key areas that need to be covered include psychotropic medication, visiting by the partner and baby, emergency situations, early parenting and screening.
Acknowledgements
All authors certify that they have no affiliations with or involvement in any organisation or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript. The authors have no financial or proprietary interests in any material discussed in this article.
Author contributions
All authors contributed to the study conception and design. The literature search and review of material for the purposes of inclusion were performed by Eimear McGuire and Sean Murray, overseen by Richard M Duffy. Material preparation and analysis were performed by all authors. The first draft of the manuscript was written by Eimear McGuire and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
The authors have no competing interest to disclose.
Ethical standard
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committee on human experimentation with the Helsinki Declaration of 1975, as revised in 2008.