The one-hundred women study: characteristics of New Zealand women with severe mental illness

People with severe mental illness (SMI) have high rates of chronic physical illness, complex health and welfare needs¹ and experience significant healthcare disparities.² Women have specific healthcare needs across the life trajectory,³ and parents require explicit considerations.^4,5,6,7,8 Despite awareness of the significance of perinatal mental illness, the high rate of peripartum suicide and the significance of children’s early years, their needs may not be adequately addressed, and health initiatives focus on high-prevalence conditions, primarily anxiety and depression. The subset of mental health conditions that constitute SMI remain poorly understood.

SMI affects 5% of the population over 18 years.⁹ It is a non-diagnostic term, a subset of mental illnesses defined as resulting in functional impairment that substantially interferes with or limits one or more major life activities. While suboptimally defined, in New Zealand this serves as a proxy determination for who can access specialist mental health services. This includes, but is not limited to, schizophrenia, bipolar disorder and major depression (with or without psychosis) and other diagnoses.

Few studies describe the demographics, backgrounds and needs of women with SMI, and there is no New Zealand research of this type. Internationally, women with co-occurring mental health, substance use disorders and trauma experiences report varying symptom severity and use of support services.^10,11,12 Women with SMI experience higher rates of violence in adulthood, which, alongside socio-economic deprivation, has a compounding and detrimental impact on their lives, relationships and ability to access services.^13–15

Abuse and household dysfunction in early life are associated with lifelong disadvantage including adult risk behaviours and poor health outcomes.¹⁶ A description of 10 common early life adversities, which includes parental mental illness, are well characterised and researched. Known as Adverse Childhood Experiences (ACEs), they have a strong, cumulative impact, reflected in poor adult health outcomes,¹⁷ wellbeing measures and life satisfaction.¹³ ACEs are associated with adult health risk behaviours,¹⁸ increased risk of poor obstetric, physical and psychological health, higher rates of hospitalisation and lower socio-economic status in the course of an individual’s life.^16,17 In the first comprehensive study of childhood adversity in New Zealand, ACEs were more prevalent in those who were young, of lower socio-economic status and identifying as Māori.¹⁹ Data sourced from the New Zealand Family Violence Survey revealed 55% had one ACE and 11% reported four ACEs before age 18.¹⁹ Little is known about trauma or ACEs in women with SMI in New Zealand.^4,5,6,20

Our research team of clinicians and researchers work in infant and maternal mental health services based at Health New Zealand – Te Whatu Ora Counties Manukau, a catchment area with a youthful, multi-ethnic population who live in low socio-economic environments.²¹ Females comprise 46% of the population seen by mental health services. In this exploratory study, we aimed to identify the health characteristics of women with SMI referred to community mental health clinics (CMHCs). We proposed that women presenting to community mental health services had significant adverse childhood experiences. In this study, we describe aspects of their mental illness, trauma experiences, physical and sexual health, parenting role and service utilisation. The study was designed to emphasise the importance of identifying women’s needs to inform interventions and the roles of primary and specialist health services in supporting optimal health. The delivery of efficient and effective services depends upon accurate information about service users and their social matrix. This paper attempts to define the problems faced by women who bear the concurrent load of responsibility for the next generation along with their own immediate physical and mental health needs.

Methods

Study design and methodology

In this cross-sectional retrospective study, data were collected from information contained in the clinical notes of women with SMI assessed by community mental health services from two hospital electronic databases (Health Care Community [HCC] and Concerto).

Sample

In our estimation of sample size, we chose 100 women to reflect the volume and in-depth nature of data collection in a qualitative format, as well as the number of variables, the time and resources available and the feasibility of statistical analysis.²² A sample of 120 adult women between ages 18 to 65 were randomly selected by a data analyst to allow for a certain percentage of files to be discarded due to incomplete information. If data were less than 80% complete, the file was discarded. To obtain 100 cases, 115 were screened. A higher proportion of Māori (17%) and Pacific (8%) women were sampled. This was recommended in cultural consultation and prioritised to reflect the locality’s demographic, in line with current utilisation of services.²¹ Inclusion criteria were women who attended assessment during an episode of care at one of the five community mental health services clinics in the Counties Manukau catchment area within a 1-year timeframe from 1 December 2022 to 30 November 2023. The study involved the file review of health data contained in clinical files, approved by the New Zealand Health and Disability Ethics Committee. We did not seek informed consent or to re-interview women about information they had already provided to mental health services.

Data collection

RedCAP is a secure, web-based research data capture tool that enables the creation of a survey or data to track information. We constructed a RedCAP survey to compile qualitative data collected across the electronic databases (HCC and Concerto) within broad categories to encapsulate the large volume of information, where women’s histories could be collated in a binary format to aid statistical analyses. Categories included demographic information, psychiatric diagnoses, use of mental health legislation, medical conditions, health screening (for example, cervical smears), medication, contact with sexual health services, social circumstances, parenting role, pregnancies and number of children, substance use, adverse childhood experiences, adult trauma, referrals to governmental and community agencies, registration with primary care and documented cultural concerns. The tool was tested and adjusted as it became apparent that some categories required expansion (for example, family relationships and making a distinction between adult and childhood trauma). The final iteration of the survey (version nine) was used to draw data (see Appendix).

Data analysis

Data distributions were summarised using mean ± standard deviation (SD), median (interquartile range [IQR]) or n (%), as appropriate. Statistical analyses were performed (using R [version 4.4.0]) to determine differences between ethnic groups as categorised by Statistics New Zealand Level 1 codes: Māori, Pacific, European, Asian and Middle Eastern/Latin American/African (MELAA). Comparisons among ethnicity groups were performed using the Chi-squared test, Fisher’s exact test, or analysis of variance (ANOVA), as appropriate. All analyses were conducted using a complete-case approach, with no imputation performed. The sample size for each variable is clearly reported in the results tables to ensure transparency about missing data. As this was a descriptive study with exploratory comparisons, we reported both unadjusted p-values and false discovery rate (FDR)–adjusted q-values using the Benjamini-Hochberg procedure to account for multiple testing.

Results

Demographics and psychiatric variables (Table 1)

This sample comprised 20 women who identified as Māori, 12 Pacific, 47 European, 17 Asian and four MELAA of mean age 34.8 years (± 12.9). There were 98 who identified as cis-gender and two as transgender. The median duration of contact with mental health services was 9 years. While 18% did not have an annotated psychiatric diagnosis, 32% had two or more, and these are described under broad categories (Table 1). The majority (56%) were taking two or more psychiatric medications; 13% were not taking any. Thirty percent had been placed under mental health legislation (MHA) on at least one occasion. Eighty-one percent had suicidal ideation across their lifetime and 38% had a documented attempted suicide. Thirteen percent had a forensic history. Sixty-six percent had a family history of mental illness. Ethnicity was not significantly different statistically for these variables.

View Table 1–5.

Psychosocial variables (Table 2)

The majority (84%) of the women were living with other people, more commonly with family members (41%) than partners (30%). Pacific women were more likely to live with family (Pacific, 75% P=0.03). Twenty-four percent were living with children. Fifty-one percent were not in paid work and 82% had accessed supports from community agencies. Six percent had been adopted or were whāngai (an informal arrangement whereby a child is raised by a relative and birth parents retain legal guardianship) and 12% were raised by a caregiver who was not their parents. Twenty-seven percent were migrants or refugees (Asians 76.5%, P<0.001) and 14% had English as a second language (Asians 64.7%, P<0.001).

Health variables and service utilisation (Table 3)

The majority (97%) had a listed general practitioner on their file. Thirty-eight percent had one or more documented physical illnesses. Fifty-five percent reported cigarette smoking or vaping (Māori, 75%; Pacific, 83.3%; European, 48.9%; Asian, 23.5% P=0.004). Ninety percent reported alcohol use (Māori, 90% P=0.009) and 50% reported illicit substance use (Māori, 75% P=0.002). Cannabis use (42%) was statistically significant in Māori and MELAA (70% and 100%, P=0.001).

Forty-five percent had given birth, and 14% had three or more live births. Thirteen percent reported having terminations of pregnancy and this was higher in Pacific women (33.3%, P=0.046). Contraception was utilised by 24%, the majority of which were NZ Europeans (14/24), although Pacific women had a higher contraception-use rate (33%). Eighty-one percent had utilised services for sexual health (including primary care), 26% had utilised obstetric services, 56% had utilised gynaecology services and 79% attended cervical screening. Notably, 13% of eligible women (45+) attended mammogram screenings and none were Māori.

Traumatic experiences (Table 4)

Nearly 90% had documented trauma experiences. Eighty-four percent reported adverse childhood experiences and 32% reported four or more. Thirty-eight percent of women reported childhood sexual abuse and 30% reported domestic violence. In Māori, reported childhood sexual abuse (55%, P=0.05) and domestic violence (45%, P=0.04) were significantly higher. Nineteen percent experienced physical and/or emotional neglect, and 20% reported household substance abuse. Thirty-one percent had a parent with a mental illness. Homelessness (3%) was also statistically significant for Māori (15%, P=0.04). Adult trauma experiences included sexual abuse (19%), domestic violence (28%) and emotional abuse (9%).

Children of the 100 women (Table 5)

Forty-five percent of the sample were parents. Twenty-four percent reportedly lived with children (0–18 years) in their care, including non-biological children. The category “biological children who were not in the person’s care” were for reasons of uplift by welfare services, living with other family members, fostering, adoption or whāngai arrangements. Thirty-one percent evidently had children who were not in their care, and the rate was lower for Pacific people compared to other ethnic groups. We noted this information could not be deduced from looking at live births or children in care as there was no accounting for adult children or deceased children.

Discussion

The 100-women study yields clinically rich and relevant data about a vulnerable, hard-to-reach group with SMI. The analysis of their reported histories, collected at the time of initial assessment and contained in clinical files, is salient in revealing their diverse and varied health and psychosocial needs. There is considerable diagnostic variability within this group under CMHC care, but the severity of their mental illness is apparent in the high rates of suicidality, protracted periods of service involvement, involuntary treatment and high utilisation of community support agencies.

Broad categories of diagnosis are included. For example, the psychosis spectrum includes multiple diagnoses such as schizophrenia, schizoaffective disorder, brief psychotic disorder, psychosis due to a general medical condition, etc. An evaluation based on diagnosis would have been valuable; however, there was variability in diagnoses and a surprising lack of diagnosis for 13% of women. The high proportion of people without a listed primary diagnosis is indicative of the deficiencies in deriving data from mental health records; however, it is also clinically understandable. People may have differing aspects of their illness presenting across time, and the primary diagnosis may vary and overlap with other diagnoses, with particular regard to impaired function. Multiple diagnoses contribute to functional impairment in SMI, and the complexity brought on by traumatic life events further obfuscates statistical analysis of single variables.

There are a number of constraints in gathering information during the first comprehensive psychiatric assessment. As such, we acknowledge some areas are not explored in depth, particularly traumatic events, and may be under-reported. Equally, we acknowledge the inaccuracy of some data points; for example, we do not believe 97% are enrolled with and actively engaged with a general practitioner, despite having an organisation listed on their file. It was regrettable that data pertaining to dichotomous variables, such as pregnancy, miscarriage and children in their care, were poorly documented. A greater awareness of this would assist in providing quality information about the impact of treatments including psychiatric medications and enable consideration of the impacts on child raising.

There was relatively high utilisation of women’s health services, yet notable differences between ethnicities pertaining to use of contraception, terminations and cervical and mammogram screening. Given the higher mortality rates from cancer in people with SMI,²³ these behavioural and provider risk factors warrant further specific consideration.

Ninety percent of women in this study reported trauma experiences and 32% reported four or more ACEs. This figure of 4+ ACEs is much higher than New Zealand’s first comprehensive study of ACEs (11.6%).¹⁹ Multiple ACEs are associated with an increased risk of physical and mental illness^13,24 and suicide attempts,²⁵ and is most strongly associated with intergenerational risks of violence, substance use and psychopathology.²⁶ Despite many years of specialist mental health care, there was limited documentation that considered the longer-term impact of traumatic experiences. Therefore, women’s willingness to engage in therapeutic interventions with trauma as a focus remains unknown.

The life circumstances of our sample showed that most lived with other people, primarily family, and particularly so for Pacific women. One-quarter had children in their care and nearly half had caregiver responsibilities. More than half of the 100 women were not in paid work. There were high rates of suicidality, substance use, involuntary treatment and forensic issues. These substantial vulnerabilities as well as historic early life adversity and high ACE burden have far-reaching intergenerational impact.^27,28

Parenting, fertility and sexual health are important^29,30 given the women’s mean age of 34 years. We posit that these aspects of women’s health are not holistically considered or adequately addressed in the mental health care system.^5,27 Mothers with a severe mental illness carry additional burdens in negotiating the normative life stage of parenthood.³¹ When ill, they risk undermining of their autonomy, to the extent that childcare and custody are disputed.⁵ Hospital admissions are particularly disruptive for building attachment with children.^6,14 Poor maternal mental health has a long-term, significant effect upon child health¹⁴ and contributes to intergenerational transmission of psychopathology.¹⁵

This study highlights challenges that women with SMI face and presents compelling evidence that this vulnerable population warrants substantive investment. This would benefit the women themselves and also have impacts on their children. The formative years of a child’s life are influenced by responsive caregiving within a whānau/family environment,^27,28 and intergenerational effects of poor maternal mental health, with concomitant social adversity, reverberate during a child’s lifecourse.¹⁹ In the New Zealand government report He Ara Oranga: Report of the Government Inquiry into Mental Health and Addiction, intergenerational trauma was a repeated theme as were recommendations for whānau-centric approaches to healing for individuals and families.³² Trauma and stigma further increase health inequity. There is a need to understand the difficulties and barriers women experience in accessing services, and mental health clinicians can assist with this through undertaking sensitive enquiries about traumatic experiences, sexual health and about their roles as caregivers.

However, there is inherent complexity in providing care to women who have trauma histories. Much more is required: intervening at critical timepoints in the lifecourse where basic schemas of self, others and the world are formed; tackling stigma;^2,33 and bolstering providers’ understanding of complex trauma.³⁰ There needs to be improved access to trauma-informed care that is multi-disciplinary and multi-agency. The relational context of caregivers, children and the wider environment is highly relevant.²⁷ The 100-women study emphasises the need to screen for ACEs across services to mitigate the risk of poor health, develop policy that prioritises high need and conduct research that focusses on family, community and culturally protective factors.^19,34 Women seek help from a range of health services, but many are dissatisfied with the care they receive.¹³ When women access mental health services, there are opportunities for a more holistic approach and to integrate support for their psychological health with general healthcare, family and social concerns.

We believe this exploratory study to be largely representative of women in our locality as data were drawn from original clinical files. However, there are limitations to report, primarily missing data contained in inconsistently completed forms. Missing data may reflect the constraints of a cross-sectional assessment, the acuity of risk and the quality of rapport formed in gathering histories. A further limitation is that broad psychiatric conditions may be arbitrarily assigned with a tentative formulation and not updated on the file. The study reflects imperfections of frontline contemporary clinical practice and ambiguous documentation prone to misinterpretation.

There are several practical implications for practice and policy. More precise data gathering is likely to improve service delivery. We emphasise care with assigning and recording diagnoses, with a rigorous approach as fundamental to effective treatment. Reasons for suboptimal practice reflects the reality of clinical work and/or the complexity of service users’ presentation, as diagnoses may change over time. With many relevant and competing factors, it can be difficult to determine those of greatest importance. Record keeping should enable both the technical diagnosis and the social context to be made more explicit, and it should be provided by robust diagnostic formulations and their orderly review over time. This is fundamental to coherent service delivery.

Accurate records of the community supports accessed by women is another crucial element of coherent service delivery. Records were frequently imprecise, e.g., there were discrepancies between recorded and actual GP allocation. It would be beneficial to have an accurate depiction of “who’s who”, i.e., who has been involved in the overall care of women with recurring presentations. This could be generated and added to by case workers to guide subsequent clinical encounters.

The 100-women study raises concerns that health services provide care to patients without a clear understanding of women’s specific needs and life circumstances. As such, there is limited accountability to referring agencies and a lack of ongoing responsibility for care as it is passed to these agencies. Remedies for these defects lie outside the scope of the paper. Scarcity of resources and workforce are not the only problems. Therefore, we explicitly recommend research with a systems focus to better coordinate treatment of women’s mental and physical health needs.

Conclusion

This study highlights that women with SMI have a complex array of needs that health professionals should take into consideration at all points of health service access. We recommend that routine enquiry is made about women’s health issues, pregnancy, parenting and adverse childhood experiences, and that documentation is recorded more precisely. There are additional implications for Māori and minority-ethnic women due to the impact of trauma in forming trusting relationships with mainstream healthcare services. Women with SMI should be acknowledged as a strategic target for integrated, holistic, trauma-informed interventions with potential and significant positive intergenerational impact.

View Appendix.