Exploring psychosocial vulnerability among Dutch pregnant women: a register study

Abstract

In the Netherlands adverse perinatal outcomes are also associated with non-medical factors which vary across geographical locations. This study analyses the presence of non-medical vulnerabilities in pregnant women in two regions with high numbers of psychosocial adversity using the same definition for vulnerability in both regions. A register study was performed in 2 regions. Files from women in midwife-led care were analyzed using a standardized case report form addressing non-medical vulnerability based on the Rotterdam definition for vulnerability: measurement A in Groningen (n = 500), measurement B in South-Limburg (n = 538). Only in South-Limburg a second measurement was done after implementing an identification tool for vulnerability (C (n = 375)). In both regions about 10% of pregnant women had one or more urgent vulnerabilities and almost all of these women had an accumulation of several urgent and non-urgent vulnerabilities. Another 10% of women had an accumulation of three or more non-urgent vulnerabilities. This study showed that by using the Rotterdam definition of vulnerability in both regions about 20% of pregnant women seem to live in such a vulnerable situation that they may need psychosocial support. The definition seems a good tool to determine vulnerability. However, without considering protective factors it is difficult to establish precisely women’s vulnerability. Research should reveal whether relevant women receive support and whether this approach contributes to better perinatal and child outcomes.

Keywords:

• Vulnerability
• disadvantaged
• pregnancy
• perinatal care
• maternity care
• social care
• midwives
• midwife-led care

Introduction

Creating equality at the start of life to achieve health for all is high on the (inter)national agenda [1,2]. However, in the Netherlands significant differences in perinatal outcomes between different regions exist [3,4]. These differences are mostly due to an accumulation of medical and non-medical factors, such as poor lifestyle, mental health problems, and difficult socio-economic circumstances. Poverty and deprivation affect perinatal outcomes [5].

This insight offers opportunities to improve perinatal outcomes. However, it requires significant changes in maternity care. Therefore, the term "social maternity care" has been introduced, which refers to medical and psychosocial care for pregnant women, using expertise from maternity care, preventive youth health care (YHC) and social care during the preconception and perinatal period [5].

In the Netherlands, almost 90% of all pregnant women start in midwife-led care [6]. The primary care midwife is authorized to supervise physiological pregnancy, childbirth and puerperium. If risk factors arise or complications occur the midwife refers the pregnant woman to secondary (or tertiary) obstetrician-led care in the hospital. After birth, maternity care assistants provide care to most women and newborns at their home (up to 8 d postpartum) under the responsibility of the primary care midwife. At the end of the first week, the care for the child is transferred to the YHC where YHC nurses and - doctors provide preventive health care to children (0–18 years) and their parents. The care for the women is transferred to her general practitioner. Local governments are responsible for providing social services and for this purpose they contract with welfare organizations. For complete information of the organization of Dutch maternity and social care see supplement 1.

In 2018, the Dutch government launched the national action program “Solid Start”. This program aims to promote local collaboration between the health and social domains, focusing on children in the first 1000 days (from conception till their second birthday), especially those born in vulnerable psychosocial situations [7]. Solid Start supports the development of social maternity care, and the continuity of - especially psychosocial - care for these children.

Implementation of social maternity care is challenging for several reasons. First, differences in legislation, funding, organization and culture between the health and social domains complicate collaboration at the meso and micro levels. Second, when implementing a program to identify women in vulnerable situations and refer them for preventive or curative care in the health and/or social domains, all necessary program components should be available and the program’s effectiveness regarding the intended goals (better perinatal and child outcomes) should be scientifically substantiated [8].

So far, Solid Start’s strategy has been focused on developing local best practices. However, a need for strategies at a national level have become more prominent [9]. Part of this is the identification of women in vulnerable situations. This requires national consensus on a definition of vulnerability and on how it can be determined, which is currently not available. In 2020, the Rotterdam definition of vulnerability (Figure 1; from now on: definition) was launched to guide professionals in how to determine psychosocial vulnerability and to support a common language between professionals [10]. But the definition has not been applied and evaluated in maternity care. To successfully achieve social maternity care, it is important to explore the usability of the definition and to know the extent of vulnerability among pregnant women. The aim of this study is to gain insight into the type and accumulation of psychosocial vulnerabilities of pregnant women from two Dutch regions and the prevalence of pregnant women living in a vulnerable situation according to the definition.

Figure 1. Rotterdam definition of vulnerability (van der Meer et al. [10]).

Materials and methods

This study is part of a large research project aimed on the implementation of social maternity care. Data of pregnant women in two Dutch regions, Groningen, and South-Limburg, were collected. These regions are situated in the north and southeast of the Netherlands, both of which are known to have high numbers of families living in poverty and deprivation [11].

We conducted a register study with three measurements in three different study populations, one in Groningen and two in South Limburg (measurement A, B and C).

Measurement A was conducted in Groningen. In Groningen the Alpha-NL questionnaire was introduced in 2017 to identify vulnerability. The Alpha-NL is filled out by pregnant women as part of routine prenatal care, intending to identify risk factors for adverse pregnancy outcomes, child abuse and psychosocial problems in (future) parents at an early stage [12]. Measurement A was performed among women starting their pregnancy care in two midwifery practices in Groningen and who gave birth between July 2019 and February 2020.

Measurement B was conducted in South-Limburg among women under the care of five midwifery practices and who gave birth between July and December 2019. During that period, midwives did not use a specific tool to identify vulnerability in addition to usual care.

Measurement C also took place in South-Limburg. As part of the research project, a new method to identify vulnerability was developed and implemented after measurement B. This method includes introducing the definition of vulnerability to professionals and using an updated version of the regular midwives’ questionnaire to take the psychosocial history filled out by pregnant women. The study population of measurement C consisted of pregnant women who started their care between May and July 2021 at 9 midwife-led practices that implemented the questionnaire.

We intended to collect data from 500 medical records per measurement, a sample large enough to demonstrate a variety of vulnerabilities based on earlier estimates that in the Netherlands 1 in 6 families lives in a vulnerable situation [2]. We wanted to get an overall picture of women’s vulnerability in midwife-led care and therefore included all women attending the midwifery practices during this period. We excluded women with miscarriage or abortion after the first consultation, women who moved and women referred to obstetrician-led care in the first trimester of pregnancy because of lack of information about possible vulnerabilities.

The main outcomes of interest were types and accumulation of vulnerabilities and prevalence of vulnerability. The authors of the definition gave some examples of (non)urgent vulnerabilities in their publication [10]. In consultation with the participating professionals we added relevant vulnerabilities based on data in the medical records (Tables 2 and 3). For the accumulation of vulnerabilities we calculated per woman the amount of non-urgent vulnerabilities, the amount of urgent vulnerabilities, and the combination of both urgent and non-urgent vulnerabilities. The prevalence of vulnerability was determined in line with the definition: women having one or more urgent vulnerabilities are classified as highly vulnerable and women having one or more non-urgent vulnerabilities are classified as vulnerable. Because protective factors were not written down in the records it was not possible to differentiate between potentially vulnerable and vulnerable. To get a clear overall understanding of women who need professionals’ special attention we defined the outcome “living in a complex vulnerable situation” as being highly vulnerable or having three or more non-urgent vulnerabilities.

Table 1. Characteristics of the study populations.

Characteristics	Measurement A Groningen N = 500 N (%)	Measurement B South-Limburg N = 538 N (%)	Measurement C South-Limburg N = 375 N (%)	National Dutch population 2019a
Age
<20	12 (2.4)	8 (1.5)	4 (1.1)	1 %
20–23	21 (4.2)	25 (4.6)	30 (8.0)	NA^b
>23	467 (93.4)	505 (93.8)	341 (90.9)	NA
Parity:
nulliparous	259 (51.8)	256 (47.6)	186 (49.6)	45%
multiparous	241 (48.2)	282 (52.4)	189 (50.4)	55%
Maternity care at birth:
midwife-led care	138 (27.6)	194 (36.0)	104 (27.7)	27.3%
referred to obstetrician-led  care	362 (72.4)	343 (63.8)	265 (70.7)	72.7%
missing value	0	1 (0.2)	6 (1.6)

^a Perined, Perinatal care in the Netherlands anno 2019: national perinatal figures and interpretation, Utrecht 2020 (www.perined.nl).

^bNA: not available.

Table 2. Urgent vulnerabilities.

Urgent vulnerabilities	Measurement A Groningen N = 48 of 500 (9.6%) N*	Measurement B South-Limburg N = 58 of 500 (10.8%) N*	Measurement C South-Limburg N = 45 of 375 (12%) N*
Psychological and psychiatric problems in this pregnancy	30	38	32
Domestic violence	6	14	7
Involvement Safe Home in history**	2	7	1
Involvement Child Protection Council in history	5	7	4
Intellectual disability	15	5	6
Addiction problems (alcohol, drugs)	1	2	5
Homeless	0	2	0

* Adds up to > than N because some women have > 1 urgent vulnerability.

** Safe Home is the Dutch national organization for advice, support and reporting of domestic violence and child abuse.

Table 3. Non-urgent vulnerabilities among women in a vulnerable situation.

Non-urgent vulnerabilities	Measurement A Groningen N = 255 of 500 (51%) N*	Measurement B South-Limburg N = 265 of 500 (49.2%) N*	Measurement C South-Limburg N = 227 of 375 (60.5%) N*
Psychological problems in history**	207	196	182
Psychosocial problems partner	114	120	86
Unhealthy lifestyle	45	65	47
Financial problems	46	56	48
Women do not speak Dutch or English/German	19	47	18
Relationship problems	26	35	19
Teenage pregnancy	33	33	34
Single mother	30	29	15
Unhealthy/ unsafe living environment	33	19	25
Unplanned and unwanted pregnancy	4	11	0
Low educational level	33	5	39
Suboptimal use of care	–	–	10
Composite family problems	–	–	17

* Adds up to > than N because some women have > 1 vulnerability.

** Including nonconsensual sexual experience, domestic violence in childhood.

To systematically extract data on vulnerabilities from medical records, a case report form (CRF) was developed in cooperation with midwives, obstetricians and social pediatricians.

Midwives extracted the data anonymously from their files, while researchers put this into the CRF. Instructions were developed to guide the completion of the CRF, with several calibration sessions between researchers and midwives to ensure the quality of registration.

The medical ethics committee of Zuyderland Hospital and Zuyd University of Applied Sciences Heerlen approved the study (registry number METCZ20200180).

This study was designed to estimate the number and types of vulnerabilities in two regions. Therefore we conducted descriptive statistical analysis on the categorical variables by calculating frequencies and percentages using IBM SPSS statistics 25.0 [13].

Results

Table 1 shows the baseline characteristics of the study population. A total of 500 medical records in Groningen (measurement A), and respectively 538 records (measurement B) and 375 (measurement C) in South Limburg were analyzed. Compared to the National Dutch population of women given birth in 2019, our study populations include more teenage mothers <20 years and more nulliparous women. In the population of measurement B more women stayed in midwife-led care. Figures of 2022 with which we should compare measurement C are not yet available.

Vulnerabilities

Tables 2 and 3 show the prevalence of different types of urgent and non-urgent vulnerabilities in the three study populations. The most common urgent vulnerability is “psychological and psychiatric problems in this pregnancy”. The study population of Groningen had more women (n = 15) with a (suspected) intellectual disability. Among the non-urgent vulnerabilities “psychosocial problems in history” was most common (Table 3).

Accumulation of vulnerabilities

Almost every pregnant woman in a highly vulnerable situation had an accumulation of – both urgent and non-urgent – vulnerabilities. Most women had 2–7 different vulnerabilities, with a few having up to 10. Within the category of vulnerable women we also found a ─ less distinct ─ accumulation of non-urgent vulnerabilities. supplement 2 provides more information on these data.

Prevalence of vulnerability

Table 4 shows the prevalence of vulnerability in categories according to the definition. Considering the degree of vulnerability and the accumulation of vulnerabilities, about 20% of pregnant women in both regions were classified as living in a complex vulnerable situation. Ten percent lived in a highly vulnerable situation and another 10% had an accumulation of 3 or more non-urgent vulnerabilities.

Table 4. Prevalence of (complex) vulnerability.

*No reporting of protective factors means no distinction between women living in a potentially vulnerable and a vulnerable situation.

Bold and italic figures (18.8, 19.5 and 23.2) is to distinguish these figures from the other results. This distinction in the table supports the text under "prevalence of vulnerability".

Discussion

In our study, we found a large variation of vulnerabilities. According to the Rotterdam definition about 20% of the women in both regions can be considered as living in a complex vulnerable situation.

Prevalence data on vulnerability in the Dutch population are scarce. Studies differ in terms of the definition of vulnerability, operationalization of vulnerabilities, study population and time frame. In 2007, the Netherlands Institute for Social Research (NISR) used a definition in which having 4 or more psychosocial risk factors was defined as vulnerable and reported 15% vulnerability among families with children living at home [14]. This study showed that 50–60% of women had one or more vulnerabilities. In 2016 Molenaar et al. showed that about 40% of the women had socio-economic, psychosocial, and multidimensional vulnerabilities based on pre-pregnancy data from national data sources [15]. Despite the fact that data cannot be completely compared, the high prevalence of psychological vulnerabilities corresponds. Molenaar et al. showed that psychosocial vulnerabilities were related to a higher prevalence of SGA and hypertensive disorders. Additionally, having multiple types of vulnerabilities is associated with a variety of adverse perinatal outcomes. A concern given the rates of women with an accumulation of vulnerabilities in our study [15–17]. This influence on mother and child underlines the importance of the need for action in the two regions in the current study.

Our results showed little difference between the three measurements in the two regions and between different ways of determining vulnerability. Before (measurement B) and after (measurement C) the implementation of a program aimed at identifying vulnerability in South Limburg, the results do not differ substantially. A possible explanation is that Dutch midwives have always provided accessible and frequent care [18]. They work in the community and often know a woman’s social network, allowing a proper assessment of women’s vulnerability.

Identifying vulnerability according to the definition is a different concept than medical risk assessment used in maternity care. A widely accepted definition is necessary for a collaborative attitude of professionals, who are open to signals of psychosocial vulnerability and discuss this with women to initiate tailored support. It should support solving personal and social problems, not stigmatize women and medicalizing them. Van Blarikom warns that interpreting vulnerability primarily from a risk perspective leads to an unbalanced focus on individual mothers rather than on the social context of the problems. She suggests to approach vulnerability as a multilayered, situated, and relational concept rather than from an epidemiological perspective [19].

The literature on vulnerability emphasizes that barriers to psychosocial care experienced by women and professionals also determine vulnerability, beside risk factors and protective factors [14,20]. Except for suboptimal use of care, these barriers are not explicitly addressed in the definition. Still, they are mentioned by women and professionals: e.g. women who fear their child will be placed out of the home if they talk about problems or professionals afraid of disrupting the trusting relationship when addressing vulnerabilities. Adding barriers to the concept of vulnerability deserves attention [21].

Although the definition is useful for professionals, there are still some important questions to answer before the definition can be validated. One of the issues is what is aimed to achieve with social maternity care. If the aim is to help all women and families from a broader preventive perspective, a “sensitive” approach as in the definition might be used. If the aim is to provide more targeted support, for example to women in complex vulnerable situations, a “specific” approach might be appropriate. This implies choices on a national level that have not yet been decided in the Netherlands.

The definition suggests a broad and holistic view of vulnerability. As this research shows, numerous vulnerabilities are present in the studied population. Many referrals for support in the health and social domain are to be expected, domains already struggling with financial and staffing challenges. Additionally, there is insufficient knowledge of the effectiveness of interventions to support vulnerable women on perinatal and child outcomes. Future research should help to better understand which vulnerabilities predict adverse perinatal and child outcomes and what interventions are effective. This is essential to make the identification of vulnerability more targeted on those who experience the greatest consequences. Another option is to define what care can be guaranteed and adjust identification accordingly [8]. Given the large numbers of vulnerabilities, other ways to explore may be strengthening pregnant women’s resilience through other models of care such as centering pregnancy and parenting [22] or through informal care by peers [23].

It cannot be justified to implement structural identification of vulnerability unless all resources are in place to support women after expectations have been raised. If resources are limited, feasible choices must be made regarding identification and support [8].

To our knowledge, this is the first study that tested the Rotterdam definition of vulnerability, and it contributes to the applicability of the definition and to knowledge about the prevalence of vulnerability. Another strong point is that we do not miss women who were in care during the study period because of the chosen method of data collection. However, this study also has several limitations. Most important is the lack of registration of protective factors in the medical files, which hindered to distinguish between women in a vulnerable or potentially vulnerable situation probably resulting in an overestimation of vulnerable women. However, with the outcome “living in a complex vulnerable situation” we may have captured the group of women in need of social support. This information is useful to organize this support. The fact that current software used in midwifery practices is not optimally equipped to record routinely identification of vulnerabilities may have lead to an underreporting of vulnerabilities as may the exclusion of women who were referred to obstetrician-led care early in pregnancy.

For measurement C, the targeted 500 files were not included. To estimate how much time we needed for data collection, we asked midwives to indicate how many new registrations of pregnant women they had per month. Subsequently, all practices appeared to have fewer new registrations, possibly caused by a (national) decrease in the number of new pregnancies between May and July 2021 [24] or by underestimating the dropout of women due to miscarriage, for example. We do not think that this negatively influenced the results: based on the vulnerability rates of measurements A and B, we deduced that the size of sample C was sufficient for reliable results.

Only two practices in Groningen participated in the study. This may negatively affect the generalizability of the results. However, It involved two large practices with a diverse population regarding their demographic characteristics. They adopted the identification tool, and their population in the study represented an estimated 20% of the Groningen population of pregnant women [25].

In conclusion, in two Dutch regions with a high prevalence of Western pregnant women living in poverty, about 20% of pregnant women seem to live in such a complex vulnerable situation that they may need psychosocial support. However, as this study did not take protective factors in consideration, it is difficult to establish precisely women’s vulnerability and therefore the need for psychosocial support. Future research must contribute to the ongoing debate of the efficacy and the feasibility of the definition as a basis to identify vulnerability, to improve collaboration between maternity care and social services with the ultimate goal of improving perinatal outcomes.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Raw data were generated at Zuyd university of applied sciences. Derived data supporting the findings of this study are available from the corresponding author (DD) on request.

Additional information

Funding

This study received funding from the Dutch Taskforce for Applied research SIA grant number [RAAK.PUB06.007]