The role of women’s emotional profiles in birth outcome and birth experience

Abstract

Objective

The aim was to investigate birth outcome and birth experience in relation to women’s emotional health. An additional aim was to explore the relationship between emotional health, continuity with a known midwife, and the birth experience.

Methods

A prospective longitudinal cohort study of 243 women enrolled in a continuity of care project in a rural area in Sweden. Profiles were constructed from instruments measuring depressive symptoms, worries, fear of birth, and sense of coherence. Antenatal and birth records and questionnaires were used to collect data.

Result

Women were categorized into two cluster profiles: “emotionally healthy” vs. “emotionally unhealthy”. Women in the “emotionally unhealthy” cluster had a less positive birth experience (p = 0.006). The total score of the Childbirth Experience Questionnaire was highest in women who had had a known midwife assisting at birth. Babies born to women in the “emotionally unhealthy” cluster were more likely to have a severe neonatal diagnosis.

Conclusion

There were few differences in birth outcome between the clusters, while there were explicit differences in the childbirth experience. Having a known midwife is important to warrant women a more positive childbirth experience. Screening with validated instruments during antenatal care could be a first step to further investigate women’s emotional well-being and provide targeted psychosocial support.

Keywords:

• Birth experience
• birth outcome
• cluster analysis
• continuity model of care
• midwifery

Background

Pregnancy and birth are periods in life largely affecting women’s physical, mental, and social health and well-being. For the majority of women it is a period of joy and happiness, but some are troubled with worries, anxiety, depressive symptoms, and fear of birth [1]. In recent years there has been a stronger focus on women’s mental well-being during and after pregnancy, and in some regions in Sweden women are screened for fear of birth and/or depressive symptoms during pregnancy [2]. Fear of birth affects 14% of women worldwide [3] and 10–20% of pregnant women in Sweden, depending on the classification and measures used [4]. Counseling with midwives is the most common treatment offered to Swedish women with fear of birth [5]. Women are usually satisfied with the counseling [6,7], although the efficacy of the treatment has been questioned due to the limited effects on e.g. cesarean section on maternal request [8]. Reducing such cesarean sections is one of the goals in counseling [9]. Continuity with a known midwife during childbirth has shown promising results when it comes to assisting women afraid of birth. One feasibility study in which the antenatal midwife assisted fearful women during childbirth [10], and one study conducted in three Swedish hospitals in which the counseling midwife also assisted the women they previously met [11], showed a reduction in fear and a more positive birth experience.

Fear of birth could present in different shapes; sometimes fear is communicated, sometimes women talk more about worries. When asked, women can usually explain the differences between fear and worry, as shown in a think-aloud study performed when testing a short instrument – the Fear of Birth scale [12]. Worry is described as something ongoing and unspecific, and often negatively loaded feelings and thoughts about something unpredictable; fear is described as something more specific [12]. The most common worries during pregnancy have been rank ordered by Swedish women [13,14], and worries about the baby’s health, the risk of miscarriage, and giving birth, are the most common things women worry about during pregnancy. These findings have been confirmed in several international studies [15–18]. Depressive symptoms affect around 10% of the general population in Sweden and pregnant women are no exception [19]. Screening for depressive symptoms is recommended by some researchers [20], and a recent study from Australia revealed that undergoing a screening procedure resulted in better neonatal outcomes such as fewer pre-term births and babies with low birth weight [21].

The co-morbidity between fear of birth and lack of emotional well-being is well known [22], but usually presented as single associations in studies. One exception is a study of women with fear of birth who were recruited to a randomized controlled trial [23]. The study used several psychological instruments, such as injection and blood phobia. Moreover, performance-based self-esteem, pain catastrophizing, instruments detecting depressive symptoms and anxiety, and symptoms of post-traumatic stress disorders were included in a cluster analysis. The results showed several sub-groups of fearful women, and the authors concluded that women with fear of birth are a heterogeneous group and that psychological measures could add more to the understanding than the usual division by parity [23].

Although psychological measures are important, screening women during pregnancy with such instruments is not common in Sweden, albeit in some places screening for fear of birth and depressive symptoms have been introduced [2]. Contrary to identifying risk factors for poor emotional health, such as fear, worries, and depressive symptoms, there are also known protective factors. The construct of Sense of Coherence (SOC) [24] is described in terms of dispositions of personality [25]. High SOC, e.g. finding the world manageable, comprehensible, and meaningful, predicts well-being during and after birth [26].

In a previous study, profiles of women were investigated through a cluster analysis of four scales measuring women’s emotional well-being (depressive symptoms, major worries, sense of coherence, and fear of birth). Women recruited to the current study and women recruited to a national Swedish cohort study performed in 1999–2000 showed similar patterns, with two distinct cluster profiles [27]. The two clusters were labeled “emotionally healthy” versus “emotionally unhealthy”. Those included in the “emotionally unhealthy” cluster reported higher levels of depressive symptoms, worries, and fear of birth and lower levels of sense of coherence. A reversed pattern was found in women included in the “emotionally healthy clusters”. Women in the “emotionally unhealthy” clusters were more often single and born in a country outside of Sweden. They also had more negative attitudes about being pregnant and toward the first weeks with the new-born baby. This was found in both the study sample and the historical cohort [27].

Problem area

Previous research proposes that depressive symptoms, worries, fear of birth and Sense of coherence are conditions affecting women’s and children’s health. Few studies have created profiles of pregnant women comprising these aspects in relation to birth outcome and women’s childbirth experiences. Hence, the aim was to investigate birth outcome and birth experience in relation to women’s emotional health. An additional aim was to explore the relationship between emotional health, continuity with a known midwife, and the birth experience

Method

Design

A longitudinal cohort study of women enrolled in a continuity of care project in a rural area in Sweden. Details of the project are presented elsewhere [28,29].

Setting

An antenatal clinic in a small town in a rural area of Sweden and two labor wards within 100–120 km distance. The small labor ward situated in the same hospital as the antenatal clinic closed shortly before the study started. Four midwives provided antenatal care for the women in the study and were on-call for labor assistance during part of the day, after traveling to one of the two remaining labor wards in the region. Usually the on-call service was available between 7 a.m. to 11 p.m., with some exceptions; e.g. in case of sick leave, holidays, and internal education. There were not enough midwives to cover on-call around the clock, and the work time regulations stated that midwives were not allowed to work more than 12 h, thereafter they needed to be replaced. The midwife on-call made individual plans for the woman after the onset of labor. The midwifery group had a car equipped for emergent births and a telephone with excellent coverage, similar to those used by paramedics in the mountains. When arriving at the hospital of the woman’s choice, the midwife followed the hospital regulations regarding intrapartum care. After 11 p.m. until 7 a.m., or in cases where there was no midwife from the group available, women were taken care of by staff midwives working on the labor wards.

Participants and recruitment

Women were eligible for recruitment to the project if they had a due date between 1 August 2017 and 30 June 2019. Information about the project was given by telephone when women contacted the antenatal clinic to book a visit. Additional information was available in the clinic’s waiting area and on a website. Another inclusion criterion was to be able to communicate in Swedish by telephone. Women were allocated a personal midwife, who was responsible for antenatal and postnatal care, but they also had the opportunity to meet all midwives in the group, as the midwives rotated on the on-call schedule. The definition of “a known midwife” in the present study is to receive intrapartum care from one of the midwives in the group.

Data collection

Background data was collected by a questionnaire in mid-pregnancy. The questionnaire covered women’s socio-demographic and obstetric backgrounds (age, marital status, country of birth, level of education, and parity). From the first questionnaire a cluster analysis was performed and compared with a historical cohort of pregnant women recruited to a national survey 20 years ago [27]. The cluster analysis was built on the summed scores of the 16-item Cambridge Worry Scale (CWS) [30], the summed scores of the Edinburg Postnatal Depression scale (EPDS) [31], the sum scores of the 13-item Sense Of Coherence scale (SOC) [32], and a question previously used in the national cohort measuring fear of birth (FOC) [33], all instruments completed in mid pregnancy.

Health-related variables during pregnancy (previous medical and psychiatric conditions, composite variables of identified risk factors and pregnancy-related complications) and birth outcome (gestational length, onset of labor, mode of birth, epidural, oxytocin augmentation, composite variables of severe birth complications and severe infant complication and if there was a known midwife assisting) were collected from electronic pregnancy and birth records. The composite variable of identified risk factors consisted of any of the following; kidney problems, thyroid disease, rheumatic disease, previous cesarean section and previous intrauterine fetal death.

The composite variable of pregnancy complication included gestational diabetes and/or preeclampsia and/or pregnancy induced hypertension and/or severe anemia. In the composite variable of severe birth complications any of the following conditions were included; abruptio placentae, postpartum hemorrhage, eclampsia, sphincter rupture, manual/operative removal of the placenta, shoulder dystocia and prolapsed cord. Severe neonatal diagnosis were mainly conditions that require admission to neonatal intensive care such as asphyxia, convulsions, respiratory distress, pulmonary hypotension, severe infection, clavicle fracture, pneumothorax, prematurity, plexus injury and severe jaundice.

Data was also collected by a questionnaire distributed two months after birth. Women’s birth experience was investigated using the original version of Childbirth Experience Questionnaire [34] and its domains Own capacity, Professional support, Perceived safety, and Participation. Cronbach alpha values for the domains ranged from 0.81 to 0.91.

An additional composite variable was created to further study the importance of having a known midwife and cluster membership for the domains of the Childbirth Experience Questionnaire (CEQ). This variable consisted of four levels: 1) Women in the “emotionally healthy” cluster who did not have a known midwife during labor and birth (reference group); 2) Women in the “emotionally healthy” cluster who had a known midwife during labor and birth; 3) Women in the “emotionally unhealthy” cluster who did not have a known midwife during labor and birth; and 4) Women in the “emotionally unhealthy” cluster who had a known midwife during labor and birth.

Analysis

Statistical analysis was conducted using SPSS for Windows Chicago, IL, USA Version 25. We used the previously developed cluster analysis [27] and compared the explanatory variables by calculating crude and adjusted odds ratios with 95% confidence intervals (CI) between the clusters. T-tests or ANOVA was used to detect differences between the clusters and the domains of the Childbirth Experience Questionnaire.

Result

A total of 314 women consented to participate when approached and were sent the first questionnaire, which was returned by 280 women (89%). Before administrating the follow-up questionnaires we got information from the project midwives that 23 of the women had had a miscarriage or gave birth prematurely at a university clinic, and 13 had withdrawn from participation. The follow-up questionnaire was sent to the remaining 278 women and returned by 236 (85%). The 42 women who were eligible for a follow-up, but did not return the follow-up questionnaire, were more likely to be born outside Sweden (p 0.000) and had not returned the first questionnaire (p 0.000). In all, 243 antenatal and birth records were scrutinized and questionnaires from 229 women regarding the birth experience, was further analyzed.

The majority of the recruited women were 25–35 years old, living with a partner, and born in Sweden (Table 1). The most common level of education was high school and the majority had previous children. Data from mid-pregnancy the women who completed all four scales that formed the basis for the cluster analysis revealed that around 35% of the sample belonged to the “emotionally unhealthy cluster” and 65% of the women belonged to the “emotionally healthy” cluster. Women not living with a partner and women born in a country other than Sweden were more likely to present in the “emotionally unhealthy cluster”. From the antenatal electronic records the result also showed that women in the “emotionally unhealthy cluster” were more likely to had a history of medical and psychiatric conditions and identified risk factors (Table 1).

Table 1. Women’s background in relation to cluster belonging.

	Emotionally healthy cluster n = 157	Emotionally unhealthy cluster n = 86	p Value
	n (%)	n (%)
Socio-demographic background (data collected from questionnaires)
Age groups
25 years	24 (15.3)	13 (15.3)
25–35 years	111 (70.7)	58 (68.2)
35 years	22 (14.0)	14 (16.5)	0.873
Civil status
Living with a partner	151 (96.2)	75 (88.2)
Not living with a partner	6 (3.8)	10 (11.8)	0.028
Country of birth
Sweden	150 (95.5)	70 (82.4)
Other country	7 (4.5)	15 (17.6)	0.002
Level of education
High school or lower	100 (63.7)	54 (63.5)
University education	57 (36.3)	31 (36.5)	1.000
Parity
Primiparas	60 (38.2)	38 (44.7)
Multiparas	97 (61.8)	47 (55.3)	0.340
Health related background (data collected from antenatal records)
Previous medical condition
Yes	14 (8.9)	17 (19.8)
No	143 (91.1)	69 (80.2)	0.025
Previous psychiatric condition
Yes	15 (9.6)	32 (37.2)
No	141 (90.4)	54 (62.8)	0.000
Identified risk factors
Yes	21 (13.5)	22 (25.6)
No	135 (86.5)	63 (74.1)	0.022
Pregnancy-related complications
Yes	31 (19.9)	16 (18.8)
No	125 (80.1)	69 (81.2)	1.000

The birth outcome of women in the two clusters comprising 243 women are presented in Table 2. The majority of women had a normal length of pregnancy with a spontaneous onset of labor and a normal vaginal birth. The proportion of epidural anesthesia ranged from 30 to 45% and labor augmentation with synthetic oxytocin from 37 to 39%. Slightly more than one third of the women had a known midwife assist during childbirth. The only birth variable associated with cluster membership was that women in the “emotionally unhealthy” cluster had a higher occurrence of severe neonatal diagnosis. The finding boarded on statistical significance when adjusting for background characteristics.

Table 2. Birth outcome in relation to cluster belonging.

	Emotionally healthy cluster n = 157	Emotionally unhealthy cluster n = 86	Crude Odds Ratio	Adjusted^a Odds Ratio
	n (%)	n (%)	(95% CI)	(95% CI)
(data collected from birth records)
Length of pregnancy (weeks)
mean (SD)	39.54 (1.71)	39.22 (1.97)	0.90 (0.78–1.04)	0.94 (0.78–1.15)
Onset of labor
Spontaneously	113 (72.0)	63 (73.3)	1.0 Ref.	1.0 Ref.
Induction	44 (28.0)	23 (26.7)	0.94 (0.51–1.69)	1.35 (0.34–5.32)
Mode of birth
Normal vaginal	124 (79.0)	68 (79.1)	1.0 Ref.	1.0 Ref.
Instrumental vaginal	9 (5.7)	3 (3.5)	0.60 (0.15–2.32)	0.70 (0.14–3.36)
Elective cesarean section	8 (5.1)	6 (7.0)	1.36 (0.45–4.10)	1.47 (0.38–5.66
Emergency cesarean section	16 (10.2)	9 (10.5)	1.02 (0.43–2.44)	0.91 (0.28–2.81)
Epidural analgesia
Yes	63 (40.1)	38 (45.2)	1.23 (0.72–2.10)	1.10 (0.54–2.24)
No	94 (59.9)	46 (54.8)	1.0 Ref.	1.0 Ref.
Augementation of labor
Yes	57 (39.3)	29 (37.2)	0.91 (0.51–1.61)	0.61 (0.28–1.30)
No	88 (60.7)	49 (62.8)	1.0 Ref.	1.0 Ref.
Severe birth complication
Yes	42 (26.8)	15 (17.4)	0.57 (0.29–1.11)	0.52 (0.23–1.17)
No	115 (73.2)	71 (82.6)	1.0 Ref.	1.0 Ref.
Severe neonatal diagnosis
Yes	15 (9.6)	17 (19.8)	2.33 (1.10–4.94)*	2.53 (0.98–6.48)
No	142 (90.4)	69 (80.2)	1.0 Ref.	1.0 Ref.
Known midwife during childbirth
Yes	52 (33.1)	28 (32.6)	0.97 (0.55–1.70)	0.83 (0.42–1.63)
No	105 (66.9)	58 (67.4)	1.0 Ref.	1.0 Ref.

a Adjusted for statistically significant background characteristics and mode of birth.

*p < 0.05, **p < 0.01, ***<0.001.

Table 3 shows the mean values (SD) of the domains of the Childbirth Experience Questionnaire (CEQ) in relation to cluster membership. There were statistically significant differences between the clusters for most domains except for Professional support.

Table 3. Mean values (SD) of The Childbirth Experience Questionnaire (CEQ) in relation to cluster belonging.

	Emotionally healthy cluster n = 144	Emotionally Unhealthy cluster n = 78
	mean (SD)	mean (SD)	p Value
Own Capacity	2.80 (0.67)	2.60 (0.71)	0.048
Professional support	3.60 (0.63)	3.44 (0.67)	0.082
Perceived safety	3.24 (0.66)	2.87 (0.70)	0.000
Participation	3.26 (0.89)	2.98 (0.89)	0.033
Total CEQ	3.14 (0.63)	2.89 (0.64)	0.006

As one of the goals with this project was to offer women continuity with a known midwife during pregnancy and childbirth, we further explored the relationship between cluster membership, continuity, and the Childbirth Experience Questionnaire (CEQ) domains. In Table 4, the subgroups of women are presented. In general, women in Cluster 2 (the emotionally unhealthy cluster) who did not have a known midwife during labor and birth presented the lowest scores in all domains and in the total CEQ. However, women in this cluster who received continuity from a known midwife also scored highest in the domains Own Capacity and Professional support. For the other two domains and the total CEQ, women in Cluster 1 (emotionally healthy) who had a known midwife scored highest in the domains Perceived safety and Participation, as well as in the total CEQ. The effect sizes were medium for Professional support, Perceived safety, and total CEQ and small for Own capacity and Participation.

Table 4. The relationship between cluster membership and continuity and childbirth experience^a.

	Own capacity	Professional support	Perceived safety	Participation	Total CEQ
	mean (SD)	mean (SD)	mean (SD)	mean (SD)	mean (SD)
Women in emotionally healthy cluster, no known midwife (n = 99)	2.77 (0.66)	3.55 (0.68)	3.20 (0.69)	3.17 (0.97)	3.09 (0.66)
Women in emotionally healthy cluster, known midwife (n = 48)	2.85 (0.68)	3.70 (0.51)	3.31 (0.62)	3.44 (0.66)	3.27 (0.55)
Women in emotionally unhealthy cluster, no known midwife (n = 55)	2.46 (0.75)	3.29 (0.74)	2.73 (0.68)	2.93 (0.95)	2.74 (0.69)
Women in emotionally unhealthy cluster, known midwife (n = 25)	2.90 (0.51)	3.75 (0.32)	3.16 (0.67)	3.09 (0.77)	3.19 (0.42)
p Value	0.010	0.004	0.000	0.043	0.001
F-statistics	3.846	4.657	7.539	2.759	6.123
Effect size	0.05	0.06	0.094	0.04	0.07

a Data collected by questionnaires two months after birth.

Discussion

This study showed few differences in birth outcome in relation to women’s emotional wellbeing, with the exception of more neonatal complications in women belonging to the unhealthy cluster. One important finding was the difference in the childbirth experience, with higher level of satisfaction when having a known midwife, regardless of cluster belonging.

From women’s background data we found that single women and women born in a country outside Sweden were more likely to be found in the “emotionally unhealthy cluster” Similar findings have been reported elsewhere, where single status and fear of birth were associated with depressive symptoms [35,36]. Being born in a country outside Sweden has been associated with depressive symptoms [37] as well as with fear of birth [38].

Although the antenatal records showed that more women in the “emotionally unhealthy cluster” often had a history of medical and/or psychiatric conditions, the birth outcome was similar. The only exception was the higher proportion of severe neonatal diagnosis found in the cluster of more emotionally unhealthy women. Research has shown that women with severe depression face an increased risk of preterm birth, low Apgar scores and stillbirth and admission to neonatal care [39]. Smith and coworkers also concluded from a systematic review that depressive symptoms during pregnancy increase the likelihood of poor health in new-borns [40]. In a systematic review and meta-analysis based on 30 scientific papers, the result showed that depressive symptoms during pregnancy were associated with an increased risk of premature birth [41]. This finding was further confirmed in a systematic review of 39 scientific studies pointed out a strong evidence that anxiety, stress, and depressive symptoms increase the risk for spontaneous premature births [42]. Studies have also found negative consequences for the relationship between the mother and the infant if the mother has symptoms of anxiety or depression [43].

The cluster profiles in this study were somewhat similar to the results from a cluster analysis performed on psychological variables of women with fear of birth who were randomized into internet-based cognitive therapy or counseling with midwives [23]. Depressive symptoms, fear, and anxiety were prevalent in that study. The study also stressed that blood and injection phobia could be underlying reasons for fear of birth. According to the authors of the present paper (all midwives, experienced in antenatal and intrapartum care), such phobias are rare and might be more obvious when seeking psychological care. Existential reasons are the more common cause of fear of birth for the majority of pregnant women in Sweden, such as the unpredictability of birth [44,45], being harmed during labor and birth [42], worries that something will be wrong with the baby [14,46], or previous experience of sexual abuse [47].

The importance of a known midwife

One of the prerequisites of the project was to prioritize “vulnerable” women, such as those with fear of birth or other psychosocial problems. The result showed, however, no difference between the clusters in terms of having a known midwife or not, in the crude analysis. However, the subgroup analysis showed a different pattern, favoring continuity when it comes to the childbirth experience. Women in the “emotionally unhealthy” cluster, e.g. those afraid of birth, having depressive symptoms, low sense of coherence, and major worries, who did not have a known midwife during birth scored lowest in all domains of the CEQ. In contrast, having a known midwife increased the likelihood of assessing the Own Capacity and the Professional support as highest among women in this cluster. The importance of continuity was also shown in the two other clusters (Perceived safety and Participation), in which women in the “emotionally healthy” cluster scored the highest of all subgroups. This finding, although based on small groups when breaking down the sample, is very much in line with other studies. In a previous randomized controlled trial of women afraid of birth who received internet-based cognitive therapy or counseling with midwives [48], the mean scores of Own Capacity were quite similar to the values reported in the more vulnerable cluster in the present study (2.64 vs 2.68), suggesting that women who have emotional health problems do not view their capacity as high. Interestingly, it seems that continuity of a known midwife could actually help women to trust their own capacity, e.g. feel empowered when receiving continuity. In the other domains it was also obvious that continuity with a known midwife attending the birth mattered.

There was no difference between the clusters in most of the variables included in labor outcome. During the period many women, regardless of cluster, were admitted for labor induction, often due to long travel distance to hospital. Also, there were few planned cesarean sections. The association between, e.g. fear of birth and request for a cesarean section is well known [8,9,13]. In a previous publication from the current study [27], 9% in the “emotionally unhealthy” cluster reported a cesarean section preference in mid pregnancy, versus 4% in the “emotionally heathy” cluster; this did not yield a statistically significant difference. The follow-up results showed that few women, regardless of cluster category (7.5% vs 4%) actually had an elective cesarean section.

Methodological considerations

This study has some limitations that will compromise the findings. First, the non-randomized design makes it difficult to evaluate any cause and effects. Due to the nature of the study, with limited access to midwives and few pregnant women in the area that fulfilled the inclusion criteria, randomization would have challenged both the development of and the time allocated to the project. Another thing to consider is the fairly small sample size and the limited number of women who actually received continuity because of circumstances beyond our control due to the lack of midwives. Nevertheless, the findings are in accordance with international studies, such as a better birth experience when having a known midwife [49]. The contextual circumstances also limit the applicability of the findings outside of Sweden. Another strength is the use of birth records that was almost complete. The exception was some few women recruited to the study who gave birth prematurely on a university hospital outside the region and some women who chose to give birth in a hospital outside the region.

Conclusion

This study showed few differences in birth outcome between the clusters, while there were explicit differences in the childbirth experience. Having a known midwife is important to warrant women a more positive childbirth experience. Screening with validated instruments during antenatal care could be a first step to further investigate women’s emotional well-being and provide targeted psychosocial support.

Disclosure statement

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

Additional information

Funding

The study was funded by grants from the Kamprad Family Foundation for Entrepreneurship, Research and Charity [Grant number 20190008].