Fear of childbirth in obstetrically low‐risk nulliparous women in Sweden and Denmark

Abstract

The prevalence of troublesome fear of childbirth (FOC) in Western countries is about 20%, of which approximately 6–10% suffer from severe FOC that impacts daily life. The countries of Sweden and Denmark are quite alike as far as childbirth culture is concerned. However, to some extent they differ in the organisation of midwifery care during the antenatal and labour period, respectively, and this may influence women's FOC. The aims of this study were to compare FOC among Danish and Swedish nulliparous women and to investigate a possible difference in FOC between women who, during pregnancy, had met the midwife who they were subsequently coincidentally allocated to receive labour care from and women who had not previously met the midwife. In total 165 women participated, comprising 55 Swedes and 110 Danes, of whom 55 among the latter group had met the midwife during pregnancy. Severe FOC was found in 10%. There were no differences between the Swedish women and the Danish women who had or had not met the midwife. Fear of childbirth measured in gestational week 37 correlated positively with fear at admission to the labour ward.

Keywords:

• Fear of childbirth
• nulliparous women
• cross‐national comparison
• Sweden
• Denmark

Introduction

Childbirth is a significant life event for most women. Although women in developed countries face minimal risks of adverse outcomes as compared to women in developing countries, many experience fear of childbirth (FOC) in response to the unknown, unpredictable and uncontrollable nature of the birth event (Zar, Wijma, & Wijma, 2001). FOC is a continuum ranging from negligible to extreme fear (Wijma, Wijma, & Zar, 1998). About 20% of pregnant women in Western countries report troublesome FOC (Hofberg & Ward, 2003; Zar et al., 2001), whereas approximately 6–10% suffer from severe FOC that seriously influences daily life (Areskog, Uddenberg, & Kjessler, 1981; Geissbuehler & Eberhard, 2002; Saisto & Halmesmäki, 2003; Waldenström, Hildingsson, & Ryding, 2006a; Wijma, 2003).

Many reasons for FOC have been reported such as fear of pain (Eriksson, Westman, & Hamberg, 2006; Geissbuehler & Eberhard, 2002; Sjögren & Thomassen, 1997), of losing control (Melender, 2002; Saisto, Ylikorkala, & Halmesmäki, 1999; Sjögren & Thomassen, 1997), of rupturing (Sjögren, 1997), of operative delivery (Sjögren, 1997; Szeverényi, Póka, Hetey, & Török, 1998), and of having an impaired or stillborn child (Eriksson et al., 2006; Saisto et al., 1999; Sjögren, 1997; Sjögren & Thomassen, 1997; Szeverenyi et al., 1998). Women whose previous experience of giving birth is negative are at risk of FOC (Ryding, Wirfelt, Wängborg, Sjögren, & Edman, 2007). Other important risk factors for developing FOC are psychological background factors such as general anxiety and depression (Zar, Wijma, & Wijma, 2002), experiences of a problematic childhood of their own (Ryding, Wirfelt, Wängborg, Sjögren, & Edman, 2007) and adverse socio‐economic conditions (Laursen, Hedegaard, & Johansen, 2008; Melender, 2002; Saisto, Salmela‐Aro, Nurmi, & Halmesmäki, 2001). Moreover, studies have shown that childbirth fear is related to distrust of delivery ward staff (Saisto et al., 1999; Sjögren & Thomassen, 1997).

The support and management of labouring women suffering from FOC are discussed intensively, as yet with no conclusive results. One subject of these discussions is the handling of severe FOC by means of caesarean section (Lavender, Hofmeyr, Neilson, Kingdon, & Gyte, 2007), FOC being an important basis for women's requests for elective caesarean sections (Melender, 2002; Nerum, Halvorsen, Sørli, & Ølan, 2006; Wiklund, Edman, & Andolf, 2007). Antenatal education to prepare parents‐to‐be for childbirth is widespread and attended by a good number of pregnant women. However, the effects of such classes on women's anxiety, sense of control, experience of pain during labour and birth support are still unknown (Gagnon & Sandall, 2008).

Although FOC is associated with fear of pain, studies have shown that pharmacological pain relief does not automatically imply increased satisfaction with the labour process (Alehagen, Wijma, & Wijma, 2006; Lind & Hoel, 1989; Waldenström, 1999). In general, support from the midwife, the quality of the caregiver–patient relationship and involvement in decision‐making figure significantly in women's evaluation of their childbirth experiences (Hodnett, 2002; Persson & Dykes, 2007; Waldenström, Rudman, & Hildingsson, 2006b). To our knowledge, none to date has studied midwifery care of nulliparous women with FOC during childbirth.

The primary aim of this study was to compare FOC among Danish and Swedish nulliparous women. This comparison is of interest as Denmark and Sweden are fairly similar as far as childbirth culture is concerned. Every woman sees one midwife during antenatal visits unless the midwife is ill or otherwise unable to be present. However, the two countries do differ in the organisation of midwifery care during the antenatal and labour period. In Sweden, antenatal care and labour care are separated, i.e. the woman, when about to give birth, meets another midwife than the one she knows from her antenatal visits. Thus the midwife attending the labour is almost always unknown to the labouring woman. In Denmark, on the other hand, midwives' work schedules comprise working days at the antenatal clinic as well as shifts at the delivery ward. In consequence, the same midwife may coincidentally assist the woman both during pregnancy and labour. Although this is generally not a systematically scheduled effort to ensure continuity of carers, it may be a way to strengthen the psychological support and care of the woman giving birth and reduce her FOC. In the comparison of FOC we added women's experiences of pain as a variable as studies have shown association between FOC and pain (Eriksson et al., 2006; Geissbuehler & Eberhard, 2002; Sjögren & Thomassen, 1997).

Our secondary aim was to investigate a possible difference in FOC between women who, during the pregnancy, had met the midwife assisting at the birth and those who had not.

Method

In two studies performed independently in Denmark and Sweden by two distinct research groups, the recruitment strategies and data collection instruments employed were almost identical. This allowed for the first ever comparison of data on FOC from both countries, and is a starting point for cross‐national comparisons focusing on the psychosocial aspects of childbirth.

Procedures in Sweden

The Swedish data are part of a broader study on the interaction of FOC and the output of stress hormones before, during and after childbirth and the use of pain relief and obstetrical variables during labour (Alehagen, Wijma, Lundberg, & Wijma, 2005). Inclusion into the Swedish study took place in 1996.

Written information was given to nulliparous women who, in gestational weeks 33–35, visited antenatal clinics of the Department of Obstetrics and Gynaecology, Linköping University Hospital, Linköping, with approximately 2500 annual deliveries. Potential participants were telephoned by the research leader who offered supplemental information, invited potential respondents to participate and obtained written informed consent. In pregnancy week 37, participants completed the Wijma Delivery Expectancy/Experience Questionnaire (W‐DEQ version A) and the Delivery Fear Scale (DFS) (Wijma, Alehagen, & Wijma, 2002) and a pain score was filled in at the woman's admission to the labour ward. The items of the DFS were read aloud to the woman and her reply was recorded by a midwife/research leader. The pain score was filled in by the midwife/research leader, after having asked the woman to indicate her perception of the pain intensity. The study was approved by the Ethics Committee at the Faculty of Health Sciences, Linköping University.

Procedures in Denmark

The Danish data stem from the Danish Dystocia Study, a population‐based multi‐centre study using prospectively collected data from a fixed cohort with open entry of nulliparous women followed from gestational week 37 through to 2 weeks after delivery (Kjærgaard, 2007). Inclusion into the study took place between May 2004 and July 2005 from antenatal clinics of nine obstetric departments in Denmark. Four of these units belonged to university hospitals and three to county hospitals, while two were local district departments. The annual birth rates per clinic varied between approximately 850 and 5400. Written information was given in gestational week 33 and written informed consent was obtained at the antenatal visit in gestational week 37. At this point of time the participants received a questionnaire comprising W‐DEQ version A (Wijma et al., 1998) along with a password to an electronic version of the questionnaire. Participants were free to complete either the electronic or the paper version.

Following the admission procedures, items of the DFS were read aloud to the woman and her reply was recorded on the scale by the midwife or nurses‐aid. The pain score was registered by the midwife according to the woman's indication of her perception of the pain intensity. The participants completed a postnatal questionnaire within the first two weeks of the delivery. This questionnaire comprised information on whether, during pregnancy, the woman had met the midwife who assisted her during admittance and during labour and delivery.

In preparation for the study, items of both the W‐DEQ and DFS were translated and back‐translated by four bilingual persons (English–Danish and Swedish–Danish). In a pilot study, all steps of the data collection were tested and compliance, potential misinterpretations and response rates were examined. Forty‐two women were interviewed about their responses to the W‐DEQ. Seventy‐nine women were asked to reply to the DFS. Some adjustments were made according to the findings of the pilot study prior to the implementation of the main study (Kjærgaard, 2007).

No Ethics Committee approval was necessary according to Danish law, since no invasive procedures were applied. Permission to establish the database was obtained from the Danish Data Protection Agency (j.nr. 2004‐41‐3995).

Participants

In both countries, the inclusion criteria were age ≥18 years, nulliparity, singleton pregnancy, no planned elective caesarean section or induction of labour. Exclusion criteria at delivery were <37 or >42 weeks of gestation, induction, elective caesarean delivery and breech presentation. Participants were required to be able to communicate in Swedish or Danish, respectively.

The Swedish sample

In the Swedish sample, 189 nulliparous women were invited to participate. Forty‐two women did not meet the inclusion criteria and 74 declined to participate. Although 73 women participated in the pregnancy measurement, collection of later measurements for 18 of these was incomplete. Thus, the full measurements of 55 women were available. The Swedish data set on these 55 women constituted the basis for the selection of the Danish participants.

The Danish sample

From the database of the Danish Dystocia Study, we initially identified a group of women who had completed W‐DEQ version A, the DFS and pain scores (n = 2067). Of these, we identified those who in a postnatal questionnaire stated that during pregnancy they had met the midwife who subsequently by coincidence was the one who assisted them during labour and delivery. As one of the measurements of FOC was carried out briefly after the woman's admittance to the labour ward, it was necessary to ensure that when responding to this measurement, she was aware that she would be assisted by the midwife whom she already knew. We therefore carried out a further search to identify women who were assisted only by this midwife from admittance through labour (n = 101). Among these we selected 55 women who were closest in age to and had a cervix dilatation at admission similar to the women in the Swedish sample. We chose these two latter selection criteria as we consider age and dilatation of cervix to influence the perception of FOC at admission to labour.

In addition, we selected another sample of 55 Danish women who had not met the midwife during pregnancy. These women were also selected on age and cervix dilatation. This sample was used for comparison of FOC with those Danish women who had met the midwife during pregnancy and the Swedish group, respectively, in order to examine a potential national influence.

In summary, the final study sample comprised 165 nulliparous women almost identical on age and dilatation of cervix at admission; 55 Swedish women, 55 Danish women who during pregnancy had met the midwife who by coincidence assisted them during labour and delivery, and 55 Danish women who had not met the midwife during pregnancy. All participants had completed the W‐DEQ version A in gestational week 37 and were admitted to hospital in spontaneous labour at term, with a singleton infant in cephalic presentation. After admission procedures, all participants had replied to DFS and indicated a pain score.

Measures

We chose to focus on measures that disclose FOC in late pregnancy and at admission to labour ward, as FOC identified at these points in time may substantiate potential preventive interventions in late pregnancy as well as from admittance to labour ward onwards through delivery. Thus we consider that measuring at these points in time may have the greatest clinical implications.

The W‐DEQ version A (Wijma et al., 1998) measures FOC by means of a woman's cognitive appraisal of the delivery. Version A concerns the woman's expectations of the pending delivery. The questionnaire has 33 items which are rated on a 0–5 Likert scale from ‘not at all’ ( = 0) to ‘extremely’ ( = 5). The sum‐score may vary from 0 to 165. The higher the score, the more intense the fear of childbirth (Wijma et al., 1998). A sum‐score ≥85 is considered as severe fear of childbirth (Ryding, Wijma, & Wijma, 1998).

The Delivery Fear Scale (Wijma et al., 2002) measures FOC during labour and delivery. This instrument comprises 10 items to be answered by a number from 1 to 10 (1 = ‘do not agree at all’, 10 = ‘agree totally’). A cut‐off level for severe fear of childbirth was set at a sum score ≥70.

Pain intensity was assessed by a one‐item pain intensity scale at admission: ‘How did you experience the last contraction?’ (Swedish sample) and ‘How do you experience the pain?’ (Danish sample). The woman's appraisal of pain was measured on a scale from 1 to 10 in the Swedish sample and a scale from 0 to 10 in the Danish sample (0/1 = ‘no pain’, 10 = ‘totally unbearable pain’). Before calculating the sum scores of the Danish sample the values 0 and 1 were pooled, as these values comprised 0.5% and 0.7%, respectively, of the total scores. The pooled value was given the value 1. A cut‐off level for severe pain was set at a sum score ≥7.

We dealt with missing values in the W‐DEQ by excluding those with ≥7 missing items. For those missing ≤6 items, missing items were replaced by the mean value of the respondents' remaining items. Cases with ≥4 missing values in DFS were excluded. For those with ≤3 missing items, missing values were replaced by the mean value of the respondents' remaining items.

The Danish women's acquaintance with the midwife was assessed by the following items from the postnatal questionnaire: ‘I knew her from antenatal visits, home visit or antenatal classes’ (n = 44), or ‘I knew her from admission to hospital during the pregnancy’ (n = 2), or ‘I knew her from other encounters’ (n = 9).

The Swedish and the Danish questionnaires differed in their registration of educational level. Information on Swedish women's educational level was based on one item, the women choosing one of four alternatives describing four levels of education, whereas information from the Danish sample was obtained from two different items in the questionnaire, one comprising basic schooling up to 12 years and the second further training/study after basic schooling.

Analyses

Prior to analyses, the following variables were categorised according to pre‐defined categories: age, educational level and pain. Sum scores of W‐DEQ and DFS were calculated and grouped. Items relating to acquaintance with the midwife were dichotomised into ‘yes’ and ‘no’.

A possible difference in age was tested with Student's t‐test. Mann–Whitney U‐test was applied to test possible differences in FOC between Danish and Swedish nulliparous women during pregnancy as well as in FOC and pain at admission to labour ward (aim 1). Spearman's rank correlation was applied to test for a relation between FOC in gestational week 37 and at admission to the labour ward, as well as a relation between FOC and pain at admittance to the ward (aim 1). The Kruskal–Wallis test was used to ascertain possible differences in FOC during pregnancy, and FOC and pain at admission to the labour ward, between women who, during their pregnancy, had coincidentally met the midwife assisting at the birth and those who had not (aim 2).

Calculations were performed with SPSS 14.0 software (SPSS Inc., Chicago, IL). Statistical significance was accepted at p<0.05.

Results

Social and obstetrical characteristics are presented in Table 1. The mean age in the total sample was 26.9 (SD 3.9). At admission to the labour ward 146 of the women had a cervix dilatation of 3–5 cm, 17 women had a cervix dilatation of 6–9 cm and two women had a fully dilated cervix. Methods for registration of educational level differed too much in the two samples to pool data or make comparisons. The educational level in the Swedish population was as follows: Elementary school: 7%; high school: 46%; and university degree: 47%. In the Danish sample, 7% had completed elementary school (<10 years in school), 62% had completed upper secondary school/high school (3 years following 9 or 10 years in elementary school), and 39% had completed a university study or other study/training ≥3 years after having completed upper secondary school/high school. The sum of the Danish figures exceeds 100%, as there is some overlap between the upper secondary school/high school group and the group that completed university study or other study/training after upper secondary school/high school.

Table 1. Social and obstetric characteristics in Swedish and Danish samples of nulliparous women.

Characteristics	Swedish women n = 55	Danish women n = 110	P
Age mean (SD)	26.4 (4.1)	27.2 (3.7)	0.269*
Married or cohabiting
Yes n (%)	53 (96)	106 (96)
No n (%)	2 (4)	4 (4)
Dilatation of cervix at admission to labour ward
3–5 cm n (%)	50 (91)	97 (88)
6–9 cm n (%)	5 (9)	12 (11)
10 cm n (%)	0	1 (1)
Note: *Student's t‐test.

There were no differences between the Swedish and Danish women either in FOC in late pregnancy (W‐DEQ) or in fear during childbirth (DFS) and pain at admission to labour ward (Table 2). The median value of the W‐DEQ scores in the total sample as assessed in gestational week 37 was 59.0 (min.: 15, max.: 135). The median DFS score at admission to the labour ward was 41.0 (min.: 10, max.: 89) and pain showed a median of 7.0 (min.: 1, max.: 10). Eleven percent of the women had a sum‐score of ≥85 on the W‐DEQ and 10% scored ≥70 on the DFS, which is considered severe fear of childbirth. Sixty percent of the women scored pain ≥7.

Table 2. Fear of childbirth in late pregnancy (W‐DEQ version A), and fear (DFS) and pain at admission to the labour ward in the Swedish and Danish samples of nulliparous women.

Characteristics	Swedish women n = 55	Danish women n = 110	p*
W‐DEQ version A^1
median (min–max)	64 (21–135)	58.5 (15–101)	0.661
⩾85† n (%)	8 (14)	10 (9)
DFS^2
median (min–max)	40 (14–89)	41.5 (10–83)	0.977
⩾70† n (%)	6 (11)	10 (9)
Pain^3
median (min–max)	7 (2–10)	7 (1–10)	0.478
⩾7† n (%)	35 (64)	64 (58)
Notes: ^1W‐DEQ = Wijma Delivery Expectancy/Experience Questionnaire version A; ^2DFS = Delivery Fear Scale; ^3Pain = a pain intensity scale; † values represent severe fear and severe pain; * Mann–Whitney U‐test.

FOC measured in gestational week 37 correlated positively with fear at admission to the labour ward in the Swedish (r = 0.33, p<0.05) and the Danish (r = 0.22, p<0.05) women. The experiences of pain at admission to the labour ward also correlated positively with FOC in both groups (Swedish women: r = 0.34, p<0.05 and Danish women: r = 0.40, p<0.05).

None of the Swedish women had previously met the midwife who attended the delivery. In accordance with the design, 55 of the Danish women had met the midwife during pregnancy and 55 had not. There were no differences between these three groups regarding FOC in gestational week 37; FOC at admission, or pain at admission, neither were there differences between these three groups regarding severe FOC in gestational week 37 or at admission, or severe pain at admission.

Discussion

This study is the first comparison of FOC in samples of pregnant women from Denmark and Sweden. Like previous studies (Areskog et al., 1981; Geissbueler & Eberhard, 2002; Saisto & Halmesmäki, 2003; Waldenström et al., 2006a; Wijma, 2003), we found that 10% of our sample (Danish and Swedish women combined) had severe FOC, i.e. suffering from FOC that seriously influences their daily lives. The same prevalence of severe FOC was found in a recent study based on a large Danish cohort of nulliparous women (Laursen et al., 2008).

According to clinical experience, it is generally encouraging for women in labour to encounter a familiar face when they arrive at the labour ward. However, we did not find any difference in FOC between women who during pregnancy had met the midwife assisting at the birth and those who had not. The Danish and Swedish women did not differ in FOC either in late pregnancy or at admission to the labour ward.

According to Lundgren and Berg (2007), there are central conditions that should be fulfilled to ensure a sustainable contact between the woman and the midwife. It is essential that the childbearing woman surrenders to and trusts the midwife and has a willingness to participate. Midwives should respond with availability, mediation of trust, confirmation, and meaningful and unique support. Neither the Danish subgroup of women nor the midwives knew until admission to the labour ward that they would meet again, and this may constitute a barrier for establishing some of the dimensions described by Lundgren and Berg. Probably therefore we could not expect to find a difference in FOC between the groups who had or had not met the midwife before and our findings do not constitute a true test of the impact of ‘knowing’ the midwife. However, our findings suggest that a chance encounter at admission with a midwife previously acquainted during pregnancy does not lead to a benefit that reduces FOC. The quality of the encounter may be more important than the continuity of the carer (Green, Renfrew, & Curtis, 2000). Our results show a correlation between FOC in late pregnancy and at admission to labour ward, supporting the so‐called ‘vicious circle principle’ (Zar et al., 2001), i.e. that women who fear childbirth during pregnancy also experience fear during birth. This highlights the possibility of identifying women with FOC already during pregnancy.

Data on the two samples were collected in 1996 and in 2004–2005, respectively. It could be argued that some cohort effect may have influenced the findings of FOC, presuming that the prevalence of FOC had changed over time. We do not consider the time span of the data collection to play a major role in the prevalence of FOC based on findings from studies in the two countries. FOC has recently been studied in a Danish cohort of 30,480 nulliparous women. The authors found that the prevalence of FOC was stable in the period 1997–2003 and that self‐rated health was the most important risk factor for FOC followed by lack of social network, having an unskilled job or vocational education, being a current smoker, of young age and unemployed (Laursen et al., 2008). Swedish studies have also reported a stability of FOC over time (Areskog et al., 1981; Waldenström et al., 2006a).

Our study has limitations. We conducted two separate studies, one in Sweden and one in Denmark. The Swedish sample comprised 55 women and, as we chose to select twice the number for comparison, the final study sample consisted of 165 women. This sample size was suitable for descriptive analyses but did not allow analyses that could provide identification of risk estimates. Likewise, the lack of statistically significant differences in FOC may reflect low power. As FOC was measured at admission, we do not know if those Danish women who had previously met the midwife may have had some benefit of their previous acquaintance later in labour. We have no information on the level of FOC later in labour neither in the Swedish nor in the Danish groups. Moreover, although inclusion criteria were identical in the samples from both countries, we failed to have similar registrations of some of the social variables, which made comparison at that level impossible. We selected the Danish sample on the basis of cervix dilatations similar to the Swedish sample and this may not be the most relevant factor. Time spent in labour before admission may be associated with fear but unfortunately we did not have information on this from the Swedish sample. Finally, registration of acquaintance with the midwife was by a quantitative measurement. To gain a more thorough understanding of what constitutes a meaningful relationship between the labouring woman and her midwife, this kind of measure should be supplemented with qualitative data.

Clinical and research implications

Our findings add to the existing body of knowledge of FOC. Further studies on the correlation between FOC in late pregnancy and FOC at admission may open up avenues for development of preventive strategies. Studies on the use of WDE‐Q version A and DFS as screening tools may indicate the need for psychological assistance to women in late pregnancy and during labour. An acquaintance with the midwife during pregnancy does not in itself influence the woman's perception of FOC at admission. Qualitative studies to elucidate women's needs for FOC‐reducing support are recommended, as well as studies using quantitative designs to examine the effect of such support.

Conclusions

According to our findings, Swedish and Danish nulliparous women have similar levels of fear of delivery, both in late pregnancy and at admission to labour ward. Coincidentally being allocated to receive labour care from a midwife already met during pregnancy is not associated with less fear of delivery at admission to the labour ward.

Acknowledgements

The Danish Dystocia Study was funded by grants from the following institutions and foundations: Copenhagen Hospital Corporation, The Lundbeck Foundation, Aase and Ejnar Danielsen's Foundation, The Augustinus Foundation, The Health Insurance Foundation, The Danish Midwifery Association, King Christian X's Foundation, Lund University, Faculty of Medicine. The Swedish study was funded by research grants from The Swedish Foundation for Health Care and Science and Allergy Research, and The Bank of Sweden Tercentenary Foundation.