Fear of childbirth according to parity, gestational age, and obstetric history

Authors


Dr T Saisto, Department Obstetrics and Gynaecology, Helsinki University Central Hospital, Jorvi Hospital, PO Box 800, Helsinki FIN-00029 HUS, Finland. Email terhi.saisto@hus.fi

Abstract

Objective  To examine fear of childbirth according to parity, gestational age, and obstetric history.

Design  A questionnaire study.

Population and setting  1400 unselected pregnant women in outpatient maternity clinics of a university central hospital.

Methods  Visual analogue scale (VAS) and Wijma Delivery Expectancy/Experience Questionnaire (W-DEQ) and preferred mode of delivery.

Main outcome measures  W-DEQ and VAS scores according to parity, gestational age, obstetric history, and preferred mode of delivery.

Results  The W-DEQ and VAS scores were higher in nulliparous (W-DEQ 72.0 ± 20.0 [mean ± SD] and VAS 4.7 [median]) than parous women (65.4 ± 21.9; 3.2, P < 0.001 for both W-DEQ and VAS). Higher W-DEQ and VAS scores were found for those beyond 21 weeks of gestation compared with those before (W-DEQ 71.6 ± 23.0 versus 66.6 ± 20.0, P < 0.001; VAS 4.7 versus 3.2, P < 0.001). Caesarean section was preferred mode of delivery for 8.1% and these women scored higher on fear (W-DEQ 87.6 ± 26.5, VAS median 7.0) than those who preferred vaginal delivery (W-DEQ 61.8 ± 18.7, VAS 2.7, P < 0.001, respectively). Those with a previous caesarean scored higher on fear (W-DEQ 73.2 ± 23.5, VAS 5.1) than parous women without previous caesarean (W-DEQ 63.3 ± 20.8, VAS 2.9, P < 0.001, respectively). Those with a history of a vacuum extraction (VE) (W-DEQ 70.6 ± 19.7, VAS 5.0) had higher fear scores than those without (W-DEQ 64.8 ± 22.0, P < 0.05 and VAS 3.0, P < 0.001).

Conclusion  Severe fear of childbirth was more common in nulliparous women, in later pregnancy, and in women with previous caesarean section or VE. Caesarean section as a preferred mode of childbirth was strongly associated with high score in both W-DEQ and VAS.

Introduction

In previous studies, mostly conducted in the Nordic countries, it has been estimated that 6–10% of all pregnant women suffer from severe fear of childbirth.1–4 Some studies have shown a rising tendency for fear and anxiety later in advancing pregnancy.5,6 The fear may overshadow the entire pregnancy, complicate labour, lead to difficulties in the mother–infant relationship, and to postpartum depression.7 Fear is manifested mainly by anxiety, nightmares, and physical symptoms.8,9 Fear of childbirth often lies behind the mothers’ request for caesarean section 10 and could, if untreated, lead to unnecessary caesarean section without medical indication.11 Parity influences the contents of fear. Nulliparous women fear the unknown, pain, and loss of control. In parous women, fear arises from previous experiences.2,12,13 Earlier studies have shown that emergency caesarean section, vacuum extraction (VE), and untreated or unbearable pain during labour can lead to post-traumatic stress disorder (PTSD) and to fear of childbirth in subsequent pregnancies.2,14,15

The Wijma Delivery Expectancy/Experience Questionnaire (W-DEQ) has been developed to measure women’s feelings and fear before delivery by the means of the woman’s cognitive appraisal regarding the delivery process.16 It is a self-assessment scale and it has been used both in scientific and in clinical work.

Patient-rated visual analogue scales (VAS) are both reliable and valid tools in the measurement of pain, mood, and other subjective feelings.17 As far as we know, the VAS has not, however, been used before to evaluate fear. In addition to the W-DEQ, we wanted to add the VAS to our questionnaire to obtain more information, and to examine the suitability of the method in measuring fear of childbirth.

The primary aim of this study was to explore the fear of childbirth in relation to parity, gestational age, and previous obstetric experiences in a representative pregnant population. The second aim was to compare information from the VAS with a more traditional method of measuring fear of childbirth, the W-DEQ. A further aim was to study the relation between fear of childbirth and the preferred mode of delivery.

Materials and methods

Subjects

After approval from the local ethics committee, 2000 questionnaires were administered during October and November 2006 to consecutive and unselected Finnish- and Swedish-speaking maternity patients irrespective of gestational age, parity, or the reason for the visit the maternity outpatient clinics of the Women’s hospital and Jorvi hospital, both within the Department of Obstetrics and Gynaecology in the Helsinki University Central Hospital. Together these hospitals have approximately 8500 births per year. The women were asked to participate and given the questionnaire when they signed in at the outpatient clinics. They were asked to return the questionnaires before leaving the clinic. As the surveys were anonymous, we have no information on the women who did not complete the questionnaires.

Measurements

W-DEQ is self-assessment scale16 containing 33 items on childbirth (questions like, ‘How do you think you will feel in general during the labour and delivery?’ Extremely weak—not at all weak; extreme panic—not at all panicked; extreme trust—no trust at all). Answers are given on a 6-point scale ranging from ‘not at all’ (0) to ‘extremely’ (5), yielding a maximum score of 165 and a minimum score of 0. A higher score indicates more severe fear of childbirth. We used the W-DEQ A, which is a prepartum version of the scale. The internal consistency reliability (the Cronbach alpha reliability) coefficient of the W-DEQ A was 0.926. A W-DEQ score of >100 is regarded as indication of a clinical problem.18 After approval from the copyright holder (K Wijma, personal communication), the W-DEQ was translated into Finnish.

After filling in the W-DEQ, women were asked to indicate, on the visual analogue acale (VAS), how afraid they were of childbirth (scale from 0 to 10). They were also asked which mode of delivery (vaginal or caesarean section or probably vaginal or probably caesarean section) they would prefer and for information concerning their previous deliveries (any caesarean section or VE), miscarriages, or terminations of pregnancy (TOP).

Statistics

Before the analysis, gestational age at the time of filling in the questionnaire was classified into two groups: those who filled in the questionnaire before or at the 20th gestational week (n = 867) and those who filled it in after 20 weeks (n = 480). The first group was much larger because many women were there for routine ultrasound visits at approximately weeks 12 and 19.

Statistical analysis was carried out using SPSS 15.0 (Statistical Package for the Social Science for Windows, Chicago, IL, USA). Student’s t test for parametric continuous data and the chi-square test for skewed data were used to test the significance of differences between groups. P < 0.05 was considered significant.

Results

Of the 1400 (70%) forms returned, 1276 (91.1%) were filled so that W-DEQ score was able to count, and 1393 (98.3%) had a readable mark on VAS line. After excluding those who already had booked for caesarean section in this pregnancy (n = 44), those who were not pregnant (n = 6), and those who had previously had two or more caesarean sections but not yet booked for repeat caesarean section (n = 3), the final sample consisted of 1348 women (aged 16–47 years, mean 30.5 years) with mean gestational age of 22.2 (range 7–42) weeks. Nulliparous women accounted for 43.3% (n = 582) and parous women 56.7% (n = 763) of the total, which is representative of the normal Finnish pregnant population (www.stakes.fi, STAKES, Medical Birth Register, Finland). A second child was expected in 70.9% (n = 541) of the parous women, while 29.1% (n = 222) had already 2–9 children.

The W-DEQ scores were normally distributed (mean [±SD] W-DEQ score was 68.3 [±21.1], range 9–160, skewness 0.467, kurtosis 0.879). The effect of parity and gestational age on the W-DEQ score is shown in Tables 1 and 2. Nulliparous women were significantly more fearful than parous women both in early and late gestation. Those at more advanced gestational age scored higher on fear in the W-DEQ than those who filled in the questionnaire before or at 20 weeks of gestation. This difference was more significant in parous than in nulliparous women, as shown in Table 2. Maternal age had no significant effect on the W-DEQ score.

Table 1.  Fear of childbirth measured by W-DEQ and VAS according to parity and gestational age
 W-DEQ, n (%)W-DEQ score, mean ± SD (range)P95% CI for the difference between the meansVAS, n (%)VAS score, median (range)P
  1. Total number is not same in every part because of the incomplete answers of some participants.

Parity
0537 (43.3)72.0 ± 20 (17–160)<0.0014.1–8.9580 (43.2)4.7 (0–10)<0.001
≥1702 (56.7)65.4 ± 22 (9–150) 762 (56.8)3.2 (0–10) 
Gestational age at the time of answering the questionnaire (weeks)
7–20796 (64.9)66.6 ± 20 (17–136)<0.001−7.5 to −2.5846 (63.8)3.2 (0–10)<0.001
21–42430 (35.1)71.6±23 (9–160) 479 (36.2)4.7 (0–10) 
Table 2.  Effect of parity and gestational age to fear of childbirth measured by W-DEQ and VAS
 W-DEQ mean score ± SD at gestation weeks ≤20W-DEQ mean score ± SD at gestation weeks >20P*VAS median value ± SD at gestation weeks ≤20VAS median ± SD at gestation weeks >20P*
  • NS, no significance.

  • *

    P value for the difference of means in W-DEQ and the medians of VAS according to gestational age.

  • **

    P value for the difference of means in W-DEQ and the medians of VAS according to parity.

Parity
070.7 ± 1974.7 ± 21<0.054.3 ± 2.34.7 ± 2.4NS
≥163.0 ± 2069.7 ± 23<0.0013.5 ± 2.54.4 ± 2.5<0.001
P**<0.001<0.05 <0.001NS 

The severity of fear of childbirth was also estimated by VAS. The VAS scores (mean 4.1 [±2.5], median 3.7, range 0–10) were not normally distributed but rather showed a bimodal distribution (skewness 0.3112, kurtosis −0.979). Therefore, we used medians and chi-square test to compare the different groups. The nulliparous women had a median (range) score of 4.7 (0–10), which was significantly higher than that of the parous women with the median score of 3.2 (0–10; Table 1). The gestational age did not change the difference (Table 2). Again, a significant difference was found between women who answered the questionnaire at or before and those who answered it after gestational age of 20 weeks (Table 1). The difference was more significant in parous women (Table 2). Maternal age did not affect the VAS scores. The correlation between VAS and the W-DEQ was quite good (Pearson correlation coefficient r = 0.7, with significance at P = 0.01).

We were also interested in finding how well the VAS correlated with the W-DEQ and to see how well the questionnaire detected mother’s fear in our maternal population. A W-DEQ score of 100 or above was recorded in 38 nulliparous (7.0%) and in 55 parous women (7.7%). The VAS showed a sensitivity of 97.8% in screening for fear of childbirth (W-DEQ ≥100), when the VAS threshold was 5.0. At that value, the specificity of the scale was 65.7%. When the VAS threshold was 6.0, sensitivity was 89.2% and specificity 76.3%.

To determine the possible interconnectedness between fear of childbirth and the preferred mode of delivery, we compared the mean scores of W-DEQ and VAS from those who preferred vaginal delivery to those who preferred caesarean section or were ambivalent about their preference (Table 3). Altogether 8.1% of women reported a preference for caesarean section or probable preference for caesarean section. The women who preferred vaginal delivery scored lower on the W-DEQ and VAS than the women who preferred caesarean section or who were ambivalent about their preference. The difference between the other groups to the vaginal preference group was significant at the level of P < 0.001 (see Table 3).

Table 3.  Fear of childbirth (W-DEQ score and VAS) according to preferred mode of delivery
Preferred mode of deliveryW-DEQ, n (%)Mean W-DEQ score ± SD (range)P in comparison with the vaginal group 95% CI for the difference between the meansVAS, n (%)Median VAS ± SD (range)P in comparison to the vaginal group
  1. Total number is not same in every part because of the incomplete answers of some participants.

Vaginal783 (63.3)61.8 ± 18.7 (14–133) 847 (63.3)2.7 ± 2.2 (0–10.0) 
Probably vaginal356 (28.8)77.1 ± 17.9 (17–141)<0.001−17.6 to −12.9382 (28.6)5.2±2.3 (2.0–10.0)<0.001
Caesarean section48 (3.9)91.4 ± 31.5 (9–160)<0.001−35.4 to −23.955 (4.1)7.4 ± 2.7 (0.5–10.0)<0.001
Probably caesarean section49 (4.0)83.8 ± 20.0 (48–132)<0.001−27.4 to −16.654 (4.0)6.7 ± 2.3 (0.7–9.8)<0.001

Among the parous women (n = 763), there were 172 who had previously had one caesarean section (22.5%). The women who previously have had one caesarean section scored higher on fear in the W-DEQ and also gave a higher VAS count than those who had not had a caesarean section (Table 4). Altogether 87 women (11.4% of parous women) had previously given birth with the aid of VE. They scored higher on fear than those without previous VE both in W-DEQ (P < 0.05) and VAS (P < 0.01) (Table 4).

Table 4.  Previous mode of delivery in parous women and its effect on fear of childbirth
Previous caesarean sectionW-DEQVAS
n (%)Mean W-DEQ score ± SD (range)P95% CI for the difference between means n (%)Median VAS ± SD (range)P
  1. Total number is not same in every part because of the incomplete answers of some participants.

Previous caesarean section
None545 (78.0)63.3 ± 20.8 (14–136)<0.001−13.8 to −6.1587 (77.3)2.9 ± 2.5 (0–10)<0.001
One154 (22.0)73.2 ± 23.5 (9–150) 172 (22.7)5.1 ± 2.6 (0–10) 
Previous VE
None620 (88.7)64.8 ± 22.0 (9–150)<0.05−10.9 to −0.7673 (88.8)3.0 ± 2.5 (0–10)<0.001
One79 (11.3)70.6 ± 19.7 (34–134) 85 (11.2)5.0 ± 2.4 (0–9.8) 

The W-DEQ scores or VAS in those who had gone through miscarriage or TOP did not differ from those who had not.

Discussion

The main goal of our study was to study how severe fear of childbirth is distributed in a representative pregnant population categorised according to parity and gestational age. We found that nulliparous women had higher scores for fear of childbirth than parous women and that scores are higher in later pregnancy. Previous delivery experiences had a great impact on the amount of fear. We believe our results give a reasonable picture of primarily maternity patients in Finland. As the methods we used to study fear of childbirth (W-DEQ and VAS) measure emotions and feelings related to childbirth, rather than phenomena dependent on the culture in which the birth will occur (such as the use of pain relief, having a doula or partner attending the labour and delivery, or the possibility to perform caesarean section or VE), we believe that the results could be generalised to other pregnant women in the developed world, as well.

The W-DEQ questionnaire was developed over 10 years ago to measure fear of childbirth. The W-DEQ has been studied and used before and found to be a valid and reliable measurement of fear of childbirth.16,18–20 Other questionnaires that focus more on the actual labour and delivery process3,8,21 do not take into account the thoughts, beliefs, and emotions of pregnant women. However, personality, self-esteem, and mood have great impact on fear of childbirth among women.8,22,23 Thus, it can be assumed that the W-DEQ would better screen severe fear than the questionnaires previously used. The W-DEQ questionnaire was again shown to have good reliability in our study.

In addition to the W-DEQ, we added the VAS to our questionnaire to obtain more information about the fear of childbirth. The bimodal distribution of VAS scores was interesting, and not described before, although nonparametric tests for statistical analyses are recommended.17 The simplicity of VAS promotes high compliance and this was shown also in our study where VAS was answered by 98.3%, while W-DEQ was sufficiently answered by 91.1% of women. Combining the VAS and the W-DEQ gives us more information about the fear of childbirth. The VAS could also easily be used in follow-up testing of the fear during pregnancy.

We were also interested in seeing how the W-DEQ questionnaire would fit in our maternal population and how well it would identify the mothers with the fear of childbirth as a clinical problem. The W-DEQ score was 100 or above in 7.0% of nulliparous and in 7.7% of parous women. This is in line with previous studies, where 6–10% of pregnant women have been found to suffer from severe fear of childbirth.1–4 According to these numbers, fear of childbirth is one of the most common problems experienced during pregnancy, and it would be interesting to know how much effort is spent on treating it compared with other pregnancy-related problems.

Previous studies have shown a strong association between previous delivery experiences and fear of childbirth in subsequent pregnancies.2,12,13 Our results support these findings; fear of childbirth was significantly stronger in subsequent pregnancies if mothers had previously gone through caesarean section or VE when compared with those women who had previously not given birth through caesarean section or VE.

Fear of childbirth has previously been studied mostly in late pregnancy.2,3,19 Our study found that fear was milder in early compared with late pregnancy. The same is known about the anxiety during pregnancy; it is at its lowest between 22 and 26 weeks of gestation and increases thereafter.6 Accordingly, this would be ideal time to start treatment for fear of childbirth and avoid an undesirable unsuccessful outcome from the treatment.9

Preferred mode of delivery was significantly associated with fear of childbirth. Women who preferred caesarean section had stronger fear measured by both the W-DEQ and VAS. Our results support the clinical finding that fear of childbirth is often behind the request for caesarean section. Many studies have shown that fear of childbirth is treatable and approximately two-thirds of fear-related requests for caesarean section may be withdrawn after an opportunity has been offered to talk about fears and to obtain support, information, psycho-education, or relaxation exercises relating to fear of childbirth.9,24–27

There are many reasons for the worldwide rise in caesarean section rates, but a significant, and potentially modifiable, reason is maternal request. Concerns about rising caesarean section rates include the increased morbidity, not only of the operation itself, but consequences in subsequent pregnancies.28 The declaration of the International Federation of Gynecology and Obstetrics from January 2007 states that, ‘CS should be undertaken only when indicated to enhance the wellbeing of mothers and babies and improve outcomes (http://www.figo.org/Caesarean.asp)’. In Italy, the rates of caesarean section on maternal request have risen from 7%29 to 17%30 in 2 years. Previous studies have shown that, together with fear of childbirth, older age (≥35 years), previous experiences (birth, TOP, and miscarriages), smoking habits, higher educational level, and lack of psychosocial support are associated with maternal request for caesarean section.10,21,24,29,30 In the Nordic countries, the total rates of caesarean sections have risen up to 15%,24 but remain far below the rates in Italy30 (33%). Our study also showed, in line with previous research,24 that in the Nordic countries fewer women request caesarean section (8.5–9.8%). We speculate that this is because of both the permissive atmosphere on labour wards where parturients are encouraged to express their feelings and wishes openly and thence influence their care, and the history of 10–20 years of studying and treating fear of childbirth as well as accepting the fear of childbirth as a reason for specialist consultation.

Our study has certain limitations, which must be taken into consideration before generalising the results. First, we did not ask the women about indications for caesarean section in previous deliveries or the health of their previous newborns. These factors clearly can influence the amount of fear in subsequent pregnancies, for example, emergency caesarean section has been shown to cause PTSD and fear of childbirth.2,12,14 Furthermore, the psychological stress after elective caesarean section is highly comparable with that after vaginal delivery.31 If the infant was compromised during an earlier delivery (vaginal or caesarean section), it is very natural that mothers are afraid of the next delivery as they worry about what might happen to their baby. Secondly, the women filled in the questionnaire only once. Thus, we do not know how fear varies during a specific pregnancy, for example, if in some women it decreases and in others increases with advancing pregnancy. It is known, however, that anxiety during pregnancy is at its lowest between 22 and 26 weeks of gestation and increases thereafter.6 Thirdly, we did not ask the women about the reason why they had an appointment at the outpatient maternity clinic in our maternity hospitals. As far as it is possible to assume from their gestational ages, most of the women were attending for routine screening ultrasounds performed at 12 and 19 weeks of pregnancy. However, those women who filled in the questionnaire later may have had specific questions or worries about their pregnancy as they had been referred for consultation. This worry might have resulted in their scoring higher in both the W-DEQ and VAS.

Our results could be of value when estimating the best way and time to screen for fear of childbirth. W-DEQ scores of 100 and above detected about 7.5% of pregnant women with severe fear of childbirth as a clinical problem. The W-DEQ was a specific and useful way to screen. The VAS was also shown to be a possible screening method. It is not as accurate as the W-DEQ, but it is much easier to use and evaluate. The importance of screening for fear of childbirth early enough lies in being able to relieve the anxiety as early as possible, potentially avoid a caesarean section, and help with preparation for normal childbirth. In clinical practice, women with fear of childbirth and a request for caesarean section are sent often to maternity clinics too late for adequate treatment of their fears. If screening for fear could be performed earlier, we might have a better opportunity to treat women with fear. We suggest that one possibility for screening for fear of childbirth is to ask every pregnant woman to evaluate their fear by VAS in mid-pregnancy, and if the count is 5.0 or below, it is very unlikely the patient has significant fear. Using the ceiling of 5.0 in the VAS gives a sensitivity of 97.8% and specificity 65.7% to identify women with W-DEQ ≥100. For patients whose count is over 5.0, the W-DEQ could be used to estimate which of them need special attention. These patients could then be offered treatment for their fear.

Conclusions

Severe fear of childbirth was shown to be more common in nulliparous women, in later pregnancy, and in women with previous caesarean section or VE. Caesarean section as a preferred mode of childbirth was not very common but was strongly associated with a high fear score both in the W-DEQ and VAS.

The W-DEQ and VAS were both appropriate methods to measure fear of childbirth. As a screening method, VAS is easy and fast and has very good sensitivity. Using a VAS cutoff value of 5.0, 38.9% of the women were classified as having fear of childbirth. This group included 97.8% of all women with severe fear of childbirth (limit W-DEQ ≥100).

Disclosure of interests

No conflicts of interests exist.

Contribution to authorship

All the authors planned the study and the questionnaire was used to collect data. The collection of data was supervised by T.S. and E.H. H.R. saved the data, analysed it, and wrote the first version of the manuscript. K.S-A. and T.S. helped her with the analysis. All the authors have equally contributed to the further revision of the manuscript, and they all approve the final version.

Details of ethics approval

This study has an approval from the Ethics Committee for Gynaecology and Obstetrics, Otology, Ophthalmology, Neurology and Neurosurgery of the Helsinki University Central Hospital (376/E9/05 from 27 October 2005).

Funding

Funding was arranged by grants from Emil Aaltonen Foundation and Helsinki University Central Hospital.

Acknowledgements

Emil Aaltonen Foundation and Helsinki University Central Hospital supported this study.

Ancillary