https://orcid.org/0000-0001-5002-7496
Abstract
Purpose
Post-traumatic stress disorder (PTSD) after childbirth causes severe and lasting effects. Screening of childbirth experience may expedite early PTSD recognition. Systematic reviews have not provided consensus on how and when to measure childbirth experience and the clinical implications of such measurements. We aimed to identify a reliable and simple scale for screening the childbirth experience with minimum risk of missing PTSD.
Materials and methods
This cohort study evaluated the childbirth experience of 1527 unselected women with Wijma Delivery Experience Questionnaire (W-DEQ-B), Delivery Satisfaction Scale (DSS), and Visual Analogue Scale (VAS). VAS was measured first <1 week (VAS1) and then, together with the other scales, a few months after childbirth (VAS2). The scales’ ability to identify PTSD (measured with Traumatic Event Scale) was evaluated and compared with receiver operating characteristic (ROC) analysis. Diagnostic accuracy and clinical usefulness were used to suggest cutoff values for scales.
Results
W-DEQ-B showed highest recognition of partial or full PTSD (area under the ROC curve 0.96 in W-DEQ-B, 0.92 in VAS2, 0.91 in DSS and 0.82 in VAS1).
Conclusions
All included scales recognized partial or full PTSD well. Although W-DEQ-B performed best, VAS (measured twice) with cutoff value of 50 mm is most suitable for screening in clinical circumstances.
Introduction
Childbirth is a positive experience for most, but unfortunately not all women. Even a medically uncomplicated delivery can be experienced negatively [Citation1]. A negative childbirth experience should be prevented and recognized, as this can lead to several adverse effects. This negative experience can influence mother–infant bonding [Citation2], postpartum depression [Citation2,Citation3], fear of childbirth [Citation2], request for cesarean section in next pregnancy [Citation4], and unwillingness for more children [Citation5]. At worst, childbirth may lead to post-traumatic stress disorder (PTSD) (estimated incidence 3–4% of all childbirths) [Citation1,Citation2].
Detecting PTSD can be difficult, as a traumatized woman can live daily life avoiding everything that reminds her of childbirth [Citation6]. According to diagnostic criteria, a PTSD diagnosis can only be made at the earliest 1 month after birth [Citation6]. Thus, screening of the childbirth experience may expedite early recognition of women at risk for PTSD. Besides length of time, traumatic experience, and symptom criteria of avoidance, PTSD diagnosis also requires fulfilling other symptom criteria (persistent reexperiencing of the traumatic event and increased arousal in Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV)), and that symptoms cause significant impairment in daily life functioning [Citation6].
However, previous systematic reviews about childbirth experience scales have shown considerable variance [Citation3,Citation7]. Consequently, the prevalence of negative childbirth experience varies from 4.5% to 20% [Citation8–10]. A gold standard method to measure childbirth experience is warranted [Citation3]. Although a systematic review ranked the Wijma Delivery Experience Questionnaire (W-DEQ-B) as best [Citation7], this scale is laborious to complete and is thus not optimal for screening. Moreover, only a few studies applied a cutoff value for a negative childbirth experience in W-DEQ-B [Citation11,Citation12].
We aimed to identify a reliable and simple childbirth experience scale for screening with a justified cutoff value with minimum risk of missing PTSD. Furthermore, we studied how often both negative and traumatic childbirth experience leads to posttraumatic stress symptoms (PTSS) or PTSD.
Materials and methods
An unselected pregnant population attending routine ultrasound in maternal clinics in Helsinki and Uusimaa Hospital district over 17 months volunteered to participate in this study. Altogether, 1527 women answered the survey less than a year after childbirth. The survey collected information about the childbirth experience and mental wellbeing after childbirth. Inclusion criteria were having a singleton pregnancy and a living newborn seven days after birth. Background characteristics were collected on average at gestational age 16 + 4 weeks (SD 3.9, range 6–40) (). Altogether, 1445 (94.6%) women evaluated their childbirth experience within a week after childbirth. Obstetric data were available for 1479 (96.9%). The mailed or online survey was completed on average 158 days (median 144, SD 55.9; range 67–361) after birth. The study was approved by Helsinki University Hospital and the local ethics committee (250/13/03/03/2013). Informed consent was collected from all participants.
Table 1. Background characteristics of the study population.
The childbirth experience was measured with W-DEQ-B [Citation13], Delivery Satisfaction Scale (DSS) [Citation14], and Visual Analogue Scale (VAS) [Citation10]. Moreover, Traumatic Event Scale (TES) was used to evaluate PTSS/PTSD related to childbirth [Citation6]. In addition, the questions “Did the childbirth experience differ from your expectations?” and “Has the thought of having more children appealed to you” were asked. VAS was measured within a week (VAS1) and some months after childbirth (VAS2) together with the other questionnaires.
For VAS, women marked “X” on the 100-mm VAS line describing overall childbirth experience, ranging from 0 (terrible) to 100 (excellent). The W-DEQ-B scale contains 33 statements about childbirth rated on a Likert scale from 0 (not at all) to 5 (extremely) [Citation13]. Total score ranges from 0 to 165, with higher scores indicating a more fearful childbirth experience. Cronbach’s α coefficient was 0.95. In the eight-question DSS, each question is rated on a Likert scale from 1 (not at all) to 5 (very much) [Citation14]. Total score ranges from 8 to 40, with higher scores indicating better delivery satisfaction. Cronbach’s α coefficient was 0.77. Question 5 “Did you get efficient pain relief during labor” was used in the analysis to describe if the pain relief during labor was effective enough.
TES was developed in line with the DSM-IV criteria for PTSD [Citation6]. TES includes four statements about trauma/stressor (criterion A) and 17 statements concerning symptoms (criteria B, C and, D). The subjects report the frequency of each symptom rated on a Likert scale from 1 (never/not at all) to 4 (very much/often). A statement is considered positive if answered with 3 or 4. Trauma criterion A (traumatic childbirth experience) is fulfilled if at least one of statements 1–3 and statement 4 are positive. To meet symptom criteria, at least one positive statement for criteria B, three for criteria C, and two for criteria D are required. The symptom criteria determine whether a woman has full, none, or partial PTSS [Citation15,Citation16]. Cronbach’s α coefficient was 0.86.
Statistical analyses
Statistical analyses were performed using SPSS 27.0 (SPSS Inc., Chicago, IL). Descriptive statistics were calculated for childbirth experience scales and stratified by traumatic event and PTSS. Between-group differences were assessed via χ2 (exact) tests for categorical variables and Mann–Whitney’s U-tests for all continuous scales, as they were all abnormally distributed. Logistic regression was performed to predict different trauma subgroups. All possible data were included in the analyses. If there were missing values, those examinees were excluded from that index analysis. Receiver operating characteristic (ROC) [Citation17] analysis was performed to evaluate and compare VAS1, VAS2, W-DEQ-B, and DSS. The best cutoff points for each scale were examined by the ability to recognize partial PTSD. This was first done by reporting the best balance in the sensitivity and specificity with utilization of Youden’s index [Citation17]. The discriminative and predictive potential of the most promising and clinically sensible cutoff values were then quantified by measures of diagnostic accuracy [Citation18].
Results
Of all parturients, 58.9% were nulliparous (). We identified the following four traumatic experience subgroups: traumatic experience (criteria A in TES fulfilled without considering symptoms, 9.7% of total sample); traumatic experience without symptoms (criterion A without any PTSS criteria fulfilled, 36.2% of those who had a traumatic experience); partial PTSD (criterion A with some or all PTSS criteria fulfilled, 63.8%); and full PTSD (all criteria A–D fulfilled, 11.3%). If trauma criterion A was not met, only 17.7% suffered from partial and 0.04% from full PTSS. Altogether, 90 (6.2%) women appeared to suffer from partial and 16 (1.1%) from full PTSD. The most common PTSSs were intrusive thoughts (such as recurrent distressing recollections and nightmares); 53.2% of those with traumatic experience experienced these symptoms. The proportions of traumatic experience, partial, and full PTSD were similar between primipara (in 10.3%, 7.1%, and 1.2%, respectively) and multipara (in 8.8%, 5.5%, and 1.3%, respectively).
In different subgroups of traumatic experience, VAS1, VAS2, W-DEQ-B, and DSS all indicated significantly more negative childbirth experience than if the trauma criterion A was not fulfilled (). Although primipara evaluated their experience more negatively than multipara with all scales in total, in partial or full PTSD their experiences did not differ significantly. The more negative the childbirth experience, the greater the risk was for any kind of trauma (Supporting Information Table S1, which shows evaluation of risks for different trauma subgroups when compared with those without traumatic experience) (p < .001 in all ORs).
Table 2. Childbirth experience in different trauma subgroups based on Traumatic Event Scale (TES) [Citation6].
Altogether, 1132 (74.4%) women experienced their childbirth more positively or as expected, 938 (62.1%) women found the thought of having more children appealing, and 1171 (78.5%) considered their pain relief as effective enough after this index childbirth. In general, in different trauma subgroups, the childbirth experience differed from expectations rather negatively than positively, the thought of having more children was less appealing, and pain relief was considered rather ineffective than effective enough (). The specificity of a more negative childbirth experience than expected, finding the thought of having more children unappealing, and ineffective pain relief for all trauma subgroups were 79.6%, 64.0% and 82.1%, respectively. The answer of the experience being more negative than expected had highest sensitivity. Sensitivity for recognition of partial PTSD was 77.8%. The corresponding sensitivity values for thought of having more children unappealing and in ineffective pain relief were 63.3% and 53.3%, respectively.
Table 3. Answers to simple questions in different trauma subgroups based on Traumatic Event Scale (TES) [Citation6].
To identify traumatic childbirth experience, partial PTSD, and PTSD, the area under ROC curve (AUC) was highest in W-DEQ-B and lowest in VAS1 ( and Supporting information Table S2, which shows pairwise comparisons of AUCs of different childbirth experience scales). W-DEQ-B discriminated best both traumatic experience and partial PTSD. W-DEQ-B and DSS were similar in discrimination of full PTSD profile. The discrimination potential of VAS2 and DSS was similar. The discrimination power of VAS1 was the lowest.
Figure 1. Receiver operating characteristic (ROC) curve for identifying partial PTSD in Traumatic Event Scale (TES) [Citation6]. VAS1: Visual Analogue Scale measured within a week after childbirth; VAS2: Visual Analogue Scale measured a few months after childbirth; W-DEQ-B: Wijma Delivery Experience Questionnaire; DSS: Delivery Satisfaction Scale; AUC: area under ROC curve; J: Youden Index.
![Figure 1. Receiver operating characteristic (ROC) curve for identifying partial PTSD in Traumatic Event Scale (TES) [Citation6]. VAS1: Visual Analogue Scale measured within a week after childbirth; VAS2: Visual Analogue Scale measured a few months after childbirth; W-DEQ-B: Wijma Delivery Experience Questionnaire; DSS: Delivery Satisfaction Scale; AUC: area under ROC curve; J: Youden Index.](/cms/asset/d4f77650-83e0-45d9-859b-fdd9584449cd/ipob_a_2210750_f0001_c.jpg)
Different scales possessed Youden’s indexes for identifying partial PTSD (). shows the different scales’ discriminative and predictive potential for partial PTSD with different cutoff values. The prevalence of negative childbirth experience depended on the chosen scale and a cutoff value. The positive predictive value (representing how often a negative experience led to partial or full PTSD) increased along with more negative childbirth experience evaluations. Both positive and negative likelihood ratio and the proportion of correctly classified subjects among all subjects (overall diagnostic accuracy) increased with more negative childbirth experience evaluations.
Table 4. Discriminative and predictive potential for partial post-traumatic stress disorder (PTSD) according to Traumatic Event Scale (TES) [Citation6] with the suggested cutoff values for Visual Analogue Scale (VAS), Wijma Delivery Experience Questionnaire (W-DEQ-B) [Citation13], and Delivery Satisfaction Scale (DSS) [Citation14].
Discussion
This large study on an unselected pregnant population revealed that a traumatic childbirth experience very often led to PTSS (63.8%). The prevalence of the negative childbirth experience depended on a chosen childbirth experience scale and the cutoff value. All studied scales recognized partial or full PTSD well. W-DEQ-B had the highest recognition of traumatic childbirth. However, this scale is laborious to complete in clinical circumstances, such as screening. VAS2 and DSS shared the second ranking. VAS is simple, short, and had higher compliance (98.9%) than DSS (95.5%) and especially higher compliance than W-DEQ-B (91.9%). Therefore, this study suggests that VAS administered some months after childbirth would be the best measure in clinical circumstances to assess childbirth experience with the goal of screening for traumatic childbirth experience.
VAS1, W-DEQ-B, and DSS all showed more negative childbirth experience along with tighter TES criteria. Women with partial or full PTSD had similar VAS2 but lower VAS2 than those without traumatic experience. This means that VAS2 identifies women with traumatic experience with PTSS but cannot be used in PTSS grading. Furthermore, VAS2 seems suitable for screening traumatic childbirth experience with PTSS but not for PTSD diagnosis. TES is recommended for further evaluation of PTSD, and women with partial or full PTSD should be diagnosed and examined for the need for psychotherapeutic care.
Choosing a cutoff value for clinical use is always a compromise. Previous studies have examined negative childbirth experience using VAS with cutoff values of 3–5, scale 0/1–10 [Citation8–10] and W-DEQ-B with cutoff value ≥66 [Citation11]. However, no adequate testing of cutoff values have been previously reported. Previous cutoff values were based on each country’s or hospital’s own clinical recommendations [Citation8,Citation9] or on quartile scores [Citation11,Citation12]. Thus, the present study is the first to statistically examine the cutoff values of VAS, W-DEQ-B, and DSS.
Our results showed that Youden’s indexes in all scales settled down to positive childbirth experience evaluations. Choosing Youden’s indexes as cutoff values would mean poorer specificity, positive predictive value, positive likelihood ratio, and overall diagnostic accuracy than more negative childbirth experience evaluations. Likelihood ratios are recommended as optimal in diagnostic accuracy, as they are independent of disease prevalence, tell about the probability of a disease, and can be transferred beyond individual studies [Citation18]. The results showed that negative predictive values were at least 95% and changes in negative likelihood ratios were less than 1%. Based on this, for clinical usefulness without compromising on sensitivity, we suggest setting the cutoff values to more negative evaluations, such as VAS to 50 mm, W-DEQ-B to 80 and DSS to 25. With these suggested cutoff values, the prevalence of negative childbirth experience varied from 5.7% to 14.6% (the latter in VAS2) and is consistent with previous studies [Citation10]. This led to partial or full PTSD in 33.3%. With same cutoff value in VAS2, the prevalence of non-negative childbirth experience was 85.4% and only 1.5% developed partial or full PTSD (calculated 1 – negative predictive value).
Due to the clinical difficulty in detecting PTSD and considering the potentially harmful consequences of a PTSD and negative childbirth experience [Citation1–5,Citation19], early screening of the childbirth experience is arguably warranted even with sub-optimal quality. As an early screening test, it is not necessary for the AUC values of VAS1 to match those of diagnostic tests. Regardless, VAS1 possessed at least moderate accuracy in recognition of traumatic childbirth experience with and without PTSS [Citation20]. Thus, in clinical circumstances, we suggest the childbirth experience screening protocol as shown in . A midwife- or clinician-led brief and early counseling intervention after traumatic childbirth reduces PTSS [Citation21].
Figure 2. Childbirth experience screening protocol with Visual Analogue Scale (VAS) to identify women at risk for post-traumatic stress disorder (PTSD) based on Traumatic Event Scale (TES) [Citation6].
![Figure 2. Childbirth experience screening protocol with Visual Analogue Scale (VAS) to identify women at risk for post-traumatic stress disorder (PTSD) based on Traumatic Event Scale (TES) [Citation6].](/cms/asset/720797e5-d57a-46e1-931b-6c07d426adc3/ipob_a_2210750_f0002_b.jpg)
The answers to questions about the difference of the childbirth experience from expectations, the willingness for more children in the future, and the effectiveness of pain relief during labor corresponded with traumatic experience measured with all childbirth experience scales. In particular, further PTSD screening is warranted if the experience is more negative than the expectations. Additionally, unwillingness for more children should be considered as a possible sign of traumatic experience. Finally, if pain relief is reported as effective enough, a traumatic experience is unlikely.
This large questionnaire study executed the further research on already existing childbirth experience scales asked in previous studies [Citation7] by evaluating and comparing W-DEQ-B and two other scales in a large unselected cohort. W-DEQ-B and VAS have been previously validated and performed well in measuring childbirth experience [Citation7,Citation10,Citation13]. DSS has also been used for this purpose previously [Citation14]. Additionally, we even studied the suitability of three simple questions. This is the first study to assess a cutoff value for DSS, and the cutoff values we suggested for W-DEQ-B and VAS are more justifiable than in previous studies. To our knowledge, this is the first study to report how often negative childbirth leads to PTSD.
Nevertheless, this study has limitations. First, the negative childbirth experience also has consequences other than PTSD that could be screened for. However, as PTSD is the most severe potential consequence after a negative childbirth experience, we chose this as the most important outcome that childbirth experience screening should identify. Second, previous studies have recognized numerous factors that affect the subjective childbirth experience [Citation2,Citation4,Citation8–12,Citation14,Citation22] and PTSD development [Citation1,Citation2,Citation15,Citation16,Citation23], such as nulliparity or previous negative childbirth experience. However, our aim was to identify a screening method for an unselected population that can be used in clinical settings. Because of this and the observation that childbirth experiences in primipara and multipara were similar in different trauma subgroups, ROC analyses were performed on the entire study population. Third, the prevalence of PTSD after childbirth depends on the study population. In one systematic review from the United States, the prevalence of 3% in a community sample is consistent with our results, although up to 15.7% was observed in at-risk samples [Citation1]. Another systematic review revealed a PTSD prevalence of 26.2% in Iranian women [Citation24]. An earlier study also recommended cross-cultural comparisons between developing and developed countries [Citation19]. Therefore, our results may apply at least in a community sample in Western populations. Fourth, we used DSM-IV criteria for PTSD instead of DSM-V, as TES is validated for DSM-IV. Finally, measuring TES together with the childbirth experience scales may have led to bias.
Conclusions
Early screening of childbirth experience is important for identifying women with traumatic experience and to prevent later problems in mother–infant bonding and the mother’s mental wellbeing. VAS was the most suitable in clinical circumstances to measure childbirth experience. Although VAS worked better some months after than only a few days after childbirth, it worked at least sufficiently well for the purpose of primary screening within a week after childbirth. This would allow targeted, brief and early interventions and hopefully prevent later PTSS or PTSD. Early screening alone may miss PTSS or PTSD. Accordingly, this study recommends screening the childbirth experience with VAS twice, within a week and a few months after delivery. A cutoff value of 50 mm is recommended for PTSD screening. In addition to VAS, it may be useful and easy to also ask if the childbirth experience differed from expectations, if pain relief was adequate, and if the thought of having more children is appealing.
Supplemental Material
Download MS Word (14.1 KB)Supplemental Material
Download MS Word (18.4 KB)Acknowledgements
The authors are grateful for comments from Jenny Miettinen, Lauri Hietajärvi, and Katja Upadyaya.
Disclosure statement
The authors report there are no competing interests to declare.
Data availability statement
The datasets generated and/or analyzed during the current study are not publicly available due to limitations of ethical approval involving the patient data and anonymity but are available from the corresponding author on reasonable request.
Additional information
Funding
References
- Grekin R, O'Hara MW. Prevalence and risk factors of postpartum posttraumatic stress disorder: a meta-analysis. Clin Psychol Rev. 2014;34(5):389–401.
- McKelvin G, Thomson G, Downe S. The childbirth experience: a systematic review of predictors and outcomes. Women Birth. 2021;34(5):407–416.
- Bell AF, Andersson E. The birth experience and women’s postnatal depression: a systematic review. Midwifery. 2016;39:112–123.
- Jenabi E, Khazaei S, Bashirian S, et al. Reasons for elective cesarean section on maternal request: a systematic review. J Matern Fetal Neonatal Med. 2020;33(22):3867–3872.
- Shorey S, Yang YY, Ang E. The impact of negative childbirth experience on future reproductive decisions: a quantitative systematic review. J Adv Nurs. 2018;74(6):1236–1244.
- Wijma K, Soderquist J, Wijma B. Posttraumatic stress disorder after childbirth: a cross sectional study. J Anxiety Disord. 1997;11(6):587–597.
- Nilver H, Begley C, Berg M. Measuring women’s childbirth experiences: a systematic review for identification and analysis of validated instruments. BMC Pregnancy Childbirth. 2017;17(1):203.
- Adler K, Rahkonen L, Kruit H. Maternal childbirth experience in induced and spontaneous labour measured in a visual analog scale and the factors influencing it; a two-year cohort study. BMC Pregnancy Childbirth. 2020;20(1):415.
- Falk M, Nelson M, Blomberg M. The impact of obstetric interventions and complications on women’s satisfaction with childbirth a population based cohort study including 16,000 women. BMC Pregnancy Childbirth. 2019;19(1):494.
- Larsson C, Saltvedt S, Edman G, et al. Factors independently related to a negative birth experience in first-time mothers. Sex Reprod Healthc. 2011;2(2):83–89.
- Zar M, Wijma M, Wijma B. Pre- and postpartum fear of childbirth in nulliparous and parous women. Scand J Behav Ther. 2001;30(2):75–84.
- Ulfsdottir H, Nissen E, Ryding EL, et al. The association between labour variables and primiparous women’s experience of childbirth; a prospective cohort study. BMC Pregnancy Childbirth. 2014;14:208.
- Wijma K, Wijma B, Zar M. Psychometric aspects of the W-DEQ; a new questionnaire for the measurement of fear of childbirth. J Psychosom Obstet Gynaecol. 1998;19(2):84–97.
- Saisto T, Salmela-Aro K, Nurmi JE, et al. Psychosocial predictors of disappointment with delivery and puerperal depression. A longitudinal study. Acta Obstet Gynecol Scand. 2001;80(1):39–45.
- Soderquist J, Wijma B, Thorbert G, et al. Risk factors in pregnancy for post-traumatic stress and depression after childbirth. BJOG. 2009;116(5):672–680.
- Soet JE, Brack GA, DiIorio C. Prevalence and predictors of women’s experience of psychological trauma during childbirth. Birth. 2003;30(1):36–46.
- Akobeng AK. Understanding diagnostic tests 3: receiver operating characteristic curves. Acta Paediatr. 2007;96(5):644–647.
- Eusebi P. Diagnostic accuracy measures. Cerebrovasc Dis. 2013;36(4):267–272.
- Ayers S, Joseph S, McKenzie-McHarg K, et al. Post-traumatic stress disorder following childbirth: current issues and recommendations for future research. J Psychosom Obstet Gynaecol. 2008;29(4):240–250.
- Swets JA. Measuring the accuracy of diagnostic systems. Science. 1988;240(4857):1285–1293.
- Taylor Miller PG, Sinclair M, Gillen P, et al. Early psychological interventions for prevention and treatment of post-traumatic stress disorder (PTSD) and post-traumatic stress symptoms in post-partum women: a systematic review and meta-analysis. PLOS One. 2021;16(11):e0258170.
- Rouhe H, Salmela-Aro K, Toivanen R, et al. Group psychoeducation with relaxation for severe fear of childbirth improves maternal adjustment and childbirth experience – a randomised controlled trial. J Psychosom Obstet Gynaecol. 2015;36(1):1–9.
- Söderquist J, Wijma K, Wijma B. Traumatic stress after childbirth: the role of obstetric variables. J Psychosom Obstet Gynaecol. 2002;23(1):31–39.
- Abdollahpour S, Khadivzadeh T. Prevalence of traumatic childbirth and post-traumatic stress after delivery in Iran: a systematic review and meta-analysis. J Obstet Gynecol Cancer Res. 2019;4(3):86–92.