Translation and validation of the Childbirth Experience Questionnaire (CEQ) into Sinhala Language in Sri Lanka

The Childbirth Experience Questionnaire (CEQ) was initially produced to assess women’s perceptions regarding their first birth experience. A robust validated instrument to evaluate first birth experience is currently lacking in Sri Lankan obstetric practice. Main objectives of our study were to adapt the CEQ among Sri Lankan Sinhala speaking women and to determine the psychometric properties of CEQ. This would yield an opportunity to evaluate childbirth experience among our women. Method Validation study was carried out by translating the English CEQ into Sinhala language following standard methodology comprising of forward and back translation. Total of 390 postpartum women were recruited through convenience sampling method at University Unit, North Colombo Teaching Hospital, Ragama, Sri Lanka, from February 2017 to January 2018. Initially CEQ was given to be completed four weeks after delivery and again two weeks postnatal. Mothers were asked to post the completed CEQ. Face validity was tested among twenty-five postnatal mothers and with seven labour ward staff members. Psychometric properties; internal consistency and validity of the CEQ were evaluated. Known-groups validation was used for construct validity. “participation”, respectively). Women with spontaneous onset of labour (except for “professional support” in women with spontaneous onset of labour) and women with a normal delivery showed significantly higher CEQ scores. However, oxytocin augmentation could not yield a difference in CEQ scores. A weighted kappa of 0.61-0.80 for all 22 items in CEQ demonstrated a good test-retest reliability. This study shows that Sinhala version of CEQ can be used as a reliable and valid tool of childbirth experience among Sri Lankan Sinhala speaking women.

Since there is no properly validated instrument in Sinhala language to assess childbirth experience in Sri Lankan Sinhala speaking women, Sinhala CEQ may possibly be a worthwhile effort. Latest NICE guideline on intrapartum care has also highlighted that any research on the effectiveness of interventions in latent stage of labour needs measuring woman's satisfaction with childbirth experience 5 . In a similar manner for any study investigating woman's satisfaction with her birth experience in Sri Lanka needs to have a standardized method like CEQ. Validation of this tool is important in future research, audit projects as well as in service evaluation. Different tools to assess maternal childbirth experience are essential to identify those who need psychological and moral support.
Therefore, this tool can be used to assess different dimensions of childbirth experience among primiparous women 1 . The birth of a child is a significant event in a woman's life and family. Labour room staff should always try to make her labour experience as comfortable and dignified as possible.
Although women seem to be satisfied with a healthy baby, the childbirth also has a psychological dimension 6,7 . Address of these hidden aspects is important for a positive childbirth experience and mother's long-term well-being 6,7 .
Midwives are doing a wonderful job with labouring women and helping them to achieve a satisfactory birth e x p e r i e n c e 7 . Higher workload in the hospitals resulted a sub-standard level of intrapartum care during childbirth 7 . Studies have reported that mothers often felt alone and unsupported during childbirth process and some of them would not get pregnant again due to the negative experience they had during last childbirth 7 . In the latest Cochrane review on continuous support for women during childbirth has reported that continuous labour support with the help from a labour companion (doula) who is a member of the labouring woman's social network, has been proved to effective in improving laboring women's satisfaction and hence their childbirth experience 8 . Thus it elaborates the significance of woman's psychological well-being and achieving a satisfactory childbirth experience in addition to birth of her baby.
The quality of intrapartum care that a particular woman gets has an impact on her physical and emotional well-being 5 . It has both short-term and longer-term effects including care towards the newborn 5 . Good communication with woman, support from labour ward staff and having woman's wishes respected, can enhance her confidence contributing to make a positive childbirth experience 5 . Negative experiences coming from their first childbirth can increase the risk of postpartum depression. It may give rise to a validity and test-retest reliability were not measured 25 . In India there is a validated Hindi-translated scale for measuring maternal satisfaction is available 12 . In Sri Lankan context, currently we do not have a validated tool to assess birth satisfaction of our women.
General Objective was to translate and validate the CEQ from English language to the cultural context and lifestyle related to the Sinhala language. Sinhala language is one of the mainly used languages in Sri Lanka. Specific objectives were to conduct a linguistic translation, to conduct a validation study and to assess the psychometric properties of the

Study instrument
Sinhala version of the CEQ was used as the study instrument. The CEQ consists of 22 items referring to the childbirth experience (Table 2). Women's responses to 19 items in CEQ were rated on a 4-point Likert scale and remaining 3 items were evaluated by means of a visual analogue scale (VAS). These VAS scores are then converted into categorical variables facilitating interpretation and analysis; 0-40 = 1, 41-60 = 2, 61-80 = 3 and 81-100 = 4. As mentioned in Table 2, negatively worded items were reversed. All 22 items in CEQ are divided into 4 groups as 4 main domains as mentioned above; "own capacity" (8 items about sense of control, their feelings during birth process and labour pain; 1,2,4,5,6,19,20,21 items), "professional support" (5 items regarding information and the midwifery care they received; 13,14,15,16,17 items), "perceived safety" (6 items about sense of security and memories from their labour and childbirth; 3,7,8,9,18,22 items), and "participation" (3 items about own possibilities to influence maternal position, their movements and pain relief in labour; 10,11,12 items).

Data collection procedure
Eligible women were identified in the postnatal ward by the investigator and were invited to participate. They were given the information sheets and consent forms. Those who agreed to join recruited by signing an informed written consent form. Women were received a clear description of the procedure by the investigator. Initially basic demographic and clinical details were collected into a data collection sheet.
Two copies of CEQ with two stamped-envelopes were given to each mother and were asked to fill each in one moth postnatal and six weeks postnatal. Soon after childbirth, mothers were supposed to have gone through their first labour experience and having relatively fresh memories of their childbirth experience. As well as, mothers have minimum contact with their childbirth caregivers at hospital after four weeks. Therefore chances of giving socially desirable responses were minimal and minimizing potential bias.
Recruited women were offered the option of completing the CEQ at home and sending it via post and also to ask for clarifications by making a phone call to the investigator.
Reminders were sent after one moth and six weeks as a phone call or a post card.
Demographic details and their delivery details were collected from the clinical notes.

Data analysis
Basic demographic characteristics of the data was analysed to see any deviations.
Continuous variables were presented as means and standard deviations. Pearson Chi-Square test was applied to any difference between categorical variables. Mann-Whitney U test was applied to compare any skewed continuous variables. P value < 0.05 was considered as statistically significant.

Face validity
Face validity was done for the Sinhala CEQ version to see the items were well understood and easy to answer to. This was done by giving questionnaire to 25 primiparous mothers to find any difficult questions and asking their comments on questionnaire items. The CEQ was also given to a group of labour ward staff members at NCTH (n=10) consisting of three experienced nurse-midwives, three midwives and one medical officer who was also a female in reproductive age group.

Internal consistency
This was evaluated using Cronbach's coefficient alpha. Usually, Cronbach's alpha ≥ 0.7 is considered as a satisfactory level of internal consistency.

Construct validity
Construct validity was evaluated as follows. The first was whether the questionnaire was able to differentiate between complicated (operative) and uncomplicated (normal delivery) deliveries, and secondly to test questionnaire against the delivery outcome. The method of known-groups validation was used to measure construct validity. This method can be used to evaluate the ability of the CEQ to distinguish between above known groups within the sample and which were known to differ on known key socio-demographic or clinical v a r i a b l e s . The CEQ has shown good construct validity and reliability for assessing childbirth experience between its known groups 1 .
A comparison was done between CEQ subscale scores and overall CEQ score for different key characteristics.
For example,
Women with duration of labour; more than 12 hours versus less than 12 hours 3. Women with augmentation of labour; with oxytocin versus without oxytocin 4. Women who had a; normal delivery versus operative delivery.
Here the overall CEQ score means average of the 4 individual CEQ subscale scores. It is well-documented that oxytocin augmentation of labour with oxytocin, operative delivery (instrumental or caesarean delivery) and long duration of labour (> 12 hours) known to cause a negative influence on woman's childbirth experience 16,17,18 .
Since the subscale scores were not normally distributed, non-parametric Mann-Whitney U test was applied to compare subscale scores between subgroups.

Test-retest reliability
To assess stability (test-retest reliability) a two-week test-retest analysis of the sample was done. This agreement was further assessed by using the weighted Kappa (k) statistic 19 . To represent a substantial agreement between the two scores between one month and six weeks questionnaires a value of weighted Kappa between 0.61 -0.80 was taken 20 .

Ethical considerations
Eligible patients were identified and counseled by investigator at the study setting. Before entry into the study the investigator explained each patient the objectives, methods as well as potential benefits and risks of participating in the study. Patient information sheets were distributed to eligible patients. After sufficient information, informed written consent was obtained.
We have also taken written permission from the original authors of CEQ Advanced Level passed and above) are have a statistically significant better response rate compared to women educated up to G.C.E. Ordinary Level (p value < 0.05). Unemployed women were also showed a statistically significant better response rate compared to unemployed women (p value < 0.05). However, level of income, mode of delivery and mode of onset of labour did not show any significant impact on the response rate. Table 2 shows the CEQ item description with mean score for each 22 items and also test-retest reliability (weighted Kappa scores) between first and second CEQ.  mentioned that the questionnaire was easy to understand and complete, in addition to suggesting some modifications. No respondents mentioned that any of the CEQ items should be removed and all agreed that none of the questions were upsetting or offensive concluding a good face validity.

Internal consistency
Internal consistency of our CEQ was 0.65, 0.80, 0.70 and 0.83 for respective 4 domains such as "own capacity", "professional support", "perceived safety" and "participation".
The overall Cronbach's alpha was 0.85 as in Table 4. Cronbach's alpha for all of the four subscales in the initial Swedish study validating the original CEQ in a Swedish population has mentioned for comparison in same Table 3.

Construct validity
Known-group validation was applied to assess construct validity and details have been summarized in Table 4. Only two women had labour duration more than 12 hours and therefore labour duration was not taken into account in known-group validation. Women with spontaneous onset of labour and normal delivery groups obtained significantly higher CEQ scores, when compared to women whose labour had been induced and who had an instrumental delivery, respectively. Women who had a spontaneous onset of labour obtained statistically significantly higher scores in each of the following subscales: "own capacity", "perceived safety", "participation" and overall CEQ score, except for "professional support". Women who had a normal delivery obtained statistically significant higher scores than mothers with an instrumental delivery, for all four subscales including "own capacity", "professional support" "perceived safety" and "participation", as well as for the overall CEQ score.
However, no statistically significant differences were seen for any of CEQ subscales and overall score with regards to labour augmentation with Oxytocin as shown in Table 4.

Discussion
This study describes a successful transcultural adaptation of the CEQ into Sinhala language in Sri Lankan context. According to our study, the Sinhala version of the CEQ has shown most of the psychometric properties to the original Swedish as well as adapted English and Spanish versions 3 . It is therefore suitable to assess childbirth experience in Sinhala speaking Sri Lankan postnatal mothers after their first childbirth.
To Research on birth experience is limited. This assessment tool of childbirth experience (CEQ) may help to provide a positive impact in Sri Lanka. With postal method of sending the questionnaires, we could achieve a 226 women giving rise to a response rate of 57.9%. However, this is well below what was seen at Swedish study (70%) and Spanish validation (62%), but comparable to UK study (59%) 1,2 . In Malaysian study the response rate has not mentioned 20 .
Our CEQ demonstrated a high face validity. We found a good internal consistency with Cronbach's alpha ≥ 7 for three out of four domains and overall CEQ score as shown in Table 3. Dencker et al., in their original Swedish study, found significant differences when comparing women with labour lasting less than 12 hours, women with no labour augmentation using oxytocin and normal delivery, versus women with more than 12 hours of labour, augmentation with oxytocin and instrumental delivery, respectively, for all four domains 1 . However in our study, this known-group validation showed validity of the CEQ in normal delivery versus instrumental delivery and spontaneous onset of labour versus induced labour groups. These were almost in the similar way as for the initial Swedish s t u d y 1 and similar to the UK as well as Spanish adaptations 2,3 . However, we could get only two women with more than 12 hours of labour duration. Therefore, this was not included in analysis although others have done this 2,3 . Our study also could not demonstrate any significant difference in CEQ scores between augmentation of labour with oxytocin versus no oxytocin augmentation. It worthwhile to mention that oxytocin infusions are commonly used in Sri Lankan labour wards for the majority of women in labour. No difference between oxytocin augmentation and no oxytocin possibly may be due to this high oxytocin use. Small effect sizes were seen in overall CEQ score across all known groups (Table 4). In Malaysian adaptation of CEQ, construct validity has been evaluated using confirmatory factor analysis, not with known group validation as mentioned above 4 . We had only two third/fourth degree perineal tears and therefore it was not included in construct validity.
Requesting to send the questionnaires back through email is difficult in our setting which was done in UK study 2 , so we decided to have them posted instead which was also used in Spanish study 3 . According to the latest perinatal health reports in Europe; Sweden, UK and Spain have far superior maternal and perinatal statistics compared to Sri Lanka 28 .
These might also have influenced on the maternal birth experience and giving quite lower values in our setting comparatively. Some authors propose hypotheses that excessive interventions could compromise birth satisfaction leading to the difficulty of making a more accurate comparison between studies 29 . Although original CEQ was developed for mothers after their first childbirth, Spanish CEQ validation study found a novel thing that CEQ can also be used for multiparous women 3 . They have also showed that the Spanish adaptation could discriminate birth experience between primiparous and multiparous women as a test for construct validity 3 . We found the lowest mean CEQ score as 1.6 for the item 20 ("Experienced level of labour pain"). It also had a good stability (kappa 0.66).
This signals that more emphasis has to be paid on labour analgesia.
CEQ comprises of 22 items and a tool measuring women's' perception and experience of birth should ideally be a simplest possible way. The PPS which developed in UK in 2009 is a 3-item simple tool 10 . Being a simple short tool would largely have an impact on response rate, feasibility and generalizability of the measuring tool. However this PPS has been developed and validated only for the operative deliveries. Even though a relatively long questionnaire, our CEQ address the entire multidimensional picture of labour and childbirth. It is obvious that it might take about ten to fifteen minutes to complete the CEQ, but it still has a value in measuring maternal satisfaction not only in any type of delivery but also in a multidimensional picture regarding the process of labour and childbirth. The PCACL-R is also a valid tool to measure women's perception regarding labour and childbirth and it comprises of 16 items. However, as mentioned earlier this P C A C L -R content validity, criterion validity and test-retest reliability has not been computed 11 . Maternal Satisfaction for Caesarean Section (MSCS) was designed to assess satisfaction of women undergoing caesarean deliveries under regional anaesthesia and its

Limitations
This study is not without limitations. As previously mentioned, we could not get an adequate sample for women with more than 12 hours of labour duration and liberal use of oxytocin augmentation was an issue. Another limiting factor is getting accreted translated for this type of health-related studies. A recall of four to five months had to be done for some women even after monthly reminders were done as phone calls leading to the 57.9% response rate. We also had to exclude 12 questionnaire from analysis due to incomplete data. Furthermore, we could not compute criterion validity for our validation study due to unavailability of a 'gold standard' assessment of childbirth experience in Sri Lanka to be compared with. Time lapses between childbirth and questionnaire (from four weeks to four-five months in some cases) could possibly give negative influence on birth experience of women as argued by some authors 33,34,35 .

Conclusions And Recommendations
As a conclusion, this study indicates that use of the Sinhala translation of the CEQ as a reliable and valid tool with well-preserved psychometric properties to assess childbirth experience among Sinhala speaking women in Sri Lanka. And also it could be used as a screening or a monitoring tool to identify those women with negative birth experience in different interventions or implementation programmes in maternal care settings. However, need of Tamil translation and validation of the CEQ is also important for betterment of assessing childbirth experience in Sri Lanka since Sinhala and Tamil languages are the two main languages in Sri Lanka.