An investigation of women's involvement in the decision to deliver by caesarean section

Authors


Correspondence: Dr W. J. Graham, Dugald Baird Centre, Aberdeen Maternity Hospital, Cornhill Road, Aberdeen AB25 2ZL, UK.

Abstract

Objective To assess the degree and nature of women's involvement in the decision to deliver by caesarean section, and women's satisfaction with this involvement.

Design Observational study.

Setting The maternity unit in a large teaching hospital.

Sample One hundred and sixty-six women undergoing caesarean section.

Methods Interviews with the women on the third or fourth day postpartum, questionnaires sent to the women at 6 weeks and at 12 weeks postpartum, and extraction of information from the women's medical records.

Main outcome measures Women's knowledge, satisfaction, and involvement in making the decision concerning their caesarean section.

Results The majority of the women were satisfied with the information they received during pregnancy on caesarean section and with their involvement in making the decision, but the proportions were significantly higher for elective than emergency sections. For 7% of the women, maternal preference for caesarean section was a direct factor in making the decision. Just over half of the 166 women reported that they were not debriefed on the reasons for their caesarean section before their discharge from hospital. Almost a third of the women undergoing emergency caesarean section expressed negative feelings towards their delivery, compared with 13% of those undergoing elective caesarean sections.

Conclusion Women are not a homogeneous group in terms of their requirements for information, nor their desire to be involved in the decision on mode of delivery. Health professionals need to be responsive to this variability and to agree on standards for communicating with women during pregnancy about the possibility of operative delivery and for debriefing women after caesarean section. sarean section, and women's satisfaction with this involvement.

INTRODUCTION

Current policies on the provision of maternity care reflect the increasing importance attached both to women's views and to women being at the centre of decisions about their intrapartum care1,2. As regards caesarean section, there is a marked imbalance between the few studies investigating the views of women on this mode of delivery and the numerous studies of its physiological, pathological and epidemiological aspects1,3. One exception is a survey of 300 women, all delivered by caesarean section5, which revealed that 14% had requested the section; this rose to 20% for those with at least one previous caesarean delivery. It is not possible from this survey, nor the more recent study by Mould et al6, to ascertain the extent to which each request for caesarean section was primarily the woman's decision or how much it was influenced by the attending obstetrician7. Other studies have noted the expectation for women to be passive in the decision-making process8,9.

There may be a greater risk to obstetricians of litigation for not undertaking a caesarean section rather than for conducting one judged as unnecessary10, and as a result doctors may sometimes authorise sections where there is uncertainty of benefit. This may be one reason for the high rates of caesarean delivery among primiparous women reported for Scotland11. It is well-established that women's lack of involvement in making decisions is associated with litigation12,13, but little empirical data exist on how decisions are currently being taken. The purpose of this study was to measure women's involvement in the decision to deliver by caesarean section.

METHODS

The study was conducted in the maternity unit of a large teaching hospital in Scotland, which at the time was the only unit undertaking caesarean sections in the Grampian Region. In 1995, the caesarean section rate for singleton pregnancies was 16.95%. Permission from the local Ethics Committee was obtained to conduct the study, which took place between October 1995 and February 1996. At the hospital, a research midwife who was not wearing a uniform, approached women after their caesarean section and gave them verbal and written information concerning the study. Women who agreed to participate gave their written consent.

As multiparous women are more likely to have confidence and experience in negotiating with health professionals (obstetricians, midwives and general practitioners), and as consultations before childbirth regarding caesarean section can be more extensive than consultations during labour, it was decided to consider three groups:

  • 1Elective caesarean section,
  • 2Emergency caesarean section in multiparae,
  • 3Emergency caesarean section in primiparae.

The study set out to recruit 165 women, with an equal number of women in the three categories. We defined all caesarean sections where the decision was taken before labour and not in the context of a clinical emergency as ‘elective’. We did not analyse elective caesarean section according to parity because most of these sections occur among multiparae.

Exclusion criteria were: women receiving private care; women who were staff at the hospital; women transferred from outside the Grampian Region; and women whose babies died in the perinatal period. Two hundred and forty women underwent caesarean section between October 1995 and February 1996, of whom 169 were approached to participate in the study. Two women declined to participate and one woman was found after recruitment to have private care, and so 166 women were therefore included in the study. Of the remaining 71 women undergoing caesarean section during the recruitment period, 13 had been ineligible: private care (5); staff of the hospital (3); transfers from outside the Grampian Region (3); and perinatal death (2). A further 23 women were lost to the study owing to early discharge and interruption in recruitment during the Christmas holidays, and 4 for miscellaneous reasons (psychiatric disorder 2, movement out of the area 1, and unable to speak English 1). Finally, 31 primiparae undergoing emergency caesarean section were not approached since alternate cases were recruited in order to achieve some balance in the number of multiparae and primiparae undergoing emergency caesarean section.

Information for the study was gathered from three sources:

  • 1Interviews with the women. These were carried out in the hospital between the third and fourth postnatal day by the research midwife. For each woman a structured questionnaire was used to gather data on her: demographic, social and economic characteristics; antenatal and intrapartum care; the methods by which the decision to deliver by caesarean section was made; the information she received; and her postnatal recovery. The schedule drew upon questionnaires developed and validated in previous studies at the hospital14 and elsewhere15,16, and was pretested on ten women in the hospital before the study commenced. The questionnaire comprised open and closed questions. Responses to the former question type were coded after data collection, while for the majority of the latter, a range of closed response categories were offered which were based on the findings of a pilot phase involving two focus groups of women and the use of the questions in an open format during the pretest interviews. Responses to the questions on adequacy of information and on satisfaction with involvement were sought using Likert scales, whereby the woman was shown a card offering five categories on the continuum from ‘extremely adequate/satisfied’ to ‘extremely inadequate/dissatisfied’. Copies of the questionnaire are available from the authors.
  • 2Self-completion questionnaires. These were sent to women at 6 and 12 weeks postpartum to capture their views at two other points in the period after delivery, since women's feelings about the birth experience are well-known to change over time17. These questionnaires included similar topics to those mentioned above.
  • 3Medical records. Data were extracted for each woman on her medical and obstetric history, complications in her current pregnancy, and indications for her caesarean section, including discussions between the woman and the doctors and midwives attending her.

The results reported in this paper are based on the first and principal source of information, namely the interviews with women; the findings from the other sources will be reported elsewhere. These data were entered into a database and analysed using SPSS-Windows18. Simple descriptive techniques were used to characterise the study population, and bivariate analysis (primarily χ2 test) undertaken to look for associations among categorical variables. The kappa statistic19 was used to measure the extent of agreement between women's reports of the main reasons for their caesarean section and the indication given in their medical records.

RESULTS

Social and obstetric characteristics

Tables 1 and 2 show the characteristics of the women. Of the 166 women, 63 had elective and 103 had emergency caesarean sections The groups were similar, although statistically significant associations were noted between the type of caesarean section and abnormal obstetric history and abnormal medical history (Table 2). For the emergency sections, a statistically significant association was also noted between parity and whether the decision to conduct a section was made before labour (owing to a clinical emergency) or during labour. The groups also differed in the primary reason for caesarean section, as defined by the Ontario scale2′. The main reason for both elective and emergency caesarean section in multiparae was previous section, while for primiparae an emergency caesarean section was most commonly carried out for dystocia.

Table 1.  Characteristics of the women. Values are given as mean [SD] and n (%). CS = caesarean section.
 Elective CS max. n= 63Emergency multiparae max. n= 38Emergency pnmiparae max. n= 65
  1. *Excluding uemployed partments.

Age31 [4.8]29 [5.9]28 t5.51
Parity
 Primiparous18 (29)65 (100)
 Para 129 (46)27 (71)
 Para > 116 (25)11 (29)
Social class*n= 62n= 37n= 61
 I12 (19)6 (16)7 (11)
 II14 (23)8 (22)18 (30)
 III16 (26)13 (35)16 (26)
 IV15 (24)6 (16)8 (13)
 V5 (8)4 (11)12 (20)
Highest level of education achieved
 None8 (13)6 (16)6 (9)
 ‘O’ Grades13 (21)13 (34)13 (20)
 Highers/further education (HND, Diploma)22 (35)10 (26)24 (37)
 Degrees/vocational20 (32)9 (24)21 (32)
 Missing data001 (2)
Living with partner or husband61 (97)33 (87)60 (92)
Table 2.  Obstestric and medical characterstics of the women. Values are given as mean (SD) and n (%).
 Elective max. n= 63Emergency multiparae max. n= 38Emergency primiparae max. n= 65χ2P
  1. *Excluding transverse lie owing to small numbers.

Gestational age (weeks)38 [1.4]39 [2.6]39 [3.1]
Timing of decision for emergency caesarean sections   3.74; df = 1; P= 0.0530
 Before labour12 (32)10 (15) 
 During labour26 (68)55 (85) 
Presentation at delivery   2.15*; df = 2; P= 0.3412
 Cephalic45 (72)31 (81)56 (86) 
 Breech14 (22)6 (16)9 (14) 
 Transverse4 (6)1 (3) 
Reason for caesarean section (as recorded by medical staff)   
 Previous caesarean section30 (48)14 (37) 
 Breech17 (27)5 (13)9 (14) 
 Dystocia10 (26)30 (46) 
 Fetal distress5 (13)18 (28) 
 Other16 (25)4 (11)8 (12) 
Abnormal obstetric history   39.98; df = 2; P < 0.0001
 Yes45 (71)28 (74)14 (22) 
 No18 (29)10 (26)50 (78) 
Abnormal medical history   7.69; df = 2; P= 0.0213
 Yes16 (25)18 (47)30 (47) 
 No47 (75)20 (53)34 (53) 

Women's knowledge and information received

There was a significant association (P < 0.0002) between the main source of information about caesarean section received during pregnancy and the type of caesarean undertaken (Table 3). Women who had elective caesarean sections had mostly received information from individual health professionals (41%), while this was the case for only 20% of those undergoing emergency sections (95% confidence interval for difference between proportions = 6%, 36%, χ2= 7.14; P= 0.0075). Overall, no association was found between social class and the main source of information women received during pregnancy. In terms of adequacy of information, the data were analysed as a simple dichotomy (adequate/adequate) owing to insufficient responses in the extreme categories of very adequate and very inadequate. The majority of women who did receive information about caesarean section felt the amount to be adequate (Table 4), although there was a significant difference by type of section (P= 0.0139). Almost three-quarters of the women delivering by elective caesarean section felt the amount of information they received was adequate, while this was the case for half (50%) of those undergoing emergency section (95% confidence interval for difference between proportions = 6%, 36%, χ2= 6.91; df = 1; P= 0.0086). Moreover, 13% of all the women reported that they did not receive any information and the majority (90%) of these were emergency caesarean sections, many of whom had not desired information during pregnancy as they were expecting a normal delivery.

Table 3.  Type of caesarean section by main source of information on caesarean section received during pregnancy. Values are given as n (%).
Main source of information*ElectiveEmergency multiparaeEmergency primiparaeTOTAL
  1. *Source of information and type of caesarean section, χ2= 26.89; df = 6; P < 0.0002.

  2. Proportion receiving information from health professionals (obstetricians, midwives, general practitioners) by type of caesarean section, χ2= 12.72; df = 2; 0.0017.

  3. 21 women reported not receiving any information, and the data are missing for 5 women.

Antenatal classes4 (7)1 (4)18 (35)23 (16)
Self learning24 (39)12 (44)18 (35)54 (39)
Friends and family8 (13)4 (15)10 (19)22 (16)
Health professionals25 (41)10 (37)6 (11)41 (29)
TOTAL61 (100)27 (100)52 (100)140 (100)
Table 4.  Adequacy of information received during pregnancy. Values are given as n (%).
Adequate information*ElectiveEmergency multiparaeEmergency primiparaeTOTAL
  1. *Adequacy of information and type of caesarean section, χ2= 12.52; df = 4; P= 0.0139.

  2. Proportion receiving adequate information by type of caesarean section, χ2= 6.91; df = 1; P= 0.0086.

  3. Women not receiving any information.

  4. §Data are missing for 4 women.

Yes44 (71)16 (42)34 (55)94 (58)
No16 (26)13 (34)18 (29)47 (29)
Not applicable2 (3)9 (24)10 (16)21 (13)
TOTAL62 (100)38 (100)62 (100)162 (100)§

Just over half of the women (51%) reported that they did not receive any debriefing from medical or midwifery staff on the reasons for their caesarean section before leaving hospital. There was a statistically significant association between type of caesarean section and whether or not the woman was debriefed (χ2= 8.97; df = 2; P= 0.011). Interestingly, 14% of the women who were not debriefed also felt this was not necessary, and three-quarters of these women had undergone elective caesarean section.

There was good agreement21 between what the woman reported as the main reasons for caesarean section and those stated in her casenotes (70%, K= 0.625). There was no significant difference between elective and emergency caesarean sections in the level of agreement (elective sections 67%, emergency sections 73%).

Involvement in the decision to have a caesarean section

The association between satisfaction, involvement and type of caesarean section is shown in Table 5. Although the questionnaire enabled women to indicate their degree of satisfaction or dissatisfaction, there were so few responses in the extreme categories that it was necessary to revert to a dichotomous variable: satisfied and dissatisfied. As expected, women undergoing elective caesarean section were significantly more likely to be involved in the decision-making process. Overall 59% of women reported they were involved in the decision-making process and were satisfied with this, and a further 25% were not involved but still satisfied. In all groups, the most common response given by women to an open question on why they were not involved was expressed in terms of ‘doctor knows best’ (Table 6). Women undergoing elective caesarean section and primigravidae undergoing emergency sections stated the next most common response in terms of ‘own feelings were not considered by staff’.

Table 5.  Satisfaction with involvement in making the decision to deliver by caesarean section. Values are given as n (%).
 Women involved in making the decision 
 YesNoχ2P
  1. *P value using Fisher's exact test (2 sided).

  2. Information missing for one woman who had an elective caesarean section, and three who had an emergency caesarean section.

TOTALS
 Elective (n= 62)49 (79)13 (21)P= 0.0004
 Emergency (n= 100)51 (51)49 (49) 
Elective (n= 62)49 (79)13 (21) 
 Unsatisfied1 (2)4 (31)P= 0.0056
 Satisfied48 (98)9 (69) 
Emergency multiparous (n= 36)17 (47)19 (53) 
 Unsatisfied1 (5)7 (37)P= 0.0438
 Satisfied16 (95)12 (63) 
Emergency primiparous (n= 64)34 (53)30 (47) 
 Unsatisfied2 (6)11 (37)P= 0.0039*
 Satisfied32 (94)19 (63) 
TOTALS (n= 162)
 Unsatisfied4 (4)22 (35)P < 0.00001
 Satisfied96 (96)40 (65) 
Table 6.  Main reasons given by women for not being involved in the decision to deliver by caesarean section. Values are given as n (%).
Main reasonElectiveEmergency multiparaeEmergency primiparaeTOTAL
Previous caesarean section1 (8)1 (5)2 (3)
‘Doctor knows best’7 (54)6 (32)8 (28)21 (34)
Women's feelings not considered by staff5 (38)1 (5)7 (23)13 (21)
Unable to make decisions due to drugs or complications4 (21)4 (13)8 (13)
Sudden emergency6 (32)7 (23)13 (21)
Risk to baby's health4 (13)4 (6)
Not specified1 (5)1 (2)
TOTAL13 (100)19 (100)30 (100)62 (100)

Maternal preference

Twelve women (7%) directly reported ‘maternal preference’ as a factor influencing the decision to deliver by caesarean section: eight were elective and four were emergency sections. In five instances the preference was reported by the woman herself but not in the medical records, and in another five, the reverse was true. Further indirect evidence of maternal preference arose during the interviews, with an additional 10 women expressing a preference for caesarean section when describing their labour and delivery. Thus 13% of the 166 women showed direct or indirect evidence of maternal preference for delivery by caesarean section. During the interviews, women were also asked to report their feelings towards their caesarean section. Just less than half (44%) indicated positive feelings and nearly a quarter (24%) of the women expressed negative feelings. The degree of feelings varied by caesarean section type. Women in the elective group were significantly (95% confidence interval for difference between proportions = 6%, 30%χ2= 6.76; P= 0.0093) less likely to have such negative feelings (13%) compared with women undergoing emergency caesarean section (30%), of whom 48% were multiparae and 52% primiparae.

DISCUSSION

This investigation shows that women are dissatisfied if they are not involved in the decision to deliver by caeasarean section, and confirms the findings of previous studies on women's satisfaction with their childbirth experience15,22. An association between women's lack of involvement in decisions and subsequent litigation against attending staff has long been established12,23. Women who feel in control of events during labour and delivery are more satisfied and have greater emotional wellbeing postnatally15. However, women also clearly vary in the extent to which they desire to feel in control and some are happy to leave making decisions entirely to obstetricians and midwives- ‘they know best’24. Thus the challenge to the providers of care is to be responsive to the varying degrees to which individual women want to be actively involved in decision-making. The key question is whether the new and evolving models of maternity care in the UK facilitate or constrain such flexibility.

In this study, women undergoing elective caesarean section generally received adequate information; however, with emergency caesarean section, half of the women had not received enough information during pregnancy. This proportion is greater than that found by Francome et al.5 In their study, one in six of the women who had emergency caesarean sections said they lacked information concerning the operation, and its anaesthetic and post-operative complications. Giving this sort of information routinely at antenatal classes might cause unnecessary anxiety among the majority of women who do not deliver by caesarean section. However, only a minority of emergency caesarean sections have to be performed with such immediacy that there is inadequate time to involve the woman in the decision. Our results also highlight both the variation in information needs among women, with a proportion not desiring any information, and the importance of health professionals, being aware of this variability. Green et al.16 highlighted the lack of communication between health professionals and pregnant women, and the report of the Audit Committee of the Royal College of Obstetricians and Gynaecologists25 agreed with these findings. However, further studies are needed on how to meet the maternity care needs of individual women in a health service designed to treat populations26.

In 70% of the cases in our study the reasons women reported for their caesarean section agreed with those in the medical records. Agreement would perhaps have been even higher if the women had been given exactly the same closed categories (based on the Ontario scale) in the questionnaire as are used in the medical records. Francome et al6. found 92% agreement, Mould et al.6 89%, and Hillan27 73%. We did not observe a difference in agreement between the elective and emergency caesarean sections, unlike Hillan, who reported greater agreement with elective section (81%) than with emergency section (68%; P < 0.025). In our study, we found that medical and midwifery staff do not often document discussions with the women regarding the reasons for their caesarean section and its implications for future childbirth. Although this does not necessarily mean that these discussions did not take place, it is disappointing to note that just over half of the women in this study stated that they had not been debriefed prior to discharge. Of course at one level, whether these debriefings did or did not actually occur is irrelevant-the key point is that the women interviewed did not perceive or recall them to have taken place. Accepting that the numbers here are too small to draw statistically reliable conclusions, this is nevertheless a key area of care warranting further and larger-scale investigation.

The Scottish Caesarean Section Audit11 found for 1994 that 7.7% (623/8098) of all singleton caesarean sections were associated with maternal request, according to medical records; this number represents only about 1% of all singleton deliveries. However, data are not available on the proportion of women delivering vaginally but whose preference was for a caesarean section, and this is a limitation of many studies which give some insights into maternal preference. Francome et al.5 asked only women delivering by caesarean section the direct question ‘Did you ask to have a caesarean section?’ and found 14% of women responded positively. In our study, from both the women's interviews and the medical records, 7% of women requested a caesarean section. Francome et al.5 also found that only 1 in 50 consultants regarded maternal request as a valid indication. Obstetricians' attitudes towards caesarean section have recently received much attention in the literature28,29 and popular press30, with calls31 made for a rethink about the conventional unfavourable medical response to women who ask for elective caesarean section with no medical indication. Such positive attitudes towards caesarean delivery held by some women should however be balanced against negative feelings; in our study a significant proportion of women experienced such negative feelings, particularly with emergency caesarean section (30%).

The extent to which women feel able and willing to report their dissatisfaction is highly sensitive to who and how the question is asked. In our study, the interviewer was not approaching the women as a serviceprovider but as a researcher. This approach differs from that reported by Mould et al.6 where doctors conducted the interviews. Response bias is clearly a possibility with either approach and reaffirms that ‘satisfaction is difficult to measure and even more difficult to interpret’.

This investigation describes women's involvement in the decision to deliver by caesarean section. It did not include a comparison group of women who managed to avoid caesarean section and so cannot comment on their contribution to the decision-making process. The findings reported here are based on comparatively small numbers, consistent with the exploratory nature of the study, and need to be confirmed with larger samples and in populations in other parts of the UK. The questionnaire we have developed to ascertain women's involvement could be adapted and re-validated for use with all types of delivery, and in the context of randomised trials. The questionnaire yields primarily categorical data which could usefully be complemented by insights from qualitative methods applied in in-depth interviews with a small subsample of women.

CONCLUSION

Throughout the National Health Service, the views of the patient are increasingly valued in the assessment of quality care. Health information systems need to capture patients' perspectives, accepting that these may differ from providers' perspectives as presented in medical records. Maternity care is one area trying to integrate women's views and reconcile these differences of perspective. Clearly a conscious woman must be able to refuse medical intervention32, but should she be able to demand it? In the case of caesarean delivery, there is no current agreement on the appropriate section rate, nor on the appropriate level of a woman's say given the overriding need to preserve her health interests as assessed by the doctorJ3 and to use NHS resources to achieve maximum health gain.

This paper describes a study of women's involvement and satisfaction with the decision to deliver by caesarean section. It shows that women are not homogeneous in their requirements for information nor their desire to be involved, and the challenge to providers is to be responsive to this variability. The extent to which needs are met as perceived by the women themselves has implications for further research and for practice. Intervention research is needed to identify appropriate ways to improve women's preparedness for complicated deliveries, and further methodological studies are needed on measuring women's involvement and satisfaction. In terms of practice, the study highlights the need for standards both for communicating with women during pregnancy about the possibility of operative delivery and for labour debriefing of women after caesarean section. The findings of this study are also directly relevant to the recent concern for rising caesarean section rates and obstetricians' fear of litigation. Potentially opposing forces are at work here. Not surprisingly, fear of litigation is likely to increase the doctor's wish to control decisions surrounding operative delivery. On the other hand, women's lack of involvement in the decision is known to be a significant precipitating factor for complaints and legal action. Changing Childbirth34 gives a high profile to partnerships in care. Not all women will want equal partnerships as regards the decision to deliver by caesarean section, but they should have the opportunity to be involved.

Acknowledgements

The authors would like to thank all the women who participated in this study; the medical and midwifery staff at Aberdeen Maternity Hospital; Ms L. Henderson for helping with data coding; Ms E. Stirton and Ms R. Mitchell for secretarial support; and Mrs A. Fitzmaurice for statistical advice. Funding for this study was received from Aberdeen University Research Committee and Aberdeen Royal Hospitals NHS Trust.

Ancillary