4 BIRTH 21:l March 1994

Satisfaction Wdh Care in labor and Birth: A Survey of 790 Australian Women

Stephanie Brown, BA(Hons), and Judith Lum/ey, PhD

ABSTRACT: Data on satisfaction with care in labor and birth were gathered in a suruey conducted in conjunction with a review of maternity services in Victoria, Austra- lia. All women who gave birth in one week in 1989 (>1000) were mailed questionnaires eight to nine months after the birth, with a response rate of 790 (71.4%). When adjusted for parity in a logistic regression model, the following factors were highly related to dissatisfaction with intrapartum care: lack of involvement in decision making ( p < 0.001), insuficient information ( p < 0.001), a higher score for obstetric intervention ( p = 0.015), and perception that caregivers were unhelpful ( p = 0.04). No association was found between satisfaction and maternal age, marital status, total family income, country of birth, or health insurance status. The survey results were influential in shapingJina1 recommendations of the Ministerial Review of Birthing Services by coun- tering stereotypes about women who become dissutis$ed with their care, providing evidence of far greater dissatisfaction with intrapartum than antenatal care, and dem- onstrating the importance of information, participation in decision making, and rela- tionships with caregivers to women’s overall satisfaction with intrapartum care. (BIRTH 21:1, March 1994)

Since Cartwright’s seminal study of women’s views about induction of labor was published in 1979 (l), increasing efforts have been made to find out what women think about maternity services. Most re- ports are based on smaller, selective samples, but several population-based studies have also been published (2-9).

Stephanie Brown is a Research Fellow at the Centre .for the Study of Mothers’ and Children’s Health, Monash University, Australia. She was a member o f t h e Study Group which con- ducted the Ministerial Review of Birthing Services in Victoria.

Judith Lumley is the Director of the Centre for the Study of Mothers’ and Children’s Health, Monash University, Australia. She was the Chairperson of the Ministerial Review of Birthing Services in Victoria.

Address correspondence to Stephanie Brown, Centre for the Study of Mothers’ and Children’s Health, Faculty of Medicine, Monash University, 463 Cardigan Street, Carlton, Victoria, Aus- tralia 3053.

This study was supported by a grant from the Victorian Health Promotion Foundation.

0 1994 Blackwell ScientiJic Publications, Inc.

This study was based on a postal survey of a representative sample of recent mothers living in the state of Victoria, Australia. It was conducted in conjunction with an independent review of birthing services, chaired by one of the authors, that re- ported to the Victoria Minister for Health in 1990 (10). A primary reason for undertaking the study was to explore the extent to which issues and con- cerns women expressed about maternity services in public meetings and in submissions to the review of birthing services represented the views of women who had recently given birth. Details about the sam- ple, study design, and findings concerning satisfac- tion with antenatal care are reported elsewhere (1 1). This article presents findings on women’s satisfac- tion with their care during labor and birth, and dis- cusses the role the survey played in shaping final recommendations made by the review.

Maternity services in Australia combine ele- ments of both United States and Canadian systems. Around 40 percent of women who give birth in Vic- toria are public patients who receive standard pub- lic hospital care, which includes all antenatal care and is free at the point of service. Another 45 per-

BIRTH 21:l March 1994 5

cent are cared for privately by specialist obstetri- cians, and 11 percent are private patients of family physicians (10). A further 1 to 2 percent register with a hospital birth center, and are cared for as public patients by a small team of midwives or as private patients of a family physician or obstetri- cian. Fewer than 1 percent of Victoria women plan to give birth at home, and they make private ar- rangements for midwifery and medical care. Women choose among these options, depending on their knowledge of what is available, whether or not they have private health insurance or can otherwise cover the costs of private care, and their proximity to services.

Methods

Sample

Questionnaires were sent to a representative sam- ple of 1193 women living in Victoria, eight to nine months after they had given birth. All state mater- nity hospitals and home birth practitioners were asked to facilitate the mailing of questionnaires to women who gave birth in a one-week period. All but one country hospital agreed to participate, although the institutional ethics committee of a large teaching hospital delayed making a decision for eight months. As a result, the sample comprised all women who gave birth in the second week of Febru- ary 1989, except at one hospital where the sampled week was the second week in September.

Procedure

Most hospitals opted to send questionnaires out di- rectly; a few mostly smaller hospitals provided us with a mailing list. Women were excluded from the survey if they had experienced a stillbirth or their baby was known to have died. Three mailings took place for each sample at two-week intervals. Each one included the questionnaire, an invitation to par- ticipate in the study, a brief explanation of the study in six community languages, and a postpaid enve- lope.

Of 799 completed questionnaires, 9 could not be used because they duplicated an earlier response or the baby’s birth fell outside the study dates. An- other 77 were returned unknown at the mailing ad- dress, and one woman’s baby was stillborn. This left an eligible sample size of 1107, with a final re- sponse rate of 790 (71.4%).

A comparison of respondents with all women who gave birth in the sampled weeks using the state’s perinatal data collection showed that the re-

spondents were largely representative in terms of their method of delivery, infant birthweight, parity, and residence in rural or metropolitan communities. As predicted by other postal surveys of recent mothers, women who were under age 25 years, sin- gle, or non-English-speaking were underrepre- sented.

Questionnaire

The questionnaire and survey design were modeled on studies of satisfaction with maternity care car- ried out by the Institute for Social Studies in Medi- cal Care in England and Wales (12-14). Questions covered information about the baby; reproductive history; antenatal, intrapartum, and postnatal care; and sociodemographic characteristics. Women were asked for factual information (e.g., who was present at the birth, whether labor was induced), and for their views about what happened and their preferences if they were to have another baby.

The questions about women’s satisfaction with care included some asking for overall ratings of an- tenatal and intrapartum care, and others about spe- cific aspects. An overall rating of care during labor and birth was derived from the following question:

Overall do you feel your labor and delivery were: managed as you liked? managed as you liked in some ways, but not in others? not managed as you liked?

Checking either response 2 or 3 was taken to indi- cate dissatisfaction; women who said their care was as they liked were assessed as being satisfied over- all. Specific questions were asked about the helpful- ness of caregivers, whether caregivers had suffi- ciently explained what was happening, and how kind and understanding women felt caregivers had been.

Since many women in consultations and written submissions to the review of birthing services men- tioned wanting to have an active say in decisions regarding their care, we asked the following ques- tion:

Do you feel you were given an active say in making deci- sions about what happened during your labor?

yes, in all cases yes, in most cases only sometimes no, not at all uncertain

These questions were not asked of women who had an elective cesarean section. This analysis is there- fore restricted to the 690 women who had a vaginal birth or emergency cesarean.

6

Statistical Analysis

BIRTH 21:l Murch 1994

Intrapartum Events

Data were analyzed using Epi Info version 5.01 and Egret. Statistical tests carried out were the chi- square test and Fisher’s exact test, odds ratios, and chi-square for linear trend. Logistic regression was conducted using Egret.

Results

Overall Rating of Intrapartum Care

Two-thirds of women (67%) reported that their la- bor and the birth of their infant were managed as they liked, 27 percent said their care was mixed, and 6 percent not as they liked.

Associations With Social Characteristics

Only three sociodemographic variables, parity, level of school-based education, and place of resi- dence, were associated with differences in women’s overall ratings of care. Primiparas were more likely to be dissatisfied than multiparas (odds ratio = 1.62 [ 1.16-2.28]), women who had not completed sec- ondary school were more likely to be dissatisfied than those with a higher level of education (OR = 1.43 [1.01-2.01]), and rural women were more often dissatisfied than city women (OR = 1.49 [ 1 .O-2.11). Sociodemographic variables not associated with satisfaction were the mother’s age, marital status, country of birth, total family income, level of ter- tiary education, and health insurance status.

Pregnancy and Reproductive History

Very few factors related to the current pregnancy or reproductive history were associated with satisfac- tion with care. A strong association was present between dissatisfaction with antenatal and intrapar- tum care (OR = 29.26 [15.0-57.81). No other factors related to the current pregnancy were associated with satisfaction, such as whether or not women were pleased to be pregnant, attendance or nonat- tendance at childbirth preparation classes, and hos- pitalization during pregnancy. Three reproductive history factors were associated with dissatisfaction: cesarean section (OR = 2.10 [1.0-4.4]), preterm birth (OR = 1.93 [1.0-3.7]), and breech presenta- tion (OR = 2.51 [1.0-6.41). No association was found between satisfaction and a history of preg- nancy loss (stillbirth, spontaneous or induced abor- tion), previous low-birthweight infant, or prior for- ceps delivery.

Questions were asked about length of labor, use of analgesia, induction or augmentation, method of de- livery, and primary caregivers for birth. In addition, a weighted score for obstetric procedures was cal- culated based on the scoring system developed by Elliot et a1 (15).

The relationships among intrapartum events, in- cluding the total score assigned for obstetric proce- dures and women’s global rating of their care, are given in Table 1. The table shows the number and percentage of women experiencing each intrapar- turn event, and the proportion of women dissatisfied in each category. The results are stratified accord- ing to parity.

Birth events and experiences of primiparas dif- fered from those of multiparas. The former were more likely to have a longer first stage of labor, induction or augmentation, pharmacologic methods of pain relief, more pain than they expected, pain relief that was unsuccessful, operative deliveries, and episiotomies. Despite these differences, consid- erable consistency occurred across the two groups in the factors associated with dissatisfaction.

Both multiparas and primiparas were more likely to be dissatisfied if they experienced an emergency cesarean section; used pethidine, nitrous oxide, or general anesthesia; found the pain worse than antic- ipated; had a high obstetric intervention score (5-9, >lo); or could not hold the baby soon after birth. The following factors were associated with dissatis- faction in primiparas but not in multiparas: induc- tion, augmentation, epidural anesthesia, forceps, and episiotomy. A first stage lasting longer than 12 hours was associated with dissatisfaction among multiparas, but not primiparas. Satisfaction did not vary according to the primary caregiver assisting with the birth (doctor, midwife, doctor-midwife combination) in either group.

Information and Views About Caregivers

The relationship between satisfaction with intrapar- turn care and the responses of primiparous and mul- tiparous women to questions about access to infor- mation, helpfulness of caregivers, and kindness and understanding shown by caregivers are given in Ta- ble 2. Both groups were much more likely to be dissatisfied if they thought that they had not ob- tained sufficient information from caregivers, and if they thought that midwives and doctors had pro- vided only some, little, or no help. Women who described their caregivers as only fairly or not at all kind and understanding were also more likely to be

BIRTH 21:l March 1994 7

Table 1. Relationship Between lntrapartum Events and Dissatisfaction With Care in Labor and Birth by Parity

Labor induced No Yes

No Yes

<2 hrs 2- 12 hrs >12 hrs

Spontaneous vaginal Emergency cesarean

section Forceps

Episiotomy No Yes

Labor augmented

Length of labor

Birth

Obstetric procedure score 0-4 5-9 10+

Yes No

Doctor Midwife Midwife/doctor Other

Pain relief None Gas and oxygen Pethidine Epidural General anesthetic

Success of pain relief Very/partly successful Not at all successful

As expected or better Worse or much worse

Held baby soon after birth

Primary caregiver at birth

Experience of pain

than expected

Primiparas Multiparas

% Dis- % Dis- No. % satisjied OR (CI) No. % satisjied OR (CI)

-

171 90

136 121

6 174 81

160

29 71

128 I33

125 60 76

21 1 50

204 32 20 9

38 164 180 61 11

82 138

85

135

66 35

53 47

2 67 31

62

11 27

49 51

48 23 29

81 19

77 12 8 3

15 64 70 24 4

37 63

39

61

34 50

32 47

33 39 42

31

52 54

30 49

28 43 57

37 52

44 28 20 44

18 42 44 48 55

31 51

25

47

I .oo 1.95 (1.1-3.4)

1 .oo 1.93 (1.1-3.3)

1 .oo 1.3 (0.2-14.8) 1.45 (0.2-16.8)

1 .oo 2.43 (1.0-5.8) 2.61 (1.4-4.8)

1 .oo 2.26 (1.3-3.9)

1 .oo 1.97 (1.0-3.9) 3.35 (1.8-6.4)

1 .oo 1.85 (1.0-3.6)

1 .oo 0.51 (0.2-1.2) 0.32 (0.1-1.1) 1.03 (0.2-5.0)

1 .oo 3.22 (1.3-9.1) 3.46 (1.4-9.8) 4.01 (1.4-11.8) 5.3 (1 .O-29.2)

1 .oo 2.35 (1.3-4.4)

1 .oo 2.75 (1.5-5.2)

289 124

282 129

58 330 25

364

19 28

279 135

259 71 34

370 48

256 118 24 25

154 96

150 20 6

215 45

262

149

70 30

69 31

34 80 6

89

5 7

67 33

71 20 9

89 12

61 28 6 6

37 23 36 5 1

83 17

64

36

30 27

29 30

22 29 52

27

53 39

28 33

13 35 44

27 52

29 29 29 28

21 37 35 33 83

33 38

20

46

1 .oo 0.83 (0.5-1.4)

1 .oo 1.08 (0.7-1.7)

1 .oo 1.42 (0.7-3.0) 3.75 (1.2-11.4)

1 .oo 2.97 (1.1-8.2) 1.73 (0.7-4.1)

1 .oo 1.27 (0.8-2.0)

1 .oo 3.72 (1.9-7.2) 5.41 (2.3-12.5)

1 .oo 3.02 (1.6-5.8)

1 .oo 0.98 (0.6-1.6) 0.99 (0.4-2.7) 0.94 (0.3-2.5)

1 .oo 2.26 (1.4-3.8) 2.08 (1.2-3.6) 1.27 (0.3-4.0)

19.06 (2 .O-908.6)

1 .oo 1.23 (0.6-2.5)

I .oo 3.47 (2.2-5.6)

dissatisfied with their care overall, regardless of parity.

The magnitude of these associations is particu- larly striking. Lack of information was associated with a fourfold to sixfold increase in dissatisfaction. Women who did not regard their caregivers as being very kind and understanding were four to five times more likely to rate their care negatively, and the odds of dissatisfaction increased twofold to eight- fold when women assessed caregivers as only sometimes or not at all helpful.

Role in Decision Making

The survey included a general question regarding the role women saw themselves as having in deci- sion making, and a question asking women about the presence of anyone at the birth whom they did not want to be there. The relationships between women’s responses and their overall satisfaction are shown in Table 3. Not having an active say in decisions was associated with a sixfold increase in dissatisfaction among primiparas and a fifteenfold

8 BIRTH 21:l March 1994

Table 2. Relationship Between Access to Information and Views About Caregivers and Dissatisfaction With Care in Labor and Birth by Parity

Primiparas

% Dis- No. % satisfied OR (CI)

Multiparus

% Dis- No. % satisjied OR (CI)

Informed of pain relief options before labor

Yes No

doctors Information from

Enough Wanted more

Information from midwives

Enough Wanted more

Helpfulness of doctors Verylfairly

helpful Some, little, or

no help Helpfulness of

midwives Verylfairly

helpful Some, little, or

no help Kindness and

understanding of caregivers

Very kind and

Fairly kindhot understanding

very kind and understanding

229 34

185 46

205 55

200

36

246

18

206

55

87 39 1 .oo 13 44 1.22 (0.6-2.7)

80 33 1 .oo 20 74 5.76 (2.7-12.7)

79 32 1 .oo 21 67 4.33 (2.2-8.6)

85 38 1 .oo 15 64 2.95 (1.3-6.6)

93 37 1 .oo 7 83 8.52 (2.3-46.8)

79 32 1 .oo

21 73 5.78 (2.9-11.9)

277 139

272 45

36 1 57

277

48

389

28

329

91

67 24 1 .oo 33 40 2.09 (1.3-3.3)

86 23 1 .oo 14 64 6.14 (3.0-12.7)

86 24 1 .oo 14 63 5.5 (2.9-10.3)

85 24 1 .oo 15 56 4.03 (2.05-8.0)

93 26 1 .oo 7 75 8.55 (3.4-24.4)

78 22 1 .oo

22 55 4.35 (2.6-7.3)

Table 3. Relationship Between Women’s Role in Decision Making and Dissatisfaction With Care in Labor and Birth by Parity

~ ~~

Primiparas Multiparas

% Dis- % Dis- No. % satisfied OR (GI) No. % satisfied OR (CI)

Given an active say in decisions about what happened in labor

Yes, in all cases 102 39 24 1 .oo 193 46 11 1 .oo Yes, in most

Only sometimes 35 14 74 9.39 (3.6-25.2) 30 7 77 26.91 (9.5-79.4) Not at all 15 6 73 8.94 (2.3-41.1) 16 4 88 57.33 (1 1.6-533.9)

cases 99 38 38 2.02 (1.1-3.9) 163 39 33 3.95 (2.2-7.2)

Uncertain 8 3 50 3.25 (0.6-18.6) 14 3 71 20.48 (5.2-94.9)

No 238 93 39 1.00 397 95 27 1 .oo Yes 18 7 61 2.49 (0.9-7.4) 20 5 55 3.23 (1.2-8.7)

Unwanted people present at birth

BIRTH 21:l March 1994 9

increase among multiparas. Unwanted people being present at the birth was associated with a threefold increase in dissatisfaction among multiparas and a twofold increase among primiparas, although the latter just failed to reach significance.

Several women wrote about instances in which they felt their wishes had not been respected, in- cluding this mother of twins:

I had too many people in the room at the time of the births. I was forced to stay on my back for hours with monitors all around and inside my body. I was given re- peated vaginal examinations while on my back, which is agony when carrying two babies. I never got to hold my babies after the birth, so I just felt I was a machine.

Other women who felt their views had been re- spected described how much they valued this: My first child was an emergency cesarean following 14 hours labor, failed forceps, and failed epidural delivery.

The support I was given by my doctor to deliver naturally, despite the fact that he advised against it, made the birth of my second child fantastic. The respect I was shown by the medical profession in letting me determine my fate is something I was really grateful for.

Contribution of Birth Events, Birth Experiences, and Social Factors to Dissatisfaction

Ten factors reflecting major birth events, women’s experiences of caregivers, and social factors associ- ated with dissatisfaction were selected to be fitted in a logistic regression model (Table 4). Data on the helpfulness of midwives and doctors were pooled to give one measure of the helpfulness of caregivers. The same was done to create a single measure of access to information. Rather than include individ- ual procedures, we used the weighted intervention

Table 4. Odds Ratios for Associations of Panty, Education, Birth Events, and Experiences With Dissatisfaction With Care in Labor and Birth

Unadjusted OR

4.73

Given an active say 11.46 Sometimes or not

at alllalways or in most cases

Access to information from caregivers 5.7

Wanted morelenough

Worse or much worse than expectedlas expected or better

Helpfulness of caregivers

Experience of pain 3.98

Some, little, or no helplvery or fairly helpful

Obstetric procedure score 2.24

2 5 1 4 Held baby soon after

birth 2.32 Nolyes

Unwanted people

Nolyes present at birth 2.97

Length of first stage of labor 2.01

Parity 1.69

Education 1.57

2 12 hrs/< 12 hrs

0,1+

< Yr 12lcompleted year 12

( C 4

(6.4-20.4)

(3.6-8.9)

(2.7-5.8)

(2.8-7.8)

(1.5-3.3)

(1.4-3.9)

(1.4-6.3)

(1.3-3.2)

(1.2-2.3)

(1.1-2.3)

P Adjusted OR (CI)

<0.001 7.35 (3.9-14.0)

<0.001 2.55 (1.5-4.5)

<0.001 3.13 (2.0-4.9)

<0.001 1.94 ( 1 .O-3.9)

<0.001 1.38 (0.9-2.2)

<0.001 1.54 (0.8-2.9)

0.004 2.22 (0.8-6.1)

0.004 1.33 (0.7-2.5)

0.006 0.86 (0.5-1.4)

0.016 1.56 (1 .O-2.4)

P

<0.001

<0.001

<0.001

0.05

0.20

0.20

0.12

0.36

0.55

0.05

10 BIRTH 21:l March 1994

Table 5. Adjusted Odds Ratios for Associations of Parity, Obstetric Procedure Score, Role in Decision Making, Access to Information, and Helpfulness of Caregivers With Dissatisfaction With Care in Labor and Birth

Adjusted Odds Ratio (CO P

Parity

Obstetric procedure 0,1+ 1.17 (0.8-1.8) 0.48

score 2 5 1 4 1.73 (1.1-2.7) 0.015

Given an active say

not at all/yes, in all or most

Sometimes or

cases 7.31 (4.0-13.5) <0.001 Access to

information from caregivers

Wanted morelenough 3.11 (1.8-5.3) <0.001

Helpfulness of caregivers

Some, little, or no helplvery or fairly helpful 1.94 (1.0-3.7) 0.04

score, dividing the sample into two groups accord- ing to whether the score was greater than or equal to 5 or less than 5 (the point closest to the median). The only social factors included were parity and education. Place of residence was excluded because of many missing values; age, income, and insurance status were excluded because they were not signifi- cantly associated with satisfaction in univariate analyses. One hundred forty-seven respondents were excluded from the analysis because of missing values for at least one item, leaving 543 respondents in the model. Those excluded were not more or less likely to be dissatisfied than women remaining in the analysis (x2 1.5, ldf, p = 0.22).

Of the 10 selected variables, only 5 remained sig- nificantly associated with dissatisfaction after ad- justing for all other factors: having an active say in decisions sometimes or not at all; wanting more in- formation; pain being worse than expected; care- givers being some, little, or no help; and not having completed secondary school.

To give a more precise estimate of the effect of these variables, two additional models were fitted. In the second model we entered parity and the inter- vention score, and then the 5 variables from the 10- factor model that had significant odds ratios after adjusting for all other factors. In this model, primi-

parity and a higher intervention score do not exert any additional influence on satisfaction after adjust- ing for the five other variables. The odds of dissatis- faction are raised eightfold by not having an active say (OR = 8.17 14.4-15.31, p < O.OOl), and three- fold by pain being worse than expected (OR = 3.13 [2.0-4.91, p < 0.001), and insufficient information

In the third model we deliberately excluded women’s rating of pain because of its strong associ- ation with intervention (x2 for linear trend 33.27, p < 0.0000). In this model (Table 5 ) , a higher score for intervention was associated with close to a twofold increase in dissatisfaction. Primiparity continued to exert no significant additional influence on dissatis- faction after adjusting for the other factors. To ex- plore further the relationship between reporting of pain and intervention, women’s rating of their intra- partum care and the weighted intervention score were stratified according to whether pain was worse than expected, as expected, or better. Among the group who reported experiencing worse pain than expected, an intervention score of 5 or higher was associated with a fourfold increase in dissatisfaction (OR = 3.90 [1.8-8.3]), indicating that in this stratum intervention has a highly significant effect.

(OR = 3.29 [2.0-5.51, p < 0.001).

Discussion

Much has been written about the difficulties and advantages of assessing satisfaction with care using postal questionnaires (16-20). The main advantages are that it enables whole populations to be sampled, and it is relatively inexpensive. Comparison of women’s reporting of events with medical records show that women accurately recall what happened during pregnancy, labor, and birth (20-23). Ques- tions that remain unanswered concern the difficulty of defining and measuring satisfaction in a meaning- ful way (24-31), especially in the context of child- birth. Asking women to rate their satisfaction with care when they may perceive that the alternative to the kinds of procedures they experienced would have entailed unacceptable risks to their unborn in- fant is highly contradictory (25). Shearer (25) and others (19,28,30,31) warned that precoded scales asking for an overall rating of care probably provide an underestimate of the extent of dissatisfaction with particular aspects of care. In this light our find- ing that one woman in three was dissatisfied with her intrapartum care, particularly when contrasted with the much smaller proportion of women who were negative about their antenatal care, raises seri- ous questions about the ability of maternity services

BIRTH 21:l March 1994 1 1

in Victoria to provide the care that women want during labor and birth.

What Do Women Want?

Our analysis suggests that access to information during labor, relationships to caregivers, the extent of involvement in decisions about their care, and exposure to intervention are critical to women’s satisfaction with care. The high odds of dissatisfac- tion associated with wanting more information, caregivers being of only some help or not at all help- ful, and not having an active say in decisions indi- cates that these factors may be best thought about as individual components of satisfaction. The fact that each remained in our first model after adjusting for other factors suggests that women may rate their care negatively for any one of these reasons. Our data lend only partial support to the conclusion of Green et a1 that primiparity is an additional factor leading to lower satisfaction (5,6). Parity alone did not exert a significant influence on any of the logis- tic models that we developed. For certain key vari- ables, however, such as involvement in decision making and exposure to intervention, primiparas were much more likely to be dissatisfied even when they had apparently similar experiences.

The greatest increase in odds of dissatisfaction in each of the models we developed was associated with not having an active say in making decisions about what happened in labor. Some obstetricians who saw a copy of the survey in the course of our obtaining approval for the study from hospital eth- ics committees suggested omitting this question from the questionnaire. Although we included “un- certain” as a category of response, it was con- tended that women would be unable to answer the question. Another view, with which we disagree, held that if women completed the question, their responses would be biased because of its wording. These responses from practitioners highlight the de- gree of tension surrounding the issue of women’s participation in decisions regarding their care-ten- sions that from our data clearly exist for women as well as obstetricians.

Women might have meant many things when they chose among the five options concerning their participation in decisions regarding care. Having an active say might mean being given an opportunity to choose among options, being able to decide when certain action will be taken, being given information as to why decisions were made, having the right of veto, or all of these depending on the circum- stances. Green et a1 demonstrated the importance to a British population of a related concept, feeling

in control, to satisfaction, feeling of fulfillment, and postnatal emotional well-being (5,6). Not having an active say was similarly associated with less emo- tional well-being in our study (Astbury, Brown, Lumley, Small, submitted for publication). These issues require further research, in particular to ex- plore what women from different backgrounds un- derstand by the phrases “having an active say” and “feeling in control,” and what factors contribute to good and bad experiences in individual settings.

Intervention-At What Social Cost to Women?

The physiologic risks associated with many obstet- ric procedures are well documented (32), but less is known about the social and psychological conse- quences of obstetric intervention for women (29). This paper reports an association between interven- tion and dissatisfaction. Elsewhere we noted that women completing the survey who scored as poten- tially clinically depressed according to the Edin- burgh Postnatal Depression Scale were more likely than women not scoring as depressed to have had an operative delivery and a higher weighted score for obstetric procedures (Astbury, Brown, Lumley, Small, submitted for publication).

Considering the extremely high rates of interven- tion in the United States, Canada, and Australia, few studies have looked in a sophisticated way at what intervention means for women in terms of their satisfaction with the care they received, or their overall satisfaction with their experience of giving birth. Of the studies published, most focus on satisfaction with the birth itself (1,2,5-7,33,34). With one exception (7), all agreed that women were more negative about the birth when they experi- enced intervention. Satisfaction with care was in- vestigated by two British researchers using the same question asking for a global rating of intrapar- tum care as used in this study (4,8). Both report an association between exposure to intervention and dissatisfaction.

An important set of questions relates to whether the dissatisfaction accompanying more intense in- tervention is an (unavoidable?) correlate of a more complicated and therefore more painful labor, or whether it has more to do with aspects of the proce- dures themselves or the ways in which care may be altered by their use. The survey offers several clues in answer to these questions. First, the fact that having a longer labor was not associated with dis- satisfaction among primiparas suggests that the rea- sons why intervention is associated with dissatis- faction in this group are not solely related to having more complications in labor. Second, the multivari-

12 BIRTH 21:l March 1994

ate analysis confirms that many factors, and not just the amount of pain experienced, determine wom- en’s overall satisfaction with care. The five vari- ables entered into our original model that remained significantly associated with dissatisfaction (not having an active say in decisions, wanting more in- formation, caregivers not being helpful, pain being worse than expected, not having completed second- ary school) demonstrate the importance of relation- ships with caregivers. Although intervention did not add to this model, it remained significant when pain was left out of our third model. This result, together with the finding that women who rated their pain as worse than expected were much more likely to be dissatisfied if they experienced greater intervention, suggests that in the Australian context, intervention per se is important to women’s overall satisfaction regardless of their relationships with caregivers, ac- cess to information, and role in decision making. The fact that intervention carries with it an in- creased risk of dissatisfaction regardless of these factors adds further weight to an already convincing body of evidence for questioning current patterns of obstetric intervention (32).

Did the Survey Injuence Policy Development?

The aim of all research on women’s satisfaction with maternity care ultimately must be to influence what happens in practice settings. This study was unusual because it was directly linked to a process of policy review.

Having access to a reliable source of data from a largely representative sample of Victoria women enabled us to counter several stereotypes about what makes women dissatisfied, and about how rep- resentative were views expressed to us in public meetings and submissions. The circulation of stere- otypes about women who are dissatisfied with their care often operates to discredit critical feedback when it is given (35,36). The survey did not substan- tiate the idea, widely promoted by professional as- sociations during the review process, that mainly middle-class and more highly educated women (driven by “unrealistic expectations” of childbirth) are critical of maternity care. Women of higher so- cioeconomic status may have been more likely to attend public meetings and contribute written sub- missions, but in response to the survey they were not more likely to be dissatisfied than women with less formal education or lower total family income.

Although individual practitioners took issue with the survey methodology and its findings, no mem- bers of the review team or any of the professional practitioner bodies were openly critical of the evi-

dence on which the report was based. This experi- ence is in direct contrast to recent events in the United Kingdom related to the House of Commons Health Committee Report on Maternity Services (37,38).

The survey findings also influenced the delibera- tions of the review team by providing a powerful indicator that the time when women encounter diffi- culties in relation to their care tends to be during labor and birth rather than during pregnancy. The far greater dissatisfaction of women with intrapar- tum care compared with antenatal care encouraged the review team to tailor recommendations for achieving improvements in this area. This feedback is reflected in the guiding principles accompanying the recommendations regarding models of care, and underlined the priority given to models providing continuity of care, such as team midwifery, multi- disciplinary team care, birth center care, and com- munity-based midwifery.

Our findings reinforce the arguments contained in the Ministerial Review of Birthing Services that Australia’s current mix of private and public ser- vices leaves a substantial minority of women un- happy with aspects of care. The question of whether alternative models such as those suggested in the review would lead to better social outcomes for mixed populations of women requires further research. Although we are still some way from this point in Australia, acceptance is growing among bu- reaucrats and health service providers that surveys of health service users can be a useful policy tool.

Acknowledgments

We are grateful to the Victoria women who com- pleted the survey; the hospitals and home birth practitioners who assisted in distributing the ques- tionnaires; Ann Cartwright and the Institute for So- cial Studies in Medical Care for permission to adapt the institute’s national postal questionnaire on satis- faction with maternity care for use in the Victorian context; Judith Yates for assistance with prelimi- nary data analysis; Lyn Watson for statistical ad- vice; and Rhonda Small for providing comments on an earlier draft of this paper.

References

1 .

2.

Cartwright A. The Dignity of Labor? A Study of Childbear- ing and Induction. London: Tavistock, 1979. Sullivan D, Beeman R. Satisfaction with maternity care: A matter of communication and choice. Med Care 1982;10(3):321-330.

13 BIRTH 21:l March 1994

istration Press, 1991.

3

4.

5 .

6.

7.

8.

9.

10.

1 1 .

12.

13.

14.

15.

16.

17.

18.

19.

Obstetrics ’87. A report of the Canadian Medical Associa- tion on obstetrical care in Canada. Can Med Assoc J 1987; 136(suppl): 1-36. Jacoby A. Women’s preferences for and satisfaction with current procedures in childbirth: Findings from a national study. Midwifery 1987;3:117-124. Green J , Coupland V, Kitzinger J. Great Expectations: A Prospective Study of Women’s Expectations and Experi- ences of Childbirth. Cambridge: Child Care and Develop- ment Group, University of Cambridge Press, 1988. Green J, Coupland V, Kitzinger J. Expectations, experi- ences, and psychological outcomes of childbirth: A pro- spective study of 825 women. Birth 1990;17(1):15-24. Seguin L, Therrien R, Champagne F, Larouche D. The components of women’s satisfaction with maternity care. Birth 1989;16(3): 109-1 13. Martin C. How do you count maternal satisfaction? A user- commissioned survey of maternity services. In: Roberts M, ed. Women’s Health Counts. London: Routledge,

Fleissig A. Mothers’ experiences of induction of labor. J Obstet Gynecol 1991 ; 1 l(suppl):5 1 1-5 15. Lumley J, Small R, Yelland J. Having a Baby in Victoria. Final Report of the Ministerial Review of Birthing Services in Victoria. Australia: Health Department Victoria, 1990. Brown S, Lumley J. Antenatal care: A case of the inverse care law? Aust J Public Health 1993;17(2):95-103. Cartwright A. Who responds to postal questionnaires? J Epidemiol Community Health 1986;40:267-273. Cartwright A. Some experiments with factors that might affect the response of mothers to a postal questionnaire. Stat Med 1986;5:607-617. Cartwright A. Interviews or postal questionnaires? Com- parisons of data about women’s experiences with maternity services. Milbank Q 1988;66(1): 172-189. Elliot S, Anderson M, Brough D, Watson J, Rugg A. Rela- tionship between obstetric outcome and psychological mea- sures in pregnancy and the postnatal year. J Reprod Infant

Riley E. What do women want? The question of choice in the conduct of labor. In: Chard T, Richards M, eds. Bene- fits and Hazards of the New Obstetrics. London: Heine- mann, 1977:62-71. Jacoby A, Cartwright A. Finding out about the views and experiences of maternity service users. In: Garcia J, Kilpa- trick R, Richards M, eds. The Politics of Maternity Care: Services for Childbearing Women in Twentieth Century Britain. Oxford: Oxford University Press, 1990:238-255. Mason V. Women’s Experience of Maternity Care-A Sur- vey Manual. London: Her Majesty’s Stationery Office, 1989. Strasser S, Davis R. Measuring Patient Satisfaction for Improved Patient Services. Ann Arbor, MI: Health Admin-

1990:147-166.

Psycho1 1984;2: 18-22.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29,

30.

31.

32.

33.

34.

35.

36.

37.

38.

Martin C. Monitoring maternity services by postal ques- tionnaire: Congruity between mothers’ reports and their ob- stetric records. Stat Med 1987;6:613-627. Hewson D, Bennett A. Childbirth research data: Medical records or women’s reports? A m J Epidemiol

Simkin P. Just another day in a woman’s life? 11. Nature and consistency of women’s long-term memories of their first birth experiences. Birth 1992;19(2):64-81. Githens PB, Glass CA, Sloan FA, Entman SS. Maternal recall and medical records: An examination of events dur- ing pregnancy, childbirth, and early infancy. Birth

Lumley J. Assessing satisfaction with childbirth. Birth

Shearer M. The difficulty of defining and measuring satis- faction with perinatal care. Birth 1983;10(2):77. Porter M, Macintyre S. What is must be best: A research note on conservative or deferential responses to antenatal care provision. SOC Sci Med 1984;19(11): 1197-1200. Lomas J, Dore S, Enkin M, Mitchell A. The labor and delivery satisfaction index: The development and evalua- tion of a soft outcome measure. Birth 1987;14(3):125-129. Shearer M. Commentary: How well does the LADS1 mea- sure satisfaction with labor and delivery? Birth

Oakley A. Social consequences of obstetric technology: The importance of measuring “soft” outcomes. Birth

Locker D, Dunt D. Theoretical and methodological issues in sociological studies of consumer satisfaction with medi- cal care. SOC Sci Med 1978;12:283-292. Bramadat I, Dreidger M. Satisfaction with childbirth: Theo- ries and methods of measurement. Birth 1993;20(1):22-29. Chalmers I, Enkin M, Keirse M, eds. Effective Cure in Pregnancy and Childbirth. Oxford: Oxford University Press, 1989. Morgan B, Bulpitt C, Clifton P, Lewis P. Analgesia and satisfaction in childbirth (the Queen Charlotte’s 1000 mother survey). Lancet 1982;2:808-811. Hewson D, Bennett A, Holliday S , Booker E. Childbirth in Sydney teaching hospitals: A study of low-risk primiparous women. Community Health Stud 1985;10(3):195-201. Green J , Kitzinger J, Coupland V. Stereotypes of childbear- ing women: A look at some evidence. Midwifery

Bowler I. Stereotypes of women of Asian descent in mid- wifery: Some evidence. Midwifery 1993;9:7-16. Steer P. The House of Commons Health Committee Report on the Maternity Services. A personal view. Br J Obstet Gynaecol 1992;99:445-451. Lumley J, Brown S. The House of Commons Health Com- mittee Report on the Maternity Services. A personal view [letter]. Br J Obstet Gynaecol 1993;lOO: 193.

1987; 125(3):484-491.

1993 ;20(3): 136- 14 1 .

1985; I2(3): 141-145.

1987 ; 14(3): 125-129.

I983 ; 10(2):99- 108.

1990;6:125-132.