Overdue Choices: How Information and Role in Decision-Making Influence Women's Preferences for Induction for Prolonged Pregnancy

Overdue Choices: How Informationand Role in Decision-Making InfluenceWomen’s Preferences for Induction

for Prolonged Pregnancy

Gabrielle Stevens, BPsycholSc (Hons),and Yvette D. Miller, BA Psych (Hons), PhD

ABSTRACT: Background: Enabling women to make informed decisions is a crucial componentof consumer-focused maternity care. Current evidence suggests that health care practitioners’communication of care options may not facilitate patient involvement in decision-making. Theaim of this study was to investigate the effect of specific variations in health caregivercommunication on women’s preferences for induction of labor for prolonged pregnancy.Methods: A convenience sample of 595 female participants read a hypothetical scenario inwhich an obstetrician discusses induction of labor with a pregnant woman. Information providedon induction and the degree of encouragement for the woman’s involvement in decision-makingwas manipulated to create four experimental conditions. Participants indicated preference withrespect to induction, their perceptions of the quality of information received, and other potentialmoderating factors. Results: Participants who received information that was directive in favor ofmedical intervention were significantly more likely to prefer induction than those givennondirective information. No effect of level of involvement in decision-making was found.Participants’ general trust in doctors moderated the relationship between health caregivercommunication and preferences for induction, such that the influence of information provided onpreferences for induction differed across levels of involvement in decision-making for womenwith a low trust in doctors, but not for those with high trust. Many women were not aware of thelevel of information required to make an informed decision. Conclusions: Our findings highlightthe potential value of strategies such as patient decision aids and health care professionaleducation to improve the quality of information available to women and their capacity forinformed decision-making during pregnancy and birth. (BIRTH 39:3 September 2012)

Key words: consumer-focused maternity care, health care practitioner communication,induction of labor, informed decision-making

It is now widely accepted that effective health caredecision-making requires involvement by both patientsand their health care practitioners, each of whom hasunique knowledge and skills to contribute to thedecision-making process (1).

For patients, effective health care decision-makinginvolves weighing up and considering how they valueeach existing care alternative and having as great a rolein the decision-making process as they desire (2).Patient involvement in this process can range from

Gabrielle Stevens is a Doctoral candidate and Yvette D. Miller isDeputy Director, at the Queensland Centre for Mothers & Babies,School of Psychology, The University of Queensland, Brisbane,Queensland, Australia.

Address correspondence to Gabrielle Stevens, BPsycholSc (Hons),School of Psychology, The University of Queensland, Brisbane,Queensland 4072, Australia.

Accepted January 23, 2012

© 2012, Copyright the AuthorsJournal compilation © 2012, Wiley Periodicals, Inc.

BIRTH 39:3 September 2012 1

shared decision-making, in which equal patient/care-giver partnerships exist, to an informed decision-making model, where patients take on a greater portionof decisional responsibility (3). Health care practitio-ners can facilitate effective patient involvement by pro-viding quality, unbiased information relevant to allhealth care options, and by actively supporting andassisting patients’ preferred level of involvement indecision-making (4, 5).

Several population-level surveys of maternity careconsumers in Australia, the United Kingdom, and theUnited States have demonstrated that women undergo-ing obstetric procedures and treatments are often poorlyinformed, agree with practitioners’ predetermineddecisions, or both, rather than having an active say inthe care they receive (6–8). In one such Australian sur-vey, rates of informed decision-making varied consider-ably by type of procedure, with informed decisionsreportedly being made by 50 percent of women for aplanned cesarean delivery, 20 percent for an unplannedcesarean, and 30 percent for an induction of labor (7).

Caregiver behavior that may act as a barrier towomen’s involvement in maternity care decision-making and its effect on decision outcomes has beendocumented in several studies (9–11). Providing qualityand unbiased information about all available optionsand their associated outcomes is a crucial step in allow-ing women to make informed maternity care decisions(4, 5, 12). However, caregivers often use directivecommunication methods that either downplay the risksof medical interventions or emphasize the risks of alter-natives, biasing women’s decisions in favor of medicalintervention by positioning it as the only sensibleoption (9). For example, when asked to recall informa-tion given by their practitioner about the risks and ben-efits of cesarean delivery, women who report receivingless information about the risks are more likely toprefer to have a cesarean birth than those who receivemore information (10).

The caregiver’s encouragement of a woman’sinvolvement in decision-making is also likely to influ-ence decisions. Many women want to be involved inpregnancy-related decisions and are significantly moresatisfied with their care when their involvement is high(13). The use of decision aids to facilitate patientinvolvement in decision-making through nondirectiveand comprehensive information delivery has also beenshown to reduce the likelihood of patients choosinginvasive elective surgery (14). Nevertheless, outcomesare affected more by the orientation of a particularpractitioner or setting than by characteristics or prefer-ences of the individual patient (9). Even when providedwith adequate information that could support shareddecision-making, patients report feeling coerced intotreatment options that are their doctor’s preference and

find it difficult to express their own preferences ifcontrary to those of their practitioner (9).

Despite descriptive evidence for less-than-optimalinvolvement by women in decision-making and practi-tioner communication in pregnancy, labor, and birth,little is known about the potential large-scale effect ofspecific features of practitioner communication onpatient decisions. This study experimentally tested howspecific variations in practitioners’ communicationinfluenced women’s preferences for induction of laborfor prolonged pregnancy. The decision about whetheror not to have an induction of labor for a prolongedpregnancy is “preference-sensitive” (15), as it dependson the degree to which individual women value thevarious outcomes associated with induction, and itsalternative, expectant management (16, 17). Specifi-cally, we aimed to test the effect of, first, variations ininformation provision about the possible outcomes ofinduction, and second, encouragement for women’srole in the decision-making process, on women’spreferences for induction of labor. We hypothesizedthat communication not conducive to a shared decision-making model (i.e., directive information provision andnot encouraging the woman’s involvement) wouldincrease uptake of the proposed procedure. A second-ary aim was to determine the moderating effects ofwomen’s characteristics that may increase or decreasevulnerability to variations in practitioner communica-tion, including prior knowledge about induction oflabor, perceived risk of childbirth, and general trust indoctors. The relative importance of all manipulated andmeasured variables at predicting preferences for induc-tion, and perceived adequacy of information receivedon induction, were also assessed.

Methods

Design

We used a 2 (information) 9 2 (role in decision-making) experimental design, in which informationdelivery was either nondirective or directive, and therole in decision-making was characterized by thewoman’s choice or her compliance. Combinations ofthese variables resulted in four conditions (Table 1),and the primary outcome was women’s preferencesfor induction of labor. The study was approved bythe relevant university ethics review board.

Procedure

Participants were recruited by means of e-mail totake part in an online questionnaire about “Women’s

2 BIRTH 39:3 September 2012

Preferences for Labour and Birth.” E-mail addresseswere sourced from a database of women who hadagreed to be contacted for future research aboutmaternity care, various listservs, and website forums.Participants were eligible to participate if they werefemale, at least 18 years of age, and not currentlypregnant.

Women who consented to take part were askedquestions assessing demographic characteristics andprevious induction of labor knowledge, then providedwith one of four vignettes (allocated via computerrandomization) telling the story of a fictional couplewho were pregnant with their first child, and attend-ing their 41-week obstetrician appointment. Thecouple are offered an induction of labor at between41 and 42 weeks’ gestation, and are presented withadditional induction information by their obstetricianthat varied according to the experimental condition(see Table 1 and Appendix A for specific scenariodescriptions).

Information (nondirective vs directive) was manipu-lated by varying the degree of bias in communicatinginformation about possible outcomes of induction. Indirective information conditions, only information thathighlighted possible benefits of induction was pre-sented, including statements communicating anincrease in perinatal mortality after 42 weeks’ gesta-tion (18) and the reduced rate of meconium aspira-tion syndrome in pregnancies routinely inducedbetween 41 and 42 weeks’ gestation (19). In non-directive information conditions, the following infor-mation was presented in addition to the abovementionedpotential benefits: 1) quantification of estimatedoccurrence (using percent rates) of perinatal mortalityif women are not routinely induced (19); 2) the

accuracy of due date estimation (20) and the extentof normal variation from estimates; 3) more intensecontractions (21), higher rate of epidural use inwomen who undergo an induction (22), and possibleimplications of epidural use (23); 4) likelihood ofhaving continuous electronic fetal monitoring (24)and its possible implications (25, 26); and 5)potential implications of failed induction (i.e., emer-gency cesarean delivery) (24).

Role in decision-making (choice vs compliance)was manipulated by varying the degree to which theparticipant was provided with an opportunity to beinvolved in the decision-making process about havingor not having an induction of labor. In “choice condi-tions” the obstetrician provided an alternative manage-ment to induction (expectant management or “watchand wait” approach) and encouragement to consideravailable options. In “compliance” conditions theobstetrician provided no alternative option to induc-tion, and asked for consent to arrange an induction(Table 1).

After presentation of the vignette, participants wereasked questions about preference for induction onbehalf of the woman in the scenario, perceptions ofinformation quality, and their views on medicalizationof childbirth and trust in doctors.

On completion, all participants were debriefed on themain research aims, and were provided with the infor-mation included in the nondirective choice condition(the most complete set of information). Participantswere also provided with contact information for furthersupport services if they experienced distress as a resultof their participation. An external link was provided ifthey were interested in entering a prize draw to thankthem for their contribution.

Table 1. Health Practitioner Communication in Each Experimental Condition

Experimental Condition

Scenario CharacteristicsNondirectiveChoice

NondirectiveCompliance

DirectiveChoice

DirectiveCompliance

Information

Information on thepossible benefitsof induction

✓ ✓ ✓ ✓

Information on thepossible risksof induction

✓ ✓ ✕ ✕

Role in decision-making

Encouragement toconsider bothinduction andalternative

✓ ✕ ✓ ✕


Measures

Preferences for induction of labor

Participants were asked, “Knowing what you knownow, if you were in Jessica’s position, would you havean induction of labour in the next day or two?” (Yes/No).

Prior knowledge, views on the medicalizationof childbirth, and trust in doctors

Prior knowledge on induction was measured by thesingle, open-ended question “What do you knowabout induction of labour? Please type any informa-tion or opinions you may have in the box below.”Responses were coded as either “no prior knowledgeor opinions” or “any degree of knowledge oropinion.”

Participants’ views on the medicalization of preg-nancy and childbirth (perceived risk and acceptanceof technology) were measured using items originallycreated by McClain (27) and later adapted byHowell-White (28). This 6-item scale has beenpreviously shown to have adequate internal consis-tency (alpha = 0.71). The current study used anadapted version of Howell-White’s scale, whichcontained 5 items (e.g., It is best for first time moth-ers to have obstetricians and a hospital birth), towhich participants responded on a 5-point Likertscale (“1 = Strongly disagree” to “5 = Stronglyagree”). This 5-item scale had high internal consis-tency (alpha = 0.86). Responses were averaged foreach participant, resulting in a possible overall scoreof 1–5. Scale scores were categorized into twogroups: those with a medicalized view of childbirth(average score of 1.00–3.00) and those with a non-medicalized view of childbirth (average scale scoreof 3.01–5.00). Scale items were randomly presentedto prevent selective item response bias attributable toorder of presentation.

Participants’ level of trust in doctors was measuredon an 11-item scale created by Hall et al (e.g., Some-times doctors care more about what is convenientfor them than about their patients’ medical needs)(29). This scale was previously shown to have highinternal consistency (alpha = 0.89), which was main-tained in the current study sample (alpha = 0.89).Participants responded to items on a 5-point Likertscale (“1 = Strongly disagree” to “5 = Stronglyagree”), and their individual responses were averagedto create a single scale score (range of 1–5). Overallscores were categorized into two groups: those withlow trust in doctors (average score of 1.00–3.00) andthose with high trust in doctors (average score of

3.01–5.00). Scale items were presented in randomorder.

Awareness of information quantity

Participants were asked “Do you believe Jessica andher husband had enough information regarding induc-tion of labour to make an informed decision?” If theyanswered “no,” participants were asked to suggest addi-tional information that they believe should have beenprovided.

Demographic information. Participants were asked toreport demographic characteristics (age, country ofbirth, language, Aboriginal or Torres Strait Islander sta-tus, highest level of education), using items adaptedfrom Australian Bureau of Statistics census (30). Partic-ipants were also asked whether they had previouslygiven birth, and if so, whether they had experienced aninduction of labor.

Data Analysis

Participants’ demographic details were compared withthose of all women who gave birth in Australia in2007 (31) to determine the generalizability of thestudy sample. Binary logistic regression analyses wereconducted to assess the individual and interactiveeffects of information (nondirective vs directive) androle in decision-making (choice vs compliance) onparticipants’ preferences for induction of labor. Foreach potential moderator (prior knowledge, views onchildbirth medicalization, and trust in doctors), allvariables (information, role in decision-making, andmoderator variable) were entered simultaneously intoa logistic regression model to test for any significantinteraction effects. Alpha was set to 0.05 for allanalyses.

Results

Sample

The final sample of 595 participants included womenaged between 19 and 71 years. Of these women, 79percent were Australian born and 98 percent reportedspeaking the English language at home. Most partici-pants (79%) had previously given birth, 39 percent ofwhom had experienced an induction of labor. Thestudy sample was significantly older, less likely to beAboriginal or Torres Strait Islander, and more likely tohave experienced a previous birth than all birthingwomen in Australia (Table 2).


Preferences for Induction of Labor

Overall, 32 percent of participants preferred inductionof labor. Similar rates of preference for induction werefound between participants with and without at leastone previous birth, and between participants who hadand had not previously been induced. Preferences forinduction by experimental condition are detailed inTable 3. Together, information and role in decision-making significantly predicted preferences for inductionof labor (Cox & Snell R² = 0.07, v² [2] = 42.38,p < 0.001).

Information (nondirective vs directive) significantlypredicted preferences for induction of labor. The oddsof preferring induction were significantly greater forthose who received directive information (44.4%) thanfor those who received nondirective information(19.9%) (Fig. 1, Table 4). Role in decision-making(choice vs compliance) did not significantly predictpreferences for induction of labor, and there was nosignificant interaction between information and role indecision-making (Fig. 1).

Prior Knowledge, Views on the Medicalizationof Childbirth, and Trust in Doctors

Prior knowledge

Most respondents (n = 433, 72.8%) reported priorknowledge on induction. There was no significant maineffect of prior knowledge and no significant interactionamong prior knowledge, information, and role indecision-making (Table 4).

Views on the medicalization of childbirth

Most participants (n = 353, 61.2%) had a medicalizedview of childbirth, and they were significantly morelikely to prefer induction than those with nonmedical-

ized views of childbirth (Table 4). No significantinteraction effect was found in participants’ views onthe medicalization of childbirth, information, and rolein decision-making (Table 4).

Trust in doctors

Most participants (n = 425, 72.6%) had low trust indoctors, and they were significantly less likely to preferinduction than those with high trust in doctors(Table 4). A significant three-way interaction wasfound among trust in doctors, information, and role indecision-making (Table 4). For those with low trust indoctors, directive information increased the likelihoodof preference for induction compared with nondirectiveinformation for those in the compliance conditions(OR = 5.35, 95% CI = 2.56–11.17, p < 0.001), but notfor those in the choice conditions (OR = 1.75, 95%CI = 0.83–3.72, p = 0.145). For participants with hightrust in doctors, directive information resulted in signif-icantly greater preferences for induction than nondirec-tive information, and this significant difference existedfor both the compliance (OR = 5.60, 95% CI = 2.02–15.52, p = 0.001) and the choice conditions (OR =16.00, 95% CI = 4.27–59.93, p < 0.001) (Fig. 2 a and b).

Relative Contribution of Variables

When all variables with significant main effects wereentered into a multivariate logistic regression model,information, views on the medicalization of childbirth,and trust in doctors were all significant and relativelyequal predictors of preference for induction (Table 4).

Awareness of information quantity

Overall, 37 percent of participants believed that thewoman had been provided with enough information oninduction. Of those given directive information, 22.3

Table 2. Comparison of Study Sample with all Australian Birthing Women in 2007

DemographicVariable

CurrentSample

(n = 595)

Australian BirthingWomen in 2007(n = 289,496)‡ Test of Differences

Age (yr) (mean) 32.6 29.9 t (592) = 9.09, p < 0.001*

(range) 19–71 <15–56

ATSI status 0.8% 3.8% v2 (1) = 14.15, p < 0.001†

Australian born 78.8% 75.2% v2 (1) = 2.97, p = 0.085*

At least oneprevious birth

79.3% 58.4% v2 (1) = 111.15, p < 0.001*

*t test; †chi-square test; ‡Australia’s Mothers and Babies 2007 (31).ATSI = Aboriginal or Torres Strait Islander.


percent believed that enough information had been pro-vided. Among those given nondirective information,51.5 percent believed enough information had beenprovided.

Information Women Want to Receive

Many women made detailed suggestions about infor-mation on induction that they thought should have beenprovided, most of which were consistent with notionsof informed decision-making, including more detailedinformation on the risks and benefits of the procedureand possible alternatives. Some women also com-mented that they would like to have received informa-tion on less medicalized induction methods, regardlessof their proven effectiveness:

Jessica and her husband could have been informed about otherless invasive options such as acupuncture, increased safephysical activity, eating a hot curry, etc. Whether they are

proven or not to be successful, as long as they are not harm-ful, why not give them a go? (45-year-old woman, previousbirth/s)

Should be encouraged to consider alternatives, not necessarilygiven detailed information about the alternatives, but at leastbe told there are options that some people consider, i.e.,acupuncture, homeopathy, massage, emotional balancing, coun-selling, doula services. (29-year-old woman, previous birth/s)

Some participants who had personally experiencedinduction commented on the lack of information andchoice they had received themselves, suggesting thatsome women experience care that does not prioritizeinformed decision-making about induction of labor:

Unfortunately, I had my ‘membranes stripped’ without beingtold why or given any information. It was only afterwards thathe mentioned the name of what he had done (I looked it up) –and that I could go into labour soon (which I did). (43-year-old woman, previous induction/s)

My first (and currently only) child was induced. I know every-thing now, but didn’t have much of an idea before it actuallyhappened. I believe in certain circumstances it is very neces-sary, but more unbiased information needs to be available.Obstetricians need to be more forthcoming with informationespecially to first-time mothers instead of waiting for us toask the questions – it is hard to ask a question about some-thing that is so foreign to you. (30-year-old woman, previousinduction/s)

Discussion

Several studies have described patient involvement inmaternity care decision-making, yet experimental evi-dence for the role of specific aspects of the practi-tioner’s communication in women’s decision-makinghas been lacking. We found that women who weregiven directive information in favor of medical inter-vention (induction) were more likely to prefer the pro-cedure than women presented with comparativelynondirective information. These effects were consistentregardless of women’s prior knowledge of induction ortheir own views on the medicalization of childbirth.The capacity for women to make informed decisions iscompromised when they are given directive (and there-fore partial) information. In such instances, decisionoutcomes may be inconsistent with what women valueand have been led to expect of their birth experience,which is undesirable in any health scenario.

Adding to the effects of information type on prefer-ences for induction, 22 percent of the study participantswho received directive information believed that thewoman had been given enough information to make aninformed decision. These findings suggest that somewomen are unaware of what constitutes ideal nondirective

0%

20%

40%

60%

80%

100%

Choice Compiance

Pref

er In

duct

ion

(%)

Role in Decision-making

Nondirective InformationDirective Information

Fig. 1. Preferences in favor of induction by informationprovision and role in decision-making.

Table 3. Preferences for Induction by ExperimentalCondition

Experimental Condition No.Preferred

Induction (%)

Nondirective choice 150 33 (22.00)

Nondirective compliance 151 27 (17.88)

Directive choice 148 60 (40.54)

Directive compliance 146 71 (48.63)

Total 595 191 (32.10)


information provision. For these women, directive infor-mation provision by health caregivers may go undetectedand be particularly influential on decision-making.

Women with low trust in doctors were less influ-enced by directive practitioner communication thanwomen with high trust in doctors. Although low trustmay be a protective factor in the event of biased practi-tioner communication, it is also likely to be detrimentalto women’s future experiences and may lead to nega-tive interactions with health care practitioners (32).Furthermore, although less affected by directive infor-mation overall, it is interesting that women with a lowtrust in doctors who received directive informationwere susceptible to the effects of compliance-seekingapproaches to decision-making. Further research toinvestigate how women’s trust in their specific mater-nity practitioner influences decision-making and howpractitioner communication that facilitates informeddecision-making may positively influence levels ofwomen’s trust would be useful.

We were not only able to determine the independenteffects of information provision, but also the relativeimportance of both patient and health system variablesin predicting preferences for induction. Women’s viewson childbirth medicalization and their general trust indoctors (patient preferences) significantly predictedpreferences for induction after accounting for informa-tion provision, and to an equal extent. A pregnantwoman’s acceptance of technology in pregnancy and

childbirth and her degree of confidence in the medicalprofession are therefore likely to influence her decisionstoward medical intervention. Moreover, these factorsare likely to influence a woman’s decision to a similarextent as the information that is communicated to herabout her options and the possible consequences ofavailable decisions. Importantly, our findings suggestthat variations in the way information is communicatedby health practitioners can influence such decisions,even after personal preferences are accounted for.

Although participants were recruited from a varietyof sources, participation in this study was limited byonline recruitment methods, thus restricting the sampleto women with Internet access. Together with the inclu-sion of women beyond childbearing age, this biasreduces the generalizability of findings to the widerpopulation of childbearing women.

Furthermore, whereas manipulating health practi-tioner communication experimentally allowed us to iso-late the influence of specific variables, it also limits theconclusions that can be drawn from our findings.Effects of directive information provision on women’spreferences for intervention in the real world may differfrom the findings of this study, especially if the infor-mation that they receive is directive in more than oneway (e.g., selective provision of outcome informationor emphasizing one care option as more desirable thananother), or other external influencing factors arepresent (e.g., social and/or emotional factors).

Table 4. Predicting Preferences for Induction: Multiple Analyses

EffectsOdds Ratio

(OR)

95%ConfidenceInterval

WaldChi-Square p

Univariate main effects

Information 3.20 2.22–4.61 – <0.001

Role in decision-making 1.10 0.77–1.60 – n.s.

Prior knowledge 2.67 1.26–5.68 – 0.01

Views on childbirthmedicalization

13.36 5.82–30.67 – <0.001

Trust in doctors 6.87 2.88–16.39 – <0.001

Multivariate main effects

Information 5.21 3.21–8.44 <0.001

Prior knowledge 0.82 0.50–1.33 n.s.


8.02 4.96–12.95 – <0.001

Trust in doctors 4.47 2.72–7.37 – <0.001

Interaction effects

Prior knowledge – – 0.51 n.s.


– – 0.13 n.s.

Trust in doctors – – 4.63 0.031


Our finding that variations in level of involvement indecision-making did not influence women’s preferencesmay have been due to the inability for this variable tobe successfully manipulated in a hypothetical context.Future research should examine what directive informa-tion provision and compliance can look like in clinicalpractice when not experimentally standardized, toidentify aspects of health practitioner communicationthat most influence women’s decisions.

It is important to acknowledge that in order for thevignettes to be perceived as conversational (i.e., haveadequate face validity for clinical conversations) and tominimize the burden of participants reading verylengthy text, even the study condition with the mostcomplete set of information (nondirective choice) didnot satisfy universal standards for information provisionto enable informed decision-making (33). Qualitativefindings highlighted many women’s awareness of suchlimitations in information provision and suggested thatadditional information should be included, such asalternative methods of induction and additional riskoutcome information.

Clinical Implications and Future Directions

Our main finding, that variations in caregiver communi-cation influenced women’s preferences for induction,suggests that policies and routine practices in informationprovision among different practitioners may partiallyexplain differences in rates of induction for prolongedpregnancy in actual practice. We found that regardlessof women’s preferences, and even after accounting forthe significant effects of views on childbirth medical-ization and trust in doctors, there was a significant andcomparable effect of the way information was commu-nicated on preference for induction for prolonged preg-nancy. Although somewhat lacking in ecologicalvalidity, our experimental approach can be creditedfor allowing us to control for individual patient factorsto highlight the potential influence of caregivercommunication.

It would be valuable to employ a similar experimen-tal approach to investigate other preference-sensitivedecisions women may face during pregnancy and birth,such as cesarean delivery (elective/planned and after aprevious cesarean) and pain management during labor.This approach would not only add to professionalknowledge of how women make important health deci-sions during pregnancy, but may also encourage thedevelopment of further research on improving patient–caregiver communication in a wide range of clinicalsettings.

In line with this approach, it is widely accepted thatstrategies that can improve the quality of informationprovided to childbearing women are crucial to enabletheir involvement in maternity care decision-making.Patient decision aids are one strategy shown to beeffective for increasing consumer knowledge andreducing consumer decisional conflict and passivity indecision-making (14). Decision aids for specific mater-nity care decisions have already been developed andsome are publicly available, including a decision aidfor induction of labor (34). Our qualitative findingsprovide further support for the need of such tools, bycontributing to a consistently emerging finding thatwomen desire more nondirective and comprehensiveinformation on induction (6). However, both practicaland ideological barriers to their effective use (35, 36)highlight the need for adaptations and implementationstrategies that can better integrate these tools intoroutine clinical practice.

Conclusions

This study identified the potential effects of one formof biased communication, directive information provi-sion, on women’s preferences for induction for pro-

(a)

(b)

Pref

er In

duct

ion

(%)

0%

20%

40%

60%

80%

100%

Choice Compliance

Nondirective Information

Directive Information

0%

20%

40%

60%

80%

100%

Choice Compliance

Role in Decision-making

Fig. 2. The association between information provisionand role in decision-making for those with (a) low and(b) high trust in doctors.


longed pregnancy. In addition, many women appear tobe unaware of the information necessary for informeddecision-making to take place, suggesting that poorcommunication in real-life situations may be moreprevalent and perhaps more problematic than has beenpreviously thought. Together, these results highlight theneed for continued development of strategies tofacilitate women’s effective involvement in decision-making in maternity care settings. In particular, thesefindings provide a focus for improving one specificaspect of health caregiver communication—nondirec-tive information provision—to enable women to makeinformed decisions about their care in pregnancy, labor,and birth. Both women and their caregivers will benefitfrom consultations and care in which mutual trust andrespect is paramount, and women have the capacity tomake the choice that is best for them.

Acknowledgments

Our thanks are extended to the women whoparticipated in this study. We are grateful to AleenaWojcieszek and Rachel Thompson for their experthealth communication input in constructing theexperimental stimuli for this work and reviewing earlierdrafts of this paper.

References

1. Coulter A. Paternalism or partnership? Patients have grown up –and there’s no going back. BMJ 1999;319:719–720.

2. Charles C, Whelan T, Gafni A. What do we mean by partnershipin making decisions about treatment? BMJ 1999;319:780–782.

3. Trevena L, Barratt A. Integrated decision making: Definitions fora new discipline. Patient Educ Couns 2003;50:265–268.

4. Braddock CH, Fihn SD, Levinson W, et al. How doctors andpatients discuss routine clinical decisions: Informed decisionmaking in the outpatient setting. J Gen Intern Med 1997;12:339–345.

5. Elwyn G, Edwards A, Wensing M. Shared decision making:Developing the OPTION scale for measuring patient involve-ment. Qual Saf Health Care 2003;12:93–99.

6. Declercq ER, Sakala C, Corry MP, Applebaum S. Listening toMothers II: Report of the Second National U.S. Survey ofWomen’s Childbearing Experiences. New York: ChildbirthConnection, 2006.

7. Thompson R, Miller YD, Wojcieszek AM, et al. Learning toshare? The current state of decision-making in maternity care inAustralia. 30th Annual Conference of the Society for Reproduc-tive and Infant Psychology; Leuven, Belgium, 2010.

8. Redshaw M, Heikkila K. Delivered with Care: A NationalSurvey of Women’s Experience of Maternity Care 2010. Oxford,UK: National Perinatal Epidemiology Unit, 2010.

9. Stapleton H, Kirkham M, Thomas G. Qualitative study of evidencebased leaflets in maternity care. BMJ 2002;324:1–6.

10. Gamble JA, Creedy DK. Women’s preferences for cesareansection: Incidence and associated factors. Birth 2001;28:101–110.

11. Fenwick J, Gamble J, Mawson J. Women’s experience ofcaesarean section and vaginal birth after caesarean: A birthritesinitiative. Int J Nurs Pract 2003;9:10–17.

12. Bekker HL, Hewison J, Thornton JG. Understanding whydecision aids work: Linking process with outcome. Patient EducCouns 2003;50:323–329.

13. Brown S, Lumley J. Changing childbirth: Lessons from andAustralian survey of 1336 women. BJOG 1998;105:143–155.

14. Stacey D, Bennett CL, Barry MJ, et al. Decision aids for peoplefacing health treatment or screening decisions (Review). Cochra-ne Database Syst Rev 2011;Issue 10, Art. No.: CD001431; DOI:10.1002/14651858.CD001431.pub3.

15. O’Connor A, Wennberg JE, Legare F, et al. Toward the ‘TippingPoint’: Decision aids and informed patient choice. Health Aff(Millwood). 2007;26:716–725.

16. Calder A, Alfirevic Z, Baxter J, et al. Induction of Labour:Clinical Guideline. London: National Collaborating Centre forWomen’s and Children’s Health, 2008.

17. Elwyn G, Laitner S, Coulter A, et al. Implementing shareddecision-making in the NHS. BMJ 2010;341:971–973.

18. Olesen AW, Westergaard JG, Olsen J. Perinatal and maternalcomplications related to postterm delivery: A national register-based study, 1978-1993. Am J Obstet Gynecol 2003;189:222–227.

19. Gulmezoglu AM, Middleton P. Indution of labour for improvingbirth outcomes for women at or beyond term (Review).Cochrane Database Syst Rev 2006;Issue 4, Art. No.: CD004945;DOI: 10.1002/14651858.CD004945.pub2.

20. Bailey D. Management of prolonged pregnancy: Yesterday,today and tomorrow. O & G Magazine 2010;12(2):20–21.

21. Heimstad R, Romunstad PR, Hyett J, et al. Women’s experiencesand attitudes towards expectant management and induction oflabor for post-term pregnancy. Acta Obstetricia et Gynecologica2007;86:950–956.

22. Alexander JM, Mcintire DD, Leveno KJ. Prolonged pregnancy:Induction of labor and cesarean births. Obstet Gynecol2001;97:911–915.

23. Anim-Somuah M, Smyth RMD, Howell CJ. Epidural vs. non-epidural or no analgesia in labour (review). Cochrane DatabaseSyst Rev 2005;Issue 4, Art. No.: CD000331; DOI: 10.1002/14651858.CD000331.pub2.

24. Ramirez M, Ramin S. Induction of labor: ACOG PracticeBulletin No. 107. Obstet Gynecol 2009;114:386–397.

25. Lawrence A, Lewis L, Hofmeyr GJ, et al. Maternal positions andmobility during first stage labour. Cochrane Database Syst Rev2009;Issue 2, Art.No.: CD003934; DOI: 10.1002/14651858.CD003934.pub2.

26. Alfrevic Z, Devane D, Gyte GML. Continuous cardiotocography(CTG) as a form of electronic fetal monitoring (EFM) for fetalassessment during labour (review). Cochrane Database Syst Rev2006;Issue 3, Art.No.:CD006066; DOI: 10.1002/14651858.CD006066.

27. McClain CS. Perceived risk and choice of childbirth. Soc SciMed 1983;17:1857–1865.

28. Howell-White S. Choosing a birth attendant: The influenceof a woman’s childbirth definition. Soc Sci Med 1997;45:925–936.

29. Hall MA, Camacho F, Dugan E, Balkrishman R. Trust in themedical profession: Conceptual and measurement issues. HealthServ Res 2002;37:1419–1439.


30. Trewin D. 2001 Census Dictionary. Canberra, Australia: Austra-lian Bureau of Statistics, 2001.

31. Laws P, Sullivan EA. Australia’s Mothers and Babies 2007.Sydney: AIHW National Perinatal Statistics Unit, 2009.

32. Trachtenberg F, Dugan E, Hall MA. How patients’ trust relates totheir involvement in medical care. J Fam Pract 2005;50:344–352.

33. O’Connor A, Elwyn G. IPDAS 2005: Criteria for Judgingthe Quality of Patient Decision Aids. International Patient Deci-sion Aid Standards (IPDAS) Collaboration. Accessed December3, 2011. Available at: http://ipdas.ohri.ca/resources.html.

34. Thompson R. Choosing how your labour will start: A decisionaid for women with a prolonged pregnancy. Brisbane, Australia:Queensland Centre for Mothers & Babies, 2010. AccessedDecember 20, 2011. Available at: http://www.havingababy.org.au/media/pdf/labourstart.pdf.

35. Légaré F, Ratté S, Gravel K, Graham ID. Barriers and facilitatorsto implementing shared decision-making in clinical practice:A systematic review of health professionals’ perceptions. PatientEduc Couns 2008;73:526–535.

36. Watson DB, Thomson RG, Murtagh MJ. Professional centredshared decision making: Patient decision aids in practice in pri-mary care. BMC Health Serv Res 2008;8:5. doi: 10.1186/1472-6963-8-5.

Supporting Information

Additional Supporting Information may be found in theonline version of this article:

Appendix S1. Scenario DetailsPlease note: Wiley-Blackwell is not responsible for

the content or functionality of any supporting materialssupplied by the authors. Any queries (other than miss-ing material) should be directed to the correspondingauthor for the article.


Documents

Overdue Choices: How Information and Role in Decision-Making Influence Women's Preferences for Induction for Prolonged Pregnancy