In 2004, 1.2 million (29.1%) of live births in the United States were by cesarean delivery—the highest rate ever reported.1 Indeed, the rate of cesarean delivery in the United States has risen substantially over the past few decades, despite the national goal of reducing rates of cesarean delivery to 15% of births as part of Healthy People 2010.1,2 There has been considerable attention focused on how best to decrease the national cesarean delivery rate. Part of the increase in the cesarean delivery rate seems to arise from an increase in cesarean delivery requested by mothers in the absence of any medical or obstetric indications.3 Estimates of cesarean delivery on maternal request range from 4–18% of all cesarean deliveries,1,4 a rate that seems to be increasing.2 Women most commonly request cesarean delivery because of extreme tocophobia, or fear of child birth,4,5 a previous cesarean section, and previous negative birth experience6 as well as smaller family size and insurance liability concerns.7 How and whether these and other issues underlie the increase in the rate of cesarean delivery on maternal request remains to be determined.
Advocates of elective primary cesarean delivery contend that such delivery could decrease the risk of pelvic floor disorders, urinary incontinence, and sphincter damage typically associated with vaginal delivery and allow flexible timing for both mother and physician.8 Opponents stress the risks of elective cesarean delivery to the mother and fetus—increased morbidity and mortality associated with surgery and possible complications in subsequent pregnancies (eg, uterine rupture, placenta previa, and placenta accreta).8,9 An elective cesarean delivery may also increase the fetal risks of respiratory distress syndrome, persistent pulmonary hypertension, and fetal lacerations.5,8 However, as elucidated in the Conference Statement of the recent State-of-the-Science conference regarding cesarean delivery on maternal request2 there is very little consensus regarding the risks and benefits of cesarean delivery on maternal request, and very little strong data on which to base decisions.
The ethics of performing cesarean delivery on maternal request have been widely debated. The American College of Obstetricians and Gynecologists (ACOG) has stated that, “If the physician believes that cesarean delivery promotes the overall health and welfare of the woman and her fetus more than vaginal birth, he or she is ethically justified in performing a cesarean delivery. Similarly, if the physician believes that performing a cesarean delivery would be detrimental to the overall health and welfare of the woman and her fetus, he or she is ethically obliged to refrain from performing the surgery.”3. These guidelines place the onus on obstetrician–gynecologists to weigh the risks and benefits to patients on a case-by-case basis. In contrast, the International Federation of Gynecology and Obstetrics guidelines more firmly state that, “Because hard evidence of net benefit does not exist, performing cesarean sections for nonmedical reasons is ethically not justified.”10 Given that cesarean delivery has become safer over time (as have all forms of surgery), the situation is further complicated by evidence that liability insurance issues are driving the increase in cesarean delivery.7,9 As Benedetti et al11 suggest, it is possible that physicians will conduct nonmedically necessary cesarean deliveries to fend off liability issues—but they are doing so with a risk/benefit calculus that, as Minkoff et al10 state, is in “a state of dynamic change.”
It is in this context that a substantial minority of obstetricians will accede to a patient’s request for an elective cesarean delivery. Thus, a critical issue in determining the cause of the increase in elective primary cesarean delivery is examining obstetrician–gynecologists’ willingness to perform the surgery. Two recent studies found that two thirds of obstetricians surveyed are willing to perform cesarean delivery on maternal request,4,12 and there is evidence that more than one half would be willing to perform a cesarean delivery on maternal request,11 but little is known about obstetrician–gynecologists’ practical experiences surrounding elective cesarean delivery. The increasing rate of cesarean delivery at a national level and the increase in cesarean delivery without indication prompts us to explore further the reasons behind the recent trends in cesarean delivery and elective cesarean delivery.
The purpose of this study is to examine obstetrician–gynecologists’ knowledge, opinions, and practice patterns related to cesarean delivery on maternal request in an effort to better understand the factors associated with the increased rate of elective cesarean delivery, and how obstetrician–gynecologists are coping with requests for elective cesarean delivery in a climate of scientific uncertainty. The impetus for this study originated in the Office of Medical Applications of Research at the National Institutes of Health to assess current beliefs and practices before the State-of-the-Science Conference on cesarean delivery on maternal request. The final statement of this conference can be found at: http://consensus.nih.gov/2006/CesareanStatement_Final053106.pdf. A postconference survey will be administered in early 2007 to assess whether there have been any changes in knowledge, attitudes, and practices.
MATERIALS AND METHODS
Written questionnaires were modeled after the work of Wu et al5 that queried members of the American Urogynecological Society and the Society for Maternal–Fetal Medicine on their attitudes regarding elective primary cesarean delivery. Our survey was approved by the Institutional Review Board at Georgetown University.
Questionnaires were mailed to 1,031 ACOG Fellows in February 2006. Of the 1,031 recipients, 688 were members of the Collaborative Ambulatory Research Network, and 343 were a computer-generated random sample of ACOG fellows who had not received a questionnaire from ACOG in the past 2 years (non–Collaborative Ambulatory Research Network). Members of Collaborative Ambulatory Research Network are practicing obstetrician–gynecologists who have volunteered to participate in survey studies on a regular basis. The Collaborative Ambulatory Research Network was established to facilitate assessment of clinical practice patterns and aid in the development of educational materials. The non–Collaborative Ambulatory Research Network group is composed of practicing obstetrician–gynecologists who have not expressed an interest in participating in survey research activities. American College of Obstetricians and Gynecologists fellows who are not members of Collaborative Ambulatory Research Network were surveyed in an effort to obtain a sample that is more representative of the larger obstetrician-gynecologist population. Power analyses performed for similar surveys have indicated that the minimal number of surveys needed to ensure moderate effect sizes is approximately 100.13
A cover letter which stated that return of a completed questionnaire would constitute informed consent to participate in the study accompanied the questionnaire. Two additional mailings were distributed as follow-up to the nonresponders in March 2006.
The questionnaire consisted of four sections: 1) a series of items regarding the respondents’ demographic characteristics and those of their patient populations; 2) practices and attitudes surrounding cesarean (elective and nonelective) and vaginal deliveries; 3) knowledge and beliefs regarding the risks and benefits of elective and nonelective cesarean deliveries; and 4) counseling practices and department policies in the context of cesarean delivery on maternal request. It should be noted that all responses were self-report and thus implicitly rely on physicians’ recall.
Respondents were also asked to estimate the average education and income of their patient populations, and to estimate the proportion of their patients belonging to each of seven racial or ethnic groups. An Index of Racial Diversity (DeLeon R., 2003, unpublished manuscript) was computed by subtracting the probability of a patient belonging to each of the seven groups from 1.00 and multiplying the remainder by 100. A score of 100 thus indicates complete diversity (ie, a random patient drawn from the population has an equal likelihood of belonging to any of the seven groups). A score of 0 indicates no diversity (ie, only one racial or ethnic group is represented in the patient population).
Data were analyzed using a personal computer-based version of SPSS 14.0 (SPSS Inc., Chicago, IL). Descriptive statistics were computed for measures used in secondary analyses. Group differences in responses on continuous measures were assessed using independent sample t tests, and categorical data were assessed using the χ2 test. Significance tests were two-tailed, with P<.05. Data were analyzed for group differences using sex and age as between-subject factors. Each factor was evaluated separately.
Completed questionnaires received by March 29, 2006 (ie, before the National Institutes of Health [NIH] State-of-the-Science conference) were included in the data analysis. A total of 699 questionnaires (513 Collaborative Ambulatory Research Network, 186 non–Collaborative Ambulatory Research Network) of 1,031 were returned to ACOG after three mailings, yielding a response rate of 68%. Most of the questionnaires (68%) were from the first mailing, 26% from the second, and 8.0% from the third. Reflecting their greater interest in research activities, Collaborative Ambulatory Research Network members were significantly more likely to return questionnaires from the first mailing, while non–Collaborative Ambulatory Research Network recipients were more likely to return questionnaires from the second and third mailings. A total of 72.9% of Collaborative Ambulatory Research Network recipients responded to the first mailing, whereas only 47.1% of non–Collaborative Ambulatory Research Network recipients did (P<.001).
Of the 699 respondents, 591 (84.5%) perform deliveries, and 108 (15.5%) do not. Those respondents who returned questionnaires from the first mailing were more likely to deliver babies (86.6% of respondents to the first mailing deliver babies) than those who returned questionnaires from the second mailing (P<.05). The questionnaire was structured so that those respondents who do not perform deliveries completed the initial section regarding their specializations and practice structures, demographics, and demographic characteristics of their patient base but were asked to discontinue completing the remainder of the questionnaire. Those who do not deliver babies (mean age 54.81 years) were approximately 8 years older than those who do (mean age 46.52 years), a difference that was highly significant (P<.001). This difference was anticipated in light of the fact that obstetrician–gynecologists have been found consistently to cease delivering babies as they get older.9 The remainder of the analyses and discussion is limited to the group of respondents who currently deliver babies.
Table 1 presents demographic data for the respondent sample. The group of clinicians who deliver babies was evenly divided between males (50.9%) and females (48.8%). The mean age was 46.52 years (standard deviation [SD] 9.91 years), with a median of 45 years and a range of 30 years to 72 years. The mean number of years since residency was 13.67 (SD 9.73 years) with a median of 12 years and a range of 1 year to 40 years.
The majority of respondents (92.2%) practice both obstetrics and gynecology. Only a small number (3.4%) view themselves as primary care providers; the majority (53.6%) report that they are both specialists and primary care providers. Most of the respondents work in an obstetricts and gynecology partnership or group (49.8%) and within a suburban (38.0%) or non–inner city urban (27.1%) setting.
Based on respondents’ estimates, one half (49.4%) serve a middle class patient population with estimated yearly income between $25,000 and $59,999. The majority of their patients are white (mean proportion white 58.9%), Hispanic (mean proportion Hispanic 16.7%) or African American (mean proportion African American 14.9%). On average, the patient populations are moderately diverse with regard to race, with a mean diversity index of 43.01 (SD 20.51), a median of 46.50, and a range of 0 to 80.
The majority of respondents (92.2%) report that their department or practice has no policy regarding maternal request cesarean deliveries. Among the 7.8% whose department or practice has a policy and who described the policy, 72.2% stated that the policy was to support these requests (usually with informed consent, extensive counseling, or both), whereas the policy for 18.6% of those whose department has a policy was not to perform cesarean delivery on maternal request.
Clinicians were asked what their delivery choice would be if they or their spouse were pregnant, assuming a term, uncomplicated, singleton, cephalic pregnancy. Fewer than 20% (n=102, 17.8%) of respondents reported that they would choose or recommend a cesarean delivery on maternal request in this situation, due to concerns for the mother’s (64.8%) or the neonate’s (19.4%) well-being; 15.7% would choose cesarean delivery on maternal request for convenience.
The majority (58.4%) of respondents report performing between 51 and 150 deliveries per year, and for the majority (66.0%) between 10% and 25% of those are cesarean deliveries. Among those who perform more than 50 deliveries per year, cesarean delivery accounts for between 10% and 25% of deliveries for more than 65% of respondents.
Just over one half (53.0%) of respondents report having done cesarean delivery on maternal request, with the majority of those performing them once or twice per month or more rarely. Indeed, only 13 respondents reported that 10% or more of the cesarean deliveries they performed in the past year were by maternal request in the absence of medical indication. The majority of those who have done cesarean delivery on maternal request do them on a case-by-case basis (89.8%); only 10.2% report doing them routinely.
Clinicians were asked if they had changed their own practice with regard to cesarean delivery on maternal request in the past year. Ninety-eight (17.1%) of the respondents reported a change. Within this group, 66.0% had increased their use (none had decreased their use of cesarean delivery on maternal request), 41.2% now routinely discuss the topic with their patients, and 7.8% had made some other change (these proportions sum to more than 100% because respondents could choose more than one option).
Nearly three fifths (58.4%) of respondents reported noticing an increase in maternal requests for unindicated cesarean delivery over the past year. Respondents were asked to specify the factors they believed contribute to the increase. These comments were organized into rough categories, and each respondent’s comments could include more than one factor, so the proportions add to more than 100%.
A majority of those who had noted an increase and who gave an explanation attributed the increase to various sources of information, including the popular media, the internet, and child birth education (52.7%). Issues related to safety were mentioned by 47.9% of respondents. These included concerns for the welfare of the mother (eg, the risk of urinary and fecal incontinence and pelvic floor prolapse) as a result of vaginal delivery. Responses related to the welfare of the neonate were more general (eg, cesarean delivery is safer for the neonate). Convenience was mentioned by only 23.0% of respondents, the fact that women are having fewer babies (at an older age) by 3.8%, and 2.6% mentioned liability concerns.
A majority of respondents (54.6%) believe that a woman has the right to request and obtain an elective cesarean delivery without indication, and 57.4% would agree to perform one. Beliefs regarding the right to cesarean delivery on maternal request were strongly related to the likelihood of performing one. Nearly all of those who believe women have the right to request an unindicated elective cesarean delivery would perform one, whereas only a quarter of those who do not agree that women have the right would perform one (P<.001). Clinicians who do not believe women have a right to an elective cesarean delivery were also less likely to have performed one in the past year than clinicians who do subscribe to this belief (P<.001). Yet, even among those who don’t believe women have the right to request an nonmedically indicated cesarean delivery, one fifth had performed a cesarean delivery on maternal request.
About one fifth of respondents were neutral as to whether a woman has the right to request an unindicated cesarean delivery, and a similar number were neutral as to whether they would do one. As a group, the respondents were much more willing to do cesarean delivery on maternal request in the context of a history of 3rd or 4th degree laceration with her first delivery, with only about 11% likely to refuse to do one.
Clinicians’ beliefs regarding the right to cesarean delivery on maternal request, their likelihood of performing one, and their likelihood of performing one in the context of a past history of delivery complications were not related to the number of deliveries they perform, the proportion of those deliveries that were cesarean, or the proportion of cesarean deliveries that were nonmedically indicated.
Clinicians were asked to select which of 14 risks and 14 benefits to mother and neonate they believed were associated with cesarean delivery on maternal request. The list of risks and benefits were based on those examined in the NIH State-of-the-Science Conference.2 Overall, respondents endorsed more risks than benefits (mean number of risks 7.28, SD 2.69); mean number of benefits 4.47, SD 2.92; P<.001). Only three respondents (0.8%) endorsed none of the risks (ie, reported that there were no risks to cesarean delivery on maternal request), whereas 44 (11.8%) endorsed none of the benefits. Five of the 14 benefits were endorsed by more than one half of the respondents.
The most frequently endorsed benefit was the reduced risk of perineal damage (76.2%), followed by risk of dystocia or birth trauma (63.4%), pelvic organ prolapse (56.8%), long-term incontinence (54.8%), lacerations (53.9%), and complications from labor (49.6%). Fewer than 10% of respondents endorsed reduced risk of neurodevelopmental delay (9.3%), bleeding (4.8%), and improved sense of maternal well-being (9.7%). Only 5.0% endorsed some “other” reduced risk. Generally these were redundant with a risk already on the list, and for a few respondents it was reduced liability.
In contrast to the list of benefits, clinicians endorsed 8 of the 14 risks accruing from cesarean delivery on maternal request more than 60% of the time. The most frequently endorsed risk was intraoperative risks to the mother such as bleeding and damage to internal organs (93.7%). Only three risks were endorsed by fewer than 10% of respondents: higher risk of ectopic pregnancy (5.8%), higher risk of stillbirth in future pregnancy (9.5%), and “other” (3.2%). In most cases “other” was increased likelihood of cesarean delivery in future pregnancies.
Respondents who asserted that they would not perform a cesarean delivery on maternal request endorsed significantly more risks than those who would do one (P<.001), and fewer benefits (P<.001) than those who would perform one. As can be seen in Table 2, there were nine potential benefits that the three groups of respondents (those who would not do a cesarean delivery on maternal request, those who were neutral or unsure as to whether they would do one, and those who would perform a cesarean delivery on maternal request) differed on. Those who would not perform a cesarean delivery on maternal request were less likely to endorse improved sense of maternal well-being, and reduced risk of perineal damage, pelvic organ prolapse, long-term incontinence, hypoxic insult, lacerations, complications from labor, cerebral palsy, and neurodevelopmental delay. For seven of these issues, the “neutral” group is intermediate to the two other groups. For pelvic organ prolapse and improved sense of maternal well-being, responses of the neutral group were similar to the group of respondents who would not perform a cesarean delivery on maternal request.
Of the list of 14 risks to mother and neonate from cesarean delivery on maternal request, only two (intraoperative risks to neonate and problems with transition to neonatal life) were endorsed by the same proportion of respondents in each of the groups defined by willingness to perform cesarean delivery on maternal request. For 8 of the 11 items, proportions for those who would not do one and those who are neutral toward cesarean delivery on maternal request are nearly equal and for both groups are higher than for those who would perform cesarean delivery on maternal request. For three items (increased costs, high risk of ectopic pregnancy, and higher risk of stillbirth in future pregnancies) the proportion of those in the neutral group was higher than the other two groups; in one of these (increased costs) the negative group was intermediate to the other two groups.
The number of risks endorsed by these clinicians was uncorrelated with age and years in practice, but there were small positive correlations between both age (r=.12, P<.01) and years in practice (r=.15, P<.001) and the number of benefits endorsed, suggesting that older clinicians and clinicians who had been in practice longer had a tendency to endorse more benefits than younger clinicians. There was no relationship between patient education and patient income and the number of risks and benefits endorsed.
When obstetrician–gynecologists were asked which factors have an impact on the outcome of cesarean delivery on maternal request, more than one half identified the following: previous delivery complications (83.9%), maternal anxiety (71.4%), maternal age (62.4%), plans for future child bearing (59.3%), and fetal size (54.4%).
A large majority (85.6%) of respondents rated themselves as very or extremely qualified to discuss the risks and benefits of elective cesarean delivery with patients. None rated themselves as “not at all qualified.” Only 20 (3.5%) of respondents reported not discussing risks and benefits with patients who request an unindicated cesarean delivery. All of these had reported that they believed there were no benefits to cesarean delivery on maternal request; the majority (75%) of those who did not discuss risks and benefits had never conducted a cesarean delivery on maternal request.
The majority of respondents who do patient counseling in the context of a patient request for a nonmedically indicated cesarean delivery reported discussing all of the risks presented on the questionnaire: short-term risks to the mother, short-term risks to the neonate, short-term risks to the mother from attempting vaginal delivery, short-term risks to the neonate from attempting a vaginal delivery, long-term risks to the mother from a cesarean delivery, and long-term risks to the mother from a vaginal delivery (Table 3). Clinicians were more likely to discuss risks to the mother than risks to the neonate, and were more likely to discuss risks from cesarean delivery than from vaginal deliveries.
Those who would not do a cesarean delivery on maternal request discussed fewer risks with patients (P<.05) than those who were neutral and positive toward doing a cesarean delivery on maternal request. There was no gender difference in this analysis, nor was there a correlation between the number of risks discussed and age, years since residency, patient income, or patient education.
There were a few differences between males and females with regard to the attitude and risk–benefit items. Female respondents were significantly more negative toward a woman’s right to request and obtain a cesarean delivery on maternal request than males (P<.001), and reported that they were less likely to agree to do one (P<.05). There was no gender difference in attitudes toward doing a cesarean delivery on maternal request in the context of a history of previous delivery complications.
Overall, females endorsed more risks than males (P<.01), and fewer benefits (P<.01). Females endorsed more of the six risks than males: Interoperative risks to mother (91.4% of males, 96.0% of females; P<.001), interoperative risks to neonate (39.0% of males, 49.3% of females; P<.05), increased hospital stay (72.4% of males, 80.9% of females; P<.05), increased risk of uterine rupture (66.2% of males, 75.2% of females; P<.05), increased risk of placenta previa and placenta accreta (74.1% of males, 82.7% of females; P<.05), and difficulty breastfeeding (39.1% of males, 60.9% of females; P<.05). None of the risks were endorsed by more males than females.
Four of the benefits were endorsed more frequently by males than females: Reduced risk of hypoxic insult (31.8% of males, 15.0% of females; P<.001), reduced risk of cerebral palsy (14.2% of males, 7.7% of females; P<.05), reduced risk of dystocia or birth trauma (68.2% of males, 58.1% of females; P<.05), and complications from labor (54.3% of males and 44.3% of females; P<.05). None were endorsed more frequently by females than males.
With regard to factors that affect the outcome of cesarean delivery on maternal request, males and females differed on two of the items. More females (65.5%) than males (53.7%) endorsed plans for future child bearing (P<.01) and more males (10.6%) than females (5.6%) endorsed “other” (P<.01).
As noted above, the response rate for the non–Collaborative Ambulatory Research Network group was lower than for the Collaborative Ambulatory Research Network group, and the Collaborative Ambulatory Research Network group was more likely to respond to the first questionnaire mailing. There was also a difference in the groups’ self-definition. Members of the non-Collaborative Ambulatory Research Network group were more likely to consider themselves either primary care providers or specialists, while Collaborative Ambulatory Research Network members were more likely to consider themselves both primary care providers and specialists (P<.05).
The non–Collaborative Ambulatory Research Network group was also more likely to have patients with a college education, while the Collaborative Ambulatory Research Network group was more likely to have patients with high school or less education, or an advanced degree (P<.001). Indeed, 20 of the Collaborative Ambulatory Research Network group and none of the non–Collaborative Ambulatory Research Network group had practices dominated by women with an advanced degree. The non–Collaborative Ambulatory Research Network group was more likely to have patients with incomes between $40,000 and $80,000, whereas the Collaborative Ambulatory Research Network group was more likely to have patients with incomes lower than $40,000 and higher than $80,000 (P<.05).
The Collaborative Ambulatory Research Network group was more likely to practice obstetrics and gynecology than the non–Collaborative Ambulatory Research Network group (as opposed to some other specialty such as maternal-family medicine), but the cell sizes were very small. Of the 591 respondents, only 46 (27 Collaborative Ambulatory Research Network and 19 non–Collaborative Ambulatory Research Network) practice a specialty other than obstetrics and gynecology. The groups were similar on all of the other demographic or practice variables except for the number of deliveries performed. Non–Collaborative Ambulatory Research Network respondents were more likely to perform fewer than 51 or more than 200 deliveries per year than the Collaborative Ambulatory Research Network group (P<.05). There was no difference between these groups on the proportion of those deliveries that were by cesarean delivery, the proportion of cesarean deliveries that were cesarean delivery on maternal request, or on whether they had done a cesarean delivery on maternal request.
The two groups did not differ on the total number of risks they discussed with their patients who inquire about cesarean delivery on maternal request, nor was there a difference on how many risks or benefits they endorsed. There were only three (of 28 possible) differences between the two groups on ratings of the risks and benefits of cesarean delivery on maternal request. Collaborative Ambulatory Research Network members were more likely to endorse the increase in interoperative risk to mother (94.9% compared with 90.0%; P<.05), problems with transition to neonatal life (39.1% compared with 29.3%; P<.05), and increased risk of placenta previa and placenta accreta (81.4% compared with 69.3%; P<.01) than non–Collaborative Ambulatory Research Network respondents were.
Our results suggest that some part of the recent increase in the rate of cesarean deliveries can be attributed to an increase in cesarean delivery on maternal request, given that about 10% of our sample reported an increased use of the procedure. Of perhaps greater interest is the fact that about 20% of our respondents have increased their counseling for patients who have requested an unindicated cesarean delivery.
Although obstetricians endorsed the right that women have to request a cesarean delivery, and many had done a cesarean delivery on maternal request, their discomfort with the procedure was evident. Obstetrician–gynecologists are, as a group, experiencing an increase in requests for cesarean deliveries in the absence of medical indication that seems to stem as much from information in the media and popular press as it does from concerns regarding the risks of vaginal—or cesarean—delivery. As several of our respondents stated in written comments, a number of celebrities have recently undergone cesarean delivery on maternal request. One respondent noted that “patients have named celebrities.” Another respondent stated, “I think the media’s portrayal of “Hollywood personalities” having planned caesarean births has made it a glamorous option...much to the detriment to our patients.” This represents a validation of the procedure that is irrelevant to the issues that bear on the decision to undergo a cesarean delivery.
It is noteworthy that only about one half of the respondents have performed one or more cesarean delivery on maternal request and believe that a woman has the right to request and obtain a cesarean delivery on maternal request, but only about one fifth would seek or recommend one for herself or his spouse or partner. This suggests cognitive dissonance in reconciling their personal and clinical experience, training, and ACOG guidelines. This is illustrated by comments from two of our respondents: “We allow patients to have many elective procedures, ie, gastric stapling, multiple plastics procedures, [LASIK] eye surgery. Why should a well-informed woman who is given good informed consent not be allowed this option for delivery?” contrasted with this from another: “If C/S at patient request is reasonable, should not hysterectomy, appendectomy, etc. after informed consent be also acceptable?” We found consistent differences between males and females in our sample regarding their attitudes and practices surrounding cesarean delivery on maternal request. Females were more negative toward cesarean delivery on maternal request and were somewhat less likely to perform them, which is consistent with other studies.5 In contrast, we found only one small effect of clinician age and none of the education and income characteristics of the respondents’ practices.
Our findings are in contrast to other reports5,14,15 that found that the majority of obstetricians would perform cesarean delivery on maternal request. Importantly however, one5 surveyed subspecialists in maternal–fetal medicine or urogynecology who may be different from the broad representation of this survey, with only 71 identified as subspecialists. Others were international surveys14,15 that may not reflect the opinions of physicians in the United States. Alternatively, the opinions of the clinicians in our study could have been influenced by the increased media exposure of cesarean delivery on maternal request.
The results of this study should be considered within the context of survey research, which is vulnerable to response bias. We have attempted to limit response bias by eliciting as many completed questionnaires as possible from clinicians who received the questionnaire; nonresponders received a second, and in some cases, a third request to complete the questionnaire. It should be noted that the response to this survey is the highest ACOG has obtained in 11 years of survey research, and suggests a high level of interest in the topic. In addition, in the cover letter the respondents’ confidentiality was emphasized and the questionnaires were coded using a randomly assigned number. This would argue against a response bias arising from a sense of stigma associated with conducting cesarean delivery on maternal request. Finally, responses to the survey were dependent on physicians’ recall. It is possible, for example, that some respondents misremembered the number of deliveries conducted in the past year. Again, however, the salience of this topic to our respondents argues against the likelihood that their responses to the central questions was vulnerable to problems with recall.
The NIH Conference Statement concludes that there is only moderate support for only a handful of risks and benefits of cesarean delivery on maternal request as compared with vaginal delivery: maternal hemorrhage (favors planned cesarean section), maternal length of stay (favors planned vaginal delivery), and neonatal respiratory morbidity (favors planned vaginal delivery). However, given the available information, for most of the risks and benefits examined in the conference report, there is not enough information to determine which approach to delivery is optimal. This survey was undertaken before the NIH Consensus conference with the desire to examine practicing obstetricians’ knowledge, opinions, and practice patterns before the publication of the independent panel’s conclusions. To identify the ways in which the conference conclusions affect knowledge, opinions, and practice patterns, a follow-up survey is planned in early 2007. Clinicians would benefit from strong evidence regarding risks and benefits, evidence that is crucial to guiding policy making with regard to cesarean delivery on maternal request.