Data are from the Listening to Mothers III (LTM3) survey, a nationally representative sample of women who gave birth to a single infant in a US hospital between July 1, 2011, and June 30, 2012 (N = 2400). The survey was commissioned by Childbirth Connection, funded by the Kellogg Foundation, and conducted online by Harris Interactive using validated procedures.^17,24 Women aged 18 to 45 years who were participating in one of several online panels maintained by Harris Interactive formed the pool of potential respondents, with checks to ensure that each respondent only participated once. After data collection was complete, responses were weighted by propensity to be online as well as several demographic variables to enhance comparability with the national population of women who gave birth in 2010, the most recent year for which birth certificate data were available for this purpose.¹⁷

The Listening to Mothers surveys are the only nationally representative samples of childbearing women that contain information about doula care alongside selfreported clinical experiences, perceptions, and decisions about childbirth. In addition to asking whether a woman had support from a doula, the survey also asked about awareness of and level of familiarity with this type of care,

and whether women who knew about doula care would have wanted to have this type of care. The latter question is particularly useful as it may help at least partially address selection issues in who chooses to have a doula.

Variable Measurement

The 2 main predictors of interest were having doula support and, among those who did not have doula support but had a clear understanding of what a doula is, desire for doula support. Women were categorized as having doula support if they reported receiving supportive care during labor from a “doula or trained labor assistant.” Those who did not use doula support during labor were asked if they had heard of doulas and whether they had a clear understanding of this type of caregiver. Those with a clear understanding of doulas were then asked whether they would have liked to have doula support during their most recent birth; those who responded affirmatively were categorized as reporting “desire for doula support” in this analysis.

Measurement of cesarean birth was based on selfreported mode of delivery (vaginal or cesarean). Women with cesarean deliveries were asked to provide the main reason for the cesarean, which we categorized as a definitive medical indication for this procedure or a nondefinitive indication. We based these categorizations on professional standards used for accreditation measures²⁵ and confirmation by our clinician co author (DKG). The following reported reasons for cesarean were considered definitive medical indications: baby being in the wrong position for birth, problems with the placenta, fetal monitor showing fetal distress during labor, and maternal health condition that called for cesarean delivery. All other reasons cited were categorized as being potential reasons, but not definitive medical indications for cesarean; these included prior cesarean, labor taking too long, provider concern regarding the size of the baby, fear of labor and vaginal delivery, being past the due date (for women whose pregnancies are <41.5 weeks gestation at delivery), having a narrow pelvis, or citing no medical reason for their cesarean. The term nonindicated cesarean refers throughout the manuscript to this type of delivery. Detailed information about the proportion of women with each of the reasons for cesarean delivery is provided in the eAppendix (available at www. ajmc.com). We conducted multiple sensitivity analyses around the classification of reasons for cesarean as medical indications, and results were substantively unchanged when we categorized any combination of the following reasons as definitive indications: labor taking too long, provider concern regarding the size of the baby, and having a narrow pelvis.