Eligibility

We enrolled 540 self-identified black and Latina women who had delivered at a large tertiary hospital located in New York City between April 2009 and March 2010. The Icahn School of Medicine at Mount Sinai Program for Protection of Human Subjects approved this study. The patient sample was identified through an electronic documentation system for the labor and delivery unit. Eligible participants were black/African American or Latina/Hispanic, 18 years old, and had infants with birthweights ≥2500 g and 5-minute Apgar scores ≥7. Maternal race/ethnicity, maternal age, infant Apgar scores, and infant birthweight are contained in the electronic system. Bilingual clinical research coordinators asked additional questions related to race and ethnicity. Mothers were eligible if they self-identified as black/African American or Latina/Hispanic, spoke English or Spanish, and had a working telephone. Mothers who were recruited during their postpartum hospital stay gave consent, underwent a baseline interview, and were assigned randomly to intervention vs control.

Intervention and control

The behavioral educational intervention that aimed to prepare and educate mothers about postpartum symptoms and experiences (including tips on breastfeeding and breast/nipple pain) and to bolster social support and self-management skills included an in-hospital educational session with a social worker, educational materials, and a 2-week follow-up call. Mothers who were assigned randomly to the intervention arm were given a 2-step intervention. The first step occurred in the hospital when a masters-trained bilingual social worker reviewed an education pamphlet and partner summary sheet with each mother. Education materials included information on breastfeeding, breast/nipple pain, cesarean delivery, site pain, episiotomy site pain, urinary incontinence, back pain, headaches, hair loss, hemorrhoids, infant colic, and depressive symptoms. Additional information was provided on social support. A partner summary sheet spelled out the typical pattern of experience for mothers after delivery in an effort to normalize the experience. During the second step, which was the 2-week postdelivery call, the social worker assessed patients’ symptoms, skills in symptom management, and other needs. Patients and the social worker created action plans to address current needs that included assessment of community resources. Enhanced usual care participants received a list of community resources and received a 2-week control call.

The intervention was tailored culturally. The intervention was revised based on the results of a pilot-test with 50 diverse mothers during the postpartum period and a review with obstetricians, psychiatrists, and social workers. The intervention was further revised based on feedback from focus groups with black and Latina mothers during the postpartum period and our community action board. All materials were translated into Spanish and back-translated for accuracy and consistency of meaning. Content, pictures, wording, and length were revised according to these sources.

Assessments

Data were collected in person at baseline and by telephone during follow-up interviews by bilingual clinical research coordinators who were blinded to intervention status. Interviews occurred by phone at 3-week (mean, 27.6 ± 5.8 [SD] days), 3-month (mean, 83.5 ± 7.8 [SD] days), and 6-month (mean, 175.6 ± 8.9 [SD] days) intervals after delivery. Mothers were asked a series of questions about breastfeeding that included the importance of breastfeeding and breastfeeding support, physical and emotional symptoms, social support, and health care factors. The dependent variable, breastfeeding duration, was ascertained by asking mothers how long they breastfed their infants (in weeks). At each interview, mothers were asked if they currently were breastfeeding their infant. If they answered no, they were asked how long they breastfed their infant after delivery.

Variables that were selected for inclusion in our analyses were maternal sociodemographic characteristics (age, race/ethnicity, nativity, language, educational attainment, insurance coverage, marital status, and work status), maternal health and clinical characteristics (parity, delivery type, presence of comorbid conditions, and antepartum complications), maternal psychosocial characteristics (social support, partner support, and history of depression), and breastfeeding characteristics (perceived importance of breastfeeding and breastfeeding support).

Maternal depressive symptoms were measured with the Edinburgh Postnatal Depression Scale (EPDS), a 10-item self-report scale. The EPDS is a validated, widely used postpartum depression screening instrument. We used an EPDS score ≥10 to indicate possible depression. Although a number of cut-off points have been used, it is recommended that, to minimize the failed detection of depressive cases, a cutoff score of 9/10 be used. The Generalized Anxiety Disorder, 7-item scale, was used to assess generalized anxiety disorder. A cut-point of 10, which has been found to measure moderate anxiety, was used. General social support was assessed by taking the mean of 3 items that measured emotional support, instrumental support with the baby, and instrumental support with the household, responses (Cronbach’s α, .701), responses ranged from 0–4. The partner support scale was constructed with the mean of the 2 items (Cronbach’s α, .693) and had a range of 0–4.

Women’s breastfeeding characteristics were assessed by 4 questions. First, respondents were asked: “How important is breastfeeding?” Responses were coded as (1) extremely or quite a bit and (0) moderately, a little bit, or not at all. Subjects were also asked whether they received breastfeeding support from health care providers, partners, family, or friends. Each response was coded (1) yes and (0) no. In addition, each mother was asked whether her infant received breast milk only, formula only, or both at the 6-month assessment.

Statistical analyses

Randomization procedures and sample size calculations have been reported elsewhere. All statistical analyses were performed with personal computer SAS software (version 9.2; SAS Institute Inc, Cary, NC). Intervention status was the key independent variable in the current analyses. Bivariate analyses were conducted with the use of χ ² tests and t tests to assess the association between sociodemographic, clinical, and breastfeeding characteristics with intervention status. Intention-to-treat analyses examined breastfeeding duration (measured in weeks). Twenty-six women, who indicated that they were breastfeeding at baseline, were missing breastfeeding duration data at all 3 follow-up time points and were assigned the midpoint value of 2 weeks for breastfeeding duration, which represented the midpoint between the time from their baseline interview and the next scheduled follow-up assessment (mean, 27.6 days for trial participants). Kaplan Meier analysis was used to test the effect of the intervention on breastfeeding duration for up to 6 months of follow up; Kaplan Meier curves were compared with the use of the log rank test. Cox regression analysis was used to calculate hazard ratios with 95% confidence intervals. Additional analyses examined breastfeeding duration among the 451 mothers who initiated breastfeeding and results were similar. We therefore only present the results for all women randomized in the trial (intention to treat).