Hypothesis: We hypothesized that providing in-hospital, supervised paraprofessional breastfeeding consultation during the hiatus between hospital discharge and the first postpartum WIC or pediatric visit would increase the rate of exclusive breastfeeding among low income Hispanic women at 1 month postpartum.

Methods:

Eligibility: Hispanic mothers who intended to breastfeed and were scheduled for simultaneous discharge with their healthy infants were eligible for enrollment. Women were excluded if factors indicating elevated risk for hyperbilirubinemia were noted in the infant's chart (<37 weeks gestation at birth, < 48 hours of age at discharge, jaundice within 24 hours of birth, Rh incompatibility; cephalohematoma, positive coombs test, family history of disorders of RBC enzyme defects or defects of RBC shape and size, and exclusive breastfeeding in the hospital).

Setting: Subjects were recruited between during the 2004 calendar year at a large metropolitan Hospital in Houston Texas. Ninety percent of mothers delivering at the facility are first generation immigrants, and 85% are monolingual Hispanic. Infants spend > 4 hours in the newborn nursery before being reunited with their mothers. Feeding counselors assist mothers with latch-on at hospital bedside and encourage exclusive breastfeeding. Ninety-seven percent of mothers initiate breastfeeding and all but 2% of them choose to offer "las dos cosas" (both breast and formula) in the hospital. Infants at elevated risk for hyperbilirubinemia are screened for medical and breastfeeding problems at 3-5 days in a Newborn Follow-up Clinic (NBFC). All mothers receive the telephone number of the hospital's breastfeeding clinic and the WIC office with instructions to call for breastfeeding assistance if needed after discharge. For infants at low risk for hyperbilirubinemia, there is a gap of approximately two weeks between hospital discharge and the first well child exam. The first WIC visit also occurs approximately two weeks postpartum.

Intervention: The intervention consisted of scheduling mothers of infants at low risk for hyperbilirubinemia to visit the hospital-based breastfeeding clinic at 3-7 days post partum. An appointment reminder card was included with the discharge papers. At the breastfeeding clinic, mothers received counseling from bilingual, bicultural paraprofessionals under the supervision of an International Board Certified Lactation Consultant / registered nurse. A breastfeeding history, breast and nipple exam, infant oral-motor assessment, infant weight, observation of a breastfeeding episode with evaluation of latch and milk transfer, and discussion of maternal concerns and support system were included in the counseling sessions. The importance of exclusive breastfeeding was reviewed, and plans for attaining exclusivity were discussed if the mother desired to achieve that goal. Information and skills training were provided as indicated for identified deficits, concerns, and breastfeeding problems. Additional visits and / or telephone consultations were provided if deemed necessary by the mother and the clinic staff.

All clinic contact was recorded by staff. Women who missed appointments received a phone call from a clinic counselor. Visits were rescheduled if possible; if not, counseling was provided over the phone. During phone counseling, mothers were screened for breastfeeding problems and concerns regarding adequacy of milk supply. Problem management was discussed where indicated.

Breastfeeding counselors were graduates of the Texas Department of State Health Services Principles of Lactation Management course and peer counselor training programs and were cross-trained in local WIC clinics and at hospital bedside to insure that advice received in the clinic was consistent with that received in hospital and during WIC visits.

Study design: Mothers were enrolled at hospital bedside within 20-48 hrs postpartum. After obtaining informed consent, group was determined using sealed envelopes containing assignments generated by random number table. The envelope was opened by the mother. Demographic, socio-cultural, and infant feeding data were collected after group assignment. Receipt of treatment was determined from clinic records. Outcomes were determined by telephone survey at 4 weeks postpartum by interviewers blinded to group assignment. Women in the intervention group received a $15 gift card upon arrival at the clinic in lieu of transportation costs. All Participants received a $5 gift card after completing the telephone survey at 1 month postpartum. The study was approved by Baylor College of Medicine's Institutional Review Board for the Protection of Human Subjects, and by the Institutional Review Board of the Harris County Hospital District.

Objectives: The primary objective was to determine whether proactive assignment to attend a breastfeeding clinic between 3-7 days postpartum would increase the rate of exclusive breastfeeding and/or reduce the extent of formula supplementation at 1 month in a population of low income, primarily mono-lingual Mexican and Central American immigrants. Secondary objectives were to evaluate the impact of the intervention on incidence of breastfeeding problems and to compare receipt of early counseling by mothers assigned to care (proactive support) with receipt of early counseling by mothers given the telephone number of the clinic with encouragement to call for support if needed ("warm-line" support).

Baseline Survey:

Infant feeding prior to enrollment:

Mothers were asked the age of the baby (in minutes or hours) when they first attempted to breastfeed, the number of times they had attempted to breastfeed since delivery, and the number of formula bottles the baby had received in the hospital. Data was confirmed by chart review. Average breastfeeding intervals were determined by dividing the number of attempted breastfeeds by the time since delivery. The quality of breastfeeding was estimated by asking mothers how frequently they experienced feeding difficulties ( 0= "every time" ; 1= "about half the nursings"; 2= "1 or 2 nursings"; 3="never").

Prior Breastfeeding experience and intentions Mothers were asked whether they breastfeed a previous child and if so whether they breastfed "successfully" or "Unsuccessfully" (Sí, con éxito; ‮ Sí, pero sin éxito‮). "Unsuccessful" responses were categorized as "prior BF failure". In addition, women were asked whether they had breastfed a previous child longer than 2 weeks and whether they had breastfed without formula for at least 1 month. Mothers were asked how long they planned to breastfeed in months, and whether they wanted to breastfeed without giving supplements (¿Quiere dar pecho sin dar fórmula?). Maternal self efficacy was assessed with the following binary choice questions: ¿Piensa usted que tendrá suficiente leche para su bebé? (Do you think you will have plenty of milk for your baby? ): ¿Piensa usted que podrá dar pecho mientras quiera? (Do you think you will be able to breastfeed as long as you would like?). Finally, mothers were asked to rank how important it was to them to breastfeed this baby on a four point likert scale ranging from low (poco) to very important (muy importante). The mother's local support network was estimated using two probes: ¿Presta sus servicios - por ejemplo: cuidando los niños, llevándolo en su carro, o de compras) familiares o amigos en este país? (Do you ever exchange services - such as babysitting, car rides, or shopping with family or friends in this country?) and ¿Usted presta dinero o pide prestado dinero de su familia o amigos en este país? ( Do you every lend money to or borrow money from family or friends in this country? )

Outcome measures:

At 4 weeks the following information was collected on Infant feeding behavior: For the last 24 hour time period: number of at-the-breast; number of expressed breastmilk feedings; number and volume of formula, water, agua de arroz, tea, infant cereal, or other feedings. Mothers were also asked: whether the baby routinely received water, agua de arroz, tea, infant cereal, juice, or other non-human milk feedings; if so how many times / day; age at introduction of any regular (> 2/week) non-human milk feedings; age at which formula supplements were discontinued if applicable; experience of breastfeeding complications after discharge (yes or no for 6 items: engorgement, sore nipples, low milk supply, difficulty nursing, breast pain, exhaustion);

Satisfaction with and utilization of health services: Mothers were asked to rank satisfaction with hospital care (1 to 5 likert scale); whether the infant had visited the pediatrician or the emergency room; whether the mother had visited the obstetrician, WIC clinic, or other health care provider; and whether the mother had received telephone support for breastfeeding concerns from WIC.

Test-retest reliability was evaluated by re-administering the survey to 60 participants 1-2 weeks later and calculating correlation coefficients between responses.

Data analysis:

Statistics: Sample size calculations were based on the assumption that 4 week breastfeeding rates in the target population would be 85% for any, and 15% for exclusive breastfeeding. A sample size of 255 women in each group would allow an 80% probability of detecting an increase from 85% to 93% in the rate of any breastfeeding, from 15% to 25% in the rate of exclusively breastfeeding, and a difference of 3 ounces per day in the volume of formula supplements given to breastfed infants at an alpha of .05.

Data was entered into Access and analyzed using MinitabTM statistical software release 13.31 and SPSS for Windows release 11.5.0 (Chicago, Ill). Differences between treatment groups were examined by student t test. Univariate relationships between outcome measures and maternal characteristics were determined by Student t-test for any and exclusive breastfeeding, and by Pearson correlation for volume of formula supplement. Binary logistic regression was used to determine relationships between binary outcomes and group assignment.

Results are reported as odds ratios with 95% confidence intervals or means ± SD where appropriate. Because of the relatively high rate of non-compliance, secondary analyses were conducted by treatment received. Multivariate models were adjusted for potential confounding.

Inclusion Criteria:

• intent to breastfeed
• delivery at target hospital
• Hispanic
• mother and infant scheduled for joint discharge
• healthy mother and infant

Exclusion Criteria:

infant at elevated risk for hyperbilirubinemia:

• <37 weeks gestation at birth,
• < 48 hours of age at discharge,
• jaundice within 24 hours of birth,
• Rh incompatibility; cephalohematoma,
• positive coombs test,
• family history of disorders of RBC enzyme defects or defects of RBC shape and size,
• exclusive breastfeeding in the hospital