OBJECTIVE: The purpose of this study was to determine whether implementing a program based on a clinical protocol affects breastfeeding rates within a pediatric primary care setting. Increasing breastfeeding rates is an important public health initiative identified by multiple agencies.
METHODS: The Academy of Breastfeeding Medicine (ABM) clinical protocol (“The Breastfeeding-Friendly Physician’s Office, Part 1: Optimizing Care for Infants and Children”) was used as a template for the provision of breastfeeding services within a pediatric primary care clinic. There were 757 mother–infant pairs included in the study. A retrospective before-and-after study design was used. Data collection points included the hospital stay, the newborn visit, and the 2-, 4-, and 6-month health maintenance visits. The 2 groups were compared to estimate the protocol’s effectiveness as a method of increasing breastfeeding rates.
RESULTS: The results of this evaluation were positive for exclusive breastfeeding, with group comparisons showing a statistically significant increase in exclusive breastfeeding rates at all 5 time points.
CONCLUSIONS: Our diverse patient population within a pediatric practice had increased initiation rates and exclusive breastfeeding rates after implementation of the ABM’s breastfeeding-friendly protocol. Families who receive care in a pediatric primary care setting that has implemented the ABM clinical protocol may have increased rates of exclusive breastfeeding.
- breastfeeding education
- breastfeeding exclusive
- evidence-based health care
- primary health care
- AAP —
- American Academy of Pediatrics
- ABM —
- Academy of Breastfeeding Medicine
- EMR —
- electronic medical record
- IBCLC —
- International Board-Certified Lactation Consultant
- RN —
- registered nurse
- WHO —
- World Health Organization
What’s Known on This Subject:
A gap exists with lack of programs to help mothers breastfeed. The 2012 American Academy of Pediatrics' “Policy Statement on Breastfeeding and the Use of Human Milk” re-emphasized breastfeeding as an important public health initiative rather than a lifestyle choice.
What This Study Adds:
Families who receive care in a primary care setting that has implemented a “breastfeeding-friendly” office protocol may have increased rates of exclusive breastfeeding. This study evaluated an accepted clinical protocol in a large, diverse pediatric primary care setting.
Increasing breastfeeding exclusivity and duration is an important public health initiative. The Centers for Disease Control and Prevention’s 2012 Breastfeeding Report Card1 reveals 76.9% of women are initiating breastfeeding in the early postpartum period; however, exclusivity and duration of breastfeeding rapidly decline within a short time period.
The American Academy of Pediatrics (AAP),2 the World Health Organization (WHO), the United Nations Children’s Fund,3 and the US Department of Health and Human Services4 have established increasing breastfeeding rates as imperative.
The 2011 US Surgeon General’s Call to Action to Support Breastfeeding endorses breastfeeding as the best nutrition for infants and details action items for the health care system as a mechanism for supporting breastfeeding families. The Call to Action implementation strategy for health care specifically states “models should be established to integrate assistance with breastfeeding into routine practice settings.”5
The Academy of Breastfeeding Medicine (ABM) has published a clinical protocol, “The Breastfeeding-Friendly Physician’s Office, Part 1: Optimizing Care for Infants and Children.”6 This protocol provides guidance for implementing breastfeeding support in the primary care setting and was developed based on the WHO and United Nations Children’s Fund Baby-Friendly Hospital Initiative.3 The literature describes this initiative as a systematic intervention providing a positive predictor of breastfeeding initiation and duration.7–16 Education of pediatricians was also identified as effective,16–18 along with lactation consultant support16,19–24 and maternal self-efficacy.25–29 The ABM protocol was chosen for the current study because it includes all of these components.
Cardosa et al30 implemented the Brazilian Breastfeeding-Friendly Primary Care Initiative, which is based on the WHO’s “Ten Steps to Successful Breastfeeding,” in their pediatric primary care setting and found an increase in the rates of breastfeeding as well as reduced rates of disease in infants aged <1 year. The primary care setting is well situated for the extension of hospital-based breastfeeding support in combination with the established AAP’s health maintenance visits. However, there are no known studies regarding the effect of a “breastfeeding-friendly” protocol in pediatric primary care in the United States.
The goal of the current study was to determine whether the use of a breastfeeding-friendly clinical protocol in a multicultural, suburban/rural US pediatric primary care setting would increase breastfeeding rates.
A retrospective before-and-after design was used. Retrospective data were collected from 2 groups of patients. The preintervention group included families visiting the primary care setting with their infant for regular health maintenance visits before implementation of the breastfeeding-friendly clinical protocol, and the postintervention group included those families cared for after the protocol was in place. The source of feeding was collected via record review by using 5 time points: the hospital stay, the initial outpatient newborn visit, and the 2-, 4-, and 6-month health maintenance visits.
For the purpose of this study, breastfeeding definitions are described as exclusive or any breastfeeding. Exclusive breastfeeding was defined as breast milk (including milk expressed) and any drops, vitamins, or medicines. It could also include solid foods if reported at 6 months but not earlier. Formula was not included. Any breastfeeding was defined as breast milk (including milk expressed) combined with any food or liquid including any non-human milk and formula.
The study was conducted within a pediatric primary care setting at 2 locations of a single practice. Both locations are in northern Virginia, 1 in a suburban area and the other in a rural setting. The populations accessing both offices comprise a general mix of socioeconomic classes from upper-income families to poverty-level families. Because of its proximity to the Washington, DC, metropolitan area, there is a large commuter population, and families from many ethnicities are well represented. Populations accessing the rural setting also include families from West Virginia.
The clinics are staffed by 13 medical providers. The total practice population is ∼13 000 patients with a patient flow of ∼45 newborns entering the practice each month. The practice uses electronic medical records (EMRs), and specific information regarding nutrition is gathered at each health maintenance visit.
Power was calculated by using nQuery Advisor version 7.0 (Statistical Solutions Ltd, Cork, Ireland). For exclusive breastfeeding at 6 months, the anticipated 280 dyads per group would have 84% power to detect a pre–post increase in rates of 10 percentage points; that is, from the rate of 13% as estimated according to the 2010 CDC Report Card to 23% (a 77% rate increase).31 Larger samples were actually obtained, increasing the power to detect that same increase to 93%. Criteria for inclusion in the study were: healthy infants, ≥37 weeks’ gestation at time of birth, singleton birth, who entered the practice within the first week of life, remained in the practice for at least the first 6 months of age, and returned for their health maintenance visits at 2, 4, and 6 months. Exclusion criteria included infants with congenital birth defects, preterm infants, multiple birth, infants not residing with their birth mother, who left the practice before 6 months of age, entered the practice at >1 week of age, had missing EMR data, or had mothers with a known contraindication to breastfeeding.
Convenience samples were obtained. For the preintervention group, all newborn records for the year 2008 were included, resulting in 516 mother–infant dyads. After applying exclusion criteria, 376 records remained for review. For the postintervention group, record review began in September 2010 and continued through June 2011, for a total of 509 mother–infant dyads. Once exclusion criteria were applied, 381 records remained for review. There were 140 newborns excluded from the preintervention group and 128 excluded from the postintervention group. Primary reasons for exclusions were: born at <37 weeks’ gestation, multiple births, and left the practice before 6 months of age (Table 1).
It was expected that due to the nature of EMRs and strict staff training, most data would be recorded. The University of Virginia institutional review board approved the study as exempt before any data collection.
Implementation of the protocol began in December 2009 and continued for 10 months. This protocol included staff training, written policies, on-site lactation consultant support, community outreach, and data tracking (Table 2). Major components of this protocol included use of an AAP/WHO-approved breastfeeding curriculum for training medical staff32 and direct support with an on-site International Board Certified Lactation Consultant (IBCLC) for all breastfeeding dyads. At the initial outpatient newborn visit, an EMR template was used to uncover and document indicators for professional lactation support such as excessive infant weight loss, breastfeeding concerns, or inadequate output of stool (Table 3).33,34 Subsequent health maintenance visits include components of this template to stimulate medical providers to refer for lactation services through 6 months of age. Treatment fidelity monitoring was conducted by a committee of nurse practitioners who were also IBCLCs. They monitored staff training, audited charts for protocol adherence, tracked newborn scheduling to maintain consistency, and ensured integrity of data collection.
Patients enter the practice at birth from the local hospital or outlying hospitals at their initial outpatient newborn visit. All breastfeeding dyads are scheduled with a registered nurse (RN), who is also an IBCLC, at the same time as their initial newborn appointment. This scheduling results in an extended visit allowing focused time with an IBCLC, followed by the established standard of care assessment by the medical provider. Specifically, breastfeeding dyads begin their visit with the RN-IBCLC who provides up to 1 hour of breastfeeding support comprising an initial assessment, direct assistance, and plans for continued support. They are seen in a specially equipped room that includes a precision scale to measure milk transfer, a comfortable chair, breastfeeding pillow, and supplies that may be needed by the IBCLC. On completion of this portion of the newborn visit, the IBCLC presents her findings to the medical provider who subsequently performs a standard newborn visit. The medical provider includes International Classification of Diseases, Ninth Revision, Clinical Modification billing codes associated with newborn feeding to submit this paired encounter to insurance for reimbursement.35,36 Mothers are instructed to return for additional support, as needed, throughout their breastfeeding time frame.
For all subsequent health maintenance as well as acute visits, any concerns regarding breastfeeding are expected to generate a referral for lactation services by the medical provider. Follow-up lactation visits are also scheduled for 1 hour, coupled with an examination by a medical provider, and billed to insurance.
A monthly “meet and greet” session is advertised to the community as an opportunity for families to view the setting and meet the providers. Breastfeeding services are highlighted; a tour of the facility is provided, along with provision of information regarding a monthly prenatal breastfeeding class and a weekly drop-in moms’ support group.
De-identified data were collected via retrospective record review and entered into the SPSS version 19 program (IBM SPSS Statistics, IBM Corporation, Armonk, NY). Feeding data were recorded from the hospital, the initial newborn visit (defined as ≤1 week), and the 2-, 4-, and 6-month health maintenance visits. Demographic data abstracted included gender of the infant, birth weight, type of delivery, age of the mother, parity, and type of insurance. Although an attempt to obtain race/ethnicity data was made, this information was not routinely collected as part of the medical record, but the distribution was expected to be similar in both groups.
No feeding data were missing for the 1-week or 2-, 4-, and 6-month health maintenance visits. In-hospital newborn feeding data were missing for 23 (6.1%) of 376 infants in the preintervention group and 11 (2.9%) of 381 infants in the postintervention group.
The preintervention and postintervention groups were compared on demographic characteristics and on their rates of exclusive and any breastfeeding at each of the 5 time points. Comparisons used Student’s t test for continuous variables and χ2 tests with continuity correction for categorical variables.
Logistic regression analyses were conducted on each of the 10 outcomes (exclusive breastfeeding and any breastfeeding recorded at each of the 5 time points) to determine if a bivariate intervention effect, if any, would persist when other risk factors were taken into account. A factor was included in all 10 models if it was significantly different in the preintervention and postintervention or if it had a significant effect at α = .10 on ≥1 of the 10 outcomes in bivariate analyses performed on the whole sample. In all of the models, the factors were entered together.
Because race/ethnicity information was not available for 58% of the records, no imputation was done and no statistical tests were conducted on this factor.
There were no significant differences between the groups for infant gender, birth weight, type of delivery, age of the mother, or type of insurance. A pre–post increase in the proportion of primiparous mothers from 40.4% to 53.8% (P < .001) was noted (Table 4).
Rates for exclusive breastfeeding were higher in the postintervention group than in the preintervention group at every time point, and the differences in the rates were statistically significant at all 5 time points (P < .01). Of particular note, when comparing exclusive breastfeeding rates, the postintervention group rates were at least 10 percentage points higher than those in the preintervention group at all 5 time points (Fig 1).
Rates for any breastfeeding were higher in the postintervention group than in the preintervention group at every time point, but the differences in the group rates were statistically significant only at the 1-week time point (P = .021) (Fig 2).
The finding that the percentage of first-time mothers was significantly higher in the postintervention group versus the preintervention group raised the question as to whether this had an effect on the results. The χ2 tests comparing rates of exclusive breastfeeding and any breastfeeding preintervention versus postintervention, at all 5 time points, were conducted separately for primiparous and multiparous mothers. The results were almost the same as for the parity groups combined, the only exception being that at week 1 for the multiparous mothers, the increase in the exclusive breastfeeding rate was only 9.6 percentage points, which, with P = .059, fell just short of significance at the α = .05 level.
Logistic regression analyses were conducted on each of the 10 outcomes: exclusive breastfeeding (1 = yes, 0 = no) and any breastfeeding (1 = yes, 0 = no) recorded at each of the 5 time points. The risk factors included in each model were: group (1 = postintervention, 0 = preintervention), primiparous (1 = primiparous, 0 = multiparous), birth method (1 = cesarean delivery, 0 = spontaneous vaginal delivery), insurance (1 = Medicaid, 0 = private), birth weight (100-g units), and mother’s age (years). The primiparous factor was included because the rates for this differed significantly between the preintervention and postintervention groups. Each of the other 5 factors had a significant effect at the α = .10 level in bivariate analyses on the whole sample for ≥2 of the 10 breastfeeding outcomes.
The group indicator was highly significant for exclusive breastfeeding at every time point (P < .001) but not significant for any breastfeeding except at week 1 (P = .016). The estimated odds of exclusive or any breastfeeding for those in the postintervention group were higher in every case than for those in the preintervention group; odds ratios varied between 1.67 and 2.66 for the 5 exclusive breastfeeding outcomes, and they varied between 1.1 and 1.77 for the 5 any breastfeeding outcomes.
The primiparous indicator was not significant in any of the 10 regressions, with P values ranging from .163 to .862. However, the insurance indicator was highly significant in all 10 regressions (P < .004 in every case), with the estimated odds of any or exclusive breastfeeding for those receiving Medicaid being only 28% to 47% of the odds for those with private insurance. The effect of birth weight on breastfeeding was significant at 2, 4, and 6 months after birth. At 2 and 6 months, the estimated odds of exclusive breastfeeding increased by 4% to 5% with every additional 100 g of birth weight. At 2, 4, and 6 months, the estimated odds of any breastfeeding increased 4% to 6% with every additional 100 g of birth weight. The only other predictor that was significant at any time was birth method. In the hospital, the odds of exclusive breastfeeding were significantly lower for those who had a cesarean delivery compared with those with vaginal delivery (odds ratio: 0.51; P < .001). But birth method was not a significant predictor of exclusive breastfeeding at 1 week or at 2 and 4 months. It was significant again at 6 months (odds ratio: 0.661; P = .017). Birth method was not a significant predictor of any breastfeeding at any time point. Mother’s age was not significant in any of the 10 models, despite its effect (P < .10) on 5 of the 10 outcomes in bivariate analyses.
The χ2 tests and logistic regression analyses showed that rates of exclusive breastfeeding were significantly higher for those in the postintervention group. Although exclusive breastfeeding increased significantly, the increases in rates for any breastfeeding were not significant except for the 1-week time point. One possible explanation is, for mothers who were exclusively breastfeeding in the early postpartum period, breastfeeding persisted because formula was not introduced. Forster et al37 found a negative association between introduction of formula in the hospital and breastfeeding duration. Hörnell et al38 reported that the earlier an infant was given formula, the shorter the duration of breastfeeding, while introduction of solids for the exclusively breastfed infants had no association with total breastfeeding duration.
The initiation rate for the exclusively breastfeeding postintervention group was much higher compared with the preintervention group (78% and 59%, respectively). This finding might be explained by 1 of the features of the protocol, which provides a prenatal meet and greet session to expose parents to an environment that supports breastfeeding mothers. Mothers who entered the practice after the breastfeeding-friendly office protocol was in place may have had increased motivation to initiate breastfeeding because they knew they would have continued support. Alternatively, it may be that mothers who were interested in breastfeeding sought out a supportive environment, therefore creating the potential for selection bias. Measurement of predelivery intention to breastfeed could have eliminated this as a concern.
The literature suggests primiparous women have more difficulty sustaining breastfeeding and they may breastfeed longer if they are knowledgeable and self-confident.39–41 We found similar results; when the primiparous mothers were compared preintervention and postintervention for exclusive and any breastfeeding, both rates increased at every time point, and the pre–post increases for exclusive breastfeeding were statistically significant (P < .05) at every time point. Therefore, the ABM protocol may have had a positive effect on exclusivity for the primiparous mothers in this study.
The literature describes mothers’ predelivery intention to breastfeed as a strong predictor of initiating breastfeeding and continuing through early obstacles.39–44 The ABM protocol provided a mechanism for extending components that support breastfeeding into the primary care setting, creating an environment that supports a mother’s intention to breastfeed. Continued lactation consultant support is emphasized in the Surgeon General’s Call to Action and is called for as part of the breastfeeding-friendly office protocol. Direct support for breastfeeding dyads can be time intensive. Pairing IBCLC services with medical professionals who are also educated in breastfeeding creates a vehicle for access and reimbursement, and it teams health professionals with shared intentions.
This study was limited by a lack of data regarding other parameters that may affect exclusivity and duration of breastfeeding such as mother’s educational level, social support, smoking, work, and socioeconomic status.43,44
The results from this study suggest that the use of a breastfeeding-friendly clinical protocol in the primary care setting may help increase exclusive (no formula) breastfeeding rates up to 6 months of age. The risks associated with not providing breast milk are numerous.45–52 The ABM protocol provides a template that encompasses interventions supported by the literature to increase breastfeeding rates.
Although not a randomized controlled trial, implementation of the ABM clinical protocol with preintervention and postintervention data collection provided an important evaluation of this program with useful data for clinical practice. This study provides an example of how the protocol can be operationalized. Further studies should explore use of this protocol, inclusive of patient populations and/or regions known to have low breastfeeding rates, to help expand its use and address this important public health initiative.
This work was supported by Loudoun Pediatric Associates, a pediatric primary care clinic that is part of Loudoun Medical Group located in Leesburg, Virginia, and by the University of Virginia, Charlottesville, Virginia.
- Accepted January 12, 2013.
- Address correspondence to Sharon K. Corriveau, DNP, RN, CFNP, IBCLC, 2001 Lakeport Way, Reston, VA 20191. E-mail:
Drs Corriveau, Drake, and Kellams participated in the concept and design of the study, analysis and interpretation of data, and drafting of the manuscript; Dr Rovnyak participated in the analysis and interpretation of data; and all authors approved the manuscript as submitted.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
- ↵Centers for Disease Control and Prevention. Breastfeeding among US children born 2000–2009, CDC national immunization survey. Atlanta, GA: Centers for Disease Control and Prevention; 2012. Available at: www.cdc.gov/breastfeeding/data/NIS_data/index.htm. Accessed August 2012.
- ↵Eidelman AI, Schanler RJ, Johnston M, et al; American Academy of Pediatrics. Section on Breastfeeding. Breastfeeding and the use of human milk. Pediatrics. Published February 27, 2012. Available at: http://pediatrics.aappublications.org/content/early/2012/02/22/peds.2011-3552
- ↵World Health Organization and United Nations Children’s Fund. Baby-Friendly Hospital Initiative revised, updated and expanded for integrated care. Published 2009. Available at: http://whqlibdoc.who.int/publications/2009/9789241594967_eng.pdf
- ↵US Department of Health and Human Services. Maternal, infant, and child health. Healthy People 2020; 2010. Available at: http://healthypeople.gov/2020/topicsobjectives2020/overview.aspx?topicid=26
- ↵US Department of Health and Human Services. The Surgeon General’s Call to Action to Support Breastfeeding. Washington, DC: US Department of Health and Human Services, Office of the Surgeon General; 2011. Available at: www.surgeongeneral.gov/topics/breastfeeding/calltoactiontosupportbreastfeeding.pdf
- Merewood A,
- Philipp BL,
- Chawla N,
- Cimo S
- Merewood A,
- Patel B,
- Newton KN,
- et al
- Feldman-Winter L,
- Barone L,
- Milcarek B,
- et al
- Guise JM,
- Palda V,
- Westhoff C,
- Chan BK,
- Helfand M,
- Lieu TA,
- US Preventive Services Task Force
- Bonuck KA,
- Trombley M,
- Freeman K,
- McKee D
- Labarere J,
- Gelbert-Baudino N,
- Ayral AS,
- et al
- Lukac M,
- Riley JK,
- Humphrey AD
- Heinig MJ,
- Farley K
- Nichols J,
- Schutte NS,
- Brown RF,
- Dennis CL,
- Price I
- ↵Cardosa LO, Vicente AS, Damião JJ, Rito RV. The impact of implementation of the breastfeeding friendly primary care initiative on the prevalence rates of breastfeeding and causes of consultations at a basic healthcare center. J Pediatr (Rio J). 2008;84(2):147–153
- ↵Centers for Disease Control and Prevention. Breastfeeding among US children born 1999–2007, CDC national immunization survey. Atlanta, GA: Centers for Disease Control and Prevention; 2010. Available at: www.cdc.gov/breastfeeding/pdf/BreastfeedingReportCard2010.pdf
- ↵Wellstart International. Wellstart’s lactation management self-study modules, level 1. 3rd ed. Available at: www.wellstart.org/
- ↵International Lactation Consultant Association. Clinician’s breastfeeding triage tool, 2007. Available at: www.ilca.org/files/resources/HealthCareProviders/BREASTFEEDING%20TRIAGE-references-Rev2.pdf
- ↵Massachusetts Breastfeeding Coalition. Approach to early breastfeeding: a guideline for healthy term newborns, 48 hours to 2 weeks. Available at: http://massbreastfeeding.org/providers/earlyBF.html.
- ↵American Academy of Pediatrics. Supporting breastfeeding and lactation: the primary care pediatrician’s guide to getting paid. Available at: http://www2.aap.org/breastfeeding/files/pdf/coding.pdf
- Geraghty SR,
- Riddle SW,
- Shaikh U
- ↵Forster DA, McLachlan HL, Lumley J. Factors associated with breastfeeding at six months postpartum in a group of Australian women. Int Breastfeed J. 2006;1:18
- Dewey KG,
- Nommsen-Rivers LA,
- Heinig MJ,
- Cohen RJ
- Kronborg H,
- Vaeth M
- Ahluwalia IB,
- Morrow B,
- Hsia J
- ↵Duijts L, Jaddoe VW, Hofman A, Moll HA. Prolonged and exclusive breastfeeding reduces the risk of infectious diseases in infancy. Pediatrics. 2010;126(1):18–26
- Ip S, Chung M, Raman G, et al; Tufts-New England Medical Center Evidence-based Practice Center. Breastfeeding and maternal and infant health outcomes in developed countries. Evid Rep Technol. 2007;153(153):1–186
- Hauck FR,
- Thompson JM,
- Tanabe KO,
- Moon RY,
- Vennemann MM
- Grummer-Strawn LM,
- Mei Z,
- Centers for Disease Control and Prevention Pediatric Nutrition Surveillance System
- Strathearn L,
- Mamun AA,
- Najman JM,
- O’Callaghan MJ
- Bartick M,
- Reinhold A
- Copyright © 2013 by the American Academy of Pediatrics