Evidence is Clear: Baby-Friendly Hospital Initiative Increases Breastfeeding Rates in the US and Closes Breastfeeding Disparities

By Melissa Bartick, MD, MSc and Nathan Nickel, MPH, PhD

The US Preventative Services Task Force (USPSTF) published its recommendations regarding breastfeeding promotion in the October 25, 2016 issue of JAMA, recommending individual efforts at breastfeeding promotion, but pointedly noting that systemic promotion efforts, such as the Baby-Friendly Hospital Initiative (BFHI), were outside its scope. The accompanying literature review, performed by the USPSTF team, purposely looked only at two trials of BFHI and a few randomized trials of its component Ten Steps, and concluded there was mixed evidence to support BFHI. The two trials they reviewed on BFHI both supported its efficacy, at least in less educated mothers (here and here).  One of the BFHI trials they reviewed was an observational trial, and the other was a before-and-after trial, yet several other US trials with similar methodologies exist which showed positive outcomes, but these were not even mentioned in the  literature review. For example, the literature review did not include this national trial showing a correlation of BFHI with increased breastfeeding rates and excluded national data from the CDC showing rising breastfeeding rates as percentage of live births in Baby-Friendly hospitals rose. The literature review acknowledged that other studies supported the effectiveness of BFHI. However, an accompanying editorial by Flaherman and von Kohorn concluded that interventions such as BFHI “should be reconsidered until good-quality evidence that these interventions are safe and effective.”

Despite the weak literature review, the editorial’s surprising conclusion can in no way be drawn from the evidence presented by the USPSTF, let alone the evidence as a whole. Yet this editorial is what is garnering the most media attention. Interestingly, the editorial does support previous research by one of its authors, Valerie Flaherman, who found that small amounts of formula help women breastfeed longer. This finding, which contradicts previous evidence (here and here) that non-indicated supplemental formula is a strongly associated with breastfeeding failure, would negate Step 6 of the Baby-Friendly Hospital Initiative, “to give no other food or drink besides breast milk without a medical indication.” One of the co-authors of Flaherman’s study disclosed that he worked for several formula companies. Because Flaherman is still conducting similar government-funded research on formula supplementation of breastfed infants, which is incompatible with Baby-Friendly, JAMA should have chosen an editorialist who could be objective about the weight of the evidence on Baby-Friendly as well as include an editorial with an opposing viewpoint in the same publication– especially given the widespread endorsement of the Ten Steps among major US and world medical organizations.

The editorial gives readers the impression that there is no robust evidence from randomized control trials on the effectiveness of the Baby-Friendly Hospital Initiative, but we see that this is not true. In addition, literature review notes that they excluded the world’s largest randomized study of Baby-Friendly, PROBIT, published in JAMA in 2001, because it was conducted in Belarus and not in a high-income country. PROBIT, led by Canadian researcher Michael Kramer, included over 17,000 mother-infant pairs and found that exclusive breastfeeding at 3 months went from 6% to 43% in areas where the facilities were randomly assigned to the Baby-Friendly type intervention. What is more important than the income status of Belarus is whether maternity care practices there are similar to those in much of the US. Indeed, PROBIT authors state they chose Belarus rather than North America or Western Europe “because maternity hospital practices in Belarus and other former Soviet republics are similar to those in North America and Western Europe 20 to 30 years ago and thus provide a greater potential contrast between intervention and control study sites. However, Belarus resembles Western developed countries in 1 very important respect: basic health services and sanitary conditions are very similar.” Thus, the effectiveness of Baby-Friendly  is only a question of magnitude. And, notably, many parts of the United States, such as parts of the deep south, arguably still practice maternity care the way it was in the US in the 1970s and 1980’s, contributing to racial and geographic disparities in breastfeeding and health outcomes.

A national survey of US Baby-Friendly hospitals compared to hospitals that were not designated Baby-Friendly, the hospitals designated as Baby-Friendly in 2001 had elevated rates of breastfeeding initiation and exclusivity, regardless of demographic factors that are traditionally linked with low breastfeeding rates. The number of Baby-Friendly practices mothers self-report having received has been shown to correlate with breastfeeding outcome at 6 weeks; and this effect is additive (see references here and here).  Before-and-after studies in an inner city hospital has shown how implementation of BFHI has increased breastfeeding rates among populations with historically low breastfeeding rates, thereby lowering disparities (see here, here and here)

A Cochrane review of randomized trials of skin-to-skin contact, part of BFHI, have shown enhanced breastfeeding status and decrease in hypothermia in late preterm infants, and other studies have found similar effects (here and here). Comparison studies have also been done on rooming in showing positive effects (here, here, and here). By contrast, formula supplementation evidence, for example, is mostly from strong observational studies.

One way to look at the correlation between BFHI, the Ten Steps, and Breastfeeding Rates is to look at national data itself from the CDC Breastfeeding Report Cards and the CDC National Immunization Survey, for the years 2007 to 2013, the years in which we have data on the percentage of births in Baby-Friendly hospitals from the CDC. We can look at the following metrics: the number of Baby-Friendly designated hospitals, the percentage of live births at Baby-Friendly Hospitals, the rate of exclusive breastfeeding at 3 months, and the average national mPINC score. The mPINC is a survey given biannually by the Centers for Disease Control and Prevention to all US maternity facilities ever since 2008. Scored on a 100-point scale, it measures the extent of implementation of the Ten Steps of Successful Breastfeeding that comprise the BFHI. This data show that the mathematical correlation between the increase in births born at Baby-Friendly hospitals and exclusive breastfeeding at 3 months is 0.93, which is extremely high.

It is true that there is a paucity of randomized control trials both on BFHI in the US, and on components of the Ten Steps. One wonders, however: would it be feasible or ethical to conduct further randomized trials of Baby-Friendly at this point? For example, skin to skin contact, breastfeeding within the first hour of life, and keeping baby and mother in close proximity have been found to have numerous beneficial effects on mother and baby. What’s more, non-exclusive breastfeeding itself is a risk factor for SIDS and multiple pediatric infectious diseases. The ethics of randomizing infants to receive non-medically indicated supplements would be problematic, as would the ethics of randomizing hospitals to agree not to pursue Baby-Friendly status. For example, the Ten Steps are endorsed by the American Academy of Pediatrics (AAP),the American College of Obstetricians and Gynecologists (ACOG); hospitals would have to be willing to be randomized to care practices that are not considered the standard of excellence by the AAP, ACOG, the Surgeon General, and the World Health Organization, and this itself may also introduce confounders.

In sum, the best we can do, at this point, is accept those randomized trials we do have, including PROBIT, the strong observational studies, and the before-and-after studies. Together, these offer a compelling picture that Baby-Friendly is an effective means for increasing breastfeeding rates in the post-partum period and beyond. As with any scientific endeavor, we must be open to changing any component of the Ten Steps should further compelling evidence become available.

Melissa Bartick, MD, MSc is an internist at Cambridge Health Alliance and Assistant Professor at Harvard Medical School. You can follow her on Twitter at @MelissaBartick .

Nathan Nickel, PhD is a public health research scientist at the University of Manitoba and Secretary of the Breastfeeding Forum of the American Public Health Association. You can follow him on Twitter at @Nickel_NC .

Posts on this blog reflect the opinions of individual ABM members, not the organization as a whole.