In the 10 years after Switzerland introduced the Baby-Friendly Hospital Initiative (BFHI), a program designed by UNICEF to ensure that maternity facilities encourage breast-feeding, the median duration of any breast-feeding increased from 22 to 31 weeks nationwide.1 Although the change cannot be attributed entirely to the initiative, an analysis of individual and hospital data indicate that use of specific practices to promote breast-feeding on the maternity ward, hospitals’ BFHI certification status and hospitals’ level of compliance with initiative criteria all were associated with a significantly elevated likelihood of breast-feeding.

In 2003, researchers distributed surveys to a randomly selected national sample of women who had given birth in the past nine months. Mothers were asked for information about their pregnancy and birth and about the hospital where they delivered. They also were asked what they had fed their infant in the past 24 hours and how old the infant had been when they introduced various foods and liquids. Feeding practices were categorized as exclusive breast-feeding (feeding of breast milk only), predominant breast-feeding (feeding of breast milk plus water-based liquids), full breast-feeding (the sum of exclusive and predominant breast-feeding) and any breast-feeding. For comparison, the researchers used selected data from the previous national survey on breast-feeding, conducted in 1994, the year after the BFHI was introduced. In addition, hospitals’ BFHI certification status was determined, and monitoring data collected from hospitals that were certified or being evaluated for certification were analyzed to assess their level of compliance with BFHI criteria.

The study sample included 2,812 mothers and their 2,861 infants aged 0–11 months, born in 146 hospitals. The mothers had an average age of 32 years, and 89% were married. A large majority (81%) were Swiss nationals. Some 14% of mothers and 20% of fathers had college degrees. Three-fourths of the mothers were employed before their delivery. For half, the birth was their first. Nine in 10 of the infants were aged 2–9 months. Their birth weight averaged 3,210 g, but 6% weighed less than 2,500 g. Sixty-two percent of infants were born in hospitals that had achieved or were being evaluated for baby-friendly certification; 18% were born in ones with high compliance with BFHI guidelines, 26% in ones with low compliance and 18% in ones with unknown compliance.

The median duration of any breast-feeding was considerably longer in 2003 than in 1994 (31 vs. 22 weeks); the median duration of full breast-feeding increased from 15 to 17 weeks over the decade. Maternal age, education and annual income were all significantly and positively associated with the duration of exclusive, full and any breast-feeding.

A variety of practices that are among the BFHI certification criteria and that mothers experienced on maternity wards—feeding breast milk exclusively, rooming in, first suckling within one hour, breast-feeding on demand, absence of pacifiers and no provision of free infant formula—were also significantly and positively associated with median durations of exclusive, full and any breast-feeding. Conversely, in multivariate analyses, lack of each of these practices was associated with a significantly elevated risk that such breast-feeding would not occur. The association was strongest for types of liquids fed on the maternity ward: Compared with infants fed exclusively breast milk at that time, infants also fed formula had a more than doubling of the risk of not being exclusively breast-fed, fully breast-fed and breast-fed at all (adjusted hazard ratios, 2.1– 2.3). Similarly, relative to their counterparts fed only breast milk on the maternity ward, infants also fed water-based liquids had elevated risks of these outcomes (1.2–1.5).

Hospitals’ BFHI certification status also was significantly associated with breast-feeding practices. For example, among infants younger than four months of age, 60% of those born in BFHI-certified hospitals were exclusively breast-fed, compared with 51% of those born in hospitals being evaluated for certification and 49% of those born in non–baby-friendly hospitals. The corresponding proportions of infants fully breast-fed were 72%, 64% and 60%. Patterns were similar (although proportions were somewhat lower) among infants younger than six months of age.

Finally, hospitals’ level of compliance with BFHI criteria was significantly associated with breast-feeding. For example, in adjusted analyses, the median duration of exclusive breast-feeding was six weeks among infants born in non–baby-friendly hospitals, 10 weeks among those born in baby-friendly hospitals with low compliance and 12 weeks among infants born in baby-friendly hospitals with high compliance. Patterns were similar for full and for any breast-feeding. In Cox regression analyses, the risk of not being exclusively breast-fed was reduced among infants born in hospitals with low compliance (adjusted hazard ratio, 0.9) and more so among infants born in hospitals with high compliance (0.8) relative to infants born in non–baby-friendly hospitals. For full breast-feeding and for any breast-feeding, the association was significant only for infants born in hospitals with high compliance.

Commenting on the results, the investigators assert that introduction of the BFHI in Switzerland about a decade ago has been at least partly responsible for the improvements in breast-feeding practices observed nationally since then. They speculate that the initiative has indirectly influenced practices even at noncertified hospitals and has generally elevated public awareness about the benefits of breast-feeding. However, they add, further effort is needed to meet the goal of exclusive breast-feeding of infants for the first six months of life. Given the evident importance of hospitals’ compliance with the initiative’s criteria for prolonging the duration of breast-feeding, “monitoring of compliance in designated hospitals is indispensable for promoting the optimal effects of the BFHI,” they contend.—S. London


1. Merten S, Dratva J and Ackermann-Liebrich U, Do baby-friendly hospitals influence breast-feeding duration on a national level? Pediatrics, 2005, 116(5): e702–e708, < peds.2005-0537>, accessed Nov. 1, 2005.