Bacteria in Expressed Breastmilk from Mothers of Premature Infants and Maternal Hygienic Status

Dear Editor:

Feeding the premature infant with breastmilk is associated with a lower incidence of necrotizing enterocolitis (NEC) and nosocomial sepsis. ¹ However, there is concern that bacterial contamination of expressed breastmilk (EBM) may theoretically pose an infective risk to these infants. There is sufficient evidence from human milk banking that breastmilk is not sterile, but that it has various organisms ranging from Gram-positive skin colonizers to Gram-negative enteric bacteria. ² The extent of bacterial presence is variable even within the individual, and some have linked this to personal hygiene. ³ However, the definition is still dubious as to which and what extent of organisms present are considered pathogenic. Our group has previously detected bacterial contamination of EBM collected either using breast pumps or by manual expression at home without supervision that was temporally related with several outbreaks of NEC. As a result, routine surveillance of breastmilk cultures was enforced, and pasteurization of the mother's own breastmilk was adopted. We are currently challenged to review this practice as lactational support services and breastfeeding education have remarkably improved. We are also cognizant that raw breastmilk is highly protective against NEC, whereas pasteurization may reduce the immune factors and calories in the milk. ⁴ For quality improvement purposes and to strategize infection control procedures for a safer and higher-quality use of breastmilk in our neonatal intensive care unit, we carried out a Plan-Do-Study-Act cycle to identify the commonly occurring bacteria in EBM of mothers with premature births, the extent that these showed significant bacterial counts, and to identify some evident behavioral risk factors that may be related.

The study was conducted in the neonatal intensive care unit of the Universiti Kebangsaan Malaysia Medical Centre (Kuala Lumpur, Malaysia) and was completed in April 2011. The hospital provided the relevant funding (project code FF-311-2010), and the protocol was approved by the Research and Human Ethics Committee of the Universiti Kebangsaan Malaysia Medical Centre. One milk sample each from 35 randomly selected mothers who consented to the study was taken for culture according to the method of Miles and Misra. ⁵ Significant bacterial contamination of breastmilk was defined as a culture of more than 10⁴ colony-forming units/mL. The mothers were interviewed with respect to personal hygienic practices such as daily cleansing rituals, ablution after using the toilet, hand hygiene, and specific breast care (e.g., cleansing of the breast before expression and discarding the first few drops of milk). Other questions included whether a breast pump was used and its subsequent decontamination process, as well as storage and transportation of the breastmilk after expression.

Of the mothers surveyed, their infants were born at an average of 31 weeks and had a birth weight of 1,400 g. The majority of women expressed their breastmilk at home using a breast pump. In total, 31 of 35 (88%) EBM samples were culture-positive. More than half of the samples were of single growth, and a third were of mixed growth. The most commonly isolated organisms were Gram-negative enteric bacteria, namely, Acinetobacter sp. and Klebsiella sp., and the Gram-positive skin commensal, coagulase-negative Staphylococcus, as shown in Table 1. Fourteen (40%) of the EBM samples had substantial bacterial growth in excess of 10⁴ colony-forming units/mL. Overall, there were no statistically significant differences in the demographic or hygienic characteristics between mothers with heavy and low bacterial counts in EBM. Only a small minority (<10%) in this survey reported poor hand hygiene or sanitary practices. Low bacterial count in EBM was observed more frequently from mothers with higher educational status and order of pregnancies, those who delivered by cesarean section, and those with previous experience in expressing breastmilk. Conversely, heavy bacterial growth in EBM tended to appear after the first postnatal week (median, 13 days; p=0.07).

This small preliminary survey suggests that maternal hygiene and sanitary practices were not significant sources of bacterial presence in EBM and is supportive of cumulative evidence that breastmilk has significant bacterial content even of Gram-negative organisms and that behavioral modification may not diminish this. ⁶ Premature infants with increasing postconception age also showed rectal colonization with enteric bacteria when fed unpasteurized breastmilk. However, this was not associated with any significant morbidity and has been speculated to be a form of transmission of bacterial flora from mother to infant for priming of gastrointestinal immunity. Furthermore, pasteurization potentially removes the various anti-infective properties in raw breastmilk that may inhibit bacterial proliferation and translocation. In completing our Plan-Do-Study-Act cycle, we have instituted an action plan to promote increased hand and personal hygiene awareness, directly supervise breastmilk expression by trained healthcare workers, and encourage more frequent maternal visitation to escalate feeding of raw breastmilk to their infants. Only breastmilk specimens collected unsupervised at home are selectively pasteurized. These measures could impact on our NEC rates, of which our most recent continuous surveillance data seem to positively indicate. Larger-scale studies relating bacterial presence in EBM, maternal hygiene, and the incidence of NEC in premature infants are warranted.