Dear Colleagues and Neonatology Experts,

We are currently rewriting our internal protocols regarding bacteriological screening and CMV management of maternal milk (MOM) for hospitalized preterm infants.

While our primary goal is maximum safety, we are guided by the principle of non-maleficence (Primum non nocere). This creates a complex clinical challenge: how do we protect the infant from potential pathogens (infection risk) without causing harm by destroying vital bioactive and immunological components through over-processing (nutritional/developmental risk)? Furthermore, we must avoid the harm of unnecessarily discarding precious breastmilk—a practice that exhausts resources and can lead to "pumping fatigue," discouraging mothers from continuing their lactation journey.

We have observed that practices vary widely between centers and countries, often based on historical habits rather than robust data. We are seeking a secure yet pragmatic solution and would value your insights on the following:

• Patient Selection & Duration: Do you perform systematic bacteriological controls on MOM for specific populations (e.g., all VLBW, or only those <28 weeks, none of them)? For how long do you continue these controls (e.g., until a specific post-natal age)?

• Frequency & Uncertainty: How often do you test the milk (weekly, every batch)? More importantly, how do you manage the "window of uncertainty" between controls?

• Bacteriological Thresholds: How do you pragmatically define "acceptable" milk?

• The CMV Challenge: For CMV-seropositive mothers, do you mandate Holder pasteurization? If so, for which gestational age groups and for what duration ?

• Resource Management: How do you ensure your protocol doesn't lead to excessive milk disposal or discourage the mother's involvement?

We look forward to hearing about your center's experience, the evidence-base you utilize, and how you navigate these "grey areas."

Thanks in advance for your valuable contributions.

@AntoineBachy this is a critical topic, and in my view, current practices in many centers may overemphasize theoretical microbiological risks at the expense of the well-established benefits of mother’s own milk (MOM) for preterm infants.

Based on more than 40 years of experience in a tertiary NICU in southern Brazil—and systematic data collection since 2011 within the Vermont Oxford Network—we have adopted a deliberately non-interventionist approach regarding bacteriological screening and CMV management of MOM. Answering to your questions:

Patient Selection & Duration:

We do not perform routine bacteriological screening of MOM in any subgroup, including extremely preterm or VLBW infants. Screening is reserved exclusively for cases with clinical suspicion of infection, particularly sepsis-like syndromes. In our experience, routine surveillance has not demonstrated sufficient benefit to justify its risks, costs, and unintended consequences.

Frequency & Uncertainty:

We do not perform periodic microbiological testing. Instead, risk mitigation relies on strict adherence to standardized protocols for milk expression, storage, and handling, as defined by the Brazilian Human Milk Bank Network. We believe this preventive approach is more effective than intermittent culturing, which may create both false reassurance and unnecessary interventions.

Bacteriological Thresholds:

We do not routinely culture MOM in our NICU. However, when microbiological control is applied—particularly within human milk banks—it follows well-defined technical criteria rather than arbitrary thresholds.

According to Brazilian standards, screening targets Gram-negative bacilli, defined as aerobic or facultative anaerobic, non-spore-forming, oxidase-negative microorganisms capable of growing in the presence of bile salts or surfactants, and fermenting lactose with acid, gas, and aldehyde production at 35°C within 24–48 hours, often with β-galactosidase activity.

This reflects a targeted safety strategy, focusing on clinically relevant pathogens. Importantly, these criteria are embedded within a structured quality-control system for donor milk and do not support routine bacteriological screening of MOM in NICU settings.

The CMV Challenge:

We do not perform routine CMV screening in mothers and do not apply Holder pasteurization to MOM based on serostatus. This is consistent with current understanding that postnatally acquired CMV infection differs significantly from congenital disease and is not a contraindication to breastfeeding, as stated by the American Academy of Pediatrics.

Our preventive strategy is internally consistent: we ensure CMV-safe blood transfusions (CMV-negative or leukoreduced) for extremely preterm infants (<30 weeks), addressing a well-established route of transmission that is sometimes overlooked when focus is placed primarily on breast milk.

Across decades of practice, severe CMV disease associated with MOM has been exceedingly rare in our unit. When it occurred, it presented as a sepsis-like syndrome and responded well to short-course antiviral therapy.

Resource Management:

We consider preservation of MOM a priority. Milk disposal is minimized and limited to exceptional, clinically justified situations. Protocols that lead to frequent discard or excessive processing risk not only reducing the biological value of milk but also discouraging maternal participation and contributing to pumping fatigue—which, in our view, directly conflicts with the principle of primum non nocere.

It is important to highlight that Brazil has the largest human milk bank network in the world, supported by robust, standardized national regulations governing collection, processing, pasteurization, and distribution of human milk.

These include:

• Clearly defined microbiological screening criteria (focused on Gram-negative pathogens)

• Standardized Holder pasteurization (62.5°C for 30 minutes) for donor milk

• Mandatory post-pasteurization microbiological quality control

Notably, even within this highly regulated system, routine CMV screening is not performed, and pasteurization is primarily reserved for donor milk—not MOM, which is only pasteurized if the production exceeds the maximum freezing time for storage.

Our position is that routine bacteriological screening and CMV-driven pasteurization of MOM are not supported by current evidence and may result in more harm than benefit.

A strategy centered on preventive practices, standardized handling, and clinical vigilance is, in our experience, safer, more sustainable, and better aligned with the unique biological value of human milk.

We recognize that practices vary across settings, but we would strongly encourage reconsideration of protocols that prioritize theoretical risks over demonstrated benefits.

I would love to know how other NICUs face these challenges and let me know if I can help with anything else.

Thanks @AntoineBachy , important topic.

The Swedish organisation MilkNet that organises all milk banks in Sweden have a a guideline only from 2016, you can download it here and ChatGPT can probably translate it well. My colleageu is leading MilkNet so I asked her. The very short version is that MoM is used as is, and without control.

• Patient Selection & Duration: No systematic bacteriological controls for anyone.

• Frequency & Uncertainty: n/a

• Bacteriological Thresholds: n/a

• The CMV Challenge: CMV status is not checked systematically.

• Resource Management: Swedish NICUs typically have "nutrition rooms" and dedicated staff (one nurse assistant per shift) that ensures little MoM waste

Dear Antoine,

This is a very important topic, and I can say in advance that practices differ between countries. I will try to present the Estonian approach here, which applies in most centers dealing with high-risk newborns:

1. We do not perform routine bacteriological analysis at any gestational age. We only analyze breast milk if the infant develops a clinical problem – such as signs of sepsis or feeding difficulties, etc.

2. In Estonia, we try to follow Swedish donor milk bank guideline thresholds, which were originally and historically established for breast milk intended to be given raw. An important aspect here is that this approach is individualized and depends on gestational age. First, we assess the overall clinical picture.

   https://www.milknet.se/

3. We do not routinely check maternal seropositivity. We think that ~80% of them are seropositive. Holder pasteurization is performed only if there is a diagnosis of congenital or acquired infection.

4. Our protocol is relatively liberal, and we believe that a mother’s own raw milk is the best for own child. One center freezes maternal breast milk for 72 hours to reduce CMV presence, but it is known that this method does not eliminate CMV.

I hope it helped. Looking for other opinions!

Best

Annika (Tartu, Estonia)

Dear Antonine Thank you for rained this important topic !

Thank you for raising this important topic. Regarding bacteriological testing of milk, we do not perform it routinely in Poland; instead, we instruct mothers on proper milk expression and hygiene practices.

There are, however, many inconsistencies in Poland concerning CMV. Over 70% of the population is IgG positive, and even a positive IgM result does not necessarily indicate the presence of the virus in milk. We have conducted such studies in mother willing to become donors, and there is no clear correlation beetwen seroprevelence status and viremia at milk. In my opinion, Holder pasteurisation of a mother’s own milk represents a significant loss, yet many centres implement it without confirming whether the virus is actually present in the milk. This would require real-time PCR; however, I do not have sufficient experience to propose a specific protocol.