Preparation of infant milk formula

Special feed units in paediatric hospitals prepare powdered feeds for high-risk patients, who could suffer significant clinical complications from microorganism ingestion. Outbreaks of Cronobacter sazazakii have been associated with infant formula. These outbreaks have led to severe complications in infants including a rare case of meningitis, necrotising enterocolitis and sepsis. Following reconstitution of the powdered milk-feeds and other products from the special feeds unit, they are usually treated with either pasteurization or blast chilling. The purpose of these treatments is to reduce/eliminate microbial load, which may present a risk for high-risk patients.

The definition of high-risk patients varies from organization to organization but at Great Ormond Street Hospital for Children NHS Trust it includes:

• Premature infants (<37/40).
• Neonates (infants <1 month) and infants on neonatal intensive care units, with the exception of powders which are added to expressed breast milk.
• Bone marrow transplant and other immune-compromised children have all powdered milk feeds pasteurized.
• Infants (<1 year) on Locasol, jejunal feeds made up from powders or with added non-sterile ingredients are pasteurized.

Pasteurisation for the context of this study is where feeds are placed in a pasteurizer and the temperature raised to 67.5^oC for 4 minutes, followed by rapid cooling to less than 10^oC. Blast chilling involves the placement of feeds into blast chillers where rapid cooling occurs to less than 5^oC. Following both methods of preparation, feeds are placed into a holding fridge at a temperature of less than 5^oC until delivery to the ward.

There is little published evidence regarding which treatment method provides the safest product for high-risk children requiring infant formula. A two-phase research project has been conducted at Great Ormond Street Hospital. Phase one collected data on the variety of treatment methods in place across the United Kingdom and the rationale for use. The majority of hospitals that responded to the survey utilized blast chilling as their sole method for treatment.

Phase two of the study evaluated the microbial load across a variety of infant feed powders following treatment with either pasteurisation or blast chilling. There was no statistical difference between treatment conditions (p>0.3), however there was a statistical difference in microbial load between infant formula types (p<0.001), indicating that variance is a result of the feed type rather than the processing.

The most commonly used feed with the lowest microbial load was used for an inoculation study, where feeds were inoculated separately with >10⁷ cells of three different organisms following reconstitution. Feeds were then returned to the unit for treatment. The three microorganisms chosen for the study represented what could be common contaminants during feed preparation: Enterobacter species within the milk formula, Staphylococcus aureus from the person preparing the feed and Pseudomonas aeruginosa from the water used. Across all species inoculated with 10⁷ organisms, 10⁶ were detected at the pre-processing stage in all feeds.

No statistically significant change in microbial load was detected after post-processing with blast chilling. Feeds processed by pasteurisation demonstrated no detectable growth, indicating at least a 10⁶ log kill. Results were repeatable across the organisms tested.

This study has demonstrated that pasteurisation is more effective at reducing microorganisms in infant formula milk feeds. This is of particular importance as infant formula feeds are not sterile when reconstituted and there is a risk of contamination from the infant feed powder, the individual involved in reconstitution and the water source. This study demonstrates the frequent presence of Bacillus contamination within milk formula, which can have clinical consequences if it is toxin producing. It also demonstrates that many methods commonly used to reduce the risk to patients by treating infant formula have little to no effect on bacterial load, highlighting that consideration should be given with regards to the treatment of infant milk formula in order to reduce/eliminate risk for high-risk patients.