Probiotics to prevent necrotizing enterocolitis: Too cheap and easy?

As someone involved at the committee level for determining eligibility for palivizumab (PVZ) prophylaxis in British Columbia, I thought it might be interesting to contrast two recommended medical interventions: one aimed at preventing bronchiolitis admissions from the community by administering monthly injections of a monocloncal antibody, and the other aimed at the prevention of necrotizing enterocolitis (NEC) and death by administering prophylactic probiotics enterally (1) with feeds – two real risks for preterm infants, with similar incidence rates.

The evidence for both interventions is very robust. Two randomized studies investigating PVZ (2,3) included nearly 3000 at-risk infants and showed a 55% to 80% reduction in hospital readmission rate for RSV. A third study involving newborn infants with a screening diagnosis of cystic fibrosis found no difference in admission rates and was never published (4). More than 5000 infants have now been enrolled in probiotic-NEC studies, and a meta-analysis shows similar efficacy with even smaller confidence limits (5).

Although PVZ has been shown to prevent hospital readmission, it does not affect paediatric intensive care unit admission rates or the incidence of death, which is now rare in developed countries. In addition, since PVZ prophylaxis was introduced into clinical practice, there has been little change in total infant respiratory admissions because most occur in previously healthy term and near-term infants who are not eligible for prophylaxis (6). In contrast, probiotics prevent NEC, death, short-gut syndrome and – probably – neurodevelopmental disability (7). Since probiotics were introduced in Japan, NEC rates have fallen to approximately 1% (8).

Both products have no significant medical side effects. Fatal anaphylaxis has only rarely been reported with PVZ. Sepsis and bacteremia have been reported from probiotic organisms (9) but did not occur in any of the published randomized trials, which also showed an overall trend toward reduced sepsis. The cost of PVZ is prohibitively high, making it an unacceptable choice for most families; therefore, it is funded by government and insurance companies. The cost of probiotics is so low that if they are provided with the appropriate information, most families would be willing to purchase a recommended product from their own pocket (10).

Finally, PVZ has not been shown to be cost effective, yet is prescribed enthusiastically by physicians in North America, even for children in whom no benefit has been demonstrated; however, probiotics are not prescribed, although they would still be cost effective if they cost >100 times as much per pot (11). This is in contrast to other developed countries, where probiotics are now prescribed routinely, but guidelines on PVZ prophylaxis are far more restrictive (12).

Why aren’t most neonatologists in North America prescribing or recommending probiotics? Could it be that they are not expensive enough? Hospital administrators often complain that physicians like to spend money. Are they right?

Probiotics fall into a peculiar category of medical intervention because they are neither drugs nor devices. Similar to maternal milk, which also has considerable benefits both for preventing NEC and RSV admission, probiotics are not regulated, although the Natural Health Products Directorate in Canada ensures that good manufacturing practices are followed and product labels are correct. Unfortunately, health food manufacturers are not permitted to make any specific health claims, even if they are true. This means that there will be no industry representatives knocking on our doors and inviting us to sponsored education sessions with free food and drink. It is up to us to examine the evidence and our motivation, and take action.

In an editorial that accompanies the recently published cohort study from Montreal (13), William Tarnow-Mordi and Roger Soll (14) agree that it is time to change practice. They suggest involving local microbiologists, drug and therapeutics committees, institutional review boards and parent representative groups to introduce probiotics using peer-reviewed guidelines. While I agree that it is prudent to accept input from such groups (particularly parent groups), we do not need to get ‘bogged down’ with committees and subcommittees. It is not acceptable to ‘delegate’ such important decisions if those who take on administrative responsibility have little interest, time or understanding of the huge benefits available.

Probiotics are not all equal: One of the most commonly used probiotics is the baker’s yeast Saccharomyces boulardii (Florastor, Biocodex, USA), which does not prevent NEC but has been reported to cause fungal sepsis (15,16) in immunocompromised patients. Another common probiotic, Lactobacillus rhamnosis, is commonly prescribed for infant colic but has limited efficacy at reducing NEC (17). Evidence-based guidelines on the choice of probiotics have been published (18) and most experts recommend using a combination of several strains of probiotics shown to have clear benefit in clinical studies.

In Victoria, British Columbia, we agreed to promote probiotics and chose the same product as Janvier et al (13) (containing four different strains of Bifidobacteria and Lactobacillus rhamnosus providing two billion colony-forming units per dose) with similar guidelines. A parent information sheet was created. Normally, our Pharmacy Committee would have deferred any decision on new product acquisition to a subcommittee of our provincial pharmacy and therapeutics group, but they agreed to obtain probiotics for our hospital, pending review, due to the overwhelming evidence of benefit with no alternative product available in our formulary. If our Pharmacy and Therapeutics Committee had not agreed, alternative routes for prescribing probiotics would have included purchasing directly for our unit, purchasing through dietary supplies or individual purchase by families.

Unlike Janvier et al (13), we were already offering donor milk to infants at high risk when no maternal milk was available, so our NEC rate was already low. We have now been using probiotics routinely for nearly a year with no confirmed cases of NEC, even for babies with birth weights <500 g. Our new practice has been strongly endorsed by both nurses and families, who are provided with information but are not asked to sign consent forms. Perhaps not surprisingly, our provincial pharmacy committee still has not approved their use, but we do not mind – bureaucratic wheels can turn very slowly.

NEC is easily prevented by the routine use of human milk and prophylactic probiotics. It is time for all of us to advocate for our patients.