Milstone AM, Elward A, Song X, Zerr DM, Orscheln R, Speck K, Obeng D, Reich NG, Coffin SE, Perl TM et al: Daily chlorhexidine bathing to reduce bacteraemia in critically ill children: A multicentre, cluster-randomised, crossover trial. Lancet 2013(0). Nearly 5000 Children more than 2 months old who were admitted to the PICU for more than 2 days were randomized to either get a bath with chlorhexidine every day (or being wiped all over with a chlorhexidine cloth) or control. Actually it was the ICUs that were randomized and had periods with and without chlorhexidine. There was an almost significant reduction in sepsis with the ITT analysis, and a just significant reduction with the ‘actually treated as they were supposed to be according to the way they were randomized’ analysis. An accompanying editorial (Toltzis P, Goldmann D: Rethinking infection prevention research. Lancet 2013(0). Suggests that such trial should be done only when all the quality control initiatives have reduced central line infection rates to minimal levels. I am not so sure; despite the best efforts using evidence based interventions currently there are still substantial variations between hospitals. We need further evidence based interventions and good trials to show us what to do next.
Chong E, Reynolds J, Shaw J, Forur L, Delmore P, Uner H, Bloom BT, Gordon P: Results of a two-center, before and after study of piperacillin-tazobactam versus ampicillin and gentamicin as empiric therapy for suspected sepsis at birth in neonates <1500g. J Perinatol 2013. I think this is a Very Bad Idea. Giving broad spectrum antibiotics that are not indicated to infants with negative cultures for reasons that are questionable (one of the reasons given for changing empiric therapy of early onset sepsis was an increase in resistant organisms causing late onset sepsis). They didn’t show an adverse effects, but just wait a while! We should be focussing on giving fewer antibiotics, with narrower, appropriate, spectra, for shorter periods. And not giving them at all unless there is an indication.
Berardi A, Rossi C, Lugli L, Creti R, Bacchi Reggiani ML, Lanari M, Memo L, Pedna MF, Venturelli C, Perrone E et al: Group b streptococcus late-onset disease: 2003-2010. Pediatrics 2013, 131(2):e361-e368. An epidemiologic study from Italy of late onset GBS. Being preterm increased the risk, as we already know, there is a high incidence of serious brain lesions associated with late onset disease. If the mother had intrapartum antibiotics, the disease tended to be milder, and to present later.
Neonatal Research 
Keith,
none of the authors from the Chong paper would argue with you regarding the desire to diminish antibiotic exposure and duration in the first days of life. The average treatment time in our cohorts was 3 days, whether it was ampicillin and gentamcin or Zosyn. Many infants never saw antibiotics at these institutions, only those with sufficient risk factors or clinical status to warrant suspicion received antibiotics and only those patients were included in the study. Antibiotics were stopped upon verification of negative culture after 48 hours. The issue we addressed was if we must use an antibiotic, can we use one that has a comparable spectrum of coverage with fewer side effects and complications when compared to amp and gent (hearing loss, resistance, blood draws, NEC) and have comparable or better safety. The answer to that hypothesis (including the issue of resistance) was yes.
Both institutions have now been using Zosyn for more than 3 years as our initial rule out antibiotic of choice and still have not seen resistance. We continue to try to minimize antibiotic exposure to none or as little as possible. I understand your concerns and would not try to persuade you or anyone else to start using Zosyn based on just this study. I do think we need to address the fact that we have been using amp and gent for a very long time with very little data and a lot of dogma as our basis for doing so. There are very clear risks specific to these two antibiotics that we seem all too happy to ignore. If this study jump starts that conversation, then it will have served an important purpose.
Phill Gordon
Hi Phil,
Thanks for the comment,
My first question is really what is the problem that we are trying to deal with? less than 2% of admission cultures in VLBW babies are positive. Less than a third of those are with E Coli and less than half of the E Coli are resistant to gentamicin. In contrast in many nurseries being preterm is treated as a septic condition and all the babies admitted receive antibiotics, even if you try and be selective, most of the admitted premies get antibiotics, and they are often continued for many days for no very good reason. If we routinely stop antibiotics at 36 hours when the cultures are negative you don’t need to do blood levels of gentamicin, and there isn’t good data that hearing loss is a significant problem. NEC certainly may be, but has only been clearly associated with prolonged antibiotics, especially those prolonged courses given in the face of negative cultures.
Piperacillin has a significantly wider spectrum, being active also against anaerobes and other organisms that rarely cause early onset sepsis (other enterobacteriaceae for example) anaerobes such as bifidobacteria are very important of course for preventing NEC. And some E Coli are resistant to piperacillin also (about 10% in the UK, and we have had pip resistant sepsis here in Sainte Justine)
You are absolutely right that we need to rethink these issues, and I must say the articles I include in Weekly Updates usually only have a brief comment, which tends to be very opinionated, and a bit off the cuff. I always appreciate when people like your group are willing to publish what they do and try and analyze it appropriately. That is 1 million times better than all of the uncontrolled variation in practice that goes on with no adequate evaluation.
I would suggest that we need to give narrower spectrum antibiotics rather than broad spectrum (the UK suggests now penicillin and gentamicin rather than ampicillin), that we should routinely combine them with probiotic administration, and that we should very rarely continue antibiotics when cultures are negative.
Keith
Keith,
much of what you say, I agree with. But I think we have been schizophrenic about our approach to initial rule outs over the years. For example, the reason so many people use ampicillin instead of penicillin is to cover Listeria, but most people don’t use an adequate dose to cover the CNS and the incidence of Listeria is far lower than the resistant forms of E. Coli. Both pathogens are horrendously devastating once established in the CNS and are better dealt with in the periphery before they can invade the brain, so in general, adequate peripheral coverage makes sense for the initial rule out. Except when it doesn’t, because Listeria has this ability to walk through the interior of neurons, making it exponentially more virulent upon gaining CNS access (versus E. Coli which bores from the outside to the interior of nervous tissue in order to evade high MICs). Suffice to say, I have always thought the choice of ampicillin was illogical (either use it at really high doses because it is imperative to treat Listeria CNS infection, or don’t use it because it is so rare). We have to make reasonable decisions about what we should cover when we think infants might be sick (everyone else can and should get nothing – on this I think we agree). If we ignore Listeria because it is so rare, penicillin for GBS would be the logical choice.
However, in our unit, we think drug resistance E. Coli is a relevant subcategory of pathogen and the infants we choose to treat should be covered from the first dose (Zosyn does that for the blood stream, but does not cover the CNS – we add cefotaxime double coverage in infants who have evidence of meningitis). Zosyn has the advantage of having excellent tissue penetration in the lungs and intestine, which are the two most common sites of pathogen entry in the newborn. It has balanced broad spectrum coverage (and yes that includes anaerobes) which we hypothesize might leave the intestine in better shape for recolonization (based on our observation of significantly less diaper rash and NEC). The main mechanism of resistance for Zosyn is plasmid transfer and the central culprit is Enterobacteria. That is not an organism that typically colonizes the newborn during the first days of life (if even the first week) and this may explain why we are not seeing emerging resistance with this time limited protocol.
I think you may be underestimating the relationship between gentamicin and hearing loss both in the population with known polymorphic susceptibilities and in preterm infants with high noise environments. Moreover, the issue with gentamicin susceptible polmorphisms is the “do no harm” part of our jobs. These infants will acquire hearing loss because of a drug that we give them. Its rare, but it happens.
Again, I am not trying to convert you. I think open dialogue and equipoise on this topic is critical and appreciate the opportunity here. You should listen to certain people from pediatric infectious disease on this topic. You would think Zosyn is a sacred cow whom we have violated in an unspeakable manner. For my part, I would like people to have conversations that expand to include all of the issues around antibiotic use. For example we have struggled for decades with osteopenia, all the while using multiple drugs that cause calcium wasting (including gentamicin). We also know that ampicillin is a subtle inhibitor of platelet function. And then there is the one that finally precipitated the my decision to change to Zosyn. Jeff Reese’s work with aminoglycosides shows that the entire family keeps the ductus open (in a mouse model) but other antibiotic types do not. Now I know that there are a lot of differences between mice and men, but I tend to believe in evolution. If the entire family of compounds has that effect, they are likely mimicking a biological agent that all mammals have. All I can tell you is that we saw trends towards decreased ligation after Zosyn in the two center study and we are to the point now where we are hardly ever using NSAIDs anymore at Tulane.
The bottom line is, I think that there are more variables that should be weighed when we try to decide what is the best possible antibiotic for initial newborn rule out than just how narrow a spectrum can we get away with. Its not that I don’t think this is important. Of course it is. I think the dialogue should be how do we weigh these things one against another and make an informed choice.
Phill