Scrub but don’t cap the hub to reduce infections in the NICU

Scrub but don’t cap the hub to reduce infections in the NICU

I had the pleasure of meeting the author of a paper I am about to comment on this week while at the 99 NICU conference in Stockholm.  Dr. Ohlin from Orebro University in Sweden presented very interesting work on their unit’s “scrub the hub” campaign. As he pointed out, many places attempt to reduce coagulase negative staphylococcal infections by introducing central line bundles but seldom is there one thing that is changed in a bundle that allows for a before and after comparison like his team was able to do.  I was so impressed by this work and at the same time concerned about another strategy to reduce infection that I felt compelled to make a comment here.

Scrub the hub!

Dr. Ohlin and the first author Dr. Bjorkman published Scrubbing the hub of intravenous catheters with an alcohol wipe for 15 sec reduced neonatal sepsis back in 2015.  They compared a 16.5 month period in their unit when they rolled out a CLABI reduction bundle to a period of 8.5 months afterwards when they made one change.  Nurses as is done in the units I work in were commonly scrubbing the hub before they injected the line with a medication but in the second epoch the standard changed to be a specified 15 second scrub instead of being left up to the individual nurse.  With permission from Dr. Ohlin here is a picture of the hubs highlighting bacterial growth without scrubbing, then for a duration less than 15 seconds and then with 15 seconds.

In the first epoch they had 9 confirmed CLABSIs and 0 confirmed in the second after their intervention.  The rate of CLABSI then in the first epoch was 1.5% vs 0% in the second group.  As with any study looking at sepsis, definitions are important and while they didn’t do paired cultures to rule out contamination (one positive and one negative as is the definition in our hospitals) they did refer each patient to a senior Neonatologist to help determine whether each case should be considered a true positive or not.  Given that they made no changes to practice or other definitions in diagnosing infections during that time perhaps the results were indeed real.  Presumably if they had missed an infection and not treated it in the second epoch the patient would have declared themselves so I think it is reasonable to say that 8.5 months without a CLABSI after their intervention is a success.  As Dr. Ohlin points out the scrub duration may also help due to the abrasion of the hub surface removing a bacterial film.  Regardless of the reason, perhaps a 15 second scrub is a good idea for all?

The lazy person’s solution – the SwabCap

One way to get around human nature or people being distracted might be to cover each luer lock with a cap containing 70% isopropyl alcohol.  In this way when you go to access the line there should be no bacteria or labour required to scrub anything since the entry of the line is bathed in alcohol already.  This was the subject of a systematic review from the Netherlands entitled Antiseptic barrier cap effective in reducing central line-associated bloodstream infections: A systematic review and meta-analysis.  The reviews ultimately examined 9 articles that met their inclusion criteria and found the following; use of the antiseptic barrier cap was effective in reducing CLABSIs (IRR = 0.59, 95% CI = 0.45–0.77, P < 0.001). Moreover, they concluded that this was an intervention worth adding to central-line maintenance bundles. Having said that, the studies were mostly adult and therefore the question of whether minute quantities of isopropyl alcohol might be injected with medications was not a concern when they made their conclusion.

What about using such caps in ELBW infants

Sauron et al in St. Justine Hospital in Montreal chose to look at these caps more carefully after they were implemented in their NICU.  The reason for taking a look at them was due to several luer valves malfunctioning.  The authors created an in-vitro model to answer this question by creating a closed system in which they could put a cap on the end of a line with a luer lock and then inject a flush, followed by a simulated medication (saline) and then a flush and collect the injected materials in a glass vial that was sealed to prevent evaporative loss of any isopropyl alcohol.  They further estimated the safe amount of isopropyl alcohol from Pediatric studies would be 1% of the critical threshold of this alcohol and using a 500g infant’s volume of distribution came up with a threshold of 14 mmol/L.  The study then compared using the SwabCap over two different valve leur lock systems they had in their units (SmartSite and CARESITE valves) vs. using the strategy of “scrub the hub”.

The results were quite concerning and are shown below.

Circuit Type Temperature Sample 1 Sample 2 Sample 3 Mean
SwabCap on Smart Site Valve Room 49.5 58.4 46.8 51.6
Incubator 35 degrees 45.16 94.7 77.9 72.6
SwabCap on CARESITE valve Room 14.1 5.7 5.2 8.34
Incubator 35 degrees 7.0 8.1 5.9 7.0
Isopropyl alcohol pad on CARESITE Valve Room 0 0 0 0

 

Certainly, the Smart Site valve allowed considerable amounts of isopropyl alcohol to enter the line but the CARESITE while better still allowed entry compared to the control arm which allowed none.  Beyond the introduction of the alcohol into the system in all cases considerable clouding of the valves occurred with repeated capping of the system with new caps as was done with each med injection since each was single use.  In lines that were not accessed contact with the cap was left for 96 hours as per recommendations from the manufacturer and these changes occurred as well.

Conclusion

While a reduction in CLABSI is something we all need to strive to obtain, it is better to take the more difficult path and “scrub the hub” and by that for 15 seconds which incidentally is the same recommended duration for hand hygiene in both of our units.  Perhaps in larger term infant’s seepage of isopropyl alcohol into the lines would not be as concerning as their larger volume of distribution would lead to lower levels but I would ask the question “should any isopropyl alcohol be injected into any baby?”.  I think not and perhaps by reading this post you will ask the same thing if your unit is using these caps.

  • Thank you to Örebro University Hospital for their permission in using the photo for the post
The Hidden Pathogen Of Late Onset Sepsis

The Hidden Pathogen Of Late Onset Sepsis

If you work in the NICU then you have seen your fair share of septic workups for late onset sepsis.  Sepsis is such a common diagnosis that if I had to guess I would say that at least 50% of all discharge summaries would include this in a list of final diagnoses for any VLBW infant.  If you were to look through the chart though you would find that while workups are common, the recovery of a pathogenic bacterium is not as much.  This is in part due to the low threshold that many people have for doing such workups.  A little bit of temperature instability, a few more apneic events than normal or a rise in O2 requirements may all trigger such investigations.  When they come back negative we all feel good that we looked but we also are then quick to blame the etiology on something else.  Mild fluctuations in temperature are written off as overbundling, apnea due to outgrowth of caffeine and a rise in FiO2 to evolving CLD.  Maybe though the explanation at least in some cases is that there was a pathogen but we didn’t test for it.

Viruses are everywhere

Tis the season so to speak so everyone is on high alert for viruses in our homes, schools, malls etc but many of us consider the NICU to be mostly free of such pathogens.  The truth is we mostly are provided that we all wash our hands well, keep sick contacts from visiting and put on a mask when our coughing starts.  Alas, if you have done a handwashing audit as we have you would know that when looking at technique and duration of handwashing, we don’t always hit 100%.  These audits are for health care practitioners but I have often wondered what sort of results we would see were we to do the same for parents and visitors.  When we know the viruses are out there such as during outbreaks of RSV and influenza we can’t help but send off our samples for respiratory viruses more frequently but what if we did this with intention for every late onset septic workup?

Lucky For Us Someone Did Just That!

Back in 2014 the following study was published. Viral respiratory tract infections in the neonatal intensive care unit: the VIRIoN-I study.  This was a simple prospective and elegant study in which any infant in the NICU who had never been home and was greater than 72 hours had respiratory samples sent for viral panels within 72 hours of starting antibiotics for presumed late onset sepsis.  The findings were certainly interesting in that 6% of 135 sepsis evaluations tested positive for a virus.  In the analysis, the infants had the following characteristics:

  • tended to be older (41 vs 11 days; P = .007)
  • exposed to individuals with respiratory tract viral symptoms (37% vs 2%; P = .003)
  • lower total neutrophil counts (P = .02)
  • best predictor of viral infection was the caregivers’ clinical suspicion of viral infection (P = .006)

What interests me about these results are a couple things.  The first is that as I was once told, the sensitivity of asking if someone has been around sick people is low during peaks in viral outbreaks as who hasn’t?  Perhaps what this study tells us is that within the NICU environment we actually do a reasonable job of keeping such contacts away but when they slip through infections happen.  The second point worth mentioning is that a low neutrophil count is associated which is interesting given how often neutropenia is pointed to as a reason to start antibiotics.  These viruses are troublesome creatures indeed!

Further Evidence Arrives

At the end of last year a similar study was published by the same group Viral Respiratory Infections in Preterm Infants during and after Hospitalization.  They took a different approach this time out and took nasopharyngeal samples from 189 infants in the NICU (96 term and 93 preterm) within 7 days of birth and then sent samples weekly while in hospital followed by monthly for four months after discharge. In this collection of infants a mere 4 patients tested positive in NICU and all of them under 28 weeks of age at birth!  How do we account for the remarkable reduction in risk while in hospital?  To answer that you can read through the NICU environment in the full article if you have access.  In short, they had a very rigorous infection control set of precautions set up.  washing-hands-waterInterestingly only one of the infections was with RSV and the unit did not provide prophylaxis for infants in hospital.  Perhaps with precautions like theirs they felt it was unnecessary.  Once discharged a little over a third of patients acquired a viral infection in the first four months at home.  Given the potential risk for readmission and with that to a PICU this rate of viral infection is concerning.

Vision for the future!

Taken together we can state that viruses do make their way into the NICU but fortunately not as commonly as one might think.  What the last study in particular does remind us though is that we need to ensure that as part of discharge teaching parents take home many of the practices that we have used in the hospital with respect to hand hygiene, limiting visitors and not being afraid to holster some hand sanitizer for those times when soap and water are not so easy to come by.  To be sure viruses are out there but at least for the first few months after discharge for our most vulnerable babies a little paranoia about viruses could go a long way.

Your address may be the most important thing when it comes to Synagis and RSV.

Your address may be the most important thing when it comes to Synagis and RSV.

As I said in another post on this topic I have been a huge advocate of RSV prophylaxis since my days as a Pediatric resident. When I started my residency we were not using Palivizumab (Synagis) and I recall admitting 10+ patients per day at times with bronchiolitis.  With the use of passive immunization this rate dropped dramatically in Manitoba although rates in other areas of the country may have not seen such significant impacts.  Manitoba may be somewhat different from many areas due to the communities in Nunavut being so impacted when RSV enters these areas and can infect many of the children due to crowded living conditions and inability to really isolate kids from one and other.  The lack of benefit in other areas though, has no doubt led to controversy among practitioners who often wonder if giving 5 IM injections during the RSV season is indeed worth it.  The real question has not necessarily been does it work but to whom should it be given so that you get the most benefit.

A Big Change in The Last Year

In 2015 the CPS published a revised statement entitled Preventing hospitalizations for respiratory syncytial virus infection.  This statement included a significant change to the recommendation for those who should receive the product.

  • In preterm infants without CLD born before 30 + 0 weeks’ GA who are <6 months of age at the start of RSV season, it is reasonable (but not essential) to offer palivizumab. Infants born after 30 + 0 weeks’ GA have RSV admission rates that are consistently ≤7% (Figure 3), yielding a minimum number needed to treat of 18 (90 doses of palivizumab to prevent one RSV admission) if one assumes 80% efficacy and five doses per infant. Therefore, palivizumab should not be prescribed for this group.

Gone are recommendations for treating those from 30 – 32 weeks and moreover 33- 35 weeks if meeting certain conditions.  There is a provision for those in Northern communities to expand these criteria to 36 0/7 weeks if such infants would require medical transport to receive care for bronchiolitis.  What is not really clear though is what is meant by Northern communities in terms of criteria to determine suitability exactly.

Incidentally, the criteria are not so different than the AAP statement from August 2014.  In their version of the statement they state:

“The burden of RSV disease and costs associated with transport from remote locations may result in a broader use of palivizumab for RSV prevention in Alaska Native populations and possibly in selected other American Indian populations.”

The American guideline also states that it is for those infants who are “well” and under 29 weeks that RSV prophylaxis is appropriate but from 29 – 32 weeks use should be restricted to those babies who are on oxygen at 28 days of life.

AAP News Release From This Week

As stated above there are those who have always been sceptical of the true cost benefit of RSV prophylaxis and I would imagine those individuals must have latched on to the following report. Otherwise healthy premature infants 29 weeks gestation and over unlikely to benefit from palivizumab

The authors of this study found that while during the RSV season admission for RSV bronchiolitis was lower from 29 – 32 weeks in those infants who had received Synagis. This would argue that it should be given in this group except for the fact that as the authors state it was a “wash” since hospitalization for non-RSV bronchiolitis in the same population was increased by a similar amount.  In essence if you didn’t get RSV you wound up getting something else that still put you in hospital.  The conclusion here is that the decision to drop the criterion for prophylaxis to under 29 weeks is supported since from 29 – 32 weeks you can’t prevent hospitalization from viruses that take the place of RSV.

A Few Thoughts Though Before We Conclude It Has No Place

As I stated upfront I am not totally free of bias having seen a very large impact up here in Manitoba.  What I worry about though is that we have in medicine a tendency to try and capture the “gist” of a guideline rather than committing it in its entirety to memory.  We can’t help ourselves as the volume of information we are asked to remember is growing daily.  What this may lead to however is changes to our practice that may expose vulnerable infants.  The AAP guideline was designed to recommend changes for the otherwise healthy infant under 29 weeks.  What I think we are really talking about are the truly exceptional babies.  In our institution babies born at less than 29 weeks and certainly those closer to 24 weeks often spend as much as a month on CPAP. At 28 days even if the patient were on room air it might only be due to the fact that they were on distending pressure.  Put them on nasal prongs and they would qualify.

Another important consideration is the remote location point.  figure35-enHere in Canada the majority of the population lives along the border with the US.  Take a look though at our population density and you can see that for our Aboriginal (Indian) and Inuit populations as well as all those living in the north we have a significant need to protect them.  Living conditions in these places involve overcrowding and high smoking levels both ideal breeding grounds for transmission of RSV and an intolerance to handling the inflammation in your bronchioles.  The same is likely true of many parts of Alaska for my US readers.

It would be too easy to simply state “I only need to give Synagis to those babies under 29 weeks now” but that would do a disservice to the populations in our remote communities both here and in the US.

In case you are wondering, I am not employed by the makers of Synagis here or in the US nor do I have any financial or conflicts otherwise.  I am someone I suppose who has seen the difference of before and after and while I will let others debate the merits of giving Synagis from 29 – 32 weeks and above, I wrote this as a reminder that not all populations are the same and therefore we should not paint all susceptible patients with one brush.

Why do we keep treating reflux in preemies?

Why do we keep treating reflux in preemies?

Preamble

May 7, 2017

As I sit drinking my morning coffee and feeling a little sense of heartburn I began to reflect on the fact that I can’t recall the last time I prescribed ranitidine or a PPI in an infant for anything other than an acute upper GI bleed.  I know I had done so after moving to Winnipeg in 2010 at a few time points but that practice has certainly died at least for me.  You know what? I don’t think it has made one iota of difference but based on the results from this post I think it is for the best.  What has inspired my republishing of this post is my question as to whether or not you think your units practice has changed as well since the revelation that these medications are not only ineffective but harmful.  Read on and enjoy your Sunday

Choosing wisely is an initiative to “identify tests or procedures commonly used whose necessity should be questioned and discussed with patients. The goal of the campaign is to reduce waste in the health care system and avoid risks associated with unnecessary treatment.”

The AAP Section on Perinatal Pediatrics puts the following forth as one of their recommendations.

“Avoid routine use of anti-reflux medications for treatment of symptomatic gastroesophageal reflux disease (GERD) or for treatment of apnea and desaturation in preterm infants.

Gastroesophageal reflux is normal in infants. There is minimal evidence that reflux causes apnea and desaturation. Similarly, there is little scientific support for the use of H2 antagonists, proton-pump inhibitors, and motility agents for the treatment of symptomatic reflux. Importantly, several studies show that their use may have adverse physiologic effects as well as an association with necrotizing enterocolitis, infection and, possibly, intraventricular hemorrhage and mortality.”

How strong is the evidence?

The evidence for risk with acid suppression is largely based on either retrospective or in the case of Terrin G et al a prospective observational cohort study Ranitidine is Associated With Infections, Necrotizing Enterocolitis, and Fatal Outcome in Newborns.  In this study the authors compared a group of premature infants with birth weights between 401 – 1500g or 24 – 32 weeks gestation who received ranitidine for reflux symptoms to those who did not.  All told 91 were exposed while 183 were not.  The authors are to be commended for standardizing the feeding protocol in the study so that when comparing NEC between groups one could not blame differences in formula consumption or rate of feeding advancement.  Additionally, bias was controlled by having those not involved in care collect outcome data without knowing the purpose of the study.  Having said that, they may have been able to ascertain that ranitidine was used and have been influenced in their assessments.

The patients in terms of risk factors for poor outcome including CRIB and apgar scores, PDA etc were no different to explain an increased risk for adverse outcome.

Fig 2

From the above table, rates of infections were clearly higher in the ranitidine group but more concerning was the higher rate of mortality at 9.9% vs 1.6% P=0.003 and longer hospitalization median 52 vs 36 days P=0.001.

Results of a Meta-Analysis

Additional, evidence suggesting harm comes from a meta-analysis on the topic by More K, Association of Inhibitors of Gastric Acid Secretion and Higher Incidence of Necrotizing Enterocolitis in Preterm Very Low-Birth-Weight Infants.  This analysis actually includes the study by Terrin and only one other retrospective database study of 11072 patients by Guillet et al.  As the reviewers point out the study by Terrin while prospective did not employ the use of multiple regression to adjust for confounders while the larger study here did.  In the end the risk of NEC with the use of acid suppression was 1.78 (1.4 – 2.27; p<0.00001).

What do we do with such evidence?

I can say this much.  Although small in number, the studies that are available will make it very difficult to ever have a gold standard RCT done on this topic.  This scant amount of evidence, backed by the biologic plausibility that raising the gastric pH will lead to bacterial overgrowth and potential aspiration of such contents provides the support for the Choosing Wisely position.

Why do we continue to see use of such medications though?  It is human nature I suspect that is the strongest motivator.  We care for infants and want to do our best to help them through their journey in neonatal units.  When we hear on rounds that the baby is “refluxing” which may be documented by gulping during a brady, visible spit ups during A&Bs or through auscultation hearing the contents in the pharynx we feel the need to do something.  The question invariably will be asked whether at the bedside or by the parents “Isn’t there something we can do?”.

My answer to this is yes.  Wait for it to resolve on its own, especially when the premature infants are nowhere close to term.  I am not sure that there is any strong evidence to suggest treatment of reflux episodes with gastric acid suppression helps any outcomes at all and as we see from the Terrin study length of stay may be prolonged.  I am all in favour of positional changes to reduce such events but with respect to medications I would suggest we all sit on our hands and avoid writing the order for acid suppression.  Failure to do so will likely result in our hands being very busy for some infants as we write orders to manage NEC, pneumonia and bouts of sepsis.

 

Do we need so many shots to prevent RSV bronchiolitis?

Do we need so many shots to prevent RSV bronchiolitis?

I have been a huge advocate of RSV prophylaxis since my days as a Pediatric resident. When I started my residency we were not using Palivizumab (Synagis) and I recall admitting 10+ patients per day at times with bronchiolitis.  With the use of passive immunization this rate dropped dramatically in Manitoba although rates in other areas of the country may have not seen such significant impacts.  Manitoba may be somewhat different from many areas due to the communities in Nunavut being so impacted when RSV enters these areas and can infect many of the children due to crowded living conditions and inability to really isolate kids from one and other.  The lack of benefit in other areas though, has no doubt led to controversy among practitioners who often wonder if giving 5 IM injections during the RSV season is indeed worth it.  The real question has not necessarily been does it work but to whom should it be given so that you get the most benefit.

A Big Change in The Last Year

In 2015 the CPS published a revised statement entitled Preventing hospitalizations for respiratory syncytial virus infection.  This statement has caused a great deal of controversy at least among those I have spoken with due to its significant departure from the previous recommendations. As per the statement:

  • In preterm infants without CLD born before 30 + 0 weeks’ GA who are <6 months of age at the start of RSV season, it is reasonable (but not essential) to offer palivizumab. Infants born after 30 + 0 weeks’ GA have RSV admission rates that are consistently ≤7% (Figure 3), yielding a minimum number needed to treat of 18 (90 doses of palivizumab to prevent one RSV admission) if one assumes 80% efficacy and five doses per infant. Therefore, palivizumab should not be prescribed for this group.

Gone are recommendations for treating those from 30 – 32 weeks and moreover 33- 35 weeks if meeting certain conditions.  There is a provision for those in Northern communities to expand these criteria to 36 0/7 weeks if such infants would require medical transport to receive care for bronchiolitis.  What is not really clear though is what is meant by Northern communities in terms of criteria to determine suitability exactly. Incidentally, the criteria are not so different than the AAP statement from August 2014.

Do We Need So Many Shots?

Just at the end of 2016 though Lavoie P et al in Vancouver, BC published a letter outlining their experience with a modified schedule of either 3 or 4 doses of palivizumab during the RSV season.  Included in the letter are their criteria for determining the number of doses and importantly pharmacokinetic data demonstrating the effectiveness of such schedules in achieving protective titres.  pharmacokineticsThe 3 dose schedule was used for those infants born between 29 0/7 and 35 weeks gestational age who had a risk factor score of 42 or more. Interestingly at the end of the RSV season, depriving such infants of 1 or 2 doses did not appear to impair the ability of the infant to maintain protective levels.

From a clinical standpoint the outcome data during this period examining 514 (3 dose) and 666 (4 dose) patients similarly suggests that they were indeed protected from disease.  In the 3 dose cohort only 1 patient was hospitalized with RSV during the dosing period and 1 infant afterwards.  In the 4 dose group, 10 were hospitalized with RSV  during the dosing schedule and a set of twins afterwards.  Aside from these known RSV infections, an additional 7 and 18 patients were hospitalized with bronchiolitis without viral identification during the dosing schedule with no cases of bronchiolitis afterwards.  Taken altogether and assuming that all cases were indeed RSV bronchiolitis the authors conclude that the overall rates are no different than those seen with a 5 dose schedule.

Is Something Rotten In The State of Denmark?

There is something peculiar here though.  There is no doubt that palivizumab must have gone through rigorous pharmacokinetic testing in order to determine the correct number of doses needed. For a 3-4 dose regimen to provide the same coverage in terms of antibody titres seems strange to me. I would love to believe the data but there is a skeptic in me. Secondly with respect to counting hospital admissions is this exhaustive in terms of including all hospitalization a in BC or at only some sites? Clarity is needed before considering such changes to practice.  Strangely it has been several months since this experience was published and there has been no discussion of it at least locally.*  Something as dramatic as this should have sparked some discussion and the absence of such leaves me questioning what am I missing?
From the standpoint of reducing interventions and pain in the neonate I am intrigued by these findings.  Parents as well would no doubt be happier with 3-4 IM injections over 5.  The additional benefit is no doubt financial as this product while effective does carry a significant cost per dose.  As you can see I have my doubts about the reproducibility of the results but it does at least offer some centres that have not been as enthusiastic about palivizumab something to consider. For some, the BC approach just might be the right thing.

  • I indicate that there has been little discussion locally of the article discussed.  There has indeed been discussion both here and in other Canadian provinces.  What I meant by that comment is that among my colleagues in Neonatology and Infectious Diseases and housestaff I have had only one discussion.
Non-sterile gloves to reduce NICU infections. Back to the Future?

Non-sterile gloves to reduce NICU infections. Back to the Future?

I don’t know about you but I have deeply rooted memories from the 1990s of donning a yellow gown and gloves before examining each and every patient on my list before rounds.  This was done as we firmly believed such precautions were needed to prevent the spread of infections in the NICU.  As time went on though the gowns were removed and not long after so went the gloves as priority was placed on performance of good hand hygiene to reduce rates of infection in our units.

You can imagine that after having it entrenched in my mind that hand hygiene was the key to success that I would find it surprising to see a paper published a few months ago suggesting that the use of gloves after hand hygiene may offer a benefit after all.  Kaufman DA et al published Nonsterile glove use in addition to hand hygiene to prevent late-onset infection in preterm infants: randomized clinical trial and given it’s challenge to a practice that is at least two decades old I thought it may be worth sharing with you the reader.

Essentially the authors hypothesized that the use of non-sterile gloves after performing hand hygiene (compared to hand hygiene alone) would reduce late-onset invasive infection (>72 hours after birth), defined as 1 or more episodes per patient of a BSI, urinary tract infection,meningitis, and/orNEC associated with clinical signs and symptoms of infection and treated with antimicrobials.  When determining the size of study needed, they used a baseline incidence of 60% and looked to find a 25% reduction in their outcome.  Unfortunately for them (although very fortunate for their patients, the incidence of LOS in the experimental arm was 32% with a 45% incidence in the control group (hand hygiene alone).  What does this mean when your expected rate is higher than your observed?  In short you need more patients to show a difference and indeed they failed to show a significant difference between the two groups. They did however find a difference in gram positive infections being 15 vs 32% p=0.03 and seem to take some comfort in this finding.  If you were to give the paper a quick read you might be impressed with the finding and might even shrug your shoulders and say the common expression “Can’t hurt but might help”  Maybe we should adopt this?

Not so Fast

There is a significant potential source of error here that needs to be addressed.  The definition of a proven blood stream infection as per the CDC is two positive cultures for the same organism.  In this study only one culture was required to be positive so the potential for diagnostic error is high.  In our own centre although unpublished we have noted since adopting a mandatory two culture collection approach for LOS that there have been a significant number of occasions where one culture was noted to be positive and the other negative.  Antibiotics in these cases have been stopped (for gram positive organisms) after 48 hours without consequence.  In this study however the findings of increased rates of positive cultures in the hand hygiene only group is heavily influenced by the presence of only one positive culture as is seen in this table.

table 1

When looking at the numbers of times there were greater than or equal to 2 positive cultures in the CoNS group one sees the vast majority were only based on one culture.  Furthermore, of the 20 infections in the hand hygiene only group, 19 were gram positve CoNS of which only 4 had more than one culture.  Based on this finding and the lack of any other significant difference in infectious outcomes the proof that gloves add anything to reducing infection rates is tough to argue.

Could Gloves Actually Make Things Worse?

Several studies have actually indicated that wearing gloves reduces hand hygiene compliance.  One such study although in adults “The dirty hand in the latex glove”: a study of hand hygiene compliance when gloves are worn. suggests that this is indeed the case with a 9% decrease in proper hand hygiene when gloves were worn.  Others such as Flores in 2006 found similar poor perfomance when gloves were used Healthcare workers’ compliance with glove use and the effect of glove use on hand hygiene.  I would speculate that although we all want to do what is best for our patients there may be a psychological trick being played here.  Perhaps knowing we will cover up with gloves leads people to take shortcuts on hand hygiene as they subconciously think they will be covered anyway.  Never mind that the “dirty” hand touches the gloves they will put on making proper hand hygiene a must.

Conclusion

It certainly was a shock to see such a paper as I saw flashes of my past yellow gowned self coming back to haunt me.  Based on my take of this paper however I would say that at least for the time being I will take my time, wash my hands before and after every patient encounter and keep the gloves around for handling those yet unbathed newborns.  Spend your energy where it counts and that is ensuring your hands are properly cleaned before touching your patient or lines.