Here we go again: Active detection and isolation, C. difficile edition

An interesting study in JAMA Internal Medicine, likely to generate a lot of discussion, addresses the use of “active detection and isolation” (ADI) for control of C. difficile disease. This quasi-experimental, single-center study employed PCR screening (tcdB detection) of all patients admitted through the emergency department (patients admitted from other locations were excluded, as were “short stay” patients), and those that were found to carry toxigenic C. difficile were admitted into a kind of “quasi-isolation”—gloves were used, but not gowns or private rooms. So all-in-all, a very pragmatic (and somewhat idiosyncratic) intervention. Healthcare-associated C. difficile disease rates declined after the intervention, which students of prior quasi-experimental studies of ADI for MRSA and VRE will find unsurprising. 

At this point, I will outsource my blog post to Jon Otter and Martin Kiernan at the Reflections IPC blog. Go on, head over there for an excellent pro-con post about this study, and vote on the question posed at the end of the post. Then come back here to read my only additional observation…..I can wait (spoiler alert: I agreed with Jon).

OK, you’re back: the only thing I have to add to Jon and Martin’s excellent post is this: we’ve been here before. Recall the persuasive quasi-experimental studies (many single-center, some multicenter) of MRSA and/or VRE ADI published over the course of a couple decades. When better designed studies were eventually performed and published (e.g. STAR*ICU, REDUCE-MRSA, MOSAR, this one by Harbarth and colleagues that doesn't have a catchy acronym)—you know, studies that included concurrent control groups (control groups are for losers!), it became evident that ADI wasn’t the key to MRSA or VRE control. I think we’re headed down that road again, this time with C. difficile. Who’s going to step up and organize the multicenter, cluster-randomized trial we need to do now? Or perhaps better to ask: who is going to pay for it?

Falling out of love with chlorhexidine?

Regular readers of this blog are aware of our on-again, off-again relationship with chlorhexidine (CHG) bathing of ICU patients. I even wrote a poem about CHG, a haiku that received no critical acclaim (see the last paragraph of this post).

Now a new study from Vanderbilt, published today in JAMA, finds that daily CHG bathing “did not reduce the incidence of health care-associated infections (HAIs)”. This single-center, pragmatic, cluster-randomized, 5 ICU study included 9340 patients and pre-specified a composite endpoint of central line associated bloodstream infection (CLABSI), ventilator-associated pneumonia (VAP), catheter-associated urinary tract infection (CAUTI) and Clostridium difficile infection (CDI). The per-patient analysis revealed rates of 2.86 and 2.90 HAIs/1000 patient days in the CHG and control groups, respectively. All told, 105 infections were included in the analysis (55 during the CHG period, 60 during the control period, excluding the “washout” periods). As anyone could have predicted, the vast majority of these infections were CAUTI, VAP and CDI (more on this later). A total of 8 CLABSI were captured, 4 in each arm.

What should we make of these data? Should institutions that have adopted CHG bathing abandon the practice? Should those planning to begin CHG bathing in their ICUs reconsider? My answers to these three questions: (1) meh, (2) no, and (3) no.

For several reasons, this study doesn’t move the needle much regarding CHG bathing. Set aside the fact that it is a single-center study, and that adherence to CHG bathing was not measured. More importantly, the composite endpoint they chose includes infections for which there is no convincing evidence (and very little biological plausibility) to suggest that CHG bathing should be effective (CAUTI, VAP, and CDI). To take VAP as an example, CHG oral care is included in the VAP prevention bundle (addressing the pathogenesis of VAP, which is entry of upper airway flora into the lung). If they practice VAP prevention at Vandy, which I’m sure that they do, then they already use daily CHG oral care in their ventilated patients. I'm not sure why anyone would think that adding CHG bathing would provide an additional detectable benefit for VAP prevention. And despite a single quasi-experimental study, I don’t know anyone who thinks CHG bathing is likely to have a significant impact on CDI rates. The more we learn about CDI epidemiology, the more it seems that antimicrobial stewardship interventions, rather than interventions to reduce transmission/acquisition, are the most likely to reduce CDI. As for CAUTI, well, CAUTI SCHMAUTI. 

The bottom line is this: the two outcomes that CHG bathing has been demonstrated to improve in controlled trials are (1) CLABSI, and (2) MRSA/VRE acquisition. This study can't claim to address either one. There were too few CLABSIs (i.e. it was way underpowered for that endpoint, given baseline CLABSI rates), and since they didn't do active surveillance during the study period they were unable to accurately measure rates of ICU acquisition of MRSA or VRE.

To sum up, CHG bathing of ICU patients should still be considered one of several adjunctive approaches to prevention of CLABSI (at least those due to common Gram-positive skin contaminants) and MRSA/VRE acquisition. What this study demonstrated, if anything, is that if your institution is practicing good HAI prevention (and has already achieved low CLABSI and MDRO rates), then adding CHG is not likely to add a measurable benefit (and that to measure any incremental benefit requires much larger studies!).

Perry Mason and REDUCE MRSA: Another "Case of the Positive Negative"

This guest post is by Marc-Oliver Wright, MT(ASCP), MS, CIC
 from NorthShore University HealthSystem
 in Evanston, IL 

"We need Perry Mason. Someone to put you in place." – Ozzy Osbourne

This week’s articles in the NEJM (Huang et al. “Targeted versus Universal Decolonization to Prevent ICU Infection”) and accompanying editorial (Edmond and Wenzel “Screening Inpatients for MRSA-Case Closed”) left me reminiscing of classic TV courtroom dramas where the well intentioned and sophisticated district attorney has finished presenting his/her case but Perry Mason has just cleared his throat and Matlock is still shining his ankle boots.


For those of you who missed it, Huang and colleagues randomized 74 intensive care units (ICUs) at 43 Hospital Corporation of America hospitals to one of 3 interventions among 2 cohorts.

1. Cohort one: States with mandatory MRSA screening (n=5)
     a. MRSA screening and isolation for ICU admissions (3)
     b. MRSA screening, isolation and decolonization for ICU admissions (2)
2. Cohort two: States without mandatory MRSA screening (38)
     a. MRSA screening and isolation for ICU admissions (13)
     b. MRSA screening, isolation and decolonization for ICU admissions (12)
     c. No screening, decolonization for all ICU admissions (13)

Decolonization included 5 days of intranasal mupirocin therapy and daily baths with chlorhexidine gluconate (CHG) for the duration of their stay in the ICU.

Adjusted hazard ratios for clinical cultures of MRSA with both decolonization strategies netted confidence intervals that were less than 1.0. Ditto for bloodstream infections with any pathogen (not just MRSA). When the authors selected out bloodstream infections caused by MRSA the findings were not significant. In the pairwise analysis universal decolonization attained statistical significance over screening/isolation/decolonization in the unadjusted analysis for bloodstream infection due to any pathogen. The authors conclude that universal decolonization is superior to either targeted decolonization or screening and isolation without decolonization.

This was a large, rather well designed, certainly well coordinated evaluation of multiple ICUs across the country. But the case is far from closed.

Defense exhibit one: Follow the blood. This article further affirms an already well conducted study by some of the same authors that demonstrated CHG bathing of ICU patients reduced bloodstream infections in ICU patients (Arch Intern Med. 2007;167:2073-9, ICHE 2009;30:959-63). There’s ample biological plausibility for this and in general, the horizontal (as opposed to vertical one MDRO at a time) approach is more logical. However, this effect in all likelihood has nothing to do with mupirocin, which has never been independently associated with a reduction in bloodstream infections. CHG reduces all-pathogen bloodstream infections in ICU patients: no objection your honor.

Defense exhibit two: An eyewitness-I see you. MRSA infections are not restricted by the boundaries of the unit. An APCHE score is not a pre-requisite to acquiring an MDRO. According to the evidence, ICU patients are at higher risk for MRSA infection, but many of prior studies were conducted in large academic medical centers where and when the case mix was different from today. Arguably, it’s operationally easier to limit intervention and research studies by geography.

Illinois was the first state to succumb to MRSA screening legislation; the legislation which reads that such screening will be performed on patients in “all intensive care units, and other at-risk patients identified by the hospital.” We, in the Land of Lincoln, have all, by design, implemented the former, but the latter is a hodge podge ranging from denial (only ICU patients are high-risk) to the assumed, evidentiary but not all-encompassing (patients from LTC are high-risk) to the near complete (70+ variable model built decision support module embedded in the EMR). The VA as well as my own organization deployed MRSA prevention strategies in ICU as well as non-ICU settings.

This brings me to: Defense exhibit three: A character witness. My organization’s strategy for the past 9 years has been a strategy most akin to Group 2 (screening, isolation and decolonization). Though we started in the ICU, we failed to see the kind of reduction we wanted (see defense exhibit 2) until we later expanded to universal screening. We used this experience to develop a robust multi-variable prediction model that was built as decision support system into our electronic medical record. This tool automatically calculates a MRSA risk score as patient data is entered and evolves within the record during the patient’s admission. When the risk score reaches a threshold the user is instructed to screen the patient. We used this to transition from universal screening to targeted screening in January 2012.

Ok. I am by no means an expert on MRSA. I am good with numbers though.

In the NEJM article, the best intervention for reducing MRSA clinical isolates is a rate of infection of 2.1 per 1,000 patient days. At NorthShore (and for comparison purposes, these measures are in our ICUs only) in the past 38 months, our unadjusted rate of MRSA infection is 0.3 per 1,000 patient days. Their best intervention netted an infection rate ratio versus my own ICUs of 7.0 (95%CI: 4.3-11.5, p=5.9 * 10-23).

Similarly, the NEJM article reported an overall, unadjusted rate of bloodstream infection from any pathogen of 3.6 per 1,000 patient days in the universal decolonization intervention arm. At NorthShore (and for comparison purposes, this is in our ICUs only) in the past 38 months, our unadjusted rate of bloodstream infection is 1.08 per 1,000 patient days. Their best intervention netted an infection rate ratio versus my own ICUs of 3.33 (95%CI: 2.54 -4.38, p=1.3 * 10-22).

I’m afraid I’ve only had 1 BSI due to MRSA in the past 38 months. My crude rate comes out as 0.018 per 1,000 patient days. If I understand Table 3 of the NEJM article correctly, the universal decolonization group experienced 48 BSIs with MRSA during the intervention and attributable ICU patient days of 101.603 for a rate of 0.47 per 1,000 patient days. This yields an infection rate ratio versus my own ICUs of 26.11 (95%CI: 4.53-150.45, p=3.1 * 10-8).

I am all too happy when a group of dedicated and brilliant people in hospital epidemiology find a new and improved way to prevent infection. And for this group, they did discover a new strategy for reducing their MRSA and bloodstream infections. I was not surprised to see that they have decided to implement the universal decolonization strategy across all HCA hospitals. But as long as their best rates are between 3 and 26 times higher than mine, I’ll be sticking with my own plan.

Defense exhibit four: If the antibiotic doesn’t fit you must acquit. Arguably, the single greatest concerning factor of this study that wasn’t addressed is the rise of mupirocin resistance. This exhibit, like the preceding, is rooted in personal experience. Until this year, our organization decolonized all MRSA-screen positive patients. During that 8 year endeavor our proportion of mupirocin resistant MRSA increased from about 3% to almost 12%. Nasal decolonization has been most convincingly demonstrated to reduce infection risk in the pre-surgical patient population and arguably only with select procedures. Our rise in resistant strains and the overwhelming desire to reserve this ammunition for the people that really need it (pre-surgical patients) gave rise to our decision to suspend mupirocin decolonization starting in 2013. We hope to see our resistance pattern return to baseline, though such a decline will likely be slow.

Justice may be blind, but we shouldn’t be to known potential adverse outcomes like antibiotic resistance.

Closing argument:  If none of the above convinced you that this case is far from closed at least consider this final argument. Living in Illinois, I am frightened awake at night with the sounds of what I imagine to be fingers typing on keyboards. The well-intentioned hands of a well-meaning lawmaker in our state or federal capital pounding out some new legislation in light of the latest and greatest publication on preventing some iteration of our nation’s bacterial scourges. Convinced that if they don’t do something those conniving hospital epidemiologists and infection preventionists won’t do anything to protect our nation’s sickest and frailest citizens. That all we’re here for is the big money check we get every week. By heaven, we must mandate! *shudder*

This paper, like all well designed and well-conducted studies deserves critical attention and consideration. Consider the methods they used and those they didn’t, the intervention processes they measured (CHG/mupirocin supply) and those they did not (isolation compliance), the settings in which the study was conducted (ICUs) and where it was not (everywhere else) and the outcomes they measured (MRSA and BSIs) and those they did not (antibiotic resistance). You must also compare their outcomes to your own performance. With such a wide variety of organizations the results may be generalizable despite HCA’s additional vertical strategies for MRSA that may have overlapped the study periods. How do your rates in your methods group compare those published here (1, 2, or 3?) versus the alternatives? Use this evidence as a tool for your ongoing risk assessment and make an educated decision before someone makes an uneducated one for you.

The defense rests.

REDUCE MRSA: Readers weigh in!

In my short post about this study last week, I promised that I’d follow up with the most likely criticisms we’ll be reading in future letters-to-the-editor. My post was going to focus on concerns about emerging resistance to topical agents, the lack of impact on MRSA bacteremia, and the outsized influence of skin commensals on the outcomes (see my post on the chlorhexidine bathing studies).

Fortunately, our readers have done this work for me. Below I’ve excerpted a comment we received today:

"We applaud investigators for their hard work in completing a high quality study. However, it has limitations and recommendations are overreaching.

--The intervention

Mupiricin/Chlorhexidine for all patients is very aggressive. Would chlorhexidine alone work in a similar fashion? (resistance is a real concern, especially for mupiricin, and if these agents become inactive they can't be used on high risk patients) We don't really know, but there will be calls to vastly increase antimicrobial exposure to mupiricin based on a study with a mixed intervention that prevented relatively few MRSA clinical cultures even over 74 ICUs I.E. A LOT OF PATIENTS WILL RECEIVE MUPIRICIN FOR A VERY SMALL ABSOLUTE DECREASE IN POSITIVE MRSA CLINICAL CULTURES!

--Outcomes

Bacteremia rates were elevated in the intervention arm during baseline period (6.1 vs. 4.8 or 4.2). Overall bacteremia rate was nearly 50% higher in ICUs that were randomized to the intervention! The intervention itself brought bacteremia rates down to what they already were in other ICUs.

The effect on bacteremia was mostly skin commensals such as coagulase negative Staphylococcus. As in Climo et al, this is probably the least important organism and may relate to decreasing contamination rates. We think this point is also important in the Climo paper, chlorhexidine bathing only had an effect on VRE acquisition and not MRSA acquisition and only had an effect on coagulase negative Staphylococci and candida bacteremia.

MRSA Clinical Cultures are a mixed group of colonization/infections that vary depending on sites. Is MRSA in the sputum the same as MRSA in a deep surgical wound? Or urine culture? or superficial wound culture? Did the frequency of culturing change during the intervention. Given data was obtained from a datapull, all of these questions could be answered which would help interpretation of this trial.

Although the trial is being hailed as a nail in the coffin of active surveillance culturing for MRSA, there are potential harms with over interpretation of the trial. If we begin using mupiricin and chlorhexidine on all ICU patients nationwide, as our European colleagues have suggested will occur, we may eventually have as many problems from the study as early studies of ASC. We also think that one cluster randomized trial should not change practice in such a sweeping manner.

From a more cautious perspective, this trial and Climo et al support that chlorhexidine bathing likely has a benefit on infections in the ICU. It isn't a huge effect and it is predominantly on markers that aren't as important (MRSA clinical cultures and coagulase negative Staphylocci AND CANDIDA (and VRE for Climo)). However, it should be considered by hospital ICUs with problems in these infections.

We should be worried that ASC may be finished only to be replaced by an equally problematic mandate (that more than swabbing every patient is actually treating every patient)."

We will also soon feature a guest blogger from a healthcare system with a long history of using active surveillance for MRSA control.  He also has quite a lot to say about REDUCE MRSA. We welcome all of our readers to weigh in, please feel free to do so in our comments section! The name of this blog begins with the word "controversies", after all.... 

REDUCE MRSA by getting horizontal

The results of this large cluster-randomized trial of different MRSA prevention strategies are now out in the New England Journal of Medicine, and are being widely (and mostly accurately) reported in all the major media outlets. Those who have followed this blog for any length of time know that we are not surprised at the results. Here’s just a sample of what we’ve written on the topics of active surveillance and the difference between horizontal and vertical infection control strategies.

I can keep this post short, because Mike Edmond and Dick Wenzel have already done my work for me in their excellent accompanying editorial (no subscription required!). I’d mostly like to congratulate and thank Susan Huang (pictured) and her excellent group of collaborators from HCA and CDC for making this very important contribution to the infection prevention literature, a contribution that should have a major impact on prevention practices.

But is it really “case-closed” on universal MRSA active surveillance, as the title of Mike and Dick’s editorial suggests? Every study that has questioned the effectiveness of MRSA screening has been followed by a barrage of letters-to-the-editor questioning the findings, and I suspect that this study will be no different. Later this week I will predict the contents of these future letters with uncanny accuracy. Stay tuned….