Hand hygiene MOST CERTAINLY prevents healthcare associated infections

I’ve spent some time trying to process Eli’s last two posts on hand hygiene. It’s tough to untangle his logic and I’m not sure what he’s trying to prove. But here’s my point-by-point response:

In the first post, he notes: “if puerperal fever was a CDC HAI and clinicians didn't wear gloves, we could still say hand hygiene prevents HAI. However, that's not the current reality.” Well, puerperal fever is a CDC HAI (see here, page 23).

Next he states that CDC defines HAI as CLABSI, CAUTI, SSI and VAP. That’s a true statement, but I think he’s implying that CDC defines HAIs as only those 4 (particularly in light of the paragraph above). Actually, CDC has defined over 30 HAIs (definitions here, page 4).

He goes on to say that the WHO hand hygiene moment #2 (hand hygiene before clean and aseptic procedures) is not part of the CLABSI bundle published by Peter Pronovost. But Pronovost’s article (to which he links) states otherwise:

Per Eli, “Do we really think that interns and nurses practicing hand hygiene on the wards prevents SSIs to any measurable extent compared to pre-operative CHG bathing or peri-operative antibiotics?” Well, to Eli I would say, the next time one of your family members has a surgical procedure, tell the surgical team that they don’t need to perform hand hygiene before touching your loved one’s fresh surgical wound.

He argues that hand hygiene is not a significant component in the causal pathway for HAI, then four paragraphs later goes a step further and states that hand hygiene is not in the causal pathway. While for some HAIs hand hygiene may not be the most important risk factor, it is nonetheless a risk factor and it is indeed in the causal pathway. Then we come to twisted logic. Per Eli, hand hygiene prevents transmission of microorganisms but it doesn’t prevent HAIs. However, transmission of microorganisms is an intermediate outcome that can lead to HAI. It’s like saying: guns don’t kill people, bullets do.

He states that interventions to improve hand hygiene are not used for outbreak control. That’s contrary to my experience. We focused on hand hygiene in almost every outbreak that I managed over two decades as a hospital epidemiologist. In some cases, it was one of our first interventions while we proceeded with the investigation.

His next post focused on how you cannot correlate ward-level hand hygiene compliance with HAIs. Given the relatively small number of observations collected, and the relatively small range of compliance, I agree.

He shows an Ishikawa diagram of factors leading to CLABSI, but it and his subsequent logic ignore the fact that CLABSIs are not just associated with catheter insertion. In fact, we rarely see insertion-related CLABSIs anymore because the CLABSI bundle (which includes hand hygiene!) is so effective. The typical CLABSI now occurs in patients who have had a central line for weeks to months. So, the important factors now are associated with line maintenance (e.g., the line dressing, skin hygiene, minimization of entry into the line for lab draws, etc).

He next estimates how many months of hand hygiene observations are necessary to witness one opportunity where the HCW touches a CVC? But his focus is on how many observations are performed not on how many opportunities exist. I’ve previously estimated that in a 700-bed hospital there are 15 million hand hygiene opportunities per year. Using Eli’s estimates (which I have no reason to doubt), that translates to 30,000 direct manipulations of a central venous catheter yearly. I sincerely hope that every one of those was preceded by hand hygiene, and I suspect almost every hospital epidemiologist and every patient hopes so, too.

I get Eli’s point that trying to precisely measure hand hygiene compliance is dumb. But I think there is value in the process of monitoring hand hygiene because it keeps hand hygiene top of mind. It’s the Hawthorne effect in action. And the past decade or so of all this measurement has made a difference. If I compare the present to when I was a house officer 30 years ago, it’s amazingly different. There was zero focus on hand hygiene in the 1980s. There were many fewer sinks and no alcohol-based products available. We have made enormous progress.

My thoughts on hand hygiene are simple: It’s important. It’s really important. Keep doing it. Keep measuring it. It eventually becomes a habit. And someday, it will be so ingrained we won’t need to talk about it anymore.


  1. Hey Mike. Interesting post. I'm not sure where you are coming from, but I have some ideas. Over the next few days, I'll respond in the comments here and maybe put them in a post. The post I did write in response was focused on reiterated my contention, which I now feel more strongly about because nothing in your post rebuts my claim that you can't associate ward-level hand hygiene compliance with ward-level HAI rates.

    But here a few thoughts on your initial concerns:

    1) You wrote: "In the first post, he notes: 'if puerperal fever was a CDC HAI and clinicians didn't wear gloves, we could still say hand hygiene prevents HAI. However, that's not the current reality.' Well, puerperal fever is a CDC HAI (see here, page 23)"

    My response: I specifically said "clinicians didn't wear gloves." Since it was an "and", my statement is correct. Thus, this comment could only mean that you (a well-noted critic of gloves), do not wear gloves when delivery babies. I will confess, I almost titled my first post "That's Just Ewww" and focused it on wearing gloves. But that would have been distracting.

    2) You mention 30 HAIs. Since you made the claim, can you further explain how ward-level hand hygiene could possibly be associated with reductions in any of them? Of course, I listed 4 and focused on one or two since I wanted to keep the post shorter than a Russian novel.

    3) Yeah, you got me on hand hygiene in the Pronovost bundle. Andrew Stewardson and Didier Pittet did the same on twitter. But of course, there is increasing evidence that you don't need to practice hand hygiene before donning non-sterile gloves. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4594854/ Hand hygiene is probably more helpful before sterile gloves, but as I showed in my second post, it is a HUGE stretch to say you would detect any of these hand hygiene moments and even if you counted EVERY ONE with some system, it would be under 1% of any change in hand hygiene compliance at the ward level. You should still teach moment #2, but don't expect to see reductions in CLABSI with a ward level hand hygiene campaign. And again, not wearing gloves when inserting a central line is ewwww.

    4) The bit about the surgical team touching a family member's wound borders on ad hominem. And I would care but because they were spreading MRSA and VRE and not because it would cause and an SSI that they got in the OR hours before. Something that comes after can't be the cause. And even if it causes 10% of SSIs you couldn't associate ward level hand hygiene with it. Oh, and the surgeons should be wearing gloves. If surgeons don't wear gloves when touching a fresh surgical wound, that's just ewww. Do you not encourage glove use among surgeons during this moment?

    You can see why I wanted to call my post "That's Just Ewww"

    more later

    1. This is so ridiculous I won't even dignify it with a response.

    2. Disappointed but can’t say I’m surprised.

  2. I've been thinking about all three of these posts for a few days now and I decided I can see Eli's point but I think it's irrelevant. The real issue here is that the CDC's definition of HAI is largely based on what hospitals *can* do instead of what they *should* do. The old adage of "you can't change what you can't measure" seems to have been originally written about hand hygiene and, thus, hand hygiene often takes a back seat to other initiatives. That said, when CLABSI, CAUTI, and SSI definitions were proposed, most hospitals didn't have adequate resources to do such detailed surveillance (some still don't) but the influence of these definitions pushed hospitals to invest in measuring and improving these metrics. Although we have the power to change healthcare, we haven't seen any big push to revolutionize infection-related safety in hospitals for a long time now. Something has led us to sacrifice innovation and forward motion for the sake of feasibility. Many of the largest infection control studies are focused on justifying and refining current guidelines instead of moving safety forward. Perhaps we have reached the limit of what we can prevent with hard stops, checklists, and bundled supplies. Changing actual behavior is next and before we can start that we need a reliable way to measure behavior, some idea of the outcomes each behavior influences, and at least a hypothesis about how to change it. We have automated aggregate hand hygiene monitoring here, by the way, and we register 40 million hand hygiene opportunities in a year for our 700 bed hospital and that's just entry and exits. That's a big elephant to ignore. Saying that hand hygiene doesn't influence HAI is basically saying that Infection Control as a field isn't ready to move beyond tinkering with decades-old standard definitions. It's failing to recognize the potential for a more enlightened view of what an HAI really is. We aren't investing in it or pushing for it. At last year's ID Week meeting there were NO talks about hand hygiene. None. about 10 posters were presented but no oral abstracts, no symposia, nothing. Is that because we have fixed the problem of hand hygiene or because we are resting on our CLABSI-reduction laurels with another RCT-style study of a known entity and it's ever diminishing returns. I can't wait to know the exact RR and NNT for CHG bathing in robotic hysterectomy. That will make a big difference in patient outcomes and in convincing our colleagues of our importance as a field. And then we can argue about the statistics, methods, and exclusion criteria in our noon conference just like the cardiologists. All while a nurse with a viral URI hands out meds to a stem cell transplant patient without washing her hands. At last she didn't touch the central line.


Post a Comment

Thanks for submitting your comment to the Controversies blog. To reduce spam, all comments will be reviewed by the blog moderator prior to publishing. However, all legitimate comments will be published, whether they agree with or oppose the content of the post.

Most Read Posts (Last 30 Days)