The Application of Science

It all seemed so easy

In 2010 an article was published in the New England Journal of Medicine, Preventing Surgical-Site Infections in Nasal Carriers of Staphylococcus aureus .  Patients were screened for Staphylcoccus aureus ( including MRSA, methicillin resistant Staphylococcus aureus) and those that were positive underwent a 5 day perioperative decontamination procedure with chlorhexidine baths and an antibiotic, mupirocin, in the nose.  The results were impressive.  Before the intervention the infection rates were 7.7 % and after the intervention it was 3.4 %.  That is an impressive drop in surgical infections.

One of the orthopedic groups approached us (us being the hospital administration, pharmacy, nursing  and infection control, of which I am Chair) to implement the protocol in their patients, citing a similar study on an orthopedic population.  Great.  It should be an easy enough intervention.  I should have known better, of course, long experience has continually demonstrated that what appears to be simple never is.

First was the question as to whether the study was applicable to our patients.  Resources were going to be devoted to an intervention, so going forward we had to demonstrate that the bang would be worth the buck.  These are financially lean times, with cutbacks and declining reimbursement, so every expenditure of time and money needs to be justified.  In the bizarro accounting of health care, not every hospital administration will include money saved in the evaluation of interventions, only the money spent.   I work in a hospital system with a remarkably strong commitment to patient safety and quality, so there was little  worry on that point.

Still, our overall surgical infection rates are less than 1%, so we are doing better at baseline than NEJM study were at the end of the intervention.  When I see surgical infection rates going from 7.7 % to 3.4 %, I have to wonder if the benefit of the intervention is a surrogate for an issue with deeper problems with infection prevention at the hospital.  Maybe there was a a decrease in infections with decolonization because they were sloppy with the compliance with other, more basic, infection prevention strategies.   My system has been very aggressive applying SCIP interventions, a series of actions that, when consisting applied, result in a large drop in surgical infections.  Perhaps they are sloppy with their SCIP.  I have no way of knowing.

Also,  I can’t help but think there is a lower limit to the infection rates under which it will be impossible to go. As long as we operate on people we will always have some infections no matter what we do for prevention.  One of our surgeons had his only infection for the year in an emergency case on a 450 lb woman with a glucose of 600 and a 2 pack a day habit.  Even after a week in the hospital she still had dirt tattooed on her palms and feet.   I think zero infections should be our goal; I am not so sure that is realistic given the co-morbidities of some patients.  Could we go lower?

I also wondered if we had infections that would be amenable to surgical screening and decolonization.  Many of the orthopedic S. aureus infections are presenting late, 4 to 8 weeks post op.  Given the virulence of S. aureus, where people usually become symptomatic the day they acquire the bacteria, I suspect the infections we had been seeing in the orthopedic population are being acquired after discharge.  If so, we may make no impact on our infections by beating down the S. aureus in the peri-operative period.

Still, we had the occasional S. aureus infection, MRSA and MSSA, Staph colonization is a risk for subsequent infections, and the orthopedic group wanted to do something so it seemed reasonable. I can count on one hand the number of times a surgeon has been the one to ask for help in reducing infection; usually we take suggestions to them.  It was an opportunity I didn’t want to miss.  We had evidence, science even, to drive change.

At first it was suggested that we not bother to screen, but treat everyone.  Everyone has MRSA, right?  Wrong.  In Washington (in Oregon, we call it Washington, not Washington State) they mandate screening for all ICU admits, and we know that about 20% of patients have S. aureus in their nose, and about 2% have MRSA.  With mupirocin resistance already at 14%  in some parts of the US and chlorhexidine resistance being described, the last thing I wanted to do was increase the use of these drugs.  Resistance, said the Borg, is inevitable, or some such.

Cost-efficacy analysis, which make my head hurt, suggests we should treat everyone but the authors minimize the threat of future antibiotic resistance. In a world that is running out of antibiotics, I cannot be quite so cavalier on the issue. Use it and lose it.   We are already sliding into the post antibiotic era, and I don’t want to drive resistance any harder than necessary.

The next step was deciding was how patients would be screened.  A nasal swab misses 30% of MRSA carriers, and to maximize the yield would probably require a swab of the nose, throat, skin folds, arm pits and stool.  5 separate swabs would make it cost prohibitive ($120 each) but if you use one swab, in what order do you swab? Always end at the anus, that’s my rule.  And the PCR, polymerase chain reaction, is not approved for every body specimen.   So in the end we decided that the published data only screened with nasal swabs, so we would as well.  We may miss a few MRSA carriers but they would be rare; the perfect is the enemy of the good and reality requires compromise.

The next roadblock was coordination of care.  The nasal swab had to be collected within 30 days of the planned surgery.  Someone had to note if the swab was positive for MRSA or MSSA, call the patient and the pharmacy for a prescription of chlorhexidine and mupirocin and start the antibiotics a few days before surgery.  When  22 in 100 have a positive PCR , initially the test was missed since there was no one person responsible for acting on the result.  Most of the people who need a new joint are old and, by definition, not mobile and getting to the pharmacy was limited by both patient mobility and transportation.  Getting the prescription to the patient was difficult to do consistently, and transitioning the outpatient prescription to the inpatient setting was equally unreliable.  Patients were not bringing in their medications  (we did not want duplicate prescriptions) and physicians were not reordering the mupirocin/chlorhexidine, thinking the patients had brought their own.

To compound the problem, many patients had to pay out of pocket for the medications, and there are two formulations of the mupirocin, the much more expensive formulation made for use in the nose.  Of course. It is weird how, when nasal mupirocin use started to be common a new, and much more expensive, formulation became available.   We elected to go with the less expensive formulation as there is no evidence for increased efficacy of the expensive formulation.

Peri-operatively we had decided to give add, rather than substitute, vancomycin to the usual cefazolin if the patient had MRSA.  That was only 2% of the patients, rare enough to be missed.  Vancomycin has many characteristics that make it a lousy drug, and I did not want to lose the efficacy of the cefazolin for MSSA.  However, initially there was no single person responsible for noting that the patient was MRSA postive and changing the perioperative antibiotics.  All too often Vancomycin was either not given, or given too late to have effect.  Vancomycin, requiring an hour infusion instead of the IV push of cefazolin, needs more lead time to give and can throw a monkey wrench into the work flow of the OR if a case has to be delayed an hour or more to give an antibiotic.

It took well over six months of trial and error to work out the kinks in the protocol with one orthopedic group to apply the evidence and and get the process to run smoothly.  Murphy’s law states that if something can go wrong, it will.  I agree with the suggestion that Murphy was an optimist.

Despite all that work, for some patients the process is grinding to a halt from an unexpected source.  The swab for Staph costs about 120 dollars, and Medicare, which many joint patients are on, does not pay for screening.  For the first several months we (not me) received irate calls from patients about the charge, and now Medicare patients have to sign a waiver when they get tested understanding that the swab will not be covered by insurance and they will be responsible for the cost of the test.

Part of the issue with instituting the protocol is that there is no health care system: one group pays, one group runs the hospital and OR and a third group does the surgery.  It would, perhaps, been much smoother to implement if we were in a unified health care system.  I dream of universal health care where the system will be totally screwed up in one consistent way, rather than the hundreds of ways it is totally screwed up now.    I really want to sit on one of them there death panels.

Has it worked? Too early to say, and I hate to jinx myself, but in the nine months since the protocol was introduced, despite all the glitches, we have had no Staphylococcal infections in that orthopedic group. I am cautiously optimistic the effort has paid off.  We hope to spread the protocol to other orthopedic groups and high risk surgeries now that we have worked out the kinks. Of course I expect a whole new collection of complications.

I read papers all the time about this or that intervention improving patient outcomes.  Reading the papers it seems ever so simple to apply the results to the real world, but it takes an amazing amount of work by a large number of people to coordinate the care that even a simple intervention can entail.  In the end we (mostly everyone else except me) accomplished what we set up to do, only to be stopped short for a large number of patients by Medicare.

This is one of many quality initiatives at my institutions that have resulted in decreased morbidity and mortality.  Last time I wrote on the issue I looked to see if there had been similar initiatives in any of the SCAMs. What are chiropractors, acupuncturists, naturopaths and their fellow travelers doing to improve patient care? At the time I had found none. I would have thought chiropractors would be interested since one of the Never Events  is “Patient death or serious disability due to spinal manipulative therapy.”  In their world a never event  never happens, not an event that can haappen but never should.   A quick search of the Googles and Pubmeds finds…

Still nothing.  Must be nice to be perfect.  It results in a lot less work.

Posted in: Basic Science, Clinical Trials, Science and Medicine

Leave a Comment (19) ↓

19 thoughts on “The Application of Science

  1. lxbxr says:

    It’s always interesting (from the point of view of somebody involved in running clinical trials, but not active in healthcare) to see articles about the implementation of trial results – things always seem more difficult outside of the trial setting!

    WRT the original paper, I’d like to know what randomization algorithm they used to get 91 more patients in the active group out of only 917 randomized (so I know to avoid it!). I wouldn’t normally be bothered but unfortunately they also ended up with excess immunocompromised patients in the placebo group (7.5% vs. 3.8%) and only present the adjusted RRs as far as I can see.

  2. rbnigh says:

    Thank you for this thoughtful post. Where I work, in Chiapas, Mexico, my health-care colleagues are facing this kind of decision constantly. Clinical conditions are extremely challenging, patients display multiple chronic problems and health-care budgets are cut to the bone and counting. With so many potential sources of problems, implementing a fancy new procedure, even when the supporting science is clear, is often a remote pipe dream.

  3. mousethatroared says:

    Really lovely post. I will admit I am one to be impatient with slowness in updating to new techniques or technology in the medical field. It an eye opener to see all the things you must think of when implementing a new system.

  4. Troyota says:

    This is a great case to make, but to be convincing, I believe you need to collect a larger set of examples. This doesn’t necessarily have to be a “random sample” (although, done properly, that would make a terrific master’s thesis!), but it should be more extensive, including, as suggested, a range of mainstream publications.

    There have been some studies of the peer-review process applicable to this problem. One great one (Schroter S, Black N, Evans S, Godlee F, Osorio L, Smith R. What errors do peer reviewers detect, and does training improve their ability to detect them? J R Soc Med. 2008 Oct;101(10):507-14. PubMed PMID: 18840867; PubMed Central PMCID:PMC2586872) found that when reviewers were given manuscripts with embedded errors, from 20% to 39% failed to detect this exact problem–discrepancy between abstract and results. In other words, this error went undetected by reviewers at least 60% of the time!

    So at least there is significant danger of this happening….Best of luck with this effort!

  5. Troyota says:

    I’m sorry, that comment was supposed to go under Dr. Gorski’s article. For a congential bumbler, the Internet provides a whole new dimension for blunder….

  6. Jann Bellamy says:

    “What are chiropractors, acupuncturists, naturopaths and their fellow travelers doing to improve patient care?”

    A few years ago I happened to be going to San Diego, where, coincidentally, the American Public Health Association was holding its annual meeting. I found out that there is a chiropractic section of the APHA and attended their section meeting. There was a presentation from a young chiropractor with an MPH. He had done research on resistance from some older chiropractors teaching in chiropractic school clinics to disinfecting between patients — forgot exactly what they were disinfecting, but something the patients have contact with during the procedure. (The table maybe?) There was an effort to get them to use alcohol but this group thought alcohol was “medical” and wanted to use (I am not making this up) yogurt instead! So I guess some credit is due for one chiropractor trying to bring the older guys up to date on the fact that germs from one patient can be transmitted to another patient. On the other hand, at this same meeting, I had a chance to talk to a young chiropractor who told me that influenza didn’t kill people.

  7. cervantes says:

    Another problem with the health care non-system: it is possible that this protocol actually saves money. In any event, that $120 for the antibiotics is at least partly subsidized by preventing infections. But there’s no way to capture that and give the customer a price break, or even free pills. That’s why we need some form of global payment.

  8. Janet Camp says:

    “I dream of universal health care where the system will be totally screwed up in one consistent way, rather than the hundreds of ways it is totally screwed up now. I really want to sit on one of them there death panels.”

    Quote of the Decade. Made me laugh and cry at the same time. If we get to return to Oregon for our “retirement” (more like impoverished final years), I’ll feel very comfortable in your hospital when I get my new hip.

  9. thatguybil81 says:

    The $120 is for the screening test.
    Muprirocin 2% ointment AWP is $65 actual wholesale is $7.34
    Muprirocin 2% cream AWP is $52.1 actual whole sale is $50.42
    Chlorhexidine 4% wash AWP is $10.09 actual wholesale is $8.91
    Premix Vancomycin 1000mg AWP is $180 actual wholesale is 45.43
    Bulk Vancomycin 1000mg AWP is $17.16 actual Wholesale is $4.39
    The bag cost ~$1.5
    Total actual cost with out mark up is 181.68 per patient.

    Based on base line surgical infection rate of 1% and an estimated reduction to 0.5% One would need to treat 200 patients at a cost of $36,336 to prevent one surgical infection.

    Just a taste of the basic cost analysis done in these kinds of studies. :)

  10. Adam Rufa says:

    I totally agree with Janet, “I dream of universal health care where the system will be totally screwed up in one consistent way, rather than the hundreds of ways it is totally screwed up now.” is a great quote!!!

  11. windriven says:


    Really nice comment (though I’d like to see your calculation arriving at 181.68 in detail). The other half of the equation would be the cost of treating that 1 case had it not been prevented.

    As an aside I wonder why the chlorhexidine is so expensive? We use it in a manufacturing process and 20% clx USP costs about $7/kg. At 4% that would be ~$1.40/kg; about $.70 for a pint. Certainly there is diluent and packaging but it’s a long way from 70 cents to nine bucks. I wonder if the other prices are stepped on at similar rates?

  12. lilady says:

    Dr. Crislip:

    What a coincidence that you are blogging about MRSA colonization testing before orthopedic surgery.

    I have a close friend who went for “pre-testing” for total knee replacement surgery and he was told that his nasal swabs had tested positive for MRSA. He was prescribed Vancomycin and mupirocin ointment.

    I wasn’t surprised, because he has a severely developmentally disabled son who lives in an intermediate care facility, where my friends spend a lot of time with him (I too, visit him and may be colonized, as well). It is very similar to a small nursing home with ~ 30 medically frail residents, many of whom have frequent hospitalizations at several area hospitals.

    The surgery has been postponed until mid April for “other” medical reasons, when he will undergo another nares testing, prior to the surgery. The orthopedist told my friends this is standard preoperative testing procedure for orthopedic surgery and if he is still colonized they will be infusing him with IV antibiotics before and during the surgery. I checked the internet and found this,

    It is my understanding that many hospitals are testing residents of nursing homes who are hospitalized, and some hospitals are testing all patients…can anyone confirm if this is so?

  13. Mark Crislip says:

    There is great variability in testing, depending on medical indications and local laws. Washington mandates it for ICU admits, but not Oregon.

  14. lizditz says:

    Asked Dr. Crislip

    What are chiropractors, acupuncturists, naturopaths and their fellow travelers doing to improve patient care?

    I went looking just for chiropractors.

    I searched for “improving patient outcome”. Zero results. I tried “patient outcome” Zero results. I tried “improving patient care” and got one result,

    J Chiropr Educ: Spr 1999(13:1): 8-1Development and design of an innovative postgraduate course in practice-based learning at the Anglo-European College of Chiropractic

    I tried “quality improvement” and got seven hits, of which four are possibly apropos

    J Manipulative Physiol Ther: Nov/Dec 2011(34:9): 627-633 22079 Human subject research: Reporting ethics approval and informed consent in 3 chiropractic journals [Journal of Manipulative and Physiological Therapeutics, Journal of the Canadian Chiropractic Association, Chiropractic and Osteopathy]

    Chiropr & Manual Ther: 2011(19:9): Online access only 35 p 21693 Creating European guidelines for Chiropractic Incident Reporting and Learning Systems (CIRLS): Relevance and structure

    J Manipulative Physiol Ther: Nov/Dec 2006(29:9): 707-725 19364 Health services research related to chiropractic: review and recommendations for research prioritization by the chiropractic profession


    On the gripping hand, searching for autism rendered 27 results, most single-subject studies finding that autistic patients treated with chiropractic & other interventions improved. More fleecing of the autism community.

    I tried Google Scholar, and found two:

    The Chiropractic Hospital-based Interventions Research Outcomes (CHIRO) Study: a randomized controlled trial on the effectiveness of clinical practice guidelines in the medical and chiropractic management of patients with acute mechanical low back pain The Spine Journal Volume 10, Issue 12 , Pages 1055-1064, December 2010

    Application of Incremental Change Strategies in Chiropractic and Multidisciplinary Clinical Settings for Quality Improvement Quality Mangaement in health care. 2000 – Volume 8 – Issue 3 Mootz, Robert D.; Hansen, Daniel T.; Souza, Thomas A.; Triano, John J.; Wiese, Barry C.

    I’m left once again wondering why chiropractic research is of such uniformly poor quality.

  15. JaneMD says:

    @that guy bill

    Of course there is going to be a markup on all of those medications. You have to pay the pharmacist to confirm the safety of the medication, weigh it for use, and send it to the right patient. Then you have to pay the anesthesiologist to administer it, pay for the IV pump, and pay for the existence of the nurses and the hospital room.

    The price would potentially go down if we had stable price controls or the current medicaid payment actually PAID for how much medical care costs. (come visit me at my blog to discuss how these things are calculated)

    Congratulations, btw on trying a patient safety protocol. Hopefully the data will bear out, particularly since institutions typically resist change like this. I’m sure you’ve read Peter Pronovost’s Safe patients; smart hospitals. The battle to replace conventional razors with electric razors in the OR was hilarious. And not universally practiced yet – I made the surgical tech find an electric razor before my c-section last year and cited the infection statistics to her.

  16. thatguybil81 says:


    If you are actualy an MD you know how this stuff works.
    Cost accounting is all that actually matters. Charge acccounts (mark up) does not matter because you can charge what every you want for your intervention the thrid party payor is only going to give you your DRG or contract agreed bundled pricing for that service.

    Actual cost account is all that matters.

    You do not need to factor in the mark up for the Pharmacy tech that makes the drug… because they are already at the hosptial. Its a fixed cost
    You do not need to factor in the mark up for the Pharmacist because they are already being paid, its a fixed cost.
    You do not need to factor in the Anes cost because they are already being paid, its a fixed cost.

    The only thing that matters is the ACTUAL cost of the intervention. The made up hypothetical charge accounting does not. You do not get paid based on what you charge.

    The $181 was arived at by looking at our wholesale cost for the drugs adding them together with the cost of the test. (Which is not a whole sale cost). I included the premix and the bulk cost of Vanco for completeness sake. That $120 for the test is what they “charge” for the test, not how much it actually costs the lab to run the test.

  17. mattyp says:

    How’s this for science being applied:

    First paragraph::Homeopaths are facing a fight to defend their practice in Australia after the National Health and Medical Research Council flagged it might declare their work baseless and unethical.”

  18. Chris says:

    Supplement spam is not science. Go away.

Comments are closed.