Is This the Answer to Hospital-Acquired Infections?

I have written quite a bit — probably far too much for the average person’s taste — about the poor state of hand hygiene in hospitals, and the resulting proliferation of bacterial infections. I often think I should shut up already about this problem. After all, it’s been 10 years since the Institute of Medicine’s damning report “To Err Is Human.”

Well, after taking at look at the latest National Healthcare Quality Report, I think I won’t be shutting up any time soon. Despite a lot of effort and innovation, despite a wise checklist approach, the problem doesn’t seem to be getting better:

Infections acquired during hospital care, also known as nosocomial infections, are one of the most serious patient safety concerns. It is unfortunate that HAI [hospital-acquired infection] rates are not declining. Of all the measures in the NHQR measure set, the one worsening at the fastest rate is postoperative sepsis (Table H.3). The two process measures related to HAIs tracked in the NHQR, both covering timely receipt of prophylactic antibiotics for surgery, are improving steadily. However, HAI outcome measures are lagging; only one shows improvement over time while three are worsening and one shows no change. This may, in part, reflect improving detection of HAI’s.

Table H.3. Measures of health care-associated infections, annual rates of improvements

DESCRIPTIONTable: 2009 National Healthcare Quality Report

Why is it so hard to get hospital personnel, doctors in particular, to do a better job with hand hygiene?

My belief, and as we wrote in SuperFreakononomics, is that it’s a question of externalities: the bacteria that a doctor may pass along via poor hand hygiene do not typically damage the doctor him/herself, but rather the next patient down the line. In this sense, deadly bacteria are a lot like our daily pollution: we do not personally pay the cost of our actions, so we have weak incentives to change our behavior.

I have often thought that if only individual doctors could be held accountable for their poor hand hygiene — if, say, an individual’s bacteria could somehow be tagged so that if a patient died from a hospital-acquired infection, the source of those bacteria could be determined — that would radically adjust the incentives at work here.

Well, the first step toward such a system may be on the way. A study published in the Proceedings of the National Academy of Sciences called “Forensic Identification Using Skin Bacterial Communities” argues that:

Recent work has demonstrated that the diversity of skin-associated bacterial communities is far higher than previously recognized, with a high degree of interindividual variability in the composition of bacterial communities. Given that skin bacterial communities are personalized, we hypothesized that we could use the residual skin bacteria left on objects for forensic identification, matching the bacteria on the object to the skin-associated bacteria of the individual who touched the object. Here we describe a series of studies demonstrating the validity of this approach. We show that skin-associated bacteria can be readily recovered from surfaces (including single computer keys and computer mice) and that the structure of these communities can be used to differentiate objects handled by different individuals, even if those objects have been left untouched for up to 2 weeks at room temperature.

If such identification worked for nosocomial infections, life would surely become more complicated for doctors (and their malpractice lawyers). And it might be one of the best things to happen for patient safety since a fellow named Ignatz Semmelweis came along.


I've often been irritated with the multitude of tests required that seem to be CYA for doctors when I visit them. Yet, the same doctors should be subjected to at least a high-level screening on a monthly or weekly basis for 'skin associated' bacteria or other easily communicable diseases.


My friend got MRSA in a top hospital in California after open heart surgery from a catherer resulting in redoing the aortic valve--infections go directly to the wound--her heart is permanently damaged.

Mike B

In defense of doctors having to decontaminate one's self between every patient interaction becomes a major undertaking with significant costs once a person realize how many different patient interactions there are. While practical in theory, washing one's hands hundreds of times a day not only takes a lot of time, it also has damaging side effects on one hands, which a doctor needs to do his job properly.

The solution here needs to be technological, not just procedural. Enforcing hand washing compliance might have an unintended consequence of doctors avoiding patient interactions. Disinfection needs to be made nearly costless to drive up use of the service.


Hospital I work at has slashed blood-culture contamination rates by doing similar. ID who signed for and took the blood sample and then retrain them once it's found to be a contaminate (i.e. taken wrong).

The externalities are that not only does the doctor signing the form have to take responsibility for their action (even if they got a junior to take the sample) and their peers will see, but they have to get retrained. Multiple times if necessary.

Word gets around, rates and numbers taken both fell.


I think you just like saying the name Ignatz Semmelweis.


Don't you know that Doctors are gods, walking amongst us mortal men? If you get infected and die from a Doctor not washing his hands, it is because the all-mighty deemed it time for you to pass. Dare yee not question the wisdom of the all knowing Doctors!

(Thank you for the article; Passed it along to all of my nursing friends and family who know all too well the problems of physician hand washing.)


Infections are caused by a small subset of the microbes on a doctor's hands, so your proposed approach wouldn't work to track an infection to an individual doctor. If hospitals were penalized strongly enough to infections, though, I bet they would come up with ways of increasing hand-cleaning, including making it easier.

Daniel Wagner MD

Identifying whose hands touched an object using a signature bacterial community is an interesting accomplishment. It would be quite another matter to find that same entire community in or on a patient with an infection so as to associate it with a source. In addition, even if you could do so with any sort of accuracy, how would you know that the source (Dr, nurse, etc) hadn't passed it inadvertently and unintentionally to a colleague who had just sanitized their hands and then touched the chart or computer keyboard that had been used (up to 2 weeks ago) by the source. Couldn't the colleague then temporarily become a surrogate source and be the one who actually infected the patient. While no one would argue that we don't need to do something about HAIs, implementing a system of culturing all hospital personnel and the hundreds of objects they touch on a daily basis in order to MAYBE find the source of a patient's infection would be a very costly exercise in futility.



Instead of a doctor washing hands every time he/she moves from patient to patient, why not just carry a box of surgical gloves and wear/change them for each patient?

Seems pretty easy to me.

marc meyer

How about requiring doctors to wear a badge with the number and percent of their patients who have HAIs?

It could be attached below their hospital namebadges, and updated daily. Bingo, externalities brought home. The doctor needs to justify his number to his patient and colleagues.


You provide no evidence that the increase in postoperative sepsis or catheter infections are due to lack of hand washing. On the contrary, surgery and catheter insertions are about the only times in medical care when hands are ALWAYS washed meticulously. These bacteria are everywhere, including all over the patients own skin, their GI tract, etc. The increase is more likely due to increasing prevalence of resistant bacteria and increasingly older/sicker patients who are less able to fight off bacterial infections.


Take a complex issue such as infection with bacterial and viral agents that have had millions of years of natural selection and evolution to perfect their survival skills, and attack it with a popular economist's simplistic solutions and miraculously another problem is solved.This is another example of the self-proclaimed Best and Brightest actually being the Worst and Dimmest.


Have you ever lived with a doctor? Looked at his hands? I have. The doctor's hands were in constant pain, rough and dry from all the hand washing he did. Using rubber gloves throughout the day didn't help. It just made it worse; even the hypoallergenic gloves didn't help.

syed mansoor hussain

Sadly I hope some of us know how Semmelweis ended up.



Mike B:

Your analysis might have some legs were it not that Docs are less likely to wash than RN's. RN's have much more patient contact and, to the extent that time and hand damage were the issue, we should expect the same poor hygiene. In fact, it is a cultural issue. Docs are high status and can get away with not washing RN's, not so much.

Wonks Anonymous

An alternative approach would be an accountable care organization where doctors work together as a team and the results of one doctor are closely identified with those of all the other doctors.

As someone once said about Kaiser Permanente: Here their doctor's name is Kaiser.

Not surprisingly everyone works just a little harder to keep things clean and the results for hospital borne infections show it.

Ian Kemmish

I suspect the forensic work you mention was not carried out in a hospital.

At a crime scene, the people touching the objects will have come from different places, probably won't have interacted much with each other (or at all, if it's a burglar), and won't have washed their hands abnormally often.

In a hospital, you have a lot of people constantly in intimate contact with each other (this is, isn't it, how superbugs spread in the first place?) and washing their hands many times a day, both with detergent and with alcohol.

From my position of total ignorance on the subject, I suspect that this would lead to everyone in the hospital having unusually similar skin bacteria "profiles". If the question you seek to answer is "which surgeon had this particular strain of C dificile on his skin twenty hand-washings ago?" I don't think this approach will help.

Gerald J Barron MD

We Docs need to start using gloves when touching our patients, just like the people serving you hot dogs at costco. It is safer for the patient and the doctor. The cost of the gloves is much less than one law suit


It seems like a situation that will require a multi-part solution.
First patients should speak up on their own behalf. "Doc, did you disinfect your hands before you walked in here?" Second perhaps would be the general consequences for the hospital for high infection rates. And then perhaps Third some sort of forensic approach to link an infection to its source person.

I really disagree with Mike B that "Disinfection needs to be made nearly costless to drive up use of the service." Many doctors are simply unwilling to face up to the consequences of their actions. The cost of Doctors not washing there hands (or using gloves, or purel) can be life and death. Why don't they consider good hygiene to be a major part of being a good doctor? Why would a doc be satisfied with saving one patients life by some brilliant diagnosis and then killing another patient with a careless infection. They all took an oath to "first do no harm." Its a cultural problem.



Why is everything so focused on the "doctors"? What about nurses, aides, therapists, etc. There are many more contacts with hospitalized patients made by ancillary personnel than the doctors. Seems like the doctors are an easy target -- maybe your article should point at "healthcare workers"?