As I wrote in a 2013 post on the subject of antibiotic overuse in the setting of treating viral infections in kids, the practice of medicine is more than a bit challenging at times:

So what is the deal with us pediatricians? Why can’t we keep our Hannah Montana brand non-latex examination mitts off of the prescriptions pad every time little Timmy has a runny nose or a cough? Why don’t we listen to the well-meaning experts from the AAP and our own inner voices? To put it bluntly, it’s because medicine is hard. It’s messy, it’s complicated, and it is practiced by humans beleaguered by the same propensity for bias and intellectual laziness as everyone else.

Nothing has changed in the past 7 years that makes the decision of whether or not to prescribe antibiotics any easier. Sure, there are a few new tests these days. And reams of data have been published on the harms that come from inappropriate use of these lifesaving medications. Something, something, something, microbiome. Unfortunately, the factors that are at the heart of this issue remain unchanged. Doctors are still human, at least until we are replaced by artificial intelligence or aliens.

Most studies on the uses and abuses of antibiotics in children come from the outpatient world. In hospitals, particularly those fancy academic pediatric facilities, things have to be better…right? Sadly, this doesn’t appear to be the case. What follows is a discussion of a study recently published in Clinical Infectious Diseases, the flagship journal for the Infectious Diseases Society of America, that demonstrates just how pervasive suboptimal antibiotic therapy is.

Appropriateness of antibiotic prescribing in U.S. children’s hospitals

In the study’s introduction, the authors point out that antibiotics are the most commonly prescribed class of medications in outpatient pediatric medicine and are ordered for more than half of hospitalized children. Most of these are for respiratory infections, which are very often viral. This doesn’t surprise me because there really seems to be a cavalier attitude regarding their use. I see this in members of the lay community, where there tends to be little awareness of the risks inherent in the use of antibiotics despite attempts at large scale educational outreach, but also far too often in those who really should know better.

Inappropriate use of antibiotics is a big problem. Antibiotics, even when used appropriately can harm patients, but prescribing them carelessly results in significant unnecessary medical costs, increases rates of infections with resistant organisms, and inches us ever closer to a return to a time when doctors were powerless against bacterial infections. Adverse effects from antibiotics, which naturally become more common with increasing numbers of prescriptions, can involve fairly benign nausea and diarrhea, but also life-threatening C. difficile infections and potentially deadly allergic reactions and immune system dysfunction. Antibiotics may have saved countless lives, and hopefully will continue to do so far into the future, but they are far from risk free.

Adult data has revealed that 30-50% of antibiotics ordered in U.S. hospitals are done so inappropriately. Pediatric data has been very limited, typically involving small numbers of children from only one study location. Using data from 32 U.S. children’s hospitals, and over a period of 18 months, the study authors attempted to answer several questions regarding antibiotic use, particularly what percentage is suboptimal, what errors are being made, and where improvement efforts might be best focused.

Another interesting aspect of this study was the involvement of what are known as antibiotic stewardship programs (ASP). With an ASP in place at a hospital, there are infectious disease experts charged with monitoring the use of specific antibiotics in certain situations. They will frequently put limits in place or require justification for continued use of these particular drugs, and they provide recommendations for medication changes and education to help reduce antibiotic related misadventures. Not all hospitals have them, and they can’t monitor and provide global assistance for every single clinical encounter where antibiotic choices are made.

In the study, each of the participating children’s hospitals completed several single-day antimicrobial use surveys, essentially one every 3 months, between July of 2016 and December of 2017. On the designated days, charts from every current patient aged 0-17 admitted prior to 8 AM were reviewed, and those with an active antibiotic order were analyzed by members of the ASP at that facility. In addition to standard demographic data, they collected the dose, route, and indication for each antibiotic being administered to the patient. They also made note of whether or not the hospital ASP would have otherwise been involved.

They determined the appropriateness of each antibiotic, which was based on the clinical judgement of the reviewing ASP member using an option from a list of predetermined reasons, and whether modification should be recommended. The designated reasons that an antibiotic might be deemed inappropriate included the need to change an antibiotic based on available lab data, unnecessary duplicate antibiotic therapies, use of an IV antibiotic when an oral agent was adequate, inappropriate surgical prophylaxis, and “other”. The selection of “other” was based on each facilities specific ASP practices.

There were 18 specific antibiotic modifications that could be recommended by the ASP reviewers at each hospital. These included options such as stopping an antibiotic, changing to a more narrow antibiotic based on either lab data or clinical judgement, and “other”. Each antibiotic was allowed only one reason for being inappropriate, if applicable, and one modification recommendation, so reviewers were instructed to pick the one that they felt was most clinically meaningful.

The results were…not so good

The primary outcome for the study was the percentage of suboptimal antibiotics, as defined by its being inappropriate, requiring modification, or both. They also looked at a few interesting secondary outcomes, which I’ll get into shortly. Because the “other” option for inappropriateness was used so often, the authors created new categories after the fact based on the provided explanations from the ASP members at each study site.

When all the surveys were collected, the study authors had antibiotic data and ASP assessments on just over thirteen thousand pediatric patients with 17,844 total antibiotic orders. Patients receiving antibiotics represented 35% of the total number of inpatients with nearly 70% getting antibiotics through a vein rather than by mouth. A lot of the data is interesting to me, such as how many patients were on more than one antibiotic and what antibiotic was most commonly prescribed (SMX/TMP, Bactrim), but I’ll spare you those details and get to the good stuff.

The most common indication for antibiotics wasn’t surprising at all: bacterial lower respiratory tract infection (LRTI) or pneumonia. The next most common indication, which reveals the difference between pediatric hospitals and community hospitals like mine that admit kids, was prophylaxis for medical problems. This means that they admit a lot of medically complex kids that are high risk for serious bacterial infections, such as patients with immune system problems or cancer.

Now the juicy stuff. The authors found that 21% of antibiotics were suboptimal, with about twice as many designated as inappropriate compared to those just needing some modification. Overall, 26% of patients received at least one suboptimal antibiotic. It’s important to point out that there were huge differences between some hospitals, with one reporting that almost half of patients on an antibiotic needed a prescription modified and one coming in at less than 10%.

Among the suboptimal antibiotic orders deemed inappropriate, the most common reason was that it was the wrong antibiotic based on available lab data. That means someone, actually multiple someones, dropped the ball on each of these 635 patients, putting them at risk of worsening of their infection. Among those orders that just needed tweaking, the ASP team recommended stopping the antibiotic almost half the time. That’s not as bad as having a sick patient on the wrong treatment, but it still risks harm and adds to resistance concerns.

Data on one particular antibiotic class, namely the oral 3rd generation cephalosporins, caught my attention even if they were not one of the top ten most prescribed hospital antibiotics. Though very commonly, and wrongly, prescribed in the outpatient setting, they didn’t crack the top ten in this study because most hospitalized kids are on IV antibiotics. But when used, they were the most likely (50%) to be suboptimal. Though they certainly have a place in medicine, they are second only to azithromycin as my least favorite antibiotic. I even wrote an entire post about them a few years ago.

There were a few additional interesting findings in this study. Of the antibiotics ordered specifically to treat pneumonia, 22% were suboptimal. This certainly fits with my experience, where pneumonia is frequently overdiagnosed in the emergency department because of overzealous imaging and discomfort with uncertainty, and the choice of initial antibiotic is almost always too broad.

The antibiotic indication most likely to be associated with suboptimal use was surgical prophylaxis. This occurred in 40% of instances where it was ordered, most commonly when the antibiotic was continued for longer than 24 hours. Current recommendations recommend one dose. Finally, almost half of the suboptimal antibiotic orders would not have been reviewed by an ASP program outside of this study. Just imagine how things might look at hospitals where there is no ASP at all.

Conclusion: We need to expand antibiotic stewardship in our hospitals

So what’s the bottom line from this study? A lot of kids admitted to pediatric hospitals are on antibiotics. And although it appears that we do better than the doctors that manage adults, we need to do better. Of the thousands of kids in this study receiving inpatient antibiotics, 1 of every 4 did so suboptimally. Half of the suboptimal antibiotics should have been stopped and another 20% needed to be changed to an antibiotic that was more narrow, both of which are risk factors for the development of resistant bacteria.

There was huge variation between hospitals in this study. Some appeared to have a much better grasp of proper antibiotic use than others. It’s important to figure out why there are such disparities and what can be done about it. Every participating hospital in this study has an antibiotic stewardship program, although the quality and effectiveness of each is unknown. That may be a factor in the differences in rates of suboptimal antibiotic use. But nearly half of the suboptimal orders wouldn’t have even been reviewed by an ASP, and many non-pediatric hospitals don’t an ASP.

As I mentioned earlier, it isn’t feasible for every hospital to have an ASP that monitors 100% of antibiotic orders. But this is an area that deserves focus and resources wherever possible because we are running out of antibiotics and we are running out of time. Expanding ASP services is just one way to reduce antibiotic misuse, but it is an important one.


Posted by Clay Jones

Clay Jones, M.D. is a pediatrician and a regular contributor to the Science-Based Medicine blog. He primarily cares for healthy newborns and hospitalized children, and devotes his full time to educating pediatric residents and medical students. Dr. Jones first became aware of and interested in the incursion of pseudoscience into his chosen profession while completing his pediatric residency at Vanderbilt Children’s Hospital a decade ago. He has since focused his efforts on teaching the application of critical thinking and scientific skepticism to the practice of pediatric medicine. Dr. Jones has no conflicts of interest to disclose and no ties to the pharmaceutical industry. He can be found on Twitter as @SBMPediatrics and is the co-host of The Prism Podcast with fellow SBM contributor Grant Ritchey. The comments expressed by Dr. Jones are his own and do not represent the views or opinions of Newton-Wellesley Hospital or its administration.