Doctors who prescribe too many antibiotics: It’s not that simple

Physicians who liberally prescribe empiric antibiotics are often maligned as irresponsible or unthinking by condescending colleagues and policy wonks. But are these doctors actually courageous and prudent, saving countless thousands of lives every year by refusing to bend to misguided pressure from antibiotic-conserving paper-pushers?

As antibiotic resistance has emerged, many hospitals have begun requiring physicians to provide a rationale for every antibiotic dose prescribed. The Centers for Medicare and Medicaid (CMS) is considering making this standard for every antibiotic dose, at every hospital, on every Medicare patient nationwide.

There’s every reason to have a rationale for antibiotics, or any other plan in patient care. But the development is part of what I perceive as a larger “pendulum swing” toward pressuring physicians to avoid prescribing empiric antibiotics whenever possible. That may not be such a great idea, since (unlike a lot of the things we do) antibiotics actually save lives on a fairly regular basis.

At academic training programs, a seductive and self-serving myth reigns: that doctors can usually know whether or not their ill-feeling patients are infected, and in what part of the body, and with what likely organisms. And all this can be done in a few moments, on rounds, with the internist’s tools of history, exam, and a few labs! This would be a miraculous feat if even one living physician could do it consistently — but in training programs, it’s presented as routine, a mundane and expected part of any physician’s job.

If a concerned resident at my training program dared to give empiric antibiotics overnight to a patient not clearly at death’s door (and showed up to rounds without a Petri dish in hand growing the responsible organism), often as not the attending’s disapproving question would be “But what are you treating?”

It wouldn’t take long until you’d hear these same residents responding to suggestions that they give a sick patient empiric antibiotics with the same magic phrase, meant to simultaneously imply their own judiciousness and the colleague’s profligacy: “But I don’t know what I’m treating!” Someone not-that-clever coined “Vosyn” to half-pejoratively describe the broad spectrum cocktail of vancomycin and piperacillin-tazobactam (Zosyn) that supposedly unthinking physicians would prescribe for patients who were very sick from unclear causes.

But what I always wondered was, how many people got better from those antibiotics?

While there are surely some penicillin-pushers among us, slinging antibiotics at anyone with a chief complaint, I’m pretty sure that the vast majority of physicians prescribe antibiotics to one group of patients: those who they think might have an infection. Maybe even an infection that could get seriously worse if untreated.

So I was glad to read Kent Septowitz at Memorial Sloan Kettering-Cornell’s editorial in the New England Journal of Medicine. He seems to agree that we physicians are overly apologetic about our supposed crimes of antibiotic overuse, and that by buying into the myth of rampant irresponsible prescribing that needs to be administratively controlled:

We also have promoted the notion that the field of clinical medicine is far simpler than it actually is. Despite our confident claims to the contrary, the diagnosis of infection is anything but an exact science. In the daily tumult that is clinical care, antibiotics have bailed us all out countless times. Blood work, radiology results, and the physical exam declare their limits to the practicing doctor every day. Often, when we are stumped and lost in caring for a patient, we turn, thankfully, to a prescription for an antibiotic. Just in case. Only hubris prevents us from admitting the number of times this approach has saved our patients’ health and our reputations.

Of course broad spectrum antibiotics like Zosyn and Cefepime need to be conserved as much as possible, and physicians should offer reassurance — not azithromycin or amoxicillin — to patients with mild, likely viral respiratory infections. But the emergence of antibiotic resistance is an insanely complex phenomenon that has been oversimplified to a story of antibiotic-hungry patients, satiated by customer-service-friendly, lawsuit-averse physicians (cast as the bad guys), with the stingy antibiotic stewards serving as humanity’s last line of defense from a superbug plague.

According to my far smarter infectious disease colleague Brad Spellberg in his book Rising Plague, it just ain’t that simple: physician overuse of antibiotics and “dirty” hospitals are not the main causes of the global rise in antibiotic resistance. (If you want to know what the causes really are, and what we should do about it, buy his book.)

My favorite sarcastic response (under my breath) to someone who argues against giving antibiotics to a sick patient who passes my test of “If this were me, would I want a shot of ceftriaxone?”

“Save the antibiotic, not the patient!”

Matthew Hoffman is a fellow in pulmonary and critical care who blogs at

Comments are moderated before they are published. Please read the comment policy.

  • Dike Drummond MD

    Great post Dr. Hoffman. One of the massive reasons doctors prescribe antibiotics inappropriately is they don’t know what to say to the patient. We evaluate – have made an internal decision NOT to prescribe – and get all stressed and tongue tied about exactly what words to use to “say no”.

    [why this is NOT a mandatory course in med school or residency is beyond me!]

    It is more than just saying no. It is aligning the patients understanding and expectations with your clinical judgement. Here is a post on my blog and subsequently published by Dr. Kevin Pho here on KevinMD that gives you a structure and specific words for that conversation.

    Dike Drummond MD

  • Lucy Hornstein

    Inpatient and outpatient: apples and oranges. Your arguments make perfect sense for patients ill enough to be hospitalized. Ambulatory patients who come in with coughs and runny noses are still getting way too many antibiotics prescribed for their “bronchitis” and “sinus infections”. It’s even worse with kids.

    No, I don’t think the odd Z-pack has “saved” anyone in my practice, given that they were destined to get better over the same 5 days anyway.

  • HemlockHouse

    Grrr, overuse of antibiotics = one of my biggest pet peeves in medicine. And beyond that is the use of broad-spectrum antibiotics when a narrow-spectrum drug will work, or when a “new” antibiotic is used instead of sticking with the “old”, still effective, antibiotic. Physicians must do the right thing for the patient, not make the patient happy by fulfilling their expectations. Throughout my 25 yr career as a pediatrician I have far too often seen antibiotics unnecessarily prescribed for children due to pressure or expectations from anxious +/or angry parents. Otitis media is overdiagnosed and overtreated (with antibiotics), then, after several of these episodes, a child is referred to an ENT MD and tubes are placed. Eustachian tube dysfunction, middle ear fluid, or a “red” TM should not be treated with antibiotics. “Tugging on ears” is not a reliable sign of AOM. Acute tonsillitis is often treated despite a negative RST (and negative TC), then, after several of these episodes, a child is referred to an ENT MD and a T & A is done. “Sinusitis” or “purulent rhinitis” or “early sinusitis” are terms used (often just because the nasal mucous has become yellow-green) to somehow justify the use of antibiotics when, in fact, a child has a viral URI (viral rhinosinusitis). Yellow-green snot is normal during a viral URI. Then with every URI the parents demand antibiotic treatment. “Bronchitis” is a term used to somehow justify the use of azithromycin. When are US docs going to get smart and tough when it comes to prescribing antibiotics?

Most Popular