More than half of all hospitalized patients receive antibiotics. The overwhelming majority of physician specialties prescribe them regularly. So, we should all know how they work and what to expect of them, right?
The unfortunate reality is that clinicians tend to treat antibiotics as a panacea, and I have consistently observed several misconceptions among healthcare providers with high potential for wasting resources and harming patients:
They work immediately
As one of my attendings in fellowship liked to say about starting an antibiotic, “it’s not magic dust.” Sure, there are cases where we start to see improvements within hours of starting an antibiotic, but I would say it’s not a fair trial until a patient has been on antibiotics for 48 to 72 hours. This is particularly true for skin and soft tissue infections. Cellulitis is known for getting worse before it gets better on appropriate therapy – probably because, when antibiotics work, they lead to increased destruction of bacterial cells. These dead cells spill their protein contents into the surrounding tissue, transiently provoking an even greater inflammatory response. I have seen it many times – a patient starts out on Ancef for non-purulent cellulitis, but the next day…”uh oh, the redness has extended beyond marked borders!” Next thing you know, the patient is on IV vancomycin, Zosyn and clindamycin with a surgical consult and MRI pending, when all they really needed was elevation of the affected body part and another 24 hours of Ancef. As long as your patient isn’t clearly and rapidly worsening (hypotension, rising lactate, increasing pain, hypoxia, etc), give that first antibiotic a chance to work before you reach for the big guns.
They eradicate the target bacteria
Wait, what? Yes, you read that correctly. Even when using a highly-effective bactericidal agent, we only ever eliminate a large proportion of a given bacterial population – not the entire thing. That’s because, as a general rule, a small portion of bacteria will be in a stationary, or dormant phase at any given time. Most of our antibiotics only work to kill bacteria while they are growing and dividing. When we hammer away with antibiotics over long periods of time, like with a bone or hardware infection, it’s not because there’s an inordinately high number of bacteria involved. It’s because we’re effectively “waiting out” the dormant bacteria – hoping they will wake up and emerge from their protective biofilm so the antibiotics can finish the job.
Ideally, between antibiotic pressure and normal immune function, we do manage to eradicate problematic bacteria from a site of concern, whether that’s an abscess under the skin or an infected hip prosthesis. Even when we clear the infection, however, bacteria tend to hang around quietly in other parts of the body without causing symptoms – a phenomenon known as colonization. Staph aureus, and particularly MRSA, is known for doing this. Patients who get a severe MRSA infection frequently continue to carry the bacterium just inside the nose or on the skin surface long after the infection has resolved. You can continue systemic anti-MRSA antibiotics as long as you want – in all likelihood, a colonized patient will remain that way. This what I mean when I say that antibiotics don’t eradicate the target bacteria. Especially with more resistant organisms, there’s a good chance that some are left alive in or on the body following antibiotic therapy, so don’t waste your time with “clearance cultures” for anything other than bloodstream infection.
The longer the course, the higher the risk of resistance
I tend to treat osteomyelitis (bone infection) with longer courses of antibiotics than many of my colleagues – that’s a post for another day. These long courses often elicit the same question from patients and even some colleagues: “aren’t you worried about causing antibiotic resistance?” In a broader sense, yes, but when I have a good reason for extended treatment of a particular bacterial infection, no. As it turns out, if we use an antibiotic that tests fully effective against the bacterium of concern, and we know that the antibiotic is reaching the target tissue, the risk of inducing resistance on therapy is quite low. Resistance is more often the product of repeated antibiotic courses over time – not a single, uninterrupted course. This is because resistance genes are more likely to emerge when bacteria are given the opportunity to grow and reproduce. In the presence of consistent, effective antibiotic pressure, resistance is much less likely to develop because there are fewer opportunities for bacterial genes to mutate.
While we generally try to limit their use, there’s a time and place for very long – even indefinite – courses of antibiotics. Orthopedic hardware infections, for example, can be very difficult to eradicate – even more so if infected metal cannot be removed. I recall one patient whose knee replacement became infected several weeks after surgery. His surgeon felt that the risk of removing the prosthesis would be greater than the benefit, and we had considerable difficulty getting the joint infection under control with IV antibiotics. Given the desire to avoid surgery and the high probability that his infection had not been completely cleared from the hardware, I recommended an indefinite course of suppressive oral antibiotics once he stabilized to keep any remaining bacteria in check. He did well for a period of months, but then his well-intentioned primary care physician advised him to stop the antibiotic out of concern for developing resistance. A little over a week later, he was back in the hospital getting his knee surgically washed out for recurrent infection. Of note, operative cultures grew the same bacterium he had with the first round of treatment, and it was still fully susceptible to both the IV and the oral antibiotic we had used previously.
“Go big or go home”
This is one of my pet peeves as an ID physician – the impulse to blast patients with the broadest-spectrum antimicrobials we can think of because, well…they’re sick.
Counterpoint: It’s a hospital. They’re all sick.
Nevertheless, the majority don’t need America’s favorite “double-barrel” combo (vancomycin and Zosyn, just in case it wasn’t clear). Non-purulent cellulitis? Ancef should work just fine. You can even throw in clindamycin for good measure, and I would consider that pretty reasonable. Pneumonia in a patient who never previously set foot in a hospital? Ceftriaxone and azithromycin, please. Granted, there are areas of the country with high overall rates of antibiotic resistance where a more aggressive empiric approach might make sense. For the rest of us, scorching the proverbial earth just drives additional antibiotic resistance, C difficile rates, higher costs of care and increased adverse outcomes for patients.
A little discretion and consideration of individual patient histories go a long way in selecting appropriate antibiotics. Just don’t pull out the “big guns” without a good rationale. Yes, we can always de-escalate later, but even one or two days of very broad-spectrum treatment can become enormously detrimental if patients are receiving it routinely. If you need more reasons to avoid the V/Z combo, see my “open letter” from 7/23/19.
IV is better than oral
Of all the erroneous ideas I strive to correct, this one might drive me the craziest. There’s a pervasive sense in US healthcare that we aren’t being aggressive enough if we use oral antibiotics in the hospital. I don’t know who gets credit for the original quote, but my favorite line to use in response to this idea is “the bacteria don’t know how the antibiotic got there.” If the patient has a working gut and is able to take pills, there’s no reason we can’t achieve the same tissue concentrations with an oral agent compared to its IV equivalent.
One caveat here is that IV antibiotics achieve those same tissue levels faster than oral therapy, so I’m not saying there’s no role for IV antibiotics. If a patient comes in acutely septic and ill-appearing, by all means, start them on IV therapy because minutes count when it comes to severe bacterial infections. Treating a highly resistant organism is another valid reason to use IV over oral therapy, since many of our most potent antibiotics do not have oral formulations. Lastly, patients may not be able to take pills for a number of reasons like malabsorption, bowel obstruction or persistent vomiting, so IV antibiotics are needed in those cases. Otherwise, if you have an effective pill that your patient can take, use it.
I know some of my colleagues would balk at this assertion, probably recalling the times that their patients purportedly failed a course of cephalexin (Keflex) for skin infection and had to come into the hospital for IV antibiotics. I find that there is usually a good explanation for these cases other than “oral therapy was inadequate.” Going back to the first misconception above, I’ve seen plenty of situations where patients were declared an “oral antibiotic failure” after just one or two doses, which, as I said before, is not really a fair trial. Another recurring theme is underdosing of the oral antibiotic. Continuing with cephalexin as our example, I typically see prescribed doses of 500 mg given three or four times a day for skin and soft tissue infection. While this is technically within labeled dosing recommendations, I don’t think it’s enough. We have medical literature showing that cephalexin’s IV equivalent, cefazolin (Ancef), often requires more than the recommended dose to achieve adequate tissue concentrations, especially in overweight patients. Even though it has not been as well-studied, it’s not hard to imagine that we are probably also underdosing cephalexin. In fact, I and others in my field prescribe 1,000 mg three times a day for skin and soft tissue infection – a cumulative dose increase of 50% over 500 mg four times a day. Patients tolerate the higher dose just fine, and I see fewer treatment failures.
IV therapy introduces a host of potential complications like clotting, line infection, renal failure, liver injury…not to mention the excess costs associated with maintaining IV access. Do your patients and the healthcare system a favor by using oral antibiotics whenever possible.
Antibiotics are probably some of the most misused drugs in healthcare today. Medications that were once antimicrobials of last resort now get prescribed “just in case.” We’re quick to reach for broad-spectrum IV therapy, largely because it just makes us feel better – like we’ve got it all covered, so nothing bad can happen to the patient now. In reality, we almost certainly expose patients to greater risk by giving them unnecessarily broad antibiotics, rather than trusting our clinical judgment and targeting our therapies. It’s our responsibility as healthcare providers to understand the treatments we recommend, and a good grasp of antibiotic properties is particularly important in this age of rising antimicrobial resistance.