Editor’s Spotlight/Take 5: Is Patient-reported Penicillin Allergy Independently Associated with Increased Risk of Prosthetic Joint Infection After Total Joint Arthroplasty of the Hip, Knee, and Shoulder?

We’ve all been there, on both sides of this one. We’d like to give the recommended first-generation cephalosporin prophylaxis to a patient undergoing major elective orthopaedic surgery, but the patient reports a serious penicillin allergy. Or, we’re treating a patient with a serious orthopaedic infection, but the chart shows a record of that allergy; we wonder whether appropriate antibiotics were used at the time of the index intervention, and so whether the disaster might have been prevented.

William F. Sherman MD

For decades, studies have shown that most patients with penicillin allergies can safely receive cephalosporin drugs [3, 6], and considerable research has called into question—or frankly, recommended against—using noncephalosporin drugs for surgical prophylaxis because of the increased frequency of infections with gram-negative organisms [12] or an increased overall risk of serious orthopaedic infections [7]. Even in the face of all this evidence, cephalosporin avoidance in this setting remains common. About one in five surgeons and anesthesiologists do not prescribe preoperative cephalosporins when a patient reports having had a penicillin allergy [9], despite the fact that the overwhelming majority of patients with documented penicillin allergies can take cephalosporins safely, and despite the substantially increased risk of serious musculoskeletal infections that occur when they are not used [14].

In this month’s Clinical Orthopaedics and Related Research®, a group of investigators from Tulane University in Louisiana, USA, report that patients labeled as having a penicillin allergy are more likely to develop prosthetic joint infections of the knee or shoulder than those who don’t carry this label, even after controlling for relevant confounding variables, like BMI, anxiety, and depression [13]. The odds of PJI increased nearly four-fold among patients having total shoulder arthroplasties (the low end of the 95% CI in that group was nearly 3, which means the estimate was fairly precise), and the odds increase likely was between 10% and 40% in patients undergoing TKA. This risk difference is no mere trifle.

The label “penicillin-allergic”—carried by so many patients—therefore is itself harmful. But there is nothing actually wrong with this patient population. The problem is on our side: Surgeons see the label, and practice a kind of defensive medicine that can maim or kill. In trying to help these patients, by “protecting” them from cephalosporins, surgeons likely prescribe less-effective antibiotic prophylaxis, according to this CORR® study from a group led by William F. Sherman, MD.

Fortunately, the label itself also is easy to spot, which is what makes this paper so important. The solutions are easy: If the allergy isn’t a serious one, cephalosporin drugs should be fine; if the patient might have anaphylaxis, a test dose [4] or a preoperative allergy consult [14], is all it takes to save a patient’s joint, and perhaps a life.

Readers often say, “this study raises more questions than it answers.” I’m glad to report that this paper answers more questions than it raises. Given the importance of the topic, that should make it required reading.

Join me as I go behind the discovery to learn a little more about how we got to a place where labels can kill, and how to implement some common-sense solutions in busy practices, with Dr. Sherman, senior author of “Is Patient-reported Penicillin Allergy Independently Associated with Increased Risk of Prosthetic Joint Infection After Total Joint Arthroplasty of the Hip, Knee, and Shoulder?” in the Take 5 interview that follows.

Take 5 Interview with William F. Sherman MD, senior author of “Is Patient-reported Penicillin Allergy Independently Associated with Increased Risk of Prosthetic Joint Infection After Total Joint Arthroplasty of the Hip, Knee, and Shoulder?”

Seth S. Leopold MD: Congratulations on this important study. It is a great reminder about how the words surgeons use and perceptions surgeons have about patients can hurt, mislead, or result in harmful treatment decisions, which are themes we’ve covered here before [10, 11]. Focusing on the label of the thing, rather than the thing itself, therefore strikes me as a perceptive approach; what caused you to go this route?

William F. Sherman MD: Our question was generated because de-labeling of incorrect self-diagnoses can itself improve health. Patients routinely list mild side-effects as “allergies” or even report being told as a child they were allergic to a particular antibiotic, specifically penicillin, resulting in surgeons not choosing first line antibiotics because they might, but probably do not, have cross reactivity.

Second-hand stories sometimes present as diagnoses, but when they matter, physicians will order specific tests for confirmation. For allergies, however, a self-report to a doctor often is treated as though it’s just as valid as a medical specialist confirming with testing. That shouldn’t be. It seemed to our group that this gap in our system was adversely affecting patient care.

Dr. Leopold: Still, attributing a biological endpoint (infection) to what I’d call a social factor (a label without any certain biological confirmation) is always a fraught exercise, which is another theme we’ve covered here recently [8]. Why are you so certain you’re drawing the right inference here?

Dr. Sherman: Research has shown that penicillin allergy labels affect antibiotic choices in nearly 80% of patients when we surveyed the available evidence [9]. This is despite the lack of cross reactivity being well established and detailed guidelines from national societies regarding antibiotic administration in this population [1, 2, 5]. While this is a social factor, it causes objective harms in treatment.

Dr. Leopold: I think the connection you’ve made is real, too. In light of it, what do you perceive to be the best-practice recommendations for a patient having arthroplasty surgery if he or she reports a penicillin allergy on your history in the office?

Dr. Sherman: Our current practice is to explore the severity of a suspected penicillin allergy by taking a more-detailed history, and if needed, getting medical records. With the rare patient who experienced documented anaphylaxis or a severe allergic reaction, testing is not done for penicillin allergy. Instead, these patients are typically given an alternative antibiotic as no current commercial testing is available for cephalosporin allergies in the United States. For patients who have self-reported allergies that are mild or unknown, we proceed with a test dose upon entering the operating room without penicillin allergy testing. If the patient tolerates the test doses of cephalosporin, we fully dose the cephalosporin. If there is an unclear history and it would potentially limit the use of a first-line antibiotic, we would send the patient to an allergy specialist for penicillin testing to determine if a test dose seems advisable.

Dr. Leopold: OK, let’s say you missed it in the office—it happens—and the patient now is in the pre-operative holding area before a total shoulder replacement, and the anesthesiologist tells you, “Your patient has a serious penicillin allergy; what antibiotic would you like me to give before surgery?”

Dr. Sherman: Similar to the office, if we cannot detect a clear path, we do err on the side of caution for cephalosporin usage. With a known infection rate of around 0.6% in most large studies in total joints and an anaphylaxis rate of 0.01% with use of a cephalosporins, we are vigilant in our approach to ensure that we make sound decisions so as not to miss opportunities for intraoperative testing. In this setting, where there is no medical documentation of a severe allergic reaction, I would discuss the risks and benefits with the patient. In the absence of documentation of a severe reaction, I would offer a test dose of a first-generation cephalosporin in the operating room, where resources are available if the patient experiences a severe reaction like anaphylaxis.

Dr. Leopold: Test doses in the operating room are free, allergy consults are more expensive, and PJIs and deaths from anaphylaxis are, well, let’s not even go there. What is the downside of just using a test dose for all patients rather than taking the time and cost of an allergy consult, and would an allergist see it differently?

Dr. Sherman: There is no identifiable delay in our system, as we give two test doses 10 minutes apart immediately after entering the operating room; costs to the patient and the system certainly are lowest with this approach. I think with the potential to decrease the risk of anaphylaxis being even lower if one takes a thorough history and performs a thorough investigation, that is a reasonable way to go. Engaging with an allergy consultant in patients whom the surgeon perceives as at higher risk for a reaction makes things safer (without increasing costs by much, since there are relatively few such patients). I think the allergist would agree that by identifying high-risk individuals, we can potentially lower the likelihood of a patient experiencing anaphylaxis to the test dose, as well. A visit with an allergist also gives the patient further opportunity to participate in the decision-making, as the visit is focused solely on antibiotic choice, which also improves the patient’s ability to make an informed decision.

Of course, the best answer would be the development of efficient, affordable, direct testing for cephalosporin allergies in the United States; however, this does not appear to be on the near-term horizon. In the meantime, I believe the use of cephalosporins for patients who do not have serious allergies, test doses for those who might, and allergy consults for those who seem likely to be at high risk is the way to go.