Cefazolin, a first-generation cephalosporin, is the most common antibiotic used for antimicrobial prophylaxis in orthopedic patients.1,2 This is due to its effectiveness against common gram-positive and aerobic gram-negative bacilli.3 However, a significant portion of the population has a self-reported diagnosis of a penicillin allergy. Penicillin shares a beta-lactam ring and various side-chains with cephalosporins. These shared structures may play a role in the cross-reactivity between these drugs.4–6 The reported rate of patients with a penicillin allergy can be higher than 10% in arthroplasty patients.2 Although early literature reported the cross-reactivity of penicillin and first-generation cephalosporins at 10%, recent studies have shown the rate of cross-reactivity to be between 0.3% and 1% in patients who report a penicillin allergy. In addition, cross-reactivity can be up to 2.55% in patients with a confirmed penicillin allergy.2,7,8 The discrepancy between the rates is attributed to older formulations of cephalosporins being contaminated with trace levels of penicillin.9–11 Despite the purification of cephalosporins in the 1970s, many providers are still hesitant to prescribe cephalosporins in patients with a documented penicillin allergy.2 Although there is evidence that cephalosporins are safe in patients with mild-to-moderate penicillin allergies, few studies emphasize this. One such study was done by Ramsey et al.12 They performed a retrospective review of 838 patients, comparing reactions to cefazolin in patients with self-reported mild to moderate penicillin allergies vs patients without penicillin allergies. They found no significant increase in adverse reactions between these 2 groups, and they had no life-threatening adverse reactions occur.12 Haslam et al2 did a similar retrospective review in arthroplasty patients and found no adverse reactions in patients who reported a non-IgE mediated penicillin allergy when they received cefazolin. Studies show that alternative antibiotics to cefazolin may lead to patients receiving prophylactic medication that is either less effective, less safe, or costlier. Vancomycin and clindamycin are 2 commonly used alternatives to cefazolin. The cost of these medications can be much higher than cefazolin at some institutions—1.6 times higher for clindamycin and nearly 9 times higher for vancomycin.2 Clindamycin use is associated with Clostridium difficile infections at an odds ratio (OR) of 9 compared with 2.6 with a first-generation cephalosporin.13 On the other hand, vancomycin use is associated with increasing vancomycin-resistant Enterococcus, and its use should be limited to prevent increased resistance.14 For these reasons, cefazolin is a better prophylactic medication to prevent surgical site infection in orthopedic populations.
Given these concerns over antibiotic selection, the purpose of this study is to evaluate the use of preoperative antibiotics in patients with penicillin allergies. In addition, it will compare the difference in perceptions and management of patients with penicillin allergies by orthopedists and anesthesiologists. The authors hypothesize that anesthesiologists and orthopedists will be more likely to prescribe cefazolin in patients with a non-anaphylactic allergy to penicillin if they know the correct cross-reactivity between cefazolin and penicillin or adhere to hospital protocols.
Materials and Methods
An institutional review board approved an exemption for this survey study. A short 16-question survey was sent to all the program coordinators at the Accreditation Council for Graduate Medical Education orthopedic and anesthesia residency programs. They were asked to forward the survey to all of the orthopedists and anesthesiologists in their practice. A second reminder email was sent 1 month later. The survey was collected over 2 months, from December 21, 2016, to February 28, 2017. The survey data were collected anonymously through Survey Gizmo. The survey was clearly identifiable as a survey and the reason and content of questions was identified. No participant was required to participate, and his or her participation in taking the survey was considered to be their consent. No reimbursement was offered or provided.
The current authors developed a questionnaire using previous literature and key informants. Orthopedic surgeons and anesthesiologists who manage patients with penicillin allergies participated in the development of the questionnaire. The current authors used a “sample to redundancy” approach by which they contacted new surgeons until no new items for the questionnaire emerged, which took 2 months. They initially pretested the questionnaire with an independent group of 5 orthopedic surgeons and anesthesiologists to evaluate whether the questionnaire appeared to adequately address the question of current practice (face validity) and whether the individual questions adequately address the objectives of the current study (content validity). These 5 physicians also commented on the clarity and comprehensiveness of the questionnaire. In addition, the current authors were able to survey 32 orthopedic (n=17) and anesthesia (n=15) residents who completed the survey twice over 30 days to assess the test-retest reliability of the questionnaire following institutional review board approval. Overall, there was an 84% agreement in their responses between time points.
The survey comprised 16 multiple choice questions. All questions had no response as an option to allow for test takers to withhold information if desired. All questions used clear and widely recognized terminology to enhance the validity of results. The survey length was kept to a minimum to maximize response rate and to limit barriers that may affect its proper completion. The survey was administered and secured by surveygizmo.com.
The current authors performed descriptive statistics and analyzed their results by performing chi-square analysis. Odds ratios were reported between each subgroup. All calculations were performed using Excel (Microsoft, Redmond, Washington). In any analysis with a result of 0, the current authors added 0.5 to the analysis.15 The OR results were rounded using the rule of 4 and the P values were rounded to 1 significant digit.
The current authors received 294 unique responses to their survey; 292 (99.3%) of these were complete. Exactly half of these respondents identified as anesthesiologists (n=146), and the other half were orthopedists. Figure 1 shows the distribution of responses across the United States. Respondents were primarily academicians (Figure 2) and represented a broad experience mix (Figure 3). Cefazolin was the first-choice preoperative antibiotic for elective orthopedic cases for 97% (n=142) of anesthesiologists and 100% (n=146) of orthopedists. In patients with a history of a non-anaphylactic penicillin allergy, cefazolin preference was 82.9% (n=121) and 78.8% (n=115), respectively. This trend of decreasing preference of cefazolin in patients allergic to penicillin continued with open fracture cases (Figure 4). However, only 57.5% (n=84) of anesthesiologists and 41.1% (n=60) of orthopedists knew the correct cross-reactivity between cefazolin and penicillin (Figure 5). The current authors found that orthopedists had higher odds of prescribing cefazolin to patients with non-anaphylactic allergies if they knew the correct cross-reactivity between cefazolin and penicillin (OR, 4.77; P<.01). This was also true in patients with open fractures and non-anaphylactic penicillin allergies (OR, 6.28; P<.01). Anesthesiologists also had higher odds of prescribing cefazolin in patients with non-anaphylactic penicillin allergies when knowing the correct cross-reactivity (OR, 3.59; P<.01). This correlation was not true for anesthesiologists when looking at patients with open fractures and non-anaphylactic penicillin allergies (Table 1).
Heat map of practice locations for survey respondents across the United States for anesthesiologists (A) and orthopedists (B).
Percent of providers in each practice setting for anesthesiologists (A) and orthopedists (B).
Years in practice of survey respondents for anesthesiologists (A) and orthopedists (B).
Percent of anesthesiologists and orthopedists who prescribe cefazolin in elective cases (A) and for open fractures (B).
Percent of providers with their believed rate of cross-reactivity between cefazolin and penicillin for anesthesiologists (A) and orthopedists (B).
Odds of Using Preoperative Cefazolin When Knowing the Correct Cross-reactivity
When assessing how the decision for which antibiotic was made, the majority of both groups believed preoperative antibiotics are chosen based on a combined decision (Figure 6). Furthermore, approximately one-third of each group (anesthesiologists, n=54; orthopedists, n=42) have an established institutional protocol about selecting antibiotics.
Percent of each specialty that makes the decision for preoperative antibiotics per anesthesiologists (A) and orthopedists (B).
When the current authors compared the likelihood of prescribing cefazolin between orthopedists and anesthesiologists, there was a strong trend toward orthopedists being more likely to choose cefazolin as a first-choice antibiotic (OR, 9.25; P=.07). However, anesthesiologists had increased odds of prescribing cefazolin in patients with a history of penicillin anaphylaxis (OR, 2.60; P=.03). Alternatively, orthopedists had higher odds of prescribing cefazolin to patients with an open fracture (OR, 2.76; P<.01). Both groups were just as likely to prescribe cefazolin in patients with a non-anaphylactic penicillin allergy (Table 2).
Odds of an Orthopedist Using Preoperative Cefazolin Compared With an Anesthesiologist
Having an established protocol led orthopedists to have lower odds of prescribing cefazolin in patients with non-anaphylactic penicillin allergy in elective cases (OR, 0.27; P<.01) and open fracture cases (OR, 0.91, P<.01). Otherwise, there was no difference in odds in prescribing cefazolin for orthopedists or anesthesiologists (Table 3). The current authors also found practice setting, experience with a patient with anaphylaxis to cephalosporin, and change in practice of prescribing cefazolin in the past decade not associated with the odds of prescribing cefazolin in both specialties. Years in practice did not increase or decrease the OR of prescribing cefazolin in patients with a non-anaphylactic penicillin allergy.
Odds of Using Cefazolin When Having an Institutional Protocol Versus No Protocol
When the current authors asked just orthopedists who made the decision about preoperative antibiotic (orthopedists, anesthesiologists, combined-decision), there was increased odds of prescribing cefazolin to patients with a history of non-anaphylactic penicillin allergy when combined decision making was used instead of orthopedic decision making (OR, 4.19; P<.01). This was also true in open fracture cases with patients with a non-anaphylactic penicillin allergy (OR, 2.52; P=.02) (Table 4). When the current authors looked at the anesthesiologist respondents, they found no significant increase or decrease in the odds of prescribing cefazolin for any group, regardless of allergy or decision maker.
Odds of Using Preoperative Cefazolin When Comparing Decision Makers
This study highlights that prescribing cefazolin is underutilized in patients with a non-anaphylactic penicillin allergy. Cefazolin is widely accepted as the ideal preoperative antibiotic for orthopedic cases given its cost-effectiveness.1,2 Previous literature strongly supports that cefazolin may be given in patients without a history of anaphylactic response to penicillin.2,7,11 This study indicates that approximately 20% of orthopedists and anesthesiologists are not using cefazolin in this patient population despite nearly all respondents prioritizing the use of cefazolin for a first-choice preoperative antibiotic. Although it is reasonable to prescribe an alternative antibiotic in the setting of a nonanaphylactic penicillin allergy, the current authors' review of the literature suggests that cefazolin is generally safe and more cost-effective. Unfortunately, only 57.5% of anesthesiologists and 41.1% of orthopedists knew that the correct cross-reactivity between cephalosporins and penicillin was 0.3% to 1%. As expected, knowledge of the correct cross-reactivity leads to higher odds of using preoperative cefazolin, including in the non-anaphylactic penicillin allergic patient. This correlation suggests that through improved education of the true cross-reactivity of cefazolin and penicillin utilization of cefazolin would increase.
Orthopedic respondents also had higher odds of using cefazolin in patients with a non-anaphylactic penicillin allergy when they used combined decision making with anesthesiologists. This may represent unease of using a cephalosporin due to reflexive decision making or unease of managing an allergic response. By engaging their anesthesia colleagues, orthopedists could relieve their apprehension due to increased dialogue and cooperation.
Unexpectedly, the current authors found that the existence of a protocol decreased the odds of an orthopedist prescribing cefazolin in non-anaphylactic penicillin allergic patients. Without being able to examine the various protocols, it is difficult to interpret this result. There are proponents of test dosing cefazolin or performing a penicillin skin test in any patient with an anaphylactic penicillin allergy.16 However, the current authors were unable to clarify whether this was occurring in their study.
The current study has many strengths. The authors were able to reach 292 physicians across a broad spectrum of the United States. This allowed them to appreciate practice patterns that have not been described previously in the literature. The authors were able to demonstrate that a surprisingly large number of orthopedists and anesthesiologists do not know the current literature of cefazolin and penicillin cross-reactivity.
However, this study had some limitations. Being a survey study, the current authors are unable to tease out nuance in answers despite allowing for free response options in the survey. In addition, many surveys were sent and only a percentage were returned, so the authors recognize a selection bias to the results. Although the current authors report associations between prescribing cefazolin and various survey responses, they are unable to directly determine the specific reason for each respondent's choice. The survey used was generated by the research team and is not validated. However, the current authors were able to show an 84% agreement when administered twice to the same group of residents after 1 month. In addition, patients are often uncertain about the type and severity of their previous allergic response to a medication. In practice, clinicians must use their clinical judgement when deciding whether a described allergic response was anaphylactic or not. The surgeon should discuss his or her concerns with the anesthesiologist and they should reach a shared decision in the best interest of the patient. Despite these limitations, this study demonstrates the need for further education. Research looking at the rate of anaphylactic response to cefazolin in patients with a penicillin allergy is needed.
Previous literature supports the use of cefazolin as a safe and cost-effective preoperative antibiotic in the setting of a non-anaphylactic allergy to penicillin. Approximately 20% of orthopedists and anesthesiologists in this study did not know the true cross-reactivity between cefazolin and penicillin. However, the current authors found that knowledge of the true cross-reactivity and use of shared decision making had the most consistent odds in predicting use of cefazolin in this patient population. This indicates that increasing physician education and encouraging shared decision making would lead to safe and more cost-effective care of patients.
- AlBuhairan B, Hind D, Hutchinson A. Antibiotic prophylaxis for wound infections in total joint arthroplasty: a systematic review. J Bone Joint Surg Br. 2008;90(7):915–919. doi:10.1302/0301-620X.90B7.20498 [CrossRef]
- Haslam S, Yen D, Dvirnik N, Engen D. Cefazolin use in patients who report a non-IgE mediated penicillin allergy: a retrospective look at adverse reactions in arthroplasty. Iowa Orthop J. 2012;32:100–103.
- Meehan J, Jamali AA, Nguyen H. Prophylactic antibiotics in hip and knee arthroplasty. J Bone Joint Surg Am. 2009;91(10):2480–2490. doi:10.2106/JBJS.H.01219 [CrossRef]
- Audicana M, Bernaola G, Urrutia I, et al. Allergic reactions to beta-lactams: studies in a group of patients allergic to penicillin and evaluation of cross-reactivity with cephalosporin. Allergy. 1994;49(2):108–113. doi:10.1111/j.1398-9995.1994.tb00809.x [CrossRef]
- Miranda A, Blanca M, Vega JM, et al. Cross-reactivity between a penicillin and a cephalosporin with the same side chain. J Allergy Clin Immunol. 1996;98(3):671–677. doi:10.1016/S0091-6749(96)70101-X [CrossRef]
- Sastre J, Quijano LD, Novalbos A, et al. Clinical cross-reactivity between amoxicillin and cephadroxil in patients allergic to amoxicillin and with good tolerance of penicillin. Allergy. 1996;51(6):383–386. doi:10.1111/j.1398-9995.1996.tb00146.x [CrossRef]
- Campagna JD, Bond MC, Schabelman E, Hayes BD. The use of cephalosporins in penicillin-allergic patients: a literature review. J Emerg Med. 2012;42(5):612–620. doi:10.1016/j.jemermed.2011.05.035 [CrossRef]
- Goodman EJ, Morgan MJ, Johnson PA, Nichols BA, Denk N, Gold BB. Cephalosporins can be given to penicillin-allergic patients who do not exhibit an anaphylactic response. J Clin Anesth.2001;13(8):561–564. doi:10.1016/S0952-8180(01)00329-4 [CrossRef]
- Kelkar PS, Li JT. Cephalosporin allergy. N Engl J Med. 2001;345(11):804–809. doi:10.1056/NEJMra993637 [CrossRef]
- Novalbos A, Sastre J, Cuesta J, et al. Lack of allergic cross-reactivity to cephalosporins among patients allergic to penicillins. Clin Exp Allergy. 2001;31(3):438–443. doi:10.1046/j.1365-2222.2001.00992.x [CrossRef]
- Saxon A, Beall GN, Rohr AS, Adelman DC. Immediate hypersensitivity reactions to beta-lactam antibiotics. Ann Intern Med. 1987;107(2):204–215. doi:10.7326/0003-4819-107-2-204 [CrossRef]
- Ramsey A, Staicu M, Liu L. Perioperative use of cefazolin in patients with reported penicillin allergy. J Allergy Clin Immunol. 2016;137(2 suppl):AB42. doi:10.1016/j.jaci.2015.12.141 [CrossRef]
- Bignardi GE. Risk factors for Clostridium difficile infection. J Hosp Infect. 1998;40(1):1–15. doi:10.1016/S0195-6701(98)90019-6 [CrossRef]
- Hospital Infection Control Practices Advisory Committee. Recommendations for preventing the spread of vancomycin resistance. Infect Control Hosp Epidemiol. 1995;16(2):105–113. doi:10.2307/30140952 [CrossRef]
- Deeks JJ, Higgins JP. Statistical algorithms in review manager 5. Statistical Methods Group of the Cochrane Collaboration. 2010:1–11.
- Park MA, Li JT. Diagnosis and management of penicillin allergy. Mayo Clin Proc. 2005;80(3):405–410. doi:10.4065/80.3.405 [CrossRef]
Odds of Using Preoperative Cefazolin When Knowing the Correct Cross-reactivitya
|Odds Ratio||P||Odds Ratio||P|
|Open fracture anaphylaxis||9.19||.01b||7.73||.03b|
|Open fracture non-anaphylaxis||1.26||.50||6.28||<.01b|
Odds of an Orthopedist Using Preoperative Cefazolin Compared With an Anesthesiologist
|Antibiotic Scenario||Odds Ratio||P|
|Open fracture anaphylaxis||0.48||.14|
|Open fracture non-anaphylaxis||1.20||.46|
Odds of Using Cefazolin When Having an Institutional Protocol Versus No Protocol
|Odds Ratio||P||Odds Ratio||P|
|Open fracture anaphylaxis||0.84||.78||0.19||.22|
|Open fracture non-anaphylaxis||1.03||.93||0.91||<.01a|
Odds of Using Preoperative Cefazolin When Comparing Decision Makersa
|Antibiotic Scenario||Combined vs Orthopedists||Combined vs Anesthesiologists||Orthopedists vs Anesthesiologists|
|Odds Ratio||P||Odds Ratio||P||Odds Ratio||P|
|Open fracture anaphylaxis||0.39||.39||0.04||<.01b||0.96||.03b|
|Open fracture non-anaphylaxis||2.52||.02b||2.00||.58||0.79||.85|