Safety of Cefazolin Test Dose in Patients With Penicillin Allergy Just Prior to Cardiac Device Implantation: A Single-Centre Experience.

Background: Cephalosporins are the cornerstone of cardiac device infection prophylaxis. Owing to fears of cross-reactivity, penicillin-allergic patients are exposed to potentially more-toxic drugs, with decreased efficacy. We evaluated the safety of a cefazolin test dose (CTD) in self-reported penicillin-allergic patients.
Methods: In this single-centre study, we evaluated consecutive patients with chart documentation of penicillin allergy undergoing cardiac device implantation, over a 2-year period. A CTD was performed if no cephalosporin allergy or severe anaphylactic reaction to penicillin had been documented. Patients were given 2 doses of 100 mg IV cefazolin, and if no allergic reaction occurred after 5 minutes, the full dose (1800 mg) was administered in the electrophysiology laboratory just before the implantation procedure.
Results: A total of 2200 patients were included. The frequency of reported penicillin allergy was 9.3% (n = 204). In 80% of cases, the type of allergic reaction was not reported in medical notes or was unknown by the patient. A CTD was performed in 67.6% of patients with a penicillin allergy (n = 138). A total of 5 adverse events occurred (3.6% of patients [95% confidence interval, 1.1%-6.1%]) - 4 skin rashes and 1 tongue edema. These 5 patients became asymptomatic after antihistaminic and corticosteroid IV treatment. Even if the test dose was negative, 79% of patients also were administered vancomycin before the procedure, as it requires a 1-hour infusion prior to the CTD in the implantation procedure room.
Conclusion: A CTD in most penicillin-allergic patients appears to be safe and allows its use per recommended guidelines.

Cardiac implantable electronic device (CIED) infection is a significant complication associated with a high level of morbidity and mortality. The pathophysiology of CIED infection is often related to contamination by local skin pathogens at the time of device implantation. Then, the infection can spread from the pocket along the electrodes, causing bacteremia and lead-related or valvular endocarditis.

The incidence of CIED infection varies from 1% to 3% in North American and European countries.2, 3, 4 This rate has been increasing because CIED indications and use of cardiac resynchronization therapy (CRT) are expanding to a population with more risk factors for infection.,

Aside from application of optimal surgical technique, preoperative intravenous (IV) antibiotic administration remains one of the most important strategies for prevention of postoperative infection. Cefazolin, a first-generation cephalosporin, is supported by guidelines, as it effectively combats most common skin pathogens associated with device infection. Recent data show that in higher-risk populations, the use of either an incremental antibiotic strategy or an antibiotic envelope also may reduce infection risk.

Penicillin Allergy

Penicillin and cephalosporins are the most widely used antibiotics. Many patients say they have an allergy to penicillin, but it is confirmed as a severe allergy in only a few patients. IgE-mediated reactions occur immediately (within 1 hour) and manifest as urticaria, angioedema, bronchospasm, and/or anaphylaxis, as opposed to non-IgE-mediated reactions, which occur days to weeks after exposure (delayed reaction). In patients with IgE-mediated penicillin hypersensitivity, the level of cross-reaction to cephalosporins varies between 1% and 10%. A review examining the level of positive response to a cephalosporin challenge estimated it to be 4.4%.

Given these findings, administration of cephalosporins to penicillin-allergic patients poses some concern. Because of the fear of cross-reactivity, the most common approach taken for penicillin-allergic patients is to select an antibiotic that does not contain the ß-lactam ring, such as vancomycin or clindamycin. However, reduced effectiveness, higher cost, and increased risk of side effects are major drawbacks of this strategy.

The aim of the current study is to report the frequency of penicillin allergy in a real-world patient population undergoing CIED implantation and to establish the safety of administering a pre-procedure incremental cefazolin test dose followed by full-dose administration in those patients without previous skin testing.

Methods

We evaluated all consecutive patients referred for a CIED procedure over a 2-year period at a single, university tertiary-care centre. Eligible procedures included new device implantation, generator replacement, and upgrade/lead-revision procedures. During this period, penicillin-allergic patients were defined as those with a self-reported pencillin allergy or those with a chart and/or pharmacy-labelled penicillin allergy. After chart screening and clinical assessment was performed, an IV cefazolin test dose (CTD) was administered to penicillin-allergic patients, unless it was refused by the attending physician. Exclusions included patients with previous anaphylaxis or severe respiratory manifestation to penicillin administration and those with confirmed allergy to cephalosporins. Even patients with a mild cephalosporin allergy were excluded, except for 7 patients with unspecified allergy. No antibacterial envelope was used in this study. The CTD consisted of the administration of 100 mg IV cefazolin, under medical supervision in the electrophysiology laboratory, with the same dose repeated 5 minutes later if no reaction occurred, and then followed an additional 5 minutes later by 1800 mg of cefazolin, completing the full dose. The protocol was approved by the institutional review committee (CER 22113).

Data collection

In penicillin-allergic patients, the collected data included the type of penicillin allergy declared, CTD administration, clinical response to the CTD, and all perioperative antibiotics received. Other data were also collected, such as anticoagulation and antiplatelet use, use of immunosuppression, presence of chronic kidney disease, temporary pacing, and duration and type of the procedure.

Statistical analysis

Continuous variables were tested for normality using the Shapiro-Wilk test and are expressed as mean ± standard deviation (SD) or median (interquartile range). Categorical variables are expressed as number and percentage and were analyzed using the Fisher exact test or the χ2 test. Quantitative data were compared using the Student t test or the Mann-Whitney U test, as appropriate. To ascertain the expected true proportion of adverse reactions related to the CTD, a Bayesian estimation was performed using a likelihood model with a Bernoulli distribution, Jeffreys prior for normal distribution, and random walk Metropolis-Hastings sampling. A P value of < 0.05 was considered statistically significant. All statistical analysis was conducted using Stata 14.1 software (StataCorp, College Station, TX).

Results

Over a 2-year period, 2200 patients underwent a CIED procedure. In 204 patients (9.3%), the medical record reported a penicillin allergy. Characteristics of the patients with a penicillin allergy are shown in Table 1. Their mean age was 72 ± 7 years, and more were female (52.5% vs 33.0%; P < 0.001). The type of allergic reaction was documented in 20.1% of cases (41 of 204). For those reported, the type of reaction was a skin rash in 53.7% (22 of 41), angioedema in 29.3% (12 of 41), anaphylaxis in 2.4% (1 of 41), and other manifestations in 14.6% (6 of 41). Having a penicillin allergy documented on the medical chart was associated with having at least one other drug allergy in 42.2% of patients (86 of 204), including 16 possible reactions to cephalosporins, 49 reactions to a different class of antibiotics, 23 reactions to iodine, 16 reactions to nonsteroidal anti-inflammatory drugs, and 24 reactions to narcotics.

Table 1

Clinical characteristics in penicillin (PNC)-allergic patients (n = 204)

Characteristic	Value for PNC-allergic patients
Age, y	72 ± 7
Male	97 (47.5)
BMI, kg/m2	27.5 (6.1)
Hypertension	146 (71.6)
Diabetes	44 (21.6)
Chronic kidney disease (MDRD GFR < 60 mL/min)	81 (39.7)
Anticoagulation	100 (49.0)
NOAC	34 (16.6)
Warfarin	65 (31.9)
Others	1 (0.5)
Antiplatelet therapy	132 (64.7)
Immunosuppressed patients	9 (4.4)
Temporary pacing	4 (2.0)
Preoperative antibioprophylaxis	204 (100)
Cefazolin alone	28 (13.7)
Vancomycin alone	66 (32.4)
Vancomycin and cefazolin	109 (53.4)
Clindamycin alone	1 (0.5)
First procedure (pacemaker or defibrillator)	140 (68.6)
Generator replacement	41 (20)
Device upgrade or lead revision	23 (11.3)
Procedure duration, min	40 ± 26

Values are mean ± standard deviation, or n (%).

GFR, glomerular filtration rate; MDRD, Modification of Diet in Renal Disease; NOAC, non–vitamin K oral anticoagulation.

The CTD was administered in 67.6% of penicillin-allergic patients (138 of 204). A CTD was not administered in 66 patients — 7.8% of patients (16 of 204) because of significant prior allergic reaction to penicillin (anaphylaxis = 1; respiratory distress = 2; significant rash = 3; angioedema = 10); 8.3% of patients (17 of 204) because of a previous cephalosporin allergic reaction; and because it was declined by the treating physician in 33 patients (16.1%), many of whom had several concomitant allergies.

Five patients (3.6% [95% confidence interval {CI}, 1.1%-6.1%]) experienced an adverse event following the CTD — 4 had skin rashes, and 1 had angioedema. The reaction occurred after the first dose of 100 mg in 2 patients, after the second dose of 100 mg in 2 other patients, and during the infusion, but prior to completion of the full dose of cefazolin, in the remaining patient. The full dose of cefazolin was not completed in these 5 patients. All 5 were successfully treated with IV corticosteroids and antihistamine drugs, with no further adverse events. No anaphylactic reaction occurred, and no prolonged hospital stay or intensive-care-unit stay was required.

In CTD-negative patients, all patients received their full dose of cefazolin, but only 21.1% of patients (28 of 133) received cefazolin alone, as antibiotic prophylaxis. No delayed allergic reactions occurred after a negative CTD. The final IV antibiotic regimen administered was cefazolin alone in 28 of 204 patients (13.7%), vancomycin alone in 66 of 204 (32.4%), both in 109 of 204 (53.4%), and clindamycin in 1 patient who was allergic to penicillin and vancomycin. Even if the test dose was negative, 79% of patients also received vancomycin before the procedure, because it requires a 1-hour infusion and had to be started prior to the CTD, which was administered in the procedure room under medical supervision.

Complications in patients who had a penicillin allergy are reported in Table 2. During the follow-up, no infection occurred in the cefazolin-alone group; and 3 infected patients required device and lead removal in the vancomycin-alone group (not statistically significant, P = 0.55).

Table 2

Adverse clinical events in penicillin-allergic patients (n = 204)

Complication	n (%)
Hematoma	7 (3.4)
Lead dislodgement	4 (2.0)
Infection	3 (1.5)

Discussion

In this study, we found the following: (i) the frequency of presumptive penicillin allergy in patients referred for a CIED procedure was 9.3%; (ii) the administration of a CTD in patients with a non-IgE-mediated penicillin allergy is safe; and (iii) a CTD can help confirm the absence of cross-reactivity to cefazolin and allow its use in the future.

The observed frequency of reported penicillin allergy in our population was 9.3%, which is consistent with the level reported previously. A CTD was performed in 67.6% of penicillin-allergic patients. A CTD was not performed in several patients, including 16.1% of patients who had either a prior significant allergic reaction to penicillin or a documented cephalosporin reaction. Given that the CTD was given at the discretion of the treating physician, some patients did not receive it. Our study confirmed that the CTD is safe, with only 4 skin rashes and 1 angioedema reported, and no cases of anaphylaxis.

The rate of confirmed penicillin-allergic patients is certainly overestimated, as pointed out by Pichichero. Adding to the confusion related to defining “an allergic reaction,” criteria are often vague, especially when the previous reaction was not severe, and allergy often is not differentiated from “intolerance.” Many patients labeled as having a penicillin allergy had non-immunologic side effects, such as diarrhea, vomiting, or a nonspecific rash. Cook et al. reported that the true rate of contraindication to penicillin was 0.9%, as determined by skin tests in the cardiac surgery population.

Conducting a thoughtful and detailed history is a key element in the context of screening for ß-lactam allergies, and it often determines the likelihood of identifying a true underlying allergy. In the absence of a detailed history and a critical evaluation of the allergic reaction, many patients will be mislabeled as allergic to ß-lactam antibiotics. Despite the importance of the allergy history, incomplete documentation of allergic reactions is widespread, occurring in up to 66% of patients. In our work, we found incomplete information about penicillin allergy in an even higher percentage of patients (80%). When assessing a patient with a drug allergy, documentation of the date and timing of the allergic reaction, the type of hypersensitivity, the treatment given, and the tolerance to other similar drugs is important. In most clinical scenarios, a patient who had a severe reaction will remember this if asked.

Physicians are often not well prepared to care for penicillin-allergic patients immediately before a surgical intervention. Indeed, Blumenthal et al. reported that physicians, residents, and nurse practitioners had limited prior drug-allergy education. Moreover, the implementation of an inpatient antibiotic standard regimen for patients with a penicillin or cephalosporin allergy was associated with an almost 7-fold increase in the number of test doses to ß-lactams, without an increase in adverse drug reactions. Given the fear of cross-reactivity, the existence of a protocol in our centre clearly encouraged physicians to perform the CTD in most penicillin-allergic patients. Our results were then used to standardize the CTD in all patients with no known angioedema to penicillin.

The 2007 Society of Thoracic Surgeons practice guidelines recommend, for non-IgE-mediated reactions to penicillin (such as a simple rash), the use of cephalosporin administration (class I, level of evidence B), without any test dose. In IgE-mediated reactions, vancomycin administration or skin-test challenges are recommended prior to readministration of cephalosporins (class I, level A). Current CIED guidelines lack clear recommendations on this issue. However, performing a CTD can reassure physicians in treating non-IgE-mediated penicillin-allergic patients and decrease use of vancomycin as a single-antibiotic prophylaxis in this population.

The additional benefit from performing a CTD is the confirmation that no cephalosporin allergy is present and that its administration for other medical conditions is safe. Li et al. reported that prescription of non-ß-lactam antibiotics in patients with an equivocal history of penicillin allergy resulted in 1.82- to 2.58-fold higher costs, compared to administration of standard first-line antibiotics.

The alternative for CIED infection prophylaxis in cefazolin-allergic patients is vancomycin. In our study, we found 3 infections in the group of 66 patients for whom vancomycin alone was administered, and 0 infections in the 28 patients in the cefazolin-alone group, a difference that is not statistically significant, although a sample-size effect cannot be excluded. Bolon et al., in a meta-analysis of 7 randomized trials, reported no difference between glycopeptides and ß-lactam for the reduction of surgical-site infection at 30 days. However, in a subanalysis, ß-lactams were found to be superior to glycopeptides for prevention of chest infection, and they approached superiority for prevention of deep chest infections caused by gram-positive bacteria. Glycopeptides were superior for prevention of surgical-site infections caused by methicillin-resistant gram-positive bacteria.

Skin testing is usually recommended to assess penicillin allergy. Some authors have reported the value of using preoperative skin testing to reduce vancomycin use in cardiac or orthopedic surgeries. However, Phillips et al. reported that the use of vancomycin in patients with a suspected history of IgE-dependent allergy in cardiovascular surgery prophylaxis was cost effective, and safer than performing a penicillin skin test or giving cefazolin to all penicillin-allergic patients.

Beltran et al. reported in a pediatric surgery prophylaxis study that the administration of cephalosporins in cases of non-IgE-mediated hypersensitivity was safe. Among 123 penicillin-allergic patients who received cefazolin, only 1 case of non-anaphylactic reaction occurred. A recent meta-analysis found that cross-reactivity between penicillin and cefazolin was rare, including in the subgroup of surgical patients.

The guidelines published in 2015 for the diagnosis of drug-hypersensitivity reaction reported that a drug-provocative test is indicated to exclude hypersensitivity when the history is unclear or to exclude cross-reactivity with related drugs. However, clinical practice often does not allow time to perform the test electively before CIED implantation. Using a CTD is cost effective and clinically applicable, and such a strategy is needed in view of the clinical importance of CIED infection prevention.

Limitations

Our study has some limitations. The administration of the CTD was not randomized, but rather was at the discretion of the treating physician. Second, the identification of penicillin-allergic patients was based on either self-report or documentation in the medical records. Patients were not screened with a skin test to confirm their penicillin allergy, a common problem reported in other trials as well. The duration of CTD observation was limited so that it could be performed within a reasonable time prior to the intervention, and no delayed reaction was observed. The overall rate of CIED infection at 30 days was too small to compare infection rates among different antibiotic strategies.

Conclusions

Reported penicillin-allergic patients represent 9.3% of a contemporary population undergoing CIED procedures. A CTD in non-IgE-mediated penicillin-allergic patients appears to be safe and useful in decreasing the use of vancomycin alone in this population, but it should be balanced against the risk and safety of administering other antibiotics. A standard operating procedure for CTD administration can be applied systematically; in this way, a CTD is safe and can prevent suboptimal antibiotic prophylaxis before CIED procedures in patients without confirmed cephalosporin allergy, previous anaphylaxis, or severe respiratory distress to penicillin.