Prevalence of penicillin allergy in adults with peritonsillar abscess.

Abstract

We noticed a seemingly high prevalence of penicillin allergy in patients who had been diagnosed with peritonsillar abscess (PTA) at our institution. To formally investigate this observation, we reviewed the emergency room (ER) records of 118 patients who had presented between Jan. 1, 1995, and Dec. 31, 1999, with suspected PTA. A diagnosis of PTA was confirmed by the presence of pus on incision and drainage in 78 of these patients (66.1%). The remaining 40 patients (33.9%) were diagnosed with peritonsillar cellulitis (PTC). Of the 78patients with confirmed PTA, 13 (16.7%) self-reported an allergy to an antibiotic, including 11 (14.1%) who claimed to be allergic to penicillin. In the 40 patients with PTC, the corresponding figures were only 3 (7.5%) and 1 (2.5%). The difference between the PTA and PTC groups with respect to the prevalence of self-reported penicillin allergy was statistically significant (p < 0.05). We also compared the prevalence of antibiotic allergies in our patients with that of 1,893 consecutively presenting patients whose records had been entered into a pharmacy database at our institution. We found that the overall prevalence of patient-reported penicillin allergy in our PTA group was similar to that of the database population, although penicillin allergy did account for a greater percentage of all antibiotic allergies (84.6%) in our PTA group than in the larger population (62.8%). In our series, patients with PTA were more likely to have reported an allergy to penicillin than were patients without an abscess. Additionally, the prevalence of patient-reported antibiotic allergy is high at our institution. Although self-reported penicillin allergy may not represent a true hypersensitivity reaction, it can influence antibiotic selection and/or compliance. Prospective studies are needed to determine what influence allergic status and antibiotic choice has on abscess development.

Introduction

When peritonsillar abscess (PTA) is suspected in the emergency room (ER) at our institution, we routinely perform immediate incision and drainage under local anesthesia. The return of frank pus confirms the diagnosis of PTA. If no pus is encountered, a diagnosis of peritonsillar cellulitis (PTC) is rendered. In rare and unusual circumstances, such as when a patient presents with severe trismus, the diagnosis may be established by computed tomography.

We noticed a seemingly high prevalence of penicillin allergy among patients who were diagnosed with PTA in our ER. To formally investigate this observation, we undertook a retrospective, cross-sectional study.

Patients and methods

Between Jan. 1, 1995, and Dec. 31, 1999, the ER at Northwestern Memorial Hospital admitted 138 patients who had presented with suspected PTA. The ER records of 118 of these patients were available for review. We examined these charts to determine (1) demographic data, (2) the definitive diagnosis (PTA or PTC), (3) the prevalence of self-reported antibiotic allergy, and (4) the type and duration of antibiotic therapy that had been prescribed (usually by a primary care practitioner) prior to ER presentation.

We divided these patients into two groups according to the definitive diagnosis (those who had been initially diagnosed with PTC but who had returned within 1 week with PTA were included in the PTA group). We also performed calculations for two subgroups within the PTA group--patients with and without self-reported penicillin allergy; we determined the number of patients who had been on antibiotics prior to ER presentation and the duration of that therapy. We also used as a basis for comparison information on 1,893 consecutively presenting patients who had been hospitalized at our institution in August and September of 1998 and whose records had been entered into our pharmacy database. (1)

Statistical comparisons between groups were performed with Microsoft Excel software. The Student's t test, the chi-square test, and Fisher's exact test were used to determine statistical significance (p < 0.05).

Results

Of the 118 patients, 78 (66.1%) had been definitively diagnosed with PTA and 40 (33.9%) with PTC. The PTA group was made up of 45 men and 33 women (mean age: 32.8 yr), and the PTC group was made up of 21 men and 19 women (mean age: 30.9 yr). The differences in sex and age distribution were not statistically significant.

Of the 118 patients, 16 (13.6%) reported an allergy to antibiotics--13 in the PTA group and 3 in the PTC group (table 1). Twelve patients (10.2%) reported an allergy to penicillin--11 in the PTA group and 1 in the PTC group, a statistically significant difference. Three patients (2.5%) reported an allergy to sulfa (2 PTA and 1 PTC), and 2 (1.7%) said they were allergic to macrolides (1 PTA and 1 PTC). One patient in the PTA group reported an allergy to both penicillin and erythromycin.

Four of the 16 allergic patients (25%) said they had developed a rash after taking an antibiotic (3 penicillin and 1 sulfa), but no specific immunologic reaction had been documented in the remaining 12 (75%).

In all, 47 patients had been prescribed an antibiotic prior to ER presentations--40 of the 78 (51.3%) in the PTA group and 7 of the 40 (17.5%) in the PTC group; the difference was statistically significantly (table 2). Likewise, the mean duration of therapy was significantly higher in the PTA group--3.9 vs. 1.0 days (table 3). Among PTA patients who did and did not report an allergy to penicillin, there was no significant difference in either parameter.

Most penicillin-allergic patients who had been prescribed an antibiotic prior to ER presentation received a macrolide (table 4). Most of those who were not penicillin-allergic received a penicillin.

Data compiled from the records of the 1,893 patients in the pharmacy database revealed that 295 patients (15.6%) claimed an allergy to penicillin, 138 patients (7.3%) were allergic to sulfa, and 44 (2.3%) were allergic to macrolides; in all, 470 patients (24.8%) were allergic to any antibiotic.

Discussion

Penicillins are the traditional antibiotics of choice for acute pharyngitis and for deep neck space infections, and indeed most of the nonallergic patients in our study (63.8%) received a penicillin. Hypersensitivity reactions to penicillins occur in approximately 5% of all patients, although the incidence may be as high as 10% with some agents in this drug class. (2) In a frequently cited report, Shepherd estimated that the prevalence of beta-lactam allergy is 2% per course of treatment. (3) The most common reaction is a maculopapular rash, but life-threatening or life-endangering anaphylaxis occurs in as many as 10% of penicillin-allergic patients. (4)

In our study, patients with PTA were more likely to have reported an allergy to penicillin than were patients without an abscess (14.1 vs. 2.5%). Additionally, the prevalence of patient-reported antibiotic allergy is high at our institution (nearly 25%).

Although the prevalence of penicillin allergy in our PTA group (14.1%) was in line with that of the large group of patients in our pharmacy database (15.6%), penicillin allergy accounted for a greater proportion of all antibiotic allergies reported by our PTA patients (84.6 vs. 62.8%). On the other hand, our PTA group had a lower prevalence of sulfa allergy (2.6 vs. 7.3%) and allergy to any antibiotic (16.7 vs. 24.8%). We can speculate then that patients who are allergic to antibiotics other than penicillin may not develop PTAs as often as penicillin-allergic patients do. Studies of larger populations of sulfa- and macrolide-allergic patients are needed to address this possibility. Of note, the frequency of sulfa and macrolide allergies observed in both the PTA and PTC groups in our study are consistent with accepted values reported in the literature. (5)

Compared with the PTA group, patients who were diagnosed with PTC were significantly less likely to have been taking an antibiotic (17.5 vs. 51.3%), and those who did take an antibiotic did so for a significantly shorter time (1.0 vs. 3.9 days). Within the two PTA subgroups (those with and without penicillin allergy), there were no significant differences in the percentage of patients who had taken a pre-ER antibiotic (63.6 vs. 49.3%) and the duration of antibiotic therapy (5.1 vs. 3.9 days). Nonetheless, it is clear that patient self-reports of penicillin allergy affect antibiotic choice because a significantly greater percentage of penicillin-allergic patients received a macrolide (57.1 vs. 17.5%). From this retrospective series, however, it remains unclear whether patient-reported allergy and its effect on antibiotic selection influence the development of abscess.

Several studies have documented the inaccuracy of patient history with respect to self-reports of antibiotic allergies. (6,7) Hung et al reported a series of 1,818 surgical patients and found that 28% reported drug allergies, half of which (14%) were associated with reactions to antibiotics. (6) In that study, however, only 75% of those antibiotic reactions had a high probability of representing a true allergy. In many of those cases, the reaction was either unknown or nonallergic (i.e., the reaction was a known side effect of the drug). Likewise in our study, 75% of the cases of self-reported antibiotic allergy had not been confirmed by a documented appearance of a specific allergic reaction. Also, the prevalence of antibiotic allergies in the report by Hung et al was similar to ours (14 and 13.6%, respectively). Therefore, we may presume that the prevalence of actual antibiotic allergy in our study was inflated. In fact, several reports have indicated that patients may interpret drug side effects (e.g., nausea, vomiting, or diarrhea) as allergies. (1,5,7) Thus, the true incidence of penicillin allergy is difficult to quantify solely on the basis of a patient self-reports in the absence of any immunologic reaction. Discrepancies in the prevalence of self-reported allergy and true allergy highlight the need to elicit the exact history of a specific self-reported antibiotic reaction.

References

(1.) Lee CE, Zembower TR, Fotis MA, et al. The incidence of antimicrobial allergies in hospitalized patients: Implications regarding prescribing patterns and emerging bacterial resistance. Arch Intern Med 2000; 160:2819-22.

(2.) Fairbanks DN, ed. Antimicrobial Therapy in Otolaryngology--Head and Neck Surgery. 9th ed. Alexandria, Va.: The American Academy of Otolaryngology--Head and Neck Surgery Foundation, 1999:1.

(3.) Shepherd GM. Allergy to B-lactam antibiotics. Immunol Allergy Clin North Am 1991;11:611-33.

(4.) Task Force on Asthma and Other Allergic Diseases. NIAID Task Force Report. Bethesda, Md.: Department of Health, Education, and Welfare, 1979. NIH publication 79-387/G.

(5.) Patterson R, Grammer LC, Greenberger PA, eds. Allergic Diseases. 5th ed. Philadelphia: Lippincott-Raven, 1997:373.

(6.) Hung OR, Bands C, Laney G, et al. Drug allergies in the surgical population. Can J Anaesth 1994;41:1149-55.

(7.) Graff-Lonnevig V, Hedlin G, Lindfors A. Penicillin allergy--A rare paediatric condition? Arch Dis Child 1988;63:1342-6.

From the Department of Otolaryngology--Head and Neck Surgery, University of Tennessee, Memphis (Dr. Chandra); and the Department of Pharmacy (Ms. Lee) and the Department of Otolaryngology--Head and Neck Surgery (Dr. Pelzer), Northwestern University, Chicago.

Reprint requests: Rakesh K. Chandra, MD, Department of Otolaryngology--Head and Neck Surgery, University of Tennessee Health Science Center, 956 Court Ave., Suite B224, Memphis, TN 38163. Phone: [phone]; fax: [phone]; e-mail: [e-mail]

Table 1. Antibiotic allergy profile in the PTA and PTC
patients

                        PTA (n = 78)               PTC (n = 40)
                            n (%)                     n (%)

No antibiotic allergy     65 (83.3)                 37 (92.5)

Antibiotic allergy        13 (16.7)                  3  (7.5)
  Penicillin allergy      11 (14.1) * ([dagger])     1  (2.5)
  Sulfa allergy            2  (2.6)                  1  (2.5)
  Macrolide allergy        1  (1.3) ([dagger])       1  (2.5)

* p < 0.05 (Fisher's exact test) for PTA vs. PTC.

([dagger]) One patient reported an allergy to both penicillin and
erythromycin.

Table 2. Antibiotic use profile in the PTA and PTC
patients

                                PTA             PTC
                               n (%)           n (%)

Used a pre-ER antibiotic   40/78 (51.3) *   7/40 (17.5)
  No penicillin allergy    33/67 (49.3) *   7/39 (17.9)
  Penicillin allergy        7/11 (63.6)     0/1

* p < 0.05 (chi-square test) for PTA vs. PTC.

Table 3. Mean duration of antibiotic use (days)

                                     PTA     PTC

All patients on pre-ER antibiotics   3.9 *   1.0
  No penicillin allergy              3.9 *   1.0
  Penicillin allergy                 5.1     N/A

* p < 0.05 (Student's t test) for PTA vs. PTC.