Morbidity in Pregnant Women Associated with Unverified Penicillin Allergies, Antibiotic Use, and Group B Streptococcus Infections

Morbidity in Pregnant Women Associated with  Unverified Penicillin Allergies, Antibiotic Use,  and Group B Streptococcus Infections

 

Shilpa H Desai, MD; Michael S Kaplan, MD; Qiaoling Chen, MS; Eric M Macy, MD, MS

Perm J 2017;21:16-080

https://doi.org/10.7812/TPP/16-080
E-pub: 01/06/2017

Morbidity in Pregnant Women Associated with  Unverified Penicillin Allergies, Antibiotic Use,  and Group B Streptococcus Infections

ABSTRACT

Context: The morbidity potentially associated with unverified penicillin allergy in pregnant women, with and without group B streptococcus (GBS) infections, is unknown. Penicillin allergy testing is safe during pregnancy but is done infrequently.
Objective: To determine morbidity associated with antibiotic use in a large cohort of pregnant women, with and without an unverified history of penicillin allergy, and with and without GBS.
Design: Retrospective. All pregnant women who delivered live infants in Kaiser Permanente Southern California between January 1, 2009, and December 31, 2014, were identified.
Main Outcome Measures: Penicillin allergy status at delivery, delivery method, maternal and infant hospital utilization, peripartum antibiotic exposures, new antibiotic-associated adverse drug reactions, and new Clostridium difficile infections.
Results: There were 170,379 unique women who had 201,316 pregnancies during the study period. There were 16,084 pregnancies in women with an active, but unverified, penicillin allergy at delivery. There were 42,524 pregnancies in GBS-positive women, and 3500 also had a penicillin allergy. Women with a penicillin allergy, with or without GBS, had significantly (about 10%) higher cesarean section rates and spent significantly more (about 0.1) days in the hospital after delivery. Among GBS-positive women, those with an unverified penicillin allergy were exposed to significantly more cefazolin, clindamycin, vancomycin, and gentamicin and had significantly higher rates of adverse drug reactions associated with all antibiotic use.
Conclusions: Unverified penicillin allergy is associated with more hospital utilization and additional morbidity. Penicillin allergy testing of pregnant women with a history of penicillin allergy may help reduce these unwanted outcomes.

INTRODUCTION

An unconfirmed history of penicillin allergy is a well-known, major public health problem.1 The Choosing Wisely initiative of the American Board of Internal Medicine Foundation has recommended that physicians minimize the use of non-b-lactam antibiotics in patients with a history of penicillin allergy, without an appropriate evaluation.2 Other than in hospitalized patients in general, there are minimal data on predicting which specific patients with a history of penicillin allergy would most benefit from having their allergy confirmed or removed from the chart if incorrectly placed there.3 Penicillin allergy testing has been shown to be safely performed in pregnant women with Group B streptococcus (GBS) infections but is rarely performed.4,5 A better understanding of the potential morbidity associated with an unverified history of penicillin allergy may support more widespread adoption of penicillin allergy testing in this population.

Our primary objective was to determine morbidity associated with antibiotic use in a large cohort of pregnant women, with and without an unverified history of penicillin allergy, and with and without GBS.

This article restricts the word “allergy” to mean a clinically significant immunoglobulin E-mediated reaction. The term “allergy” will refer to an adverse reaction or intolerance associated with the previous use of a specific medication or medication class as noted in the patient’s medical record.

METHODS

This study was reviewed and approved by the Kaiser Permanente Southern California (KPSC) institutional review board. The Southern California Kaiser Foundation Health Plan has always maintained a single comprehensive medical record for each member, and the membership pool has been shown to reflect the socioeconomic diversity of the Southern California general population.6 Since 2007, the medical record has been completely electronic.

All pregnant women who delivered between January 1, 2009, and December 31, 2014, were identified from the medical records. Those who had a positive culture for GBS before delivery were identified. Those with a history of an active penicillin allergy on the day of delivery were identified.

Maternal and infant hospital utilization within 6 months of delivery, method of delivery, antibiotic exposures within 3 days before or after delivery (a 7-day period), new antibiotic allergy reports within 30 days of peripartum antibiotic exposure, and new Clostridium difficile infections within 90 days of delivery were the primary outcome variables collected.

A course of an antibiotic was defined as any systemic exposure, oral or parenteral, to a specific antibacterial antibiotic. “Oral” exposures included delivery by mouth as well as nasogastric tube or other direct route to the stomach. Parenteral exposures included intravenous, intramuscular, subcutaneous, intradermal, intraosseous, intraperitoneal, and intravesicular injections or infusions. Topical exposures were not considered for this analysis. Any exposure to the same antibiotic, by the same route, within 36 hours was considered part of the same course. Reports of new antibiotic allergy entered into the electronic health record within 30 days of the antibiotic exposure were identified.

All cases of anaphylaxis coded (International Classification of Diseases, Ninth Revision [ICD-9] 995.0) within 3 days before or after delivery were identified. Chart reviews were then performed to verify that the clinical symptoms that occurred met the current working definition of anaphylaxis and were plausibly associated with use of a specific antibiotic.7 All cases of serious cutaneous adverse drug reactions coded (ICD-9 Codes 695.14, 695.13, and 695.15) within 30 days of receiving a peripartum antibiotic were identified. Chart review was then performed to verify that the clinical symptoms were compatible with the coded diagnosis and to identify the specific agent most likely associated with the adverse reaction noted.8

New C difficile infections within 90 days of delivery were identified by either a new report of C difficile coded (ICD-9 Code 008.45) or a positive microbiology culture or polymerase chain reaction for C difficile within 90 days of delivery.

Hypothesis testing for continuous variables was by means of Student t-test and for categorical variables by c2. Nominal statistical significance was set at p = 0.05. All statistical analyses were performed using SAS statistical software (SAS Institute Inc, Cary, NC).

RESULTS

Study Cohort

There were 170,379 unique women who had 201,316 pregnancies during the study interval (January 1, 2009, to December 31, 2014) and delivered 204,885 live infants. Among the 14,233 women with an active, but unverified, penicillin allergy at delivery, there were 16,084 pregnancies (8.0%).

Group B Streptococcus Status and Penicillin Allergy

There were 39,398 women (23.1%) who had 42,524 pregnancies and whose cultures were positive for GBS. Of these GBS-positive women, 3247 (8.8%; resulting in 3500 pregnancies) also had a penicillin allergy. There were 10,988 women (9.0%; 12,584 pregnancies) without GBS with an active penicillin allergy at delivery.

Table 1 shows the demographics of all pregnant women, with and without GBS, who delivered live infants. Demographics of all pregnant women who delivered, with and without penicillin allergy, are displayed in Table 2. The demographics of pregnant women who delivered, with GBS, and with and without penicillin allergy, are displayed in Table 3.

Peripartum Antibiotic Exposure

The specific antibiotic exposures within 3 days of delivery and new antibiotic allergies noted within 30 days of the specific antibiotic exposures in pregnant women with GBS, and with and without penicillin allergy, are displayed in Table 4. Cephalosporins were the second most commonly used antibiotic class in women with a history of penicillin allergy, after clindamycin, and virtually no penicillin allergy testing or graded cephalosporin challenges were performed before the cephalosporin exposures. The 6th through 10th most commonly used other antibiotics in women with an unverified history of penicillin allergy were metronidazole, nitrofurantoin, sulfamethoxazole, erythromycin, and doxycycline. In women without penicillin allergy, the 6th through 10th most commonly used other antibiotics were metronidazole, nitrofurantoin, erythromycin, sulfamethoxazole, and doxycycline.

Comorbidities

Of the 39,024 GBS-positive women without a history of penicillin allergy, 10,757 (27.57%) underwent a cesarean delivery and spent an average of 3.46 days (standard deviation = 2.00 days) in the hospital within 6 months of their delivery. Within 3 days of their delivery, GBS-positive women without a history of penicillin allergy received the following antibiotics, listed from most common to least common: penicillin, cefazolin, gentamicin, clindamycin, azithromycin, and vancomycin. The GBS-positive women without a history of penicillin allergy reported 15 (0.04%) new penicillin allergies, 19 (0.05%) new cephalosporin allergies, 3 (0.01%) new clindamycin allergies, 3 (0.01%) new vancomycin allergies, 3 (0.01%) new azithromycin allergies, 0 (0%) new gentamicin allergies, and 6 (0.02%) new other antibiotic allergies within 30 days of a peripartum antibiotic exposure. The GBS-positive women without a history of penicillin allergy had 20 (0.05%) new episodes of C difficile within 90 days of delivery.

The GBS-positive women with an active, but unverified, penicillin allergy (N = 3500) had a significantly higher (p = 0.0004) rate of cesarean deliveries, 30.37% (n = 1063). They also spent significantly more (p = 0.002) total days—a mean of 3.55 days (standard deviation = 1.51 days) longer—in the hospital within 6 months of their delivery, independent of the method of delivery. Within 3 days of their delivery, GBS-positive women with penicillin allergy received the following antibiotics, listed from most common to least common: clindamycin, cefazolin, azithromycin, vancomycin, and gentamicin. Those GBS-positive women with a history of penicillin allergy within 30 days of a peripartum antibiotic exposure reported higher statistically significant rates of adverse drug reactions with all antibiotics used, including 12 (0.34%) new clindamycin allergies, 10 (0.29%) new cephalosporin allergies, 22 (0.63%) new vancomycin allergies, 1 (0.03%) new azithromycin allergies, 4 (0.11%) new gentamicin allergies, and 3 (0.09%) new other antibiotic allergies. The GBS-positive women with penicillin allergy had no new episodes of C difficile within 90 days of delivery, which was not significantly different from those without penicillin allergy.

Penicillin Exposure in Women with Unverified Penicillin Allergy

During the study interval, 130 women with an active, but unverified, penicillin allergy received 262 courses of penicillin-class antibiotics at delivery without any allergy testing or graded challenge before the exposure. The specific penicillin-class antibiotics administered were as follows: amoxicillin, 2 (0.76%); ampicillin, 91 (34.74%); and penicillin, 169 (64.50%). There were 2 (0.76%) oral and 260 (99.24%) parenteral exposures to penicillin-class antibiotics.

The most commonly listed reasons for overriding the electronic health record warning not to use a penicillin-class antibiotic in a pregnant woman with a penicillin allergy history were “benefits outweigh risks” in 184 cases (47.92%), “not a true allergy” in 86 (22.4%), “previously or currently tolerated” in 47 (12.24%), “low risk” in 30 (7.81%), “not clinically significant” in 29 (7.55%), “expected side effect” in 6 (1.56%), and “unverified” in 2 (0.52%). The penicillin-class antibiotics were prescribed by the obstetrician-gynecologist in 104 cases (39.69%) or other treating physicians (but not anesthesiologists) in 158 cases (60.31%).

Chart review of the discharge summaries, first postpartum obstetrics and gynecology notes, and any follow-up medical visits 30 days after receiving the antibiotic dose revealed no adverse reactions noted with these penicillin exposures. Only 7 (5.4%) of 130 patients who tolerated a peripartum penicillin exposure had the notation of penicillin allergy removed from their electronic health record during the follow-up period through June 30, 2015.

A total of 260,755 courses of antibiotics were administered to pregnant women during the study interval. There were 77,283 courses of penicillin-class antibiotics (35,895 women; 37,439 pregnancies) given to pregnant women without a history of penicillin allergy in the 3 days before delivery, and 55 (0.15%) resulted in a new penicillin allergy notation before delivery. Only 2 cases (0.0008%; 95% confidence interval = 0 to 1 in 54,644) of possible antibiotic-associated anaphylaxis occurred within 3 days before or after delivery for the entire 201,316 pregnancies. Neither case occurred in GBS-positive women. Both cases were associated with the use of cefazolin during anesthesia induction.

There were 83,953 total exposures to cefazolin, resulting in a possible anaphylaxis rate of 0.0024% (95% confidence interval = 0 to 1 in 17,593). Neither patient had an underlying history of penicillin allergy. One episode was associated with swelling, the sensation of shortness of breath, and a slight elevation in blood pressure, and was not treated with adrenaline. The second episode occurred during a surgical procedure for removal of a retained placenta the day after delivery and was associated with facial swelling and no rash or hypotension, but was treated with adrenaline.

There was only one possible, but unlikely, case of a serious cutaneous adverse drug reaction associated with peripartum antibiotic use, and it was coded as Stevens-Johnson syndrome by one consultant. This diagnosis was given after consultation in the Allergy Department about two weeks after onset of what may have been a contact dermatitis to benzoin on her abdomen, a delayed-onset reaction to cefazolin or cephalexin, or a combination of these events. Five days earlier the patient had been seen in the Obstetrics Department for routine postpartum follow-up and was given a diagnosis of erythema multiforme minor and contact dermatitis. She had been given cefazolin at the time of her cesarean delivery and then given cephalexin four days later, at hospital discharge, for a possible abdominal cellulitis. Contact dermatitis around the surgical site was noted in the hospital discharge summary. No treatment, other than antihistamines, was given for the rashes. No percentage of body surface area affected was noted in the record, and no skin biopsy specimen was obtained. She was seen in the Obstetrics Department again two days after the allergy consultation for management of her slow-to-resolve surgical site cellulitis and was treated with topical antibiotics.

Perinatal Comorbidity

The 39,571 children born to GBS-positive women without a history of penicillin allergy spent a total of 139,738 days in the hospital within 6 months of delivery. The 3545 children born to GBS-positive women with a history of penicillin allergy spent a total of 13,109 days in the hospital within 6 months of delivery. The difference of +0.2 hospital days per infant born to a mother with penicillin allergy did not reach statistical significance (p = 0.1028).

Only 7 pregnant women with GBS had penicillin allergy testing before delivery in the KPSC Health Plan between January 1, 2009, and December 31, 2014, and 3 (42.86%) of the women were in the San Diego Service Area. Results for all 7 women were negative.

Morbidity in Pregnant Women Associated with  Unverified Penicillin Allergies, Antibiotic Use,  and Group B Streptococcus Infections

Morbidity in Pregnant Women Associated with  Unverified Penicillin Allergies, Antibiotic Use,  and Group B Streptococcus Infections

Morbidity in Pregnant Women Associated with  Unverified Penicillin Allergies, Antibiotic Use,  and Group B Streptococcus Infections

Morbidity in Pregnant Women Associated with  Unverified Penicillin Allergies, Antibiotic Use,  and Group B Streptococcus Infections

DISCUSSION

True penicillin allergy occurs much less frequently than the rate at which this allergy is placed in the patient chart.3,9,10 Patients who are labeled as penicillin-allergic receive alternate, often broad-spectrum, antibiotics, which may lead to substantial morbidity and increased health care utilization. It is notable in the current study that none of the 130 women listed with active unverified penicillin allergy who received penicillin had an adverse reaction. We found that an unverified history of penicillin allergy was associated with a higher cesarean delivery rate in all pregnant women, even those without GBS. The antibiotic utilization seen in our cohort closely matches that reported in a recent survey of American obstetricians,11 and 13.7% of our penicillin-allergic group received vancomycin, which has the potential for several adverse effects. This type of poor adherence to national guidelines was found in another recent study as well; more than 50% of women with a penicillin allergy without anaphylaxis received an antibiotic other than penicillin or cefazolin.12,13 We found that significantly more new adverse drug reactions occur with all alternative antibiotics used in GBS-positive women with an unverified penicillin allergy (see Table 3). Fortunately, there was no significant difference in the C difficile incidence in any of the subgroups. This is probably related to the very low overall use of third-generation and higher-generation cephalosporins in this population. Currently, cephalosporins are widely, safely, and appropriately used in pregnant women with an active, but unverified, penicillin allergy without any specific testing or challenge procedures.14,15

A recent study showed no increased risk of penicillin allergy in fetuses exposed during the intrapartum period.16 However, penicillin allergy testing is rarely performed in pregnant women, even in a health care system with no specific financial barrier to its performance. The cost of this test is approximately $145 per test. There have been several previous suggestions commenting on the economic utility of performing wide-scale allergy testing in patients with unverified penicillin allergy.3,17,18

Given the increase in morbidities and health care costs associated with unverified penicillin allergy, it seems quite reasonable to propose that any hospitalized patient with a clinical history of penicillin allergy should undergo allergy testing. In our case, penicillin allergy testing of all pregnant women with an unverified penicillin allergy and GBS would cost less than the 0.09 additional hospital days that these women, on average, currently utilize, which translates to a cost of at least $200 per occurrence. Although the cesarean delivery rate is increased in pregnant women with an active history of penicillin allergy, the increase in hospital days is independent of cesarean delivery compared with vaginal delivery (see Tables 2 and 3). This information, along with the known fact that penicillin allergy testing performed in pregnancy is safe and effective,4,5 argues for a systemwide adaptation of penicillin testing for unverified allergy in this population. It is quite clear that any woman who would be eligible to receive a safer, equally effective medication in the penicillin class should be receiving this drug.

CONCLUSIONS

An unverified penicillin allergy is associated with greater hospital utilization and additional morbidity. Penicillin allergy testing of pregnant women with a history of penicillin allergy may help reduce these unwanted outcomes. Any patient who does not have a true drug allergy should have the allergy listing removed from her chart.

Disclosure Statement

The author(s) have no conflicts of interest to disclose.

Acknowledgements

Kathleen Louden, ELS, of Louden Health Communications provided editorial assistance.

How to Cite this Article

Desai SH, Kaplan MS, Chen Q, Macy EM. Morbidity in pregnant women associated with unverified penicillin allergies, antibiotic use, and group B streptococcus infections. Perm J 2017;21:16-080. DOI: https://doi.org/10.7812/TPP/16-080.

References
1.    Solensky R. Penicillin allergy as a public health measure. J Allergy Clin Immunol 2014 Mar;133(3):797-8. DOI: https://doi.org/10.1016/j.jaci.2013.10.032.
    2.    American Academy of Allergy, Asthma & Immunology. Ten things physicians and patients should question [Internet]. Philadelphia, PA: Choosing Wisely, ABIM Foundation; 2012 Apr 4 [updated 2014 Mar 3; cited 2016 Sep 20]. Available from: www.choosingwisely.org/wp-content/uploads/2015/02/AAAAI-Choosing-Wisely-List.pdf.
    3.    Macy E, Contreras R. Health care use and serious infection prevalence associated with penicillin “allergy” in hospitalized patients: A cohort study. J Allergy Clin Immunol 2014 Mar;133(3):790-6. DOI: https://doi.org/10.1016/j.jaci.2013.12.559.
    4.    Macy E. Penicillin skin testing in pregnant women with a history of penicillin allergy and group B streptococcus colonization. Ann Allergy Asthma Immunol 2006 Aug;97(2):164-8. DOI: https://doi.org/10.1016/s1081-1206(10)60007-5.
    5.    Philipson EH, Lang DM, Gordon SJ, Burlingame JM, Emery SP, Arroliga ME. Management of group B streptococcus in pregnant women with penicillin allergy. J Reprod Med 2007 Jun;52(6):480-4. DOI: https://doi.org/10.1016/j.ajog.2004.10.081.
    6.    Koebnick C, Langer-Gould AM, Gould MK, et al. Sociodemographic characteristics of members of a large, integrated health care system: Comparison with US Census Bureau data. Perm J 2012 Summer;16(3):37-41. DOI: https://doi.org/10.7812/tpp/12-031.
    7.    Campbell RL, Li JT, Nicklas RA, Sadosty AT; Members of the Joint Task Force; Practice Parameter Workgroup. Emergency department diagnosis and treatment of anaphylaxis: A practice parameter. Ann Allergy Asthma Immunol 2014 Dec;113(6):599-608. DOI: https://doi.org/10.1016/j.anai.2014.10.007.
    8.    Davis RL, Gallagher MA, Asgari MM, et al. Identification of Stevens-Johnson syndrome and toxic epidermal necrolysis in electronic health record databases. Pharmacoepidemiol Drug Saf 2015 Jul;24(7):684-92. DOI: https://doi.org/10.1002/pds.3778.
    9.    Macy E, Poon K-Y T. Self-reported antibiotic allergy incidence and prevalence: Age and sex effects. Am J Med 2009 Aug;122(8):778.e1-7. DOI: https://doi.org/10.1016/j.amjmed.2009.01.034.
    10.    Macy E. Penicillin and beta-lactam allergy: Epidemiology and diagnosis. Curr Allergy Asthma Rep 2014 Nov;14(11):476. DOI: https://doi.org/10.1007/s11882-014-0476-y.
    11.    Edward RK, Tang Y, Raglan GB, Szychowski JM, Schulkin J, Schrag SJ. Survey of American obstetricians regarding group B streptococcus: Opinions and practice patterns. Am J Obstet Gynecol 2015 Aug;213(2):299.e1-7. DOI: https://doi.org/10.1016/j.ajog.2015.03.047.
    12.    Briody VA, Albright CM, Has P, Hughes BL. Use of cefazolin for group B streptococci prophylaxis in women reporting penicillin allergy without anaphylaxis. Obstet Gynecol 2016 Mar;127(3):577-83. DOI: https://doi.org/10.1097/aog.0000000000001297.
    13.    Verani JR, McGee L, Schrag SJ; Division of Bacterial Diseases, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention (CDC). Prevention of perinatal group B streptococcal disease—revised guidelines from CDC, 2010. MMWR Recomm Rep 2010 Nov 19;59(RR-10):1-36.
    14.    Macy E, Contreras R. Adverse reactions associated with oral and parenteral use of cephalosporins: A retrospective population-based analysis. J Allergy Clin Immunol 2015 Mar;135(3):745-52.e5. DOI: https://doi.org/10.1016/j.jaci.2014.07.062.
    15.    Macy E, Ho NJ. Adverse reactions associated with therapeutic antibiotic use after penicillin skin testing. Perm J 2011 Spring;15(2):31-7. DOI: https://doi.org/010.7812/tpp/11-025.
    16.    May SM, Hartz MF, Joshi AY, Park MA. Intrapartum antibiotic exposure for group B streptococcus treatment did not increase penicillin allergy in children. Ann Allergy Asthma Immunol 2016 Feb;116(2):134-8. DOI: https://doi.org/10.1016/j.anai.2015.11.013.
    17.    Sade K, Holtzer I, Levo Y, Kivity S. The economic burden of antibiotic treatment of penicillin-allergic patients in internal medicine wards of a general tertiary care hospital. Clin Exp Allergy 2003 Apr;33(4):501-6. DOI: https://doi.org/10.1046/j.1365-2222.2003.01638.x.
    18.    Rimawi RH, Mazer MA. Expanding the pool of healthcare providers to perform penicillin skin testing in the ICU. Intensive Care Med 2014 Mar;40(3):462-3. DOI: https://doi.org/10.1007/s00134-013-3196-1.

Circulation

27,000 print readers per quarter, 9,725 eTOC readers, and in 2016, 1.4 million page views on TPJ articles in PubMed from a broad international readership.

The Permanente Press

Sponsored by the National Permanente Medical Groups, The Permanente Press publishes The Permanente Journal and books related to Kaiser Permanente and health care.

Letters

Articles, editorials, letters to the editor, and other material represent the opinion of the authors. Send your comments to permanente.journal@kp.org.


Copyright 2017 Kaiser Permanente - The Permanente Journal. All Rights Reserved.