The Effect of Penicillin Allergy Testing on Future Healthcare Utilization and Morbidity: A Case-Control Study
Saturday, March 5, 2016
South Exhibit Hall H (Convention Center)
Eric M. Macy, MD FAAAAI, Yu-Hsiang Shu, MS PhD

The effect outpatient penicillin allergy testing has on future healthcare utilization (HU) and morbidity is unknown.


HU and morbidity in 500 penicillin allergic cases, only 8 penicillin allergy test-positive, seen from 6-8-2010 to 3-29-2012, were compared to non-tested controls.  Five penicillin allergic controls were matched to each case by gender, age, weighted Charlson comorbidity index (CCI), and outpatient office visits during a 2-year lead-in period.  Hospital days, new drug allergies, and antibiotic exposures were determined over a 36 to 62 month follow-up period.


Cases had a mean (95% CI) lead-in CCI of 1.65 (1.44–1.87) and 13.73 (12.76–14.7) outpatient visits per year of healthplan coverage (YHC).  Controls had a mean lead-in CCI of 1.7 (1.6–1.8) and 12.08 (11.69–12.47) outpatient visits per YHC.  For cases, hospital days per YHC increased significantly less than for control subjects, 1.62 (1.13–2.11) to 2.16 (1.00–3.31) [133%] versus 0.73 (0.59–0.87) to 2.02 (1.43–2.61) [277%], (p < 0.0001).  Antibiotic courses per year fell significantly more in cases, - 0.85 versus -0.62, (p = 0.0212).  Cases had significantly fewer new drug allergies at the end of the follow-up period, +0.044 versus +0.18, (p < 0.0001).  The top 5 antibiotics cases were exposed to were penicillins, 1st generation cephalosporins, quinolones, tetracyclines, and macrolides versus macrolides, quinolones, clindamycin, 1stgeneration cephalosporins, and tetracyclines for controls.    


Penicillin allergy testing was associated with a smaller increase in hospital days, a greater fall in antibiotic usage, fewer new drug allergies, and significantly less macrolide, quinolone, and clindamycin exposure over the next 3 to 5 years.