Healthcare Utilization and Serious Infection Prevalence Associated With Penicillin “Allergy” In Hospitalized Patients: A Cohort Study

Abstract

RationalePenicillin is the most common inaccurate drug “allergy” noted at hospital admission. Determine hospital utilization and prevalence rates of Clostridium difficile, methicillin resistant Staphylococcus aureus (MRSA), and vancomycin resistant enterococcus (VRE), in patients with and without penicillin “allergy”.MethodsA retrospective, matched cohort, study of individuals admitted to Kaiser Foundation hospitals in Southern California during 2010 through 2012.ResultsIt was possible to match 51,582 (99.6% of all possible cases) unique penicillin “allergic” hospitalized individuals to 2 unique control subjects each. Penicillin “allergic” cases averaged 0.59 (9.9%) [95% CI, 0.47 to 0.71] more total hospital days during 20.1 ± 10.5 months of follow-up, compared to discharge diagnosis category-, gender-, age-, and date of admission-matched controls. Cases were treated with significantly more fluoroquinolones, clindamycin, and vancomycin, p < 0.0001, compared to controls. Cases had 23.4% [95% CI, 15.6% to 31.7%] more Clostridium difficile, 14.1% [95% CI, 7.1% to 21.6%] more MRSA, and 30.1% [95% CI, 12.5% to 50.4%] more VRE infections than expected, compared to controls. When adding matching for the number of drug “allergies”, differences between cases and controls disappeared, but a strong positive correlation was seen for all 4 outcome variables and increasing drug “allergy” number in both cases and controls.ConclusionsA drug “allergy” history though often inaccurate, is not a benign finding at hospital admission. Drug “allergies” are associated with increased hospital utilization and increased Clostridium difficile, MRSA, and VRE prevalence. RationalePenicillin is the most common inaccurate drug “allergy” noted at hospital admission. Determine hospital utilization and prevalence rates of Clostridium difficile, methicillin resistant Staphylococcus aureus (MRSA), and vancomycin resistant enterococcus (VRE), in patients with and without penicillin “allergy”. Penicillin is the most common inaccurate drug “allergy” noted at hospital admission. Determine hospital utilization and prevalence rates of Clostridium difficile, methicillin resistant Staphylococcus aureus (MRSA), and vancomycin resistant enterococcus (VRE), in patients with and without penicillin “allergy”. MethodsA retrospective, matched cohort, study of individuals admitted to Kaiser Foundation hospitals in Southern California during 2010 through 2012. A retrospective, matched cohort, study of individuals admitted to Kaiser Foundation hospitals in Southern California during 2010 through 2012. ResultsIt was possible to match 51,582 (99.6% of all possible cases) unique penicillin “allergic” hospitalized individuals to 2 unique control subjects each. Penicillin “allergic” cases averaged 0.59 (9.9%) [95% CI, 0.47 to 0.71] more total hospital days during 20.1 ± 10.5 months of follow-up, compared to discharge diagnosis category-, gender-, age-, and date of admission-matched controls. Cases were treated with significantly more fluoroquinolones, clindamycin, and vancomycin, p < 0.0001, compared to controls. Cases had 23.4% [95% CI, 15.6% to 31.7%] more Clostridium difficile, 14.1% [95% CI, 7.1% to 21.6%] more MRSA, and 30.1% [95% CI, 12.5% to 50.4%] more VRE infections than expected, compared to controls. When adding matching for the number of drug “allergies”, differences between cases and controls disappeared, but a strong positive correlation was seen for all 4 outcome variables and increasing drug “allergy” number in both cases and controls. It was possible to match 51,582 (99.6% of all possible cases) unique penicillin “allergic” hospitalized individuals to 2 unique control subjects each. Penicillin “allergic” cases averaged 0.59 (9.9%) [95% CI, 0.47 to 0.71] more total hospital days during 20.1 ± 10.5 months of follow-up, compared to discharge diagnosis category-, gender-, age-, and date of admission-matched controls. Cases were treated with significantly more fluoroquinolones, clindamycin, and vancomycin, p < 0.0001, compared to controls. Cases had 23.4% [95% CI, 15.6% to 31.7%] more Clostridium difficile, 14.1% [95% CI, 7.1% to 21.6%] more MRSA, and 30.1% [95% CI, 12.5% to 50.4%] more VRE infections than expected, compared to controls. When adding matching for the number of drug “allergies”, differences between cases and controls disappeared, but a strong positive correlation was seen for all 4 outcome variables and increasing drug “allergy” number in both cases and controls. ConclusionsA drug “allergy” history though often inaccurate, is not a benign finding at hospital admission. Drug “allergies” are associated with increased hospital utilization and increased Clostridium difficile, MRSA, and VRE prevalence. A drug “allergy” history though often inaccurate, is not a benign finding at hospital admission. Drug “allergies” are associated with increased hospital utilization and increased Clostridium difficile, MRSA, and VRE prevalence.