Implementation of a Nursing Algorithm for Penicillin Allergy Documentation in the Inpatient Setting

Abstract

Background: Patients with a penicillin/aminopenicillin (PCN) allergy label are more likely to receive non–β-lactam antibiotics and to experience worse clinical outcomes. Given that nurses are often first to interact with patients, we pilot tested a nurse-driven quality improvement initiative to improve PCN allergy documentation and increase β-lactam use. Methods: We conducted a before-and-after study on a labor and delivery unit at The Johns Hopkins Hospital (JHH) from May 2018 to September 2019. Patients aged 18 years with a PCN allergy were included. The intervention included (1) the use of an algorithm developed by the antimicrobial stewardship team to assist nurses in obtaining accurate PCN allergy histories (Fig. 1), (2) identification of a nurse champion to facilitate implementation of the algorithm, and (3) in-person education by a stewardship physician regarding the importance and impact of adequate PCN allergy documentation on clinical outcomes. Readmissions were counted as separate encounters. The primary outcome was improved allergy documentation (either fewer blank documentations, nonspecified rash reactions, drug intolerance documentations (eg, isolated nausea), documentation of signs and symptoms of anaphylaxis not specified as such). The secondary outcome was β-lactam use. Categorical variables were compared using the ² test and continuous variables were compared with the Student t test. Severe allergic reactions were defined as anaphylaxis, severe skin reactions (eg, Stevens-Johnson syndrome), and organ involvement (eg, hepatitis). Results: Overall, 382 patient admissions were included, 305 in the preintervention (May 2018 to May 2019) and 77 in the postintervention period (June 2019 to September 2019). Mean age and length-of-stay were 30 years and 4 days, respectively, for both periods. The proportion of admitted patients with a PCN allergy label was 8% and 7% for pre- and postintervention periods. Documentation findings in the pre- and postintervention periods respectively were as follows: blank documentation 11% and 12% (P = .89), documentation of specified rashes 0.6% and 1.3% (P = .56), documentation of drug intolerance 11% and 8% (P = .39), documentation of reactions that were indicative of anaphylaxis but not documented specifically as anaphylaxis 8% and 13% (P = .20). Among patients with a documented PCN allergy who received antibiotics, 83 of 177 (47%) and 27 of 43 (63%) received β-lactams (P = .01) in the pre- and postintervention periods, with cefazolin being the antibiotic most commonly used in both periods. Conclusions: Nursing education and an algorithm did not result in significant improvements in PCN allergy documentation in the 3 months after implementation. More data collection is planned to assess the impact of the intervention.

Funding: None

Disclosures: None