Evaluation of a Computerized Decision Support Intervention to Decrease Use of Anti-Pseudomonal Carbapenems in Penicillin Allergic Patients.

Allergies to β-lactam antibiotics are commonly documented in hospitalized patients; however, true allergy is uncommon. Cross-reactivity rates for advanced generation cephalosporins and carbapenems are low; particularly for patients without a history of symptoms consistent with type 1 hypersensitivity. We observed that providers preferentially prescribed antipseudomonal carbapenems (APC) over advanced generation cephalosporins for patients with β-lactam allergy history, including those with low risk for antimicrobial-resistant infections. Information was inserted into the computerized decision support system (CDSS) to aid clinicians in assessing β-lactam cross-reactivity risk and selecting appropriate therapy. A retrospective evaluation was conducted in a small hospital to assess the impact of the CDSS changes in APC prescribing. Inpatients (n = 68) who received at least one APC dose during hospitalization over a 13 month pre-intervention period were compared to inpatients who received an APC during the 15 month post-intervention period (n = 59) for documented APC indications and β-lactam allergy history. APC initiations were measured and corrected per 1000 patient-days; interrupted time-series analysis was performed to assess changes in use before and after implementation. Aggregate monthly APC initiations decreased from 7.01 to 6.14 per 1000 patient-days after the implementation (p = 0.03). Post-intervention APC initiations for patients with low-risk β-lactam histories decreased from 92% to 83% (p = 0.17). No adverse events were observed in patients with low-risk β-lactam histories. The intervention was associated with a reduction in APC initiations.