Inpatient Penicillin Allergy Delabeling Scales Across Hospitals

Key Takeaways:

• Across 40 hospitals in 8 countries, the authors report that positive penicillin DOC results were uncommon and serious adverse events were rare within heterogeneous inpatient protocols.
• The authors report that bimonthly digital audit-and-feedback was associated with increased penicillin allergy assessment and DOC activity, alongside improving protocol fidelity over time at regularly receiving sites.
• Reported outcomes included shifts in antimicrobial prescribing patterns and post-assessment documentation updates, described alongside feasibility and sustainability measures across varied hospitals.

A prospective, international type 2 hybrid effectiveness–implementation evaluation reported that inpatient penicillin direct oral challenge (DOC) could be delivered across diverse hospital contexts when supported by a digital toolkit and bimonthly audit-and-feedback. In the iNAAN multicenter evaluation, adult inpatients with a reported penicillin allergy were assessed at 40 hospitals in 8 countries using a smartphone-enabled workflow to support clinician risk assessment (including PEN-FAST) and identify patients considered low risk and potentially eligible for direct oral challenge based on local criteria. Across sites, investigators captured 2,013 inpatient DOC episodes during the study period. The authors framed their primary focus as describing safety alongside adoption and fidelity outcomes across heterogeneous local protocols.

Patients entered the pathway when a penicillin allergy label was identified during hospitalization. Multidisciplinary clinicians then completed point-of-care assessment using the digital penicillin allergy toolkit embedded in the NAAN platform. The article reports that the toolkit incorporated a validated assessment instrument and embedded the PEN-FAST clinical decision rule to support identification of low-risk patients who could be considered for DOC if local eligibility criteria were met. Participating hospitals used site-led protocols rather than a single ward-based model, allowing variation in drug selection, dosing strategy, inpatient setting, and observation duration. This pragmatic structure situated implementation within routine inpatient workflows under local governance.

Safety and protocol execution were summarized across DOC procedures performed under local protocols, with 1,822 challenges counted as protocol-compliant in the authors’ analysis. Across protocol variants, positive direct oral challenge frequencies varied (roughly ~2% to 6% across reported strata), and serious adverse events were uncommon (generally reported around 0.0% to 0.8%). The article also reports high protocol fidelity, with 90.5% compliance when fidelity was defined as adherence to the implemented site DOC protocol elements captured in the NAAN data system. Observation periods differed by local protocol (reported as spanning roughly 60 to 180 minutes), and the authors described consistent safety patterns despite this heterogeneity. In aggregate, the authors characterized the multicenter findings as showing uncommon positive challenges alongside high fidelity across varied inpatient implementations.

Implementation analyses attributed changes in activity to the digital audit-and-feedback strategy delivered through the NAAN app via emailed, site-specific health service reports. Using interrupted time series methods, the authors report that audit-and-feedback was associated with an immediate increase in penicillin allergy assessments and DOC activity, followed by sustained activity above baseline; larger early effects were observed among pharmacists and nurses than among other discipline groups. They also report a directionally increasing trend in protocol fidelity over time at sites receiving regular feedback episodes. For longer-term adoption, the article notes growth in registered users at hospitals active for more than 12 months (from 35 in month 1 to 70 at 6 months and 98 at 12 months), while also describing attrition of 24 users who stopped entering data early. The authors interpret these patterns as consistent with audit-and-feedback supporting adoption and sustainability and with sustained improvements in protocol fidelity over time.

The authors also note antimicrobial stewardship gains reported from iNAAN overall, including increased penicillin use and reductions in World Health Organization–restricted antibiotic prescribing. They also report that documentation processes could change following assessment or challenge, highlighting Australian sites where adherence to antibiotic allergy documentation standards increased from 10.9% preassessment to 78.3% postassessment. The discussion notes pragmatic limitations, including the absence of denominators needed to calculate reach or adoption rates, low survey response for acceptability and feasibility measures, and the fact that NAAN data entry accuracy was not audited against clinical records. The reported stewardship, documentation, and feasibility findings were presented as components of a digitally supported, locally adapted inpatient implementation model.