Detection of intravenous drug physicochemical incompatibilities in clinical practice: e-Harvis compared to reference...

Physicochemical incompatibilities (PCIs) resulting from the simultaneous intravenous administration of drugs through Y-sites represent a potential source of iatrogenic adverse events in hospital settings, particularly in situations involving polymedication and limited venous access. Although several compatibility databases are available, their coverage remains incomplete and their recommendations may be inconsistent.

A prospective observational study was conducted over 30 half-day periods in four hospital departments (conventional haematology, protected haematology, adult intensive care and resuscitation (ICR) and paediatric oncology) to identify drug pairs co-administered through Y-sites and assess their compatibility using five reference tools (Handbook on Injectable Drugs, Table of Y-Site Compatibility of Injectable Drugs from the University Hospitals of Geneva, King Guide to Parenteral Admixtures, Micromedex, and Stabilis). An overall compatibility response (compatible, incompatible or unknown) was defined as the classification derived from these five tools, with incompatible assigned in case of equal classifications. This overall response was used as a comparator to evaluate the performance of e-Harvis, a decision-support tool combining bibliographic compatibility data with a predictive model used when experimental evidence is unavailable.

A total of 765 administered pairs were identified, of which 56.1% were drug/drug pairs. Only one visible PCI was observed during the observation period. According to the overall response, 60.3% of pairs were compatible, 11.4% were incompatible and 28.4% were unknown. E-Harvis concluded that 32.8% were compatible, 10.6% were incompatible and 56.6% were unknown, with a 55.8% concordance with the overall response.

The tool also provided faster access to compatibility information and broader standalone coverage than individual reference databases for drug-drug pairs. In the absence of a validated gold standard for intravenous compatibility assessment, these findings suggest that e-Harvis may constitute a useful complementary decision-support tool for synthesising heterogeneous evidence and facilitating compatibility assessment in complex infusion settings.