Adherence to double-check protocols during medication administration remains limited. Despite this, most administrations of high-risk drugs proceed without adverse events. This study explored how high-risk medications are prepared and administered, identifying real-world variations and the reasoning behind departures from the prescribed standards.Ten hospital wards in the Netherlands took part. Using the Functional Resonance Analysis Method, we produced a model reflecting the national guidance and a cross-ward model representing actual work processes. To construct the cross-ward model, each ward participated in eight semi-structured interviews focusing on the preparation and delivery of high-risk drugs. Subconscious decision processes were categorized using Efficiency-Thoroughness Trade-Off principles.A total of 77 nurses were interviewed. Six discrepancies emerged between the guideline-based model and the cross-ward model. Importantly, four distinct deviations in double-check routines appeared. Time constraints played a major role. Nurses evaluated patient condition, calculation complexity, and other risk cues to decide whether to complete the double-check. Additional tacit judgements—such as reliance on personal or colleagues’ competence—also shaped decisions.Time pressure is the dominant factor preventing consistent completion of the double-check. Instead, nurses rely on situational risk assessments to determine its necessity. As a result, the double-check may become routine for some drugs and omitted for others. Future work could use FRAM insights to redesign ward-specific, feasible double-check procedures that support safety in everyday practice.
