Using Functional Resonance Analysis Methodology to identify potential safety improvements in the process of administering medications by infusion in a veterinary hospital.

Introduction: Medication error is a major cause of preventable harm in healthcare. Superficially, solutions to medication error appear simple, however medication administration often occurs during high-risk clinical scenarios, and under varied and challenging conditions which reduce intervention efficacy. Administering medication by infusion (AMI) is particularly complex. To develop feasible, effective and sustainable medication safety, requires an in-depth understanding of clinical conditions. Resilience engineering techniques can insight into process performance and variation and can identify how frontline staff can be better supported. Methods: A Functional Resonance Analysis Method (FRAM) based process evaluation was conducted in a multidisciplinary small-animal referral hospital. Information on work-as-imagined was obtained from institutional standard operating procedures (SOPs). Work-as-done was explored using an anonymous online staff survey with open-ended questions investigating process variation. Data were analysed using the four FRAM phases: identifying process functions, characterising performance variability, assessing functional resonance, and generating resilience-informed recommendations. Result: There were 45 responses to the survey included in analysis. The SOPs described a relatively linear process consisting of 7 phases whereas the FRAM model consisted of 45 interrelated and interdependent functions. Substantial discrepancies were identified between work-as-imagined and work-as-done. Performance variability and context dependent workarounds were common, particularly in relation to communication, dose calculators, infusion sheets, and double-checking, which were often impracticable under conditions of high workload, limited staffing, or clinical urgency. Interventions aimed at improving frontline abilities to anticipate, monitor and adjust to challenging conditions, and allowing the organisation to learn from outcomes were developed from the model. Discussion: FRAM demonstrated that AMI is a highly complex and variable process which is not adequately captured by existing SOPs. Performance variability reflected necessary adaptations to local conditions rather than individual non-compliance. Resilience-informed redesign of AMI processes may better support frontline staff and improve medication safety across diverse veterinary clinical contexts.