Search for a command to run...
<h3>Background and Importance</h3> Medication management in Hospital-at-Home (HaH) requires continuous assessment of prescriptions to ensure patient safety. The broad diversity of pathologies managed in HaH makes therapeutic optimisation complex and demands high versatility from healthcare teams. Pharmaceutical interventions (PIs) are key indicators for identifying risks and improving medication safety. However, they remain underused for this purpose. <h3>Aim and Objectives</h3> This study aimed to evaluate and improve care practices by analysing all pharmaceutical interventions performed in an HaH service, in order to identify the main sources of medication related risk and define targeted corrective actions. <h3>Material and Methods</h3> A prospective registry captured all PIs between June and December 2024 at any step of the medication process (reconciliation, prescribing, preparation, administration, monitoring). For each PI we recorded: care pathway, ATC class, problem type (eg, no indication, overdose, incompatibility, drug–drug interaction...), recommended action, acceptance, and clinical impact (Cléo 0–4: minor to vital). Data were analysed descriptively to map risk clusters and prioritise improvement. Regular multidisciplinary reviews ensured consistent classification and fed back aggregated findings to prescribers. <h3>Results</h3> We recorded 188 PIs; the acceptance rate was 93%. Over half were minor (Cléo 0–1), frequently arising during medication reconciliation. However, 25% were moderate (Cléo 2), 17% major (Cléo 3), and 4% vital (Cléo 4). The most critical PIs (Cléo 3–4) occurred mainly in infectious (50%) and post-surgical (18%) pathways and involved prescriptions without indication (29%), overdosing (14%), physicochemical incompatibilities (14%), and drug–drug interactions (29%). High-leverage ATC classes included B01 (antithrombotics), B05 (nutrition/solutions), M04 (antigout), and N02 (analgesics). Corrective actions implemented, based on these findings: (1) prescriber training on decision support tools (dose adjustment, STOPP/START , PIM-CHECK ); (2) structured, multidisciplinary reviews of high risk cases; (3) focused workshops on critical ATC classes and incompatibilities; (4) local protocols for recurrent scenarios (eg, IV-to-oral switch, anticoagulant management); and (5) optimisation of interaction alerts and alert readability in the prescribing software. <h3>Conclusion and Relevance</h3> A structured PI database enabled real-time risk surveillance and targeted safety actions, with nearly half of PIs rated ≥2 on the Cléo scale. This approach strengthens safety culture in HaH and is readily reproducible. A planned follow-up will evaluate sustained effects on Cléo ≥2 incidence, alert handling, and protocol adherence. <h3>Conflict of Interest</h3> No conflict of interest