Empirical Data Driven Intensive Care Unit Drugs Inventory Policies

This paper proposes a drugs inventory policy at point-of-use level, tailored for the Intensive CareUnit (ICU) case study and aimed at relieving nurses of the time-wasting task of drugs ordering and refilling. The policy aims at jointly reducing order occurrences and imposing service regularity, while keeping stock value as low as possible. An optimization model is proposed and solved on a one-month period real instance and on a set of realistic ones derived from drugs consumption data collection at the ward. The potentially conflicting priorities of three stakeholders (nurses, administration and clinicians) have been successfully incorporated and their impact on order occurrences and stock value has been discussed. Computational results suggest that it is possible to optimize the time-consuming order process currently adopted at the ICU case study. This study is part of a more comprehensive project in which the optimization block will be integrated with a demand forecasting tool and deployed in a rolling horizon framework.