Improving Patient Outcomes by Dynamic Inventory Management of Drugs within Hospit

North Carolina State University Raleigh, Raleigh NC

Investigators

Abstract

[unreadable] DESCRIPTION (provided by applicant): The issue of ineffective inventory management affects hospitals on a national scale. Shortages have grown in recent years and the impact has been highlighted in the medical and pharmacy literature (Fox and Tyler, 2003, Baumer et al. 2004). In a national survey of 374 U.S. pharmacy directors on the impact of drug shortages in acute care hospitals, 75% of respondents indicated they were forced to purchase the drug off-contract from their current vendor, borrow the drug from another institution, or purchase the drug from an alternative vendor at an increased price (Baumer et al., 2004). In addition to the negative impact on purchasing costs, two-thirds reported delayed or canceled medical procedures due to drug shortages. We propose to address the impact of suboptimal inventory control to hospitals and patients by focusing on drug inventory policies. Our objective is to characterize the demand for a set of critical drugs, as a function of patient condition and mix. Characterizing demand enables us to create a dynamic ordering and inventory policy that results in timely fulfillment of patient demand and minimizes inventory and outdating or wastage costs. Through more accurate categorization of the patient attributes that drive pharmacy demand, the healthcare provider can more efficiently respond to a patient's pharmaceutical needs. In this proposal, we propose to use patient condition information to determine the appropriate inventory per period (i.e. day, week) minimizing wastage and holding costs while maximizing timely access to required/preferred drugs, and thereby improving patient safety and outcomes. We consider multiple inventories and the physical states (i.e. powder, solution, tablet etc...) of the inventory which contribute to its shelf life. Since drug usage is not known with certainty, the daily demand is stochastic and conditional on patient mix. We will develop a series of Markov models for drug demand as a function of patient condition and propose to focus on adult patients >50 years of age due to the complexity with polypharmacy and the shift in age of the general population. We will develop performance measures that assess the impact of shortages on patient outcomes (e.g. number of drug substitutions, number of delayed procedures) as well as the cost of overages (e.g. holding and outdating inventory costs) and compare the performance of our policies to current hospital pharmacy inventory policies. [unreadable] [unreadable] Public Health Relevance: We will conduct statistical analysis to determine what patient data (i.e. condition, demographics) effectively predicts pharmaceutical requirements in an inpatient hospital unit setting. The goal of the proposed research, using the analysis described above, is to develop a forecasting and inventory decision model. The model will allow pharmacy operations practitioners and hospital administrators to use patient data to manage the inventory of drugs more efficiently while improving the timeliness of care delivered to patients. [unreadable] [unreadable] [unreadable]