A LEAN Approach to Intravenous Antibiotic Distribution in a Tertiary Care Academic Hospital

Abstract

Background: LEAN is a quality improvement methodology that identifies and eliminates waste in processes. LEAN principles, such as value stream mapping, may be applied to the hospital IV antimicrobial distribution process to increase resource use efficiency while ensuring patient safety. Objectives: Primarily to define the current IV antibiotic distribution process at Kingston General Hospital (KGH) and to identify opportunities for improvement in resource use efficiency. Secondarily to reduce waste, categorized as non-value-added time from a patient perspective or by cost of unusable product, in the IV antibiotic distribution process. Methods: A stakeholder group of pharmacy personnel and a patient experience advisor was formed. Current state value stream maps (VSMs) were constructed with direct observation for time estimates for each of the three, not previously defined, distribution streams: (1) central intravenous admixture service (CIVAS) preparation, (2) CIVAS-prepared minibags stocked in automated dispensing cabinet (ADC), and (3) vial stock in ADC used with point-of-care activation device. Each process step was categorized into ‘value-added’, ‘non-value-added’ (pure waste), and ‘non-value-added, but essential’. Opportunities for process improvement were identified, and education of staff regarding IV antibiotic order entry into the pharmacy information system was chosen for implementation using a Plan-Do-Study-Act (PDSA) cycle. Representative product waste and adherence to standardized order entry procedures were assessed before and after education. Results: Current state VSMs of ampicillin, cefazolin and piperacillin-tazobactam distribution were constructed, representing distributions streams 1, 2 and 3, respectively. Value added time was found to be the highest with stream 3, piperacillin-tazobactam at 26% (16 min), followed by stream 1 at 21% (14 min), then stream 2 at 14% (12 min). Stream 2 had the highest proportion of non-value-added (pure waste) with 74% (62 min), followed by Stream 1 at 67% (45 min), then Stream 3 at 64% (40 min). No statistically significant difference was observed in resource waste or policy adherence before and after staff education on IV order entry processes (waste, $1804 versus $1840; adherence, 49% versus 46%, NS). Conclusions: Using LEAN methodology, the IV distribution stream associated with the most efficient resource use is ADC vial stock intended for use with point-of-care activation device, but resource waste is still substantial. A PDSA intervention focused on PIS order entry in all 3 streams failed to demonstrate waste reduction. Exploration of further opportunities identified should be undertaken.