3PC-017 Time savings and improved quality assurance of intravenous cytostatics with semi-automated dose-banding

Background With increasing numbers of cancer diagnoses, cytotoxic compounding workload continues to increase, resulting in longer patient waiting times and higher risk for compounding errors. Dose-banding has been proposed as a way to increase efficiency but there is little data available on actual time savings and efficiency improvements. Equally, dose-banded batch compounding requires increased surveillance and traceability, e.g. by using (semi-) automated systems, which may affect time performance. Purpose The aim was to assess time savings by using dose-banding, while at the same time increasing quality assurance of the finalised preparations using a semi-automated compounding system. Material and methods From 1 March 2016 to 30 April 2017, total patient waiting times were registered for all sterile cytotoxic preparations for adult patients. During the first phase (1 March 2016 to 9 October2016), baseline performance was registered and suitable molecules were selected through dose-banding simulation (logarithmic approach) on 4 years of compounding data. During the second phase (10 October 2016 to 5 February 2017), paclitaxel was prepared through dose-banding, and five selected molecules (paclitaxel, carboplatin, 5-fluorouracil, irinotecan, trastuzumab, cyclophosphamide) during the third phase (6 February 2017 to 30 April 2017). Median (MPWT) and average (APWT) patient waiting times/week were analysed using segmented regression analysis, after correction for autocorrelation. All dose-banding compounding was done using an i. v. SOFT ™ Assist (Omnicell ® ), allowing for multi-dose vial use, in-process control and traceability. Results Over 55 weeks, 10 477 preparations were time-registered (baseline: 5660/28 weeks; second phase: 2774/16 weeks; third phase: 2043/11 weeks). At baseline, MPWT was stable at 2167 s/preparation with no significant time-dependent trend (Figure 1). During the second phase, MPWT decreased with −21 s/preparation/week (p=0. 07) but without immediate gains. The third phase resulted in an immediate additional gain of −363 s/preparation (p=0. 02) but without additional time-dependent improvement. APWT (baseline 2341 s/preparation) showed similar results with a significant change in trend during the second phase (−32 s/preparation/week; p=0. 04) but with only limited immediate effect at the third phase (−176 s/preparation; p=0. 39) without further effect on trend. Associated staff cost savings/preparation varied between €−0.76 and €−1.24/preparation. Conclusion Dose-banded compounding in combination with semi-automation results in meaningful reductions in patient waiting times, together with improved quality assurance and traceability. In case of limited resources, even dose-banding with a single selected molecule can have a significant impact on overall compounding workload. No conflict of interest