Research of the most efficient model for automated dose dispensing (ADD) to be administered in a medicine unitbe

Introduction
Since 2013, an automated ward dispensing device is used to give nominative drug dispensation in our institution’s long stay units. Only dry oral forms are managed by ADD process. To secure the medication circuit, the ADD process has been extended in medicine units. The purpose of this study is to define the safest and the most adapted automated dispensing model for a geriatric medicine unit (average length of stay = 11.7 days).

Materials and methods
Two automated dispensing modes were tested and compared between them and to manual dispensation: Biweekly production and daily production. In each case, the production of doses to be administrated is in a “morning-midday-evening” order. Some indicators have been monitored: the pill dispensers’ compliance before treatments administration; the time devoted to prepare pill dispensers by nurses; the time devoted to doses dispensing by pharmacy; the rate of nonadministrated doses returned and destroyed. Results have been statistically compared by Chi² and Student tests. The significance threshold has been defined to p<0.05.

Results
The weekly time devoted by the pharmacy to treatments dispensing is not significally different neither between the two production’s models (232’ with daily mode vs. 215’ with biweekly mode; p = 0.05) nor with the manual dispensation (213’; p=0.92 vs. daily mode and p=0.22 vs. biweekly mode).
The time devoted by nurses to pills dispensers preparation is significally lower than the manual dispensing: 22’ in daily mode and 30’ in biweekly mode vs. 60’ in manual mode (p = 5.7 x 10-8 et p = 3.4 x 10-4). In the daily model, the rate of non-compliance (2.2%) between prescriptions and pills dispensers is significally lower than those obtained with biweekly production (5.0%; p = 0.011) and manual preparation (7.0%; p = 7.5 x 10-5). Futhermore, the daily production has a rate of treatments returns lower than the biweekly production (5% vs. 17%; p = 7.58 x 10-58).

Discussion/Conclusion
The daily mode is the most efficient and has been chosen to automatize the pills dispensing. Indeed, it is the safest model; it does not wasted time in pharmacy and it generated less treatments returns. So, it represents an economical advantage. However, a new risk has been observed: the persistence of treatment modification taking place after the production. A daily ADD produced in a “evening-morning-midday” order would allow to reduce this risk. Indeed, the majority of the prescription’s modifications are made during the morning medical examination. Also, it would reduce the rate of non-administrated treatments returns.

See the poster