This site is intended for health professionals only

Implementation of automated dispensing in France


Hugues Michelon, PharmD

Marie Antignac, PharmD, PhD
Pharmacy Department, Pitié Salpêtrière Hospital, Assistance Publique – Hôpitaux de Paris, France

The drug supply chain has become one of the most complex in healthcare, affecting every healthcare worker. Today, despite implementation of quality assurance systems, the different steps of medication processes workflow – from ordering to drug administration – still represent a source of potential error that can affect patient safety.1-3

Drug-related adverse events or medication errors (preventable adverse drug events) have important implications, from increased length of hospitalisation and costs to undue discomfort and disability or increased mortality.4-7

In France, the perception of this problem is relatively recent and little data was available before the results of the ENEIS national study, in which the frequency of serious adverse events was estimated to be 6.6 per 1,000 days of hospitalisation and related to medication use in 26.7% of cases.8

To address this problem, safe medication use has become a priority for public health in France, especially since the decree of 24 August 2005 announcing the advent of the contract of good use of medicines.3,9,10

In line with this contract, and in order to receive substantial credit, French hospitals have to commit themselves to implementing measures using information technology to improve patient safety, cost-effectiveness and time-optimisation. This approach also addresses the current shortage of nurses and head-nurses faced by hospitals.

Safe medication use in intensive care units
Among the several technical devices now available, automated dispensing systems (ADS) have already been shown to be a promising strategy for preventing medication errors and improving time optimisation and cost-effectiveness. 3,9-12
Several companies have developed such systems (Pyxis CareFusion, Omnicell, Mach4) and are now equivalent in the type of technology used: connection by biometric recognition, several types of drawers with different levels of safety, guidance systems to help picking, real-time visualisation of the remote stock.

To improve its safe use of medication, Pitié-Salpêtrière Hospital, a large teaching hospital with 1,600 beds, 40 healthcare buildings and one central pharmacy department, decided to install the ADS Medstation 3500 from Pyxis CareFusion.

Since 2006, six ADSs have been installed in the hospital. Implementation was made a priority for intensive care units, which were considered as having a higher risk of iatrogenic diseases and where the number and high frequency of medication orders do not help pharmaceutical interventions.13-15

ADS can be used to hold drugs at a location and dispense them only to a specific patient. As a result, introducing such systems was defined as a basic first step in improving medication security, involving a major role for the pharmacy and the technician pharmacist in patient-care units.

Evaluating implementation
Before implementation, drug storage was managed by a head nurse; afterwards, drug ordering and management was performed by a pharmacy technician. So, to assess the impact of the new ADS on the hospital’s organisation, a before-and-after analysis was conducted for each installation.

These assessments involved four criteria:

The workload and role of pharmacy-technician in patient care units
The follow-up of emergency orders (quantity for one year)
The costs of drug consumption and drug storage
Nurses’ satisfaction (assessed by self-filled questionnaire).

Automated dispensing systems appear to be a promising technology to improve the drug supply chain for multiple reasons. Since implementation, an evident shift from nurses to pharmacy-technician has been observed concerning medication-related activities such as ordering, reception, refilling, management of expired medicines and transmitting information about medicines in association with the pharmacist.

On the one hand, ADS implementation has lead to a decrease in the time per week spent by nurses on medication-related activities from about 10% to 2%. On the other hand, an average increase in technician-pharmacy weekly working time of 15% has been observed. Thus, time optimisation is clearly beneficial from an economical point of view, because expensive nursing time can be saved for clinical duties.

A qualitative and quantitative revision of drug storage was defined for each patient-care unit in agreement with the pharmacist, the physician and the head nurse before implementation, leading to a decrease from -10% to -50% and -10% to -63% of its initial value. Consequently, ADS could be an effective device to improve cost-effectiveness, especially thanks to daily management and strict monitoring of drug storage by the pharmacy department.

The before-and-after analysis of emergency orders for pharmaceuticals did not show any significant decrease since implementation of the automated dispensing system. However, interestingly, the number of emergency orders for pharmaceuticals did not increase despite the quantitative and qualitative decrease in drug storage. This could be partly due to real-time monitoring of the remote stock from the pharmacy – via a central computer connected to all ADS – in order to avoid stock-outs.

Moreover, these results directly depend on nurses’ willingness to record correctly the number of units taken from the system.

Preventing medication errors
Beyond organisational and financial objectives, ADS have been developed specially to prevent medication errors. However, such systems are today used to hold drugs at a location and dispense them to a specific patient only. Thus, prevention of medication errors is limited to the picking of medicines at present. The effectiveness of such systems can only be demonstrated when they are interfaced with computerised prescribing, which is the next step of our project.
Users’ satisfaction

Implementation of ADS technology in patient-care units is often perceived as stressful or negative by users and not really adapted to the activities of intensive care units. Thus, information and user support are necessary before and after implementation to get a better acceptability from users.

Contrary to preconceived ideas, however, our survey revealed that nurses’ opinions concerning ADS was very positive and enthusiastic: 87% were satisfied by the role of the pharmacy-technician and 60% indicated that they did not want to return to the old system.

In our experience, innovative technologies such as ADS have shown potential benefit for cost-efficiency, time-optimisation and medication safety. Similar data has also been described in intensive care units.16 Thanks to such devices, integration of the pharmacy staff into the workings of clinical units seems to have a positive impact on the quality of care due to the transfer of pharmaceutical tasks, the overview of the medication process and increased clinical pharmacy input. This impact will become more relevant in the future when computerised prescriptions and automated medication administration records are interfaced with the ADS. Without these links, the effect of such devices will be always underestimated. Moreover, data analysis and the development of integrated data analysis software will be useful to conduct pharmaco-epidemiologic studies in the sphere of cost-minimisation and cost-optimisation.

In the future, integrated data analysis software and economic data analysis will represent the only means to demonstrate the effectiveness of ADS to physicians. As a result, the next step of our project consists of developing computerised physician prescribing and data analysis software and integrating it into nine other ADS implementations planned for 2011.


  1. Koppel R et al. JAMA 2005;293(10):1197-1203.
Lazarou J et al. JAMA 1998;279(15):
  3. Poon EG et al. NEJM 2010;362(18):1698-1707. doi:10.1056/NEJMsa0907115.
  4. Bates DW et al. JAMA 1995;274(1):29-34.
de Vries EN et al. Qual Saf Health Care 2008;17(3):216-223.
  6. Leape LL et al. JAMA 1995;274(1):35-43.
Leape LL & Berwick DM. JAMA 2005;293(19):2384-2390.
Michel P et al. Qual Saf Health Care 2007;16:
  9. Bates DW BMJ 2000;320(7237):788-791.
Bates DW & Gawande AA. NEJM 2003;348 (25):2526-2534.
Barker KN Am J Health Syst Pharm 1995;52(21):2445-7.
Borel JM & Rascati KL. Am J Health Syst Pharm 1995;52(17):1875-9.
  13. Abeysekera A et al. Anaesthesia 2005;60(3):
Cullen DJ et al. Crit Care Med 1997;25(8):
Rothschild JM et al. Crit Care Med 2005;33(8):1694-1700.
Kheniene F et al. Ann Fr Anesth Reanim 2008;27(3):208-215.

Be in the know
Subscribe to Hospital Pharmacy Europe newsletter and magazine