Laurence A Goldberg
I would never build a new dispensary that was not automated. Such was the sentiment expressed by Michael Cross (Director of Pharmacy, Barts and the London NHS Trust, UK). Mr Cross described how the installation of a robot had transformed a poorly performing dispensary into an efficient workplace where staff are now keen to work.
One of the problems at the Royal London Hospital was a shortage of shelf space. Robotic storage now provides 600m of shelf space where there had previously been only 150m. In spite of a 30% increase in workload since the installation of the robot, no additional staff were required. The dispensary now functions more efficiently, partly because “almost everything” has been loaded into the robot. The machine has made two picking errors in two years, whereas there used to be 10 errors a month with manual dispensing.
Working conditions for staff have also improved. In the past it was never possible to finish the day’s work until after 18.00 hours; the average finishing time is now 17.35 hours.
Ward staff now rarely telephone to ask for missing items, explained Mr Cross. In spite of the increased throughput, the department looks and feels less busy because staff spend less time walking around, and the environment is so pleasant that staff actually request to work in the automated dispensary.
Mr Cross also installed a robot in another of his dispensaries. At Mile End Hospital, the pharmacy department suffered from an old-fashioned design, poor workflow and insufficient space. Installation of a robot allowed the store and dispensary to be combined, leading to considerable saving of space. The robot not only serves the dispensary but also picks ward stocks. There has also been considerable improvement in turnround times: whereas 75% of items were previously turned round in less than 10 minutes, this figure rose to 85% 2 months after installation.
Mr Cross is currently designing three new dispensaries with automated picking systems and is planning to automate the main pharmaceutical store. He recommends that smaller dispensaries should be merged to reap the benefits of automation.
Installation of a state-of-the-art robot enabled
Ms Pippa Roberts (Chief Pharmacist and Clinical Governance Manager, Chelsea and Westminster Healthcare NHS Trust) to meet performance targets at her hospital. She installed a robot that was capable of handling inpatient and outpatient dispensing and picking ward stocks.
The robot incorporates a fridge module and can be set to pick ward stocks overnight. The dispensary was redesigned to make a pleasant and efficient working environment. A pneumatic tube delivery system ensures that dispensed items are delivered promptly to the wards.
One important result of automation was that a number of staff – pharmacists, technicians and assistants – were released for other duties. The overall effect has been improved services, improved staff morale and fewer complaints, said Ms Roberts.
Making the case
The introduction of automated systems often relies on the chief pharmacist’s ability to make a good case, according to Keith Farrar (Chief Pharmacist, Wirral Hospital NHS Trust). There is good evidence in the published literature that medication errors can be reduced by improving the quality of medicines supply services and the quality of clinical pharmacy services, he said.
It has been estimated that more than 700,000 Americans are harmed by medication errors each year. A review of 4,031 patient records found an incidence of 6.5% actual errors and a further 5.5% potential errors.(1) The majority of errors occurred at the ordering (prescribing) and administration stages. Although only 1% were fatal, 42% of serious or life-threatening adverse drug events were considered preventable. Studies in the USA and the UK have shown that clinical pharmacy interventions can reduce the number of prescribing errors. Mr Farrar argued that automation could reduce the number of dispensing errors and release staff to improve the quality of clinical pharmacy services, thereby reducing the number of prescribing errors. A review of dispensing errors at the Wirral Hospital for the period 1994–2000 showed that the most common errors involved dispensing of the wrong dose or the wrong strength. Since the installation of a dispensing robot, such errors at the Wirral hospital have fallen from 15.7 to 7.7 per 100,000 items (a reduction of 51%). Automation frees time for clinical pharmacy services, and controlling medication errors is a key function of those services, concluded Mr Farrar.
Careful planning is important to ensure that the right robot is bought and the organisation is ready for automation, said Ann Slee (Director of Pharmacy, Glan Clwyd Hospital, Wales).
Before an order is placed, it is important to consider what the machine is intended to do and what benefits are expected. It is necessary to decide whether it will be used only for dispensing or whether it will also be used to pick ward stocks. The next step is to estimate the workload, considering the number of packs to be picked per hour and the likely peak levels of activity. Pharmacy managers also need to think about which items will be stocked in the robot – it might even be a good time to review what is kept and why, suggested Ms Slee. This information will help determine the capacity of the machine to be purchased.
Another critical issue is to consider how work will be carried out when the robot is in place. A redesign of the working processes and the dispensary layout may be necessary to get the best out of the robot. Pharmacy managers need to plan how refrigerated products and controlled drugs will be handled and how the labelling process will be interfaced.
Perhaps the most important questions relate to staff and the inevitable culture change that accompanies automation.
Mrs Slee recommended that managers should ensure that their staff know exactly what a dispensing robot is and how they will work with it. A way to achieve this is to visit working models in other departments. If installation of a robot is likely to release staff for other duties, then these duties need to be planned in advance. Similarly, ward staff need to be aware of the planned changes and their likely impact.
Three pilot sites have been set up in Wales, all of which went live between August and December 2003. The results so far show that dispensing errors have been reduced and that a large amount of staff time has been liberated, as they no longer need to put away, pick and check stock. The freed staff time has been used to improve ward services, and pharmacy-managed topping up has been introduced on an additional 17 wards. In addition, the number of internal orders has fallen by approximately 4,000 per month. Mrs Slee said that the Welsh experience had shown that the culture changes needed to be tackled a long time before anything else and that research to measure the impact of the changes should also start at an early stage.
For hospitals planning to build new pharmacies or remodel existing facilities, a robot will form the centrepiece, underpinning the medicines management process.
- Bates DW, Cullen DJ, Laird N, et al. Incidence of adverse drug events and potential adverse drug events. Implications for prevention. ADE Prevention Study Group. JAMA 1995;274:29-34.