Keith Farrar
Chief Pharmacist
Wirral Hospitals NHS Trust
Ann Slee
Chief Pharmacist
Glan Clwyd Hospital
Rhyl
UK
Why automation? There are many reasons for considering the introduction of automation into the dispensing process but perhaps the over-riding concern is to make best use of available staffing resources. Shortages of pharmacy staff are common in many countries and present a particular problem in the UK, where hospital pharmacy salaries lag a long way behind those in community. A national survey indicated that there were 547 WTE (whole time equivalent) pharmacist vacancies and 452 WTE technician vacancies.(1) More worrying was the finding that there were fewer pharmacists and technicians currently in training than the number of existing vacancies. Hence the need to find an alternative means of delivering service that would release staff time to clinical duties.
The report, Pharmacy in the future – implementing the NHS plan,(2) highlighted a number of areas of practice that needed to develop. Three of these are crucial to making the best of pharmacy staff: re-engineering supply and administration processes; making better use of skill mix; and making better use of information technology and automation. These themes were reaffirmed in the recent report by the Audit Commission, A spoonful of sugar,(3) which identified that pharmacy, as a clinical profession, needed to focus staff time on clinical issues.
Other staff could also benefit from the introduction of automation and the redeployment of pharmacy staff. A local study examining the benefits of re-engineering the supply of medicines to the bedside in original packs demonstrated a reduction in drug administration error rate of almost 400% (from 9.8% to 2.5%), as well as reducing the time nurses spent on drug administration by 7.5%. As nurses spend almost 40% of their time dealing with medication-related issues, redeploying technicians to manage medicines at the bedside could potentially release even greater amounts of nursing time.
Improving the use of staff, however, is only part of the expected benefit. A study by Spencer et al showed an error rate of 17.5 per 100,000 prescriptions dispensed in hospitals in the UK.(4) A later analysis of these reports showed that two-thirds were related to selection of the wrong product or the wrong strength of the correct product. Clearly the use of automated systems that utilise barcodes to identify products rather than manual selection has the potential to reduce dispensing errors dramatically, supporting the reductions in medication errors required by the UK Department of Health guidelines, An Organisation with a Memory.(5)
Factors affecting the selection of the system installed
The dispensing system we were to select for our department would need to meet a number of criteria: it had to utilise barcodes on input to reduce medication errors and be fast enough to improve dispensary times while retaining/improving accuracy. Loading had to use the minimum amount of time and skill so that technician could be freed up to move to ward roles and the technology had to be proven and working somewhere.
The main outcomes/objectives were to:
- Reduce staff time required for dispensing and redeploy these at ward level for clinical work.
- Reduce the turnaround time for prescriptions to improve patient care/reduce waiting times.
- Reduce the error rates within the dispensing process.
- Reduce the floor space required to store drugs within the department.
- Streamline the ordering process as shelf space should be at less of a premium.
Implementation
An automated dispensing machine (ROWA Speedcase) was installed within the main dispensary at Wirral Hospital at the end of 2000 and “went live” in January 2001. The machine is a tandem system and has the capacity to hold up to 7,000 original packs of drugs in each side while dispensing an average of one item every 20 seconds. Approximately 80% of the dispensary stock is now housed within the Speedcase, which has led to a considerable saving in space within the redesigned dispensary. The Speedcase is interfaced with the pharmacy stock control system, JAC, and picks the appropriate pack in response to a “picking message” that is linked to the production of a label.
Before the submission of a business case for automation, efforts were made to engage all the pharmacy staff and particularly the technicians with the benefits to be gained from automation. This included developing an alternative role for technicians at ward level and including technical staff in the selection process, visiting working sites in Europe and feeding back to the rest of the staff in a formal meeting. Technicians were also influential in determining the necessary changes to the design of the dispensary and the changes in workflow that might result from implementation.
Dispensary changes
The introduction of automation was a radical change and offered an opportunity for the redesign of the dispensary. Observation of dispensing processes in community practice led us to believe that a smaller dispensary could be more efficient while ensuring that we gained maximum benefit from the introduction of automation – the redesign has utilised approximately 50% less space. The redesign also included a requirement that the medicine be delivered to the terminal where the “picking message” originated.
Costs and benefits associated with the introduction of automated dispensing
The system cost £350,000 to install (£1=e. 1.50) – an investment that has represented excellent value for money. The key results that have been seen are a reduction in dispensing errors of almost 50% and the release of 31% of dispensing technician time (see Figure 1) to other duties.(6)
[[HPE07_fig1_82]]
Lessons learnt
The rapid freeing of staff time surprised us, as did the need to revisit training for all staff. Further training was required once we had discovered the problems that the machine could cause, in order to teach users how to solve them.
Summary
The benefits from implementing automation at other hospitals would be significant. Based on the results of the UK Audit Commission’s Acute Hospital Portfolio national data review, in many trusts a quarter of pharmacists’ time is spent on dispensing and supply. Reducing pharmacist time spent on dispensing and supply in line with good practice sites would release around 635 WTE pharmacists to provide clinical pharmacy services, greater than the total vacancies for trusts identified by trusts (as it stood on 31 March 2001). A reduction in technician time spent on dispensing by 25% would again release staff numbers greater than the vacancies (on 31 March 2001). Based on these figures the return on initial investment would be only two years. Releasing scarce pharmacy staff resources for more effective deployment in ward-based services would reduce the risk to patients of incorrect and ineffective use of medicines and the cost of subsequent litigation to the NHS as well as freeing up bed capacity.
It seems that automation is “catching on” in the UK as Wales announced three pilot sites on 2 January 2003, which will be evaluated before a full roll-out in Wales. In England a small number of sites have secured locally agreed funding, but no central funds have been made available as yet.
References
- Scott D, on behalf of the NHS Pharmacy Education and Development Committee. National hospital pharmacy vacancy survey 1999.
- Department of Health. Pharmacy in the future – implementing the NHS Plan. London: DoH; 2000.
- Audit Commission. A spoonful of sugar – medicines management in NHS hospitals. London: Audit Commission; 2001.
- Spencer MG, Smith AP. A multicentre study of dispensing errors in British hospitals. Int J Pharm Pract 1993;2:142-6.
- Department of Health. An organisation with a memory. London: DoH; 2000.
- Slee A, Farrar K, Hughes, D. Implementing an automated dispensing system. Pharm J 2002:268;437-8.
