Long-standing problems were encountered with a traditional one stop dispensing-based medicine supply system at a UK University Teaching Hospital. As a solution to these issues, a project evaluating closed loop medicines administration is presented
Graeme Hall BSc Dip Clin Pharm MRPharmS
Deputy Chief Pharmacist
Tim Bourne MBBS FRCA
Consultant Anaesthetist/Associate Medical Director
OptiMeD Project Manager
Jeanette Halborg RN BSc
Head of Nursing Clinical Support and Imaging,
University Hospitals of Leicester NHS Trust, UK
The University Hospitals of Leicester NHS Trust (UHL) is one of the UK’s largest hospitals and, in respect to emergency medical admissions, it is the busiest hospital in England. The hospital is based over three sites and comprises of 1791 beds. During 2013/14 the Trust had 151,600 Emergency Department attendances and 556,179 occupied bed days spread over 100 wards.
As can be seen above, our university teaching hospital in the centre of England is large and complex. This presents significant challenges to all those staff involved in medicines use processes in ensuring medicines are in the right place and the right time for patients who require them.
Adopting one stop dispensing
In 2001, the UHL pharmacy service adopted one stop dispensing (OSD) in line with the best practice of the day. Lockers were installed by each patient bed and, in 2005, ROWA speedcase – original pack picking robots – were installed in our three main dispensaries. The expected benefits of these changes were:
- Savings to the health community as hospitals were supplying a larger quantity of medicines on discharge (and hospital medicines’ prices were lower).
- A reduction in the volume of dispensing as one pack would be supplied to the ward labelled for the patient and this pack would be used for discharge.
- The introduction of robots to pick in the dispensaries, along with reduced dispensing volumes, would allow pharmacy staff to leave the department and head for the wards.
- Drug selection would be easier from the patient locker rather than a large trolley, so nursing medication rounds would be quicker and the number of administration errors would fall.
In summary, it was expected that nursing/pharmacy time would be saved and medicines would be administered faster and more accurately. Have we realised these benefits? The answer from UHL is a resounding no! Why have the expected benefits not been achieved and, further, is Leicester unique in this respect?
Since 2001, acute hospitals have seen large increases in activity. Length of stay has decreased even though patients generally have become older, frailer and with co-morbidity, requiring complex medicine regimes. The average unplanned admission patient in UHL is seen in the Emergency Department, then moves to an assessment unit and then to a base ward. Often, because of bed pressures, the base ward may be on a different site and may require more than one move before the patient goes to the discharge lounge and then home.
Once a patient has been seen by pharmacy staff and a OSD supply has been organised, ensuring this ends up in same place as the patient and remains with the patient throughout their journey has proven to be a major challenge. These challenges have effectively wiped out any reduction in dispensing activity as medicines are dispensed and then, often, re-dispensed.
The concept of OSD did not take into account that in the acute setting a medicine is often commenced, switched to an alternative, or doses escalated or reduced depending on response. In this event, departments equipped with pack picking robots struggle to cope with storage of part packs. This is a major problem in UHL. The volume of returns and the time taken to check, count and return is a huge waste of manpower which we often don’t have available. Estimates at UHL suggest we are disposing of well over £600,000 worth of medicines annually which could be saved with a different logistical model. The OSD model has resulted in a lot of re-working and wastage, not a sign of a quality, lean system. We don’t believe we are unique in this respect; most hospital pharmacists we talk to have very similar issues.
Perhaps we are able to discharge patients more quickly? Most hospitals report pharmacy is still often perceived as a blocker to discharge. In fact, sorting through the patient’s locker contents, counting part used packs and patient’s own medications, topping up supplies and re-labelling when doses have changed is, in our minds, no quicker than dispensing from scratch on the majority of occasions.
So if OSD hasn’t delivered for pharmacy, how is it doing delivering for the nursing staff on wards? At UHL, medicines rounds are taking longer. Again we are not unique in this. Studies involving the following of nurses have shown medicines rounds can take as long as 104 minutes to give medicines to as few as six patients. These studies showed that, in order to locate the required medicines, nurses were looking in up to three different places.1
Studies also report administration error rates in adult patients running at 5.6% of inpatient administrations in English hospitals in 2007,2 six years after the wide spread introduction of OSD.
UHL DATIX reports of administration errors paint a similar picture. Another big issue is the number of missed doses which are running in excess of 10% in audits we have carried out. The inability to locate medicines during the pressured medicine round often means the patient does not receive a dose at least until the next round. We do not, therefore, believe that medication errors have been reduced by the introduction of labelled patient pack dispensing, with storage in bedside lockers.
In our view, OSD has not only failed to deliver the benefits expected, it has actually led to increased problems. These include an increase in re-dispensing, vastly increased recycling activity and wastage of medicines in the pharmacy. In the ward, it has led to hard pressed nursing staff spending an increasing amount of time on medicines rounds and increased the number of occasions that medicines doses are missed due to inability to locate supplies.
What are the options?
If OSD has not worked what are the options? In the 21st century IT and robotics have the potential to offer joined-up solutions. At UHL we are taking a re-look at unit dose dispensing. The reason why we are so interested in this approach is because of what our partners, the Italian company, Ingegneria Biomedica Santa Lucia SpA a Socio Unico (IBSL), have developed. This is a fully closed loop system for medicines supply that is working successfully in over 20 hospitals in Italy.
The heart of the system is the SOFIA® software which coordinates everything from e-prescribing to re-packaging robots, logistics and administration, tracing 100% of medicines entering a ward. Integrated automated storage cabinets on wards fill medicines trolleys automatically with personal patient therapy and medicines undergo triangulated bar code identification of prescription, patient and medicine before administration.
There are many improvements in the IBSL system over systems introduced into the UK over the last 15 years, and indeed, over those currently in use in the US and Europe.
The most important difference for the acute hospital setting is the automated ward storage for unit doses (some 80% of all ward medicines can be converted into unit doses, including capsules, syringes, vials, and patches). These cabinets are stocked based on current and past usage patterns, akin to constantly reviewing stock list to know what is likely to be required the following week. In addition the medicines are ‘assigned’ to a patient immediately prior to the medicines round by the software from current prescriptions. This is done in real time, and the specific patient unit doses are placed in a pre-designated draw of the drug trolley.
At the point of administration, the patient bar coded bracelet is scanned which identifies the patient and the nominated draw. The medicines packet for the patient from the draw is then scanned and matched to the prescription giving the go-ahead to administer the medicine. All batch numbers and expiry dates for each and every medicine unit dose is recorded in the prescription record automatically. This offers a vast improvement in traceability and accountability for medicines at ward level. It also allows drug recalls down to individual unit dose level as well as having visibility of where each individual dose has been administered.
Costs have always been a concern with unit dose systems. In this respect, the IBSL system is innovative. In Italy, a central plant has been licensed, taking supply requirements from hospitals contracting with IBSL and using electronic prescribing information supplied to the SOFIA® software. The plant handles the repackaging of unit doses and logistics to the hospitals (up to 500km away). With this concept the investment is minimised and costs spread over a number of contracting hospitals. There appears to be no reason why a similar model could not succeed in the UK.
The project team at UHL includes doctors, nurses, pharmacy and IT staff. We see this as a multidisciplinary quality improvement and patient safety project.
The build of a unit dose laboratory within the pharmacy department of our Leicester General Hospital site has recently been completed. Administration ‘go live’ will commence on the four renal project wards in March. Evaluation will be vital for decisions on further roll out across UHL and potentially beyond. To achieve this, we have commissioned Loughborough University (LU) to look at this from both a business re-engineering perspective as well as a human factors change angle. We expect to have results by the end of 2015. To ensure the independence of the evaluation, the cost of commissioning at Loughborough University has been met by the East Midlands Academic Health Science Network.
For historical reasons, unit dose systems are viewed very sceptically in the UK, particularly from an affordability perspective. With what we know about the quality, risks and waste with our current supply systems, our view is that we cannot afford to continue with OSD.
- McLeod M. Medication administration processes and systems – exploring the effects of systems-based variation on the safety of medication administration in the UK National Health Service [PhD thesis]. London: University College London, The School of Pharmacy; 2013.
- McLeod M, Franklin BD, Barber N. Methodological variations and their effects on reported medication administration error rates. BMJ QualSaf 2012;10:1136.