The preparation of individual, ready-to-use doses of injectable medicines and small batches of products, such as dose-banded cytotoxic injections, is often undertaken in pharmacy departments under controlled conditions. Typically, such preparations involve experienced personnel working to validated protocols. However, in most hospitals there is also a need for nurses and doctors to prepare injections for immediate use on wards or in other clinical areas of the hospital. Accurate preparation is important, particularly with cytotoxic drugs where the therapeutic window is narrow and the consequences of both under- and over-dosage can be serious. Similar considerations apply to many other drugs including narcotics, antibiotics or insulin. Good preparation procedures should ensure that the drug and dose are correct, but in reality there are numerous opportunities for error during the preparation process1 (see Box 1).
The preparation of single, patient-specific injections is one of the few operations where final product quality control is not mandatory and is usually not performed. In the pharmacy, accurate preparation relies mainly on in-process checks. This could be a second check by another individual (often known as the ‘four eyes’ approach) either directly or through closed-circuit cameras. Another method is to use in-process gravimetric checks that confirm the quantities of drug and diluent used.
When doses are prepared on wards or in clinical areas, the only available method to determine accuracy is the second check. However, it is well-recognised that the second check method is not associated with high reliability.2
Research evidence suggests that the incidence of errors in prescribing, preparing and administering injectable medicines is higher than for other forms of medicine.3 The consequences of errors with injectable medicines can be serious. One study published in 2012 estimated that inpatient preventable adverse drug events (ADEs) associated with injectable medicines increased the annual US payer costs by $2.7 billion to $5.1 billion, averaging $600,000 in extra costs per hospital.4
Techniques are now available for rapid qualitative and quantitative checks of prepared injections. Pharmacolog AB, Sweden, has developed such a system – DrugLog®.
DrugLog® is a light-weight, portable system comprising a UV-visible light spectrophotometer (absorbing in the range 200–800nm) integrated with a computer and touchscreen (see Figure 1). The DrugLog® software includes two operation modes; a verification programme for confirming the identity and concentration of a prepared injection, and a programme for identifying an unknown injectable solution by comparing the UV spectrum of the test sample with the profiles held in the drug library.
Figure 1: The DrugLog® system
The system design allows for future versions to have the spectrophotometer and the computer separated; for example, the spectrophotometer could be placed inside a laminar flow hood or isolator and wirelessly linked to the computer/touchscreen placed outside
The DrugLog® software can also be interfaced with local networks for integration with existing patient administration systems, hospital information systems, treatment records, electronic prescribing systems and other related databases. The system is connected to the internet via the local Wi-Fi network, which enables remote service and support by Pharmacolog.
The verification programme operates by comparing the spectrum generated by the sample with the reference library of spectra held in the library. It then displays the result ‘traffic light style’ where green corresponds to correct drug and correct concentration, amber represents correct drug and incorrect concentration and red represents incorrect drug or concentration outside the clinically acceptable limits. This style of reporting means that once the library has been set up, DrugLog® can be used by people with no special knowledge of spectrometry. In addition, DrugLog® can be linked to a small printer so that when the check has been completed, a label can be printed to show the drug and the concentration actually measured (see Figure 2).
Figure 2: Printer and label
The identification programme operates by comparing the spectrum generated by the sample with the reference spectra held in the library. It then displays the result as the name of the drug and the concentration (based on the best match) or ‘unknown’ if the drug substance is not found in the library.
The sample size required for analysis is 0.2–0.4ml and the process takes two-three seconds. Several different types of closed cuvette are available including a Luer lock version and one sealed with a septum through which drug solution can be injected (Figure 3).
Figure 3: Different types of closed cuvette
DrugLog® will need to be calibrated for each drug that is added to the library. A minimum of three different concentrations – high, low and medium – must be measured. This is performed onsite by the user after a short period of training.
When using DrugLog®, a reference measurement is performed every day using the injection diluent that is to be used – typically 0.9% sodium chloride. This routine, start-up procedure takes only seconds and ensures that any variations in the lamp or spectrometer are ‘zeroed out’. A change of diluent (for example, to 5% dextrose) requires a further reference measurement using the new diluent.
DrugLog® can only be used for substances that have a unique UV-VIS spectrum. It was originally envisaged that its major use would be for cytotoxic drugs but early adopters have now investigated its use for analgesics, antibiotics and other injections. It cannot be used for intravenous emulsions as UV-VIS spectroscopy requires a clear liquid.
The DrugLog® software holds a record of each test performed and the result. It is, therefore, possible to analyse workflow patterns and individual performance for management purposes.
In-use evaluation of DrugLog®
The accuracy of the DrugLog® system has been evaluated in a study using ten commonly-used cytotoxic drugs (5-fluorouracil, doxorubicin, epirubicin, docetaxel, irinotecan, gemcitabine, paclitaxel, etoposide phosphate, carboplatin, oxaliplatin).5
In order to do this, calibration solutions of 11–20 different concentrations were made for each drug. All the calibration solutions were made using a high-precision electronic pipette. DrugLog® was then used to check the concentrations of a number of standard validation solutions and, finally, to identify and check several compounded cytotoxic infusions.
The results showed that DrugLog® was able to determine concentrations of the ten test drugs with a mean overall error of 3.7% (2.05%–9.72%).
All drugs were correctly identified and the authors commented that one major advantage of the system is that the device itself is never contaminated by the sample, so no special cleaning is needed.
Experiences and impressions
Professor Pascal Bonnabry (Director of Pharmacy, University Hospital of Geneva) and Professor Pascal Odou (Deputy Head of Pharmacy Department, University Hospital of Lille) have undertaken some of the initial evaluations of DrugLog® and they described their experiences and impressions.
At the University Hospital in Lille, France, DrugLog® has been used in the operating theatre to check the concentrations of analgesics prepared by theatre personnel. In the busy, stressful atmosphere of the operating theatre, there is always the possibility of errors when injections are prepared, Professor Odou explains. In the early stages, DrugLog® was used by a pharmacy intern but this was largely a ‘proof of concept’ exercise. In future, it is likely that nurses and doctors will use it routinely to check injections that they have prepared. “It’s the simplicity of the system that appeals to the users – simplicity of operation and simplicity of interpretation”, says Professor Odou.
“It is important to remember that error rates can be high in anaesthesiology”, comments Professor Bonnabry . “We published a study where we took unused syringes from operating theatres and we found that in one-third, the concentrations were in the range ± 10% of the stated value and 4% were ± 100% of the stated value”.
At the University Hospital of Geneva, DrugLog® is used in the pharmacy department. “It is important to have a positive check for products that do not go through a quality control process [before administration]”, explains Professor Bonnabry. When dose-banded ganciclovir injections in bags or syringes are prepared in a campaign dispensing process they are not sent for quality control. The operators now perform a final check on the prepared injections using DrugLog®. Once again, the feedback from the staff indicates that the process is “very quick and very simple”.
During the first three months of use one important error was picked up – one bag contained a double dose of ganciclovir and the next bag contained none at all. Without DrugLog®, this type of error could easily pass unnoticed because all the correct quantities of ingredients have been used.
Both Professor Bonnabry and Professor Odou foresee that DrugLog could be used for much more than checking of compounded cytotoxic doses in the pharmacy. “There is no money for centralised additive services in France and so nurses have to prepare injections on wards; DrugLog® could make this a safe process”, says Professor Odou. In addition, it could be used for training of pharmacy technicians in compounding units or nurses in simulation situations. This would be analogous to the way broth tests are used to assess aseptic technique. It is an approach that has not yet been investigated, he said.
DrugLog® is not a quality control instrument – it does not provide the accuracy of HPLC and it cannot analyse all drugs, emphasises Professor Bonnabry. However, it is ideal for situations where there is no quality control at the end of the process or even good in-process control. “It is important to see it as a device to identify errors rather than an analytical tool”, he adds.
The measurements made by DrugLog® could be accurate to a margin of ± 10%. “You can live with up to ± 15% – more than that and you start to have clinical incidents”, says Professor Bonnabry. “One area where DrugLog® could be useful is in the preparation of individual paediatric doses where dilutions are often required and the risk of errors is high”, he adds.
Pharmacolog has worked closely with Professors Odou and Bonnabry to develop and refine the DrugLog® device in line with users’ needs. “The prototype used open cuvettes and the introduction of the closed cuvette was a major step forward”, says Professor Bonnabry.
The next advance will take into account the fact that some bags of intravenous fluid are over-filled by up to 20%.6 This could result in the final, compounded drug concentration being lower than expected with consequent false negative results. The next version of the software will ask the user to specify the brand of intravenous bag being used and will then adjust the calibrations accordingly, explained Professor Odou.
DrugLog® offers quick and easy quality assurance for compounded intravenous doses. It could significantly improve patient safety when compounding hazardous drugs such as cytotoxic agents or high-risk drugs such as those used in anaesthesiology or intensive care units.
Support for the production of this editorial was provided by Pharmacolog AB
- De Giorgi I et al. Risk and pharmacoeconomic analyses of the injectable medication process in the paediatric and neonatal intensive care units. Int J Qual Health Care 2010;22:170–8.
- Keers RN et al. Understanding the causes of intravenous medication administration errors in hospitals: a qualitative critical incident study. BMJ Open 2015;5:e005948
- National Patient Safety Agency (2007). Patient Safety Alert 20: safer use of injectable medicines. www.nrls.npsa.nhs.uk/resources/type/alerts/?entryid45=59812&p=3 (accessed December 2016).
- Lahue BJ et al. National burden of preventable adverse drug events associated with inpatient injectable medications: Healthcare and medical professional liability costs. Am Health Drug Benefits 2012;5:1–10.
- Feutry F et al. Evaluation of new quality control system for cytotoxics based on UV-visible spectrophotometry (DrugLog®). Short communication, GERPAC, Hyeres, 2015.
- Feutry F et al. Evaluation in vitro des solutions de perfusion présente sur le marché Français. SFPC Congress 2014; Poster number PDM2.