Abstract
French preparation guidelines state that pharmacy staff who manipulate cytotoxic drugs have to follow specific training. In order to assess the pharmaceutical assistants’ skills and knowledge, we developed a “Cytotoxic Preparation Centralized Unit (CPCU) of errors,” derived from the Canadian concept of “Chamber of horrors.” A table listing 20 mistakes to track down was created and each pharmaceutical assistant spent 20 min in the “CPCU of errors” with the pharmacist, who wrote down the spotted mistakes in real time. Among the 21 trained pharmaceutical assistants, 15 were evaluated. On average, 11.9 mistakes on 20 were detected. The lowest score was 7 spotted errors on 20 and the highest was 16 on 20. Those results should be qualified depending on pharmaceutical assistants’ years of experience in the preparation of chemotherapy. Those results may be explained by the way the role-playing was conducted. The simulation was not conducted during an actual preparation using the usual equipment. One of the major obstacles was the difficulty to clear some time for this project because its realization required a full-time pharmacist and the referring pharmaceutical assistant in addition to the evaluated pharmaceutical assistants. Overall, the staff feedback was positive and the role-playing led to a reminder of theoretical knowledge and the good use of some devices. It would be interesting to develop this type of project through a regional oncology network to create a medium that can be used by other hospitals.
Introduction
French preparation guidelines state that “pharmacy staff who manipulate cytotoxic drugs has to be qualified and regularly trained. An initial and in-service training about the nature and the hazards of the drugs and how to protect from them is given to the staff.” 1 In our hospital, each pharmaceutical assistant (PA) is given both a theoretical and practical training session before taking part in the cytotoxic preparation activity. During one day, a pharmacist presents them the activity, the drugs prepared in the unit and the devices’ functioning. Then, during three days, the supervisor (i.e. the PA who manages the preparation activity and is in charge of the practical training) teaches them how to use safely all the preparation equipment (needles, syringes, portable infusers, etc.), without contaminating the preparation and the work environment. Over the first few months of 2015, a growing number of errors related to the preparation of chemotherapy infusions have been declared through the undesirable events declaration tool (wrong labeling, preparation in the wrong solvent, non-respect of the dressing rules, etc.). Those errors were discussed during the periodic meetings with the staff but their causes (lack of attention, lack of knowledge, practice’s drift, etc.) were not investigated. The aim was to confront the PAs with those errors in order to determine the causes through a “Cytotoxic Preparation Centralized Unit (CPCU) of errors,” derived from the Canadian concept of “Room of horrors.” This project, initiated by the Canadian Patient Safety Institute, consists in recreating a patient’s room where there are errors and potential risks for the patient everywhere. The staff (doctors, nurses, pharmacists, students, etc.) and the patients are invited to visit this room to detect the errors and suggest solutions. 2 The original concept is widely used in French hospitals to improve the quality and safety of care.3–5 This adaptation enables the assessment of the PAs’ skills and knowledge and can be part of the regular training required by the French preparation guidelines.
Material and method
Detected errors for each PA.
X = detected error; Level 1 : less than a year of experience, Level 2 : 1 to 2 years of experience, Level 3 : more than 2 years of experience.
Technical error – Jewels stick out of the sterile outfit.
Technical error – The technician does not use a swab when pulling out the needle from the vial.
Practical error – The solvent is incompatible with the drug.
Practical error – The infusion set does not match the procedure.
Cytobox© device.
Results
Among the 21 trained PAs, 15 have been evaluated from September 2015 to September 2016. Each detected error counts for 1 point. On average, 11.9 mistakes on 20 were detected. The lowest score was 7 detected errors on 20, the highest was 16 on 20. All the results are summed up in Table 1.
Discussion and conclusion
The same kind of experiment has been conducted at the hospital of Lens (France). The global results were similar to those described in this paper, with an average score of 9.4 spotted errors on 18. 6 However, those results should be qualified depending on PAs’ years of experience in the preparation of chemotherapy. The lowest results correspond to the PAs who have less than a year of experience (level 1). Levels 2 (one to two years of experience) and 3 (over two years of experience) PAs’ average score are similar (11.6 versus 11.75). However, technical errors were detected more commonly amongst level 2 PAs, while level 3 PAs tracked down more practical mistakes.
Three of the 20 items were considered as major errors and were expected to be spotted by each PA:
The incompatibility between the drug and the solvent (a), The presence of two different drugs in the same tray for one preparation (b) The presence of two unlabeled preparations in the exit airlock at the same time (c).
Indeed, these kinds of issues might lead to severe consequences for the patient (precipitation of the drug leading to a potential embolism, excessive toxicity, etc.). The detection rates for the major errors (a) and (b) were satisfactory (detected respectively by 10 and 15 PAs on 15) unlike that of the major error (c) (detected by 8 PAs on 15).
Those results may be explained by the way the role-play was conducted. The simulation was not conducted during an actual preparation using the usual equipment. The daily activities in the center where the research was conducted usually uses Cytobox© devices, with airlocks, whereas for the purpose of this study, a laminar flow, with no airlocks, was used for the simulation. The airlocks were depicted by some tape placed in the laminar flow, but still, this is a major difference that disturbed the PAs. The major drawback of the dynamic situation assessment is that mistakes have to be detected on the spot, there is no room for a playback to note further ones. Moreover, the supervisor started over the manipulation for each colleague and did not emphasize her gestures the same way. By example, considering the error number 7, sometimes she grabbed completely the plunger, while sometimes she touched it more discreetly. The PAs did not see exactly the same moves and it induces a bias in the results.
The results are acceptable since only three PAs did not reach the score of 10 on 20. Yet, they can be improved, especially considering that even the PAs who got the highest scores did not spot the three major mistakes. A catch-up session can be proposed to the PAs with the lowest scores, focusing on the mistakes they did not spot. Moreover, as they are working in pairs, one way to secure their practice would be to pair them for daily work with a PA who got a high score and who could spot the errors that their colleague would do in real life, so they can fix it right away and explain them what is wrong with their practice.
One of the major obstacles for conducting this assessment was the difficulty to clear some time for this project because its realization required a full-time pharmacist and the supervisor in addition to the evaluated PAs. Considering the needed time for each evaluation, only nine PAs completed the assessment in the “CPCU of errors” on the scheduled day and the six PAs left were randomly evaluated, when the supervisor was training new colleagues and could spend 20 min on the experiment.
Overall, the staff feedback was positive. This unusual and playful kind of evaluation was appreciated. As it is for error number 3, it enlightened some lack of knowledge about using specific devices and it led to a reminder of the good use of Chemoclave® vial Spike. The main negative comments were the discrepancy between the role-playing and the reality and the lack of immediate feedback after the exercise. This feedback was supposed to be done soon after the evaluation of all PAs during a meeting with all the staff. Yet, the schedule issues postponed it several times, and it was done only after a few weeks.
Considering the huge amount of daily work, it seems hard to monopolize one Cytobox© device during the workday so the experiment can be conducted in real-life conditions. However, as the PAs complained about the lack of realism and the inability to replay the dynamic situation, one solution could be the use of video recording through a device such as a Drugcam®. 7 Drugcam® is a system in which all the preparation process is computerized and the whole manipulation is videotaped. Following the instructions on the computer screen, the technician has to show in front of one camera the drug or the withdrawn volume to trigger the appearance of the next preparation step. Another camera is filming the work zone. Using this device, the “CPCU of errors” could take place in the real CPCU after the opening hours and the dynamic situation could be filmed. The recording could be used in addition to static situations. Then, the obligatory presence of a pharmacist and the supervisor would no longer be necessary for each evaluation and it would ease its realization. This kind of device could also be part of a random evaluation and the non-compliant practices could be debriefed with the whole staff during periodic meetings. It would be interesting to develop this type of project through a regional oncology network to create a medium that can be used by other hospitals.
Footnotes
Acknowledgments
Special thanks to Dr Ariane Blanc (Ottawa, Canada) and Dr Osadebamwen Omosigho (Reading, England) for their contribution.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.