Drug-drug cross contamination in the Swisslog fully automated medication handling system

Abstract

Objective

The risk of drug-drug cross contamination in drug dispensing robots in hospital pharmacies causes cumbersome restraints to be put on the production of the robot for example by scheduling high-risk drugs to be dispensed at the end of the day. However, we were unable to find published data on the matter, and therefore performed a worst-case scenario study to assess the magnitude of the problem.

Methods

We measured dexamethasone residue left on the suction cup after the production of 100 and 400 dexamethasone tablets, and after 20 paracetamol tablets used as a negative control.

Results

We found that 32.9 µg and 49.5 µg of dexamethasone had been transferred to the suction cup in the two experiments. This is approximately 1 per mille of the dexamethasone content in a 40 mg tablet.

Conclusion

We conclude that uncoated dexamethasone does shed measurable residue in the robot. It remains unknown to what extent this residue contaminates the subsequent production.

Introduction

The regional hospital of South-West Jutland in Esbjerg, Denmark, has recently installed a fully automated medication handling system with the brand name Swisslog (Swisslog Italia S.p.a. Maranello, Italy) (hereafter referred to as the robot). The robot carries out the process of picking (PillPicking) and packaging medicines for individual patients and delivering them directly to the ward through a tubing system. The robot is fitted with drug containers (pill boxes). Dispensing, both to individual patients and convenience packaging, is carried out by the robot by transferring tablets from the pill box to a drug sachet (plastic bag). This system is referred to as single drug dispensing. This means that a drug sachet will always only contain one tablet. The robot transfers tablets one at a time using a metal straw with a silicone suction cup at the end. Suction applied to the straw allows tablets to be lifted and transferred from the pill box and into the sachet. While pill boxes are cleaned before each tablet refill, the suction cup and straw are used repeatedly during operation regardless of type of tablets transferred, number of transactions and operation time. The robot manufacturer only vaguely describes that during tablet transfer, some tablet residue may be kept on the suction cup, the straw and its vacuum clip, and therefore recommends a cleaning of the straw and replacing the suction cup after production, especially for dangerous products. The issue of possible cross contamination from one production to the next is addressed in regulatory policies under the auspices of the Danish Medicines Agency. Herein, it is stated that in the process of automated drug dispensing, high-risk drugs like cytostatic drugs and sex hormones, and allergenic drugs like penicillin and cephalosporin must not be used if there is a risk of cross contaminating other drugs.¹ However, there are no studies in the literature, which investigated cross contamination in a system identical to the tested system. In fact, we have not been able to locate any literature at all on the matter, even for comparable systems.

We therefore aimed to quantify any cross contamination in our system from the perspective of a worst-case scenario.

Methods and materials

Four experiments were carried out, each of which were done in triplicate. In the first and second experiments, 100 units and 400 units of dexamethasone ‘Glostrup Apotek’ 40 mg uncoated tablets were dispensed sequentially using the same suction cup (Swisslog PillPicker module, Maranello, Italy). The suction cup was then detached using a pair of sterilised tweezers and put in a test tube and stored at room temperature away from direct sunlight until analysis. To simulate repeated use over a prolonged period, the pill box was vibrated with the tablets in it for 10 min using a laboratory vortex mixer immediately before the experiment. The third experiment served as a negative control and was carried out similarly to the other experiments except for dispensing 20 paracetamol tablets (500 mg) instead of dexamethasone tablets. The experiments were carried out under the recommended operating conditions of the robot. The fourth experiment served as a positive control and was carried out by preparing the suction cup by shaking it for 10 s in a sachet with two grounded 40 mg dexamethasone tablets before placing it in the test tube.

The dexamethasone content in each test tube containing a suction cup was determined using liquid chromatography and tandem mass spectrometry (LC-MS/MS). The LC-MS/MS system consisted of an Ultimate 3000 ultra-high-performance liquid chromatography (UHPLC) system connected to a TSQ Quantiva Triple Quadrupole mass spectrometer with heated electrospray ionisation (Thermo Fisher Scientific, San Jose, California, USA). Data acquisition was performed in single reaction monitoring mode. Dexamethasone was quantitated by positive ionisation at the transition from (m/z) 393.2 to 373.1, and with (m/z) 393.2 to 355.0 as a qualifier trace. Cortisol-d4 was used as internal standard and quantitated by negative ionisation and monitored at (m/z) 411.3–335.3. The analytical column was a Kinetic C18 column, 100×4.6 mm (2.6 µm) (Phenomenex, Torrance, California, USA) eluted with 0.1 M formic acid:acetonitrile (60:40 v/v%). The calibration curve and calculation of the sample concentration were based on the area ratio of the analyte/isotope labelled internal standard. The sample preparation was performed with addition of 2 mL methanol and 50 µL of 100 ng/mL cortisol-d4 (as internal standard) to each test tube containing the suction cup. The test tube was whirl-mixed for 15 min, after which the liquid was transferred to a clean test tube and placed on a heating block and evaporated to dryness at 40°C under a gentle flow of nitrogen. The sample was resuspended in 200 µL of acetonitrile:Milli-Q-treated water (40:60 v/v%). A volume of 10 µL of this resuspension was diluted further to 10 mL with acetonitrile:Milli-Q-treated water and whirl-mixed before injection of 10 µL into the LC-MS/MS system. The calibration curve ranged from 10 ng/mL to 3000 ng/mL, and was fitted with a linear regression, with a weighting index of 1/x; R² >0.996. The accuracy expressed as recovery ranged from 96% to 106%, and the interday precision was <6.5%. A volume of 10 µL of the supernatant was injected onto the LC-MS/MS system. The limit of quantification was 10 ng/mL.

Results

The amount of dexamethasone found on the suction cup used for dispensing was 28.3 µg, 33.4 µg and 36.9 µg after dispensing 100 tablets and 45.5 µg, 50.6 µg and 52.4 µg after dispensing 400 tablets (figure 1). The positive control ranged up to 515 µg and probably reflects a point of saturation, that is, how much dexamethasone that maximally can be left on the suction cup, and therefore potentially be transferred to the subsequent production. No dexamethasone was detected on the suction cup used for dispensing of paracetamol tablets in the negative control experiment. The amount of dexamethasone residue left on the suction cup compared with the content in each tablet (40.000 µg) was approximately 0.8 and 1.2 per mille for the two experiments, respectively.

Figure 1.

Absolute amount of dexamethasone found on the suction cup after transfer of a number of dexamethasone tablets (40 mg). The three measurements on the right (marked with*) are the result of the positive control experiment.

Discussion

We wondered whether cross contamination from tablets would take place in a fully automated medication handling system. We were unable to find relevant studies in the literature about this issue and therefore designed a simple experiment based on a worst-case scenario to obtain an impression of the magnitude of the problem. The experiment demonstrates that residual particles from uncoated dexamethasone tablets are indeed transferred to the suction cup in detectable amounts. Moreover, the results indicate that there might be a dose relationship with a plateau because the amount of dexamethasone on the suction cup rose from approximately 33 µg after 100 units of dexamethasone tablets to 50 µg after 400 units. This study is the first to demonstrate evidence of transfer of an active drug substance from a tablet to a suction cup with tablet contact in a fully automated drug handling system. This is important because contamination of the suction cup is a prerequisite for proper cross contamination where tablet material from one tablet is transferred to the surface of other tablets. The results of this study cannot be generalised to include other drugs than uncoated dexamethasone tablets. However, it serves the purpose of demonstrating that cross contamination may occur and cannot generally be ruled out. Contamination in the range of 1 per mille is negligible in the vast majority of clinical situations but there might be rare situations where the tolerance for contamination is particularly low. Examples of low tolerance situations include drugs with large allergenic potential such as penicillins, drugs with high potency such as digoxin, and highly toxic drugs such as anticancer drugs, and some immunosuppressive agents. Evidence from the present study is insufficient as the basis for practical general advice for users of the the Swisslog system. However, it does lend some support to a conservative approach which involves a regimen for replacing the suction cup at certain intervals, and/or after dispensing critical drugs, and possibly also for cleaning pill boxes and other surfaces which are in direct contact with the tablets being dispensed. More studies are certainly warranted—we would like to see more studies of a similar nature both from hospital pharmacies and from the industry manufacturing these systems.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statements

Patient consent for publication

Not required.