- Split View
-
Views
-
Cite
Cite
Cristina Sottani, Elena Grignani, Enrico Oddone, Beatrice Dezza, Sara Negri, Simona Villani, Danilo Cottica, Monitoring Surface Contamination by Antineoplastic Drugs in Italian Hospitals: Performance-Based Hygienic Guidance Values (HGVs) Project, Annals of Work Exposures and Health, Volume 61, Issue 8, October 2017, Pages 994–1002, https://doi.org/10.1093/annweh/wxx065
- Share Icon Share
Abstract
Antineoplastic drugs (ADs) will continue to represent a potential risk for personnel involved in the handling of these compounds and great concerns have been raised by the presence of ADs in many surveyed workplaces. Eight hospitals were investigated by means of wipe sampling for surface residue determination. Each wipe sample was tested for five ADs considered suitable exposure markers. Cyclophosphamide (CP), gemcitabine (GEM), 5-fluorouracil (5-FU), platinum-containing drugs (Pt), and epi-doxorubicin (EPI) contamination levels were measured in 85 per cent of the studied pharmacies and 93 per cent of outpatient care units (OpCUs). This study showed that 83 out of 349 samples were positive in Pharmacies, this proportion being statistically significant (χ2 = 42.9, p < 0.001). The positive samples provided evidence of at least one substance with levels greater than the limit of detection (LOD). The two most frequently detected substances were Pt (42%) and CP (30%). These accounted for 72 per cent of the whole dataset, followed by 5-FU and GEM. Based on the 90th percentile of wipe sampling data distribution, we suggest hygienic guidance values (HGVs) of 3.6, 1.0, 0.9, and 0.5 ng cm−2 for CP, 5-FU, GEM and Pt, respectively, as the best target levels of the surface contamination load in Italian pharmacies. The approach of proposing guidance values at the 90th percentile of results obtained from workplaces with good hygiene practice was found to be a simple and practical way of controlling occupational exposure. HGVs were challenged in this study as technical threshold limits to benchmark AD residual surface contamination at workplaces.
Introduction
The establishment of permissible exposure limits (PELs) is one of the most important occupational health issues for any workplace, such as hospitals and industries, where employees can potentially be exposed to harmful compounds Connor et al. (2016). Some of these powerful drugs used for chemotherapy have been known to cause cancer Ferlay et al. (2007), reproductive Connor et al. (2014) and developmental problems as well as other adverse effects even at low exposure levels (Cavallo et al., 2005; Dranitsaris et al., 2005; Fransman et al., 2007). A number of studies (Mason et al., 2005; Connor et al., 2010; Davis et al., 2011; Sottani et al. 2012; Kopp et al., 2013; Sessink 2013; Hon et al., 2014; Shcierl et al., 2014) have documented workplace contamination by antineoplastic drugs (ADs) and have resulted in the development of procedures for the safe handling of ADs (Sessink et al., 2013).
NIOSH has developed an Alert (NIOSH, 2004) declaiming that no NIOSH recommended exposure limits (RELs), OSHA PELs, or American Conference of Governmental Industrial Hygienists (ACGIH) limits have yet been established for hazardous drugs in general. For the last couple of decades, surface wipe sampling has been used in health care settings to evaluate workplace contamination. Furthermore, wipe sampling for surface residue of ADs is still the method of choice to characterize potential exposure to these drugs. Nowadays, what is an acceptable level of environmental contamination in terms of health risks for workers involved in handling ADs is still a matter of debate (Sessink et al., 2011).
Across Europe many attempts have been made by researchers to derive hygienic guidance values (HGVs) for ADs (Schierl et al., 2009; Hedmer et al. 2012; Kiffmeyer et al., 2013; Fleury-Sovereign et al., 2015). For German pharmacies, Kiffmeyer et al. (2013) developed a substance-independent guidance value based on the 90th percentile of environmental monitoring results and Schierl et al. (2009) developed HGVs based on the 75th percentile for 5-FU and Pt. In regard to Swedish outpatient care units (OpCUs) and wards, Hedmer et al. (2012) studied HGVs for CP and IF. They have developed the idea that health care workers should be informed and formed by means of HGVs so that they will be able to benchmark their own surface loads as a marker indicator of dermal exposure and, therefore evaluate if exposure had been adequately controlled. Recently, an evaluation of surface contamination during the preparation of chemotherapies in Swiss pharmacies demonstrated that 92% of the pharmacies presented chemical contamination (Fleury-Sovereign et al., 2015). The approach from the Netherlands led to a “traffic-light” model for cyclophosphamide (CP) (Sessink et al., 2011). Based on a database obtained over 20 years of monitoring studies, the authors formulated the idea that since no positive urine samples for CP were found at a contamination level <0.1 ng cm−2, this value may be considered a guidance value for environmental contamination. In Germany, Kiffmeyer et al. (2013) studied the monitoring effects of the wipe sampling strategy and developed “the MEWIP project”. These authors carefully showed that by means of the 90th percentile of the contamination values HGVs can be obtained and demonstrated that this method is a reliable and affordable system for practitioners in pharmacies. Based on the 90th percentile of the compound found in the highest concentration (5-FU with 0.117 ng cm−2), a limit of 0.1 ng cm−2 was established as an adequate HGV. These findings proved that corrective measures in German pharmacies should be taken when surface contamination is >0.1 ng cm−2. Finally, these authors challenged researchers of other European countries to perform the same project. Since surface wipe sampling has been carried out in Italian hospitals over the last 15-years, a database of biological and environmental contamination values was generated. Environmental results from past biannual monitoring programs (2009–2011) were used to generate a dataset of 771 wipe sampling data. The aim of this paper was (i) to report work area contamination in both Pharmacies and OpCUs, (ii) to evaluate the effect of improved work procedures in relation to possible contamination sources, (iii) to assess the HGVs in Italian pharmacies, and (iv) to compare Italian HGVs with those derived from other investigations.
To our knowledge, no studies in Italy have yet been carried out to record the comprehensive contamination levels in pharmacies and patient care units. The institution of a performance-based platform represents an irretrievable step to derive HGVs of the best performance for the safe handling of ADs.
Methods
Experimental design
In this study, results were obtained from monitoring surveys carried out in 2009 and 2011. Eight hospitals located in northern and central Italy and classified with letters A–H were selected for this study. In terms of environmental analyses, wipe sampling strategy was used as the method of choice for some of the most common anticancer drugs that were used as markers of overall surface contamination. These markers included CP, GEM, 5-FU, Pt, and EPI. Pharmacies (Pharm) and OpCUs were sampled to investigate workplace area contamination, in relation to the complexity of working environments, safety measures, and equipment. Pharmacies in hospitals A, C, D, and G were equipped with a room without anteroom, whereas pharmacies in hospitals B, E, F, and H were cleanrooms with anterooms. Annual preparations in the surveyed sites ranged between 2225 and 22725. For each site, the total annual amount of drugs is reported in Table 1. Monitoring programs were carried out according to the rules of the national guidelines (Italian Guidelines G.U. 236, 7.10.1999). Each participant was invited to complete a questionnaire in order to report the number of different chemotherapies prepared, equipment (biosafety cabinet, or BSC, class II or III, and/or isolator), personal protective equipment (PPE), such as types of gloves, gowns, eye or respiratory protection, and work policy. The same questionnaire had been already reported by Sottani et al. (2012) in a previous study and is always adopted in Italian surveillance programs.
. | 2009–2011 . | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | . | Cyclophosphamide . | Gemcitabine . | 5-Fluorouracil . | Platinum compounds . | Epirubicin . | |||||
Hospital . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . |
A | 134 | 795.6 | 33 | 243 | 34 | 270 | 33 | 78 | 33 | 101 | 1 | 3.6 |
B | 218 | 4864.92 | 55 | 540 | 55 | 2,905 | 48 | 1215 | 59 | 202 | 1 | 2.92 |
C | 39 | 5625 | 3 | 337 | 3 | 270 | 16 | 158 | 16 | 86 | 1 | 6.7 |
D | 49 | 502.6 | 1 | 62 | 17 | 140 | 13 | 275 | 17 | 22 | 1 | 3.6 |
E | 76 | 1795.2 | 19 | 241 | nd | 1226 | 19 | 245 | 19 | 72 | 19 | 11.2 |
F | 25 | 5191 | 8 | 641 | nd | 382 | 8 | 4122 | 8 | 38 | 1 | 8 |
G | 156 | 1077.7 | 39 | 121 | 1 | 140 | 39 | 787 | 39 | 27 | 38 | 2.7 |
H | 108 | 3714 | 27 | 1046 | 27 | 137 | 27 | 2495 | 27 | 36 | 0 | 0 |
TOTAL | 805 | 185 | 137 | 203 | 218 | 62 |
. | 2009–2011 . | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | . | Cyclophosphamide . | Gemcitabine . | 5-Fluorouracil . | Platinum compounds . | Epirubicin . | |||||
Hospital . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . |
A | 134 | 795.6 | 33 | 243 | 34 | 270 | 33 | 78 | 33 | 101 | 1 | 3.6 |
B | 218 | 4864.92 | 55 | 540 | 55 | 2,905 | 48 | 1215 | 59 | 202 | 1 | 2.92 |
C | 39 | 5625 | 3 | 337 | 3 | 270 | 16 | 158 | 16 | 86 | 1 | 6.7 |
D | 49 | 502.6 | 1 | 62 | 17 | 140 | 13 | 275 | 17 | 22 | 1 | 3.6 |
E | 76 | 1795.2 | 19 | 241 | nd | 1226 | 19 | 245 | 19 | 72 | 19 | 11.2 |
F | 25 | 5191 | 8 | 641 | nd | 382 | 8 | 4122 | 8 | 38 | 1 | 8 |
G | 156 | 1077.7 | 39 | 121 | 1 | 140 | 39 | 787 | 39 | 27 | 38 | 2.7 |
H | 108 | 3714 | 27 | 1046 | 27 | 137 | 27 | 2495 | 27 | 36 | 0 | 0 |
TOTAL | 805 | 185 | 137 | 203 | 218 | 62 |
nd = not determined.
. | 2009–2011 . | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | . | Cyclophosphamide . | Gemcitabine . | 5-Fluorouracil . | Platinum compounds . | Epirubicin . | |||||
Hospital . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . |
A | 134 | 795.6 | 33 | 243 | 34 | 270 | 33 | 78 | 33 | 101 | 1 | 3.6 |
B | 218 | 4864.92 | 55 | 540 | 55 | 2,905 | 48 | 1215 | 59 | 202 | 1 | 2.92 |
C | 39 | 5625 | 3 | 337 | 3 | 270 | 16 | 158 | 16 | 86 | 1 | 6.7 |
D | 49 | 502.6 | 1 | 62 | 17 | 140 | 13 | 275 | 17 | 22 | 1 | 3.6 |
E | 76 | 1795.2 | 19 | 241 | nd | 1226 | 19 | 245 | 19 | 72 | 19 | 11.2 |
F | 25 | 5191 | 8 | 641 | nd | 382 | 8 | 4122 | 8 | 38 | 1 | 8 |
G | 156 | 1077.7 | 39 | 121 | 1 | 140 | 39 | 787 | 39 | 27 | 38 | 2.7 |
H | 108 | 3714 | 27 | 1046 | 27 | 137 | 27 | 2495 | 27 | 36 | 0 | 0 |
TOTAL | 805 | 185 | 137 | 203 | 218 | 62 |
. | 2009–2011 . | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | . | Cyclophosphamide . | Gemcitabine . | 5-Fluorouracil . | Platinum compounds . | Epirubicin . | |||||
Hospital . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . | (n) . | (g) . |
A | 134 | 795.6 | 33 | 243 | 34 | 270 | 33 | 78 | 33 | 101 | 1 | 3.6 |
B | 218 | 4864.92 | 55 | 540 | 55 | 2,905 | 48 | 1215 | 59 | 202 | 1 | 2.92 |
C | 39 | 5625 | 3 | 337 | 3 | 270 | 16 | 158 | 16 | 86 | 1 | 6.7 |
D | 49 | 502.6 | 1 | 62 | 17 | 140 | 13 | 275 | 17 | 22 | 1 | 3.6 |
E | 76 | 1795.2 | 19 | 241 | nd | 1226 | 19 | 245 | 19 | 72 | 19 | 11.2 |
F | 25 | 5191 | 8 | 641 | nd | 382 | 8 | 4122 | 8 | 38 | 1 | 8 |
G | 156 | 1077.7 | 39 | 121 | 1 | 140 | 39 | 787 | 39 | 27 | 38 | 2.7 |
H | 108 | 3714 | 27 | 1046 | 27 | 137 | 27 | 2495 | 27 | 36 | 0 | 0 |
TOTAL | 805 | 185 | 137 | 203 | 218 | 62 |
nd = not determined.
Wipe sampling strategy and analyses
A predetermined wipe sampling scheme with selected surface areas in pharmacies and OpCUs was studied. To design a standard scheme, three pharmacy locations were selected: biosafety cabinets (BSCs), work areas, and logistic rooms. There were different types of surfaces such as stainless steel for the worktop of BSC and vitro steel™ for the work-benches. Non-smooth surfaces such as vinyl and linoleum flooring™ were used for floors in cleanrooms, OPCUs, and wards. The samples were collected using a method suitable to investigate surface contamination for CP, ifosfamide, and anthracyclines previously published by Sottani et al. (2010). The surfaces were wiped thoroughly with four Kleenex professional wipes (10 × 10 cm; Kimberly-Clark®, Irving, TX, USA), which had been wetted with 5.0 ml of pure water. This wipe media was used to sample and remove CP, GEM, 5-FU and Epi-Dox, whereas a solution of acetonitrile (ACN) and water (20:80, v/v) with 0.1% formic acid was used to sample platinum coordinated compounds (Fleury-Souverien et al., 2015). Each single wipe was tested for all five drugs. According to OSHA guidelines (OSHA, 2001), the acceptable extraction efficiency (EE) is >75%, and preferably >90%. In our study, different types of surfaces were wipe-sampled. They ranged from smooth (non-porous) to less smooth (linoleum flooring); therefore, the wipe media might have removed the drugs less effectively. In this circumstance, a EE of >75% was considered an acceptable result. These data were also reported in a previous and more detailed study by Sottani et al., (2007). Samples were collected by wiping in two different directions (up and down, right and left) inside a plastic template that was used to define the area (100 cm2) and then removed after each sampling. Large sampling areas inside the biological safety cabinet were multi sampled and wipe samples in the middle, right and left areas of the BSC bench top were always taken. Small areas such as door handles, weighing scales, and phone receivers were accurately measured. In all investigated hospitals, the pharmacy personnel were requested to wear nitrile gloves and polypropylene gowns during the compounding of ADs in BSC. Therefore, the strategy of environmental monitoring was based on wiping the external surface of the finger areas of the gloves. The list of the sampled spots for the different locations is detailed in Table 2. After sampling, the wipe samples were placed in borosilicate glass bottles (50 ml) and stored at 4°C until the analyses. The samples were collected by technicians who performed the same procedure in all investigated hospitals. Quantitative analyses for drugs were based on high-performance liquid chromatography tandem mass spectrometry (LC-MS/MS) using ions produced via collision-induced fragmentation, whereas Pt was measured by inductively coupled plasma mass spectrometry (ICP-MS). Briefly, ADs measurements in wipe samples were obtained with a lower limit of detection (LOD) of 6.0 ng per sample for CP and GEM, 100 ng per sample for 5-FU, 50 ng per sample for Epi-Doxo, and 1 ng per sample for Pt (Micoli et al., 2001; Sottani et al., 2007; Nussbaumer et al., 2010). According to previous studies, a positive Pt-sample (oxali-carbo and cis platinum compounds) was defined when there was a contamination level >0.01 ng cm−2 (10 pg cm−2) since platinum is also present in our environment and tests have shown that interferences from other sources (catalytic converter) can affect wipe sample results (Schierl et al., 2009; Kopp et al., 2013).
Pharmacy areas . | |
---|---|
Sampling spots | |
BSC | Bench surface (middle, right and left) |
Airfoil of BSC or protection glass (inner part) | |
Pass-through inside and outside BSC or pharmaceutical isolator | |
Neoprene isolator gloves (inner part) | |
Nitrile gloves (outer part) | |
Work area | Floor in front of BSC |
Floor at 1 m far BSC | |
Floor at 5 m far BSC | |
Door handle in the cleanroom | |
Floor in anteroom | |
Door handle of refrigerator | |
Weighing scale | |
Phone receiver | |
IV infusion bags | |
Drug vials | |
Logistic room | Countertops |
Storage shelves | |
Computer mouse/keyboard | |
Tray for transferring formulations | |
Outpatient care units | |
Sampling spots | Door handles |
Floor in patient room | |
Floor in restroom | |
Infusion pump | |
Pole for IV infusion bags | |
Floor in front of pole | |
Countertops | |
Furniture in patient room |
Pharmacy areas . | |
---|---|
Sampling spots | |
BSC | Bench surface (middle, right and left) |
Airfoil of BSC or protection glass (inner part) | |
Pass-through inside and outside BSC or pharmaceutical isolator | |
Neoprene isolator gloves (inner part) | |
Nitrile gloves (outer part) | |
Work area | Floor in front of BSC |
Floor at 1 m far BSC | |
Floor at 5 m far BSC | |
Door handle in the cleanroom | |
Floor in anteroom | |
Door handle of refrigerator | |
Weighing scale | |
Phone receiver | |
IV infusion bags | |
Drug vials | |
Logistic room | Countertops |
Storage shelves | |
Computer mouse/keyboard | |
Tray for transferring formulations | |
Outpatient care units | |
Sampling spots | Door handles |
Floor in patient room | |
Floor in restroom | |
Infusion pump | |
Pole for IV infusion bags | |
Floor in front of pole | |
Countertops | |
Furniture in patient room |
Pharmacy areas . | |
---|---|
Sampling spots | |
BSC | Bench surface (middle, right and left) |
Airfoil of BSC or protection glass (inner part) | |
Pass-through inside and outside BSC or pharmaceutical isolator | |
Neoprene isolator gloves (inner part) | |
Nitrile gloves (outer part) | |
Work area | Floor in front of BSC |
Floor at 1 m far BSC | |
Floor at 5 m far BSC | |
Door handle in the cleanroom | |
Floor in anteroom | |
Door handle of refrigerator | |
Weighing scale | |
Phone receiver | |
IV infusion bags | |
Drug vials | |
Logistic room | Countertops |
Storage shelves | |
Computer mouse/keyboard | |
Tray for transferring formulations | |
Outpatient care units | |
Sampling spots | Door handles |
Floor in patient room | |
Floor in restroom | |
Infusion pump | |
Pole for IV infusion bags | |
Floor in front of pole | |
Countertops | |
Furniture in patient room |
Pharmacy areas . | |
---|---|
Sampling spots | |
BSC | Bench surface (middle, right and left) |
Airfoil of BSC or protection glass (inner part) | |
Pass-through inside and outside BSC or pharmaceutical isolator | |
Neoprene isolator gloves (inner part) | |
Nitrile gloves (outer part) | |
Work area | Floor in front of BSC |
Floor at 1 m far BSC | |
Floor at 5 m far BSC | |
Door handle in the cleanroom | |
Floor in anteroom | |
Door handle of refrigerator | |
Weighing scale | |
Phone receiver | |
IV infusion bags | |
Drug vials | |
Logistic room | Countertops |
Storage shelves | |
Computer mouse/keyboard | |
Tray for transferring formulations | |
Outpatient care units | |
Sampling spots | Door handles |
Floor in patient room | |
Floor in restroom | |
Infusion pump | |
Pole for IV infusion bags | |
Floor in front of pole | |
Countertops | |
Furniture in patient room |
Statistical analysis
The 50th, 75th, and 90th percentiles of the area contamination were used to describe our results because data failed to show any normal distribution. For the determination of percentiles, values below the LOD were set as ½ LOD. The non-parametric Mann–Whitney test, analogous for unpaired t test, was applied to compare pharmacy and OpCU contamination levels. Similarly, the chi-squared test (χ2 test) or the Fisher’s exact test was used to investigate the meaning of the different percentages of positive results obtained among the hospitals. A sample was regarded as positive when at least one value of all five investigated drugs was found to exceed the LOD. Moreover, the χ2 test was applied with Bonferroni’s correction for multiple comparisons in order to determine statistical differences between pharmacies and OpCUs. Logistic regression models were used to evaluate the role of factors such as hospitals, Pharmacies, and OpCUs associated with presence/absence of contamination (response variable). All of the analyses were made using STATA® 12.
Results
We grouped our results by hospitals, pharmacies, and OpCUs. Results and statistical analysis were mainly focused on pharmacies and OpCUs because the practical arrangements, safety regulations, and working procedures were greatly different between pharmacies and OpCUs. In total, 771 results were obtained because each wipe sample was tested for all five drugs. In Table 3, the total number of values, the number of determinations above the LOD and the percentage of positive values are shown. The analysis of the overall data showed that the two most frequently detected substances in all eight investigated hospitals were Pt (42%) and CP (30%). These accounted for 72% of the whole dataset, followed by 5-FU (14%) and GEM (13%). In this Table, 349 analyses concerned pharmacies (45%) and 422 concerned OpCUs (55%). The percentages of positive samples were found to be slightly higher in OpCUs (26%) than in Pharmacies (24%). Table 4 summarizes the results for each AD in relation to the sampling locations. The number of positive results out of the total number of determinations and the percentages of positive results are shown in Table 4. The 50th, 75th, and 90th percentiles of area contamination levels (ng cm−2) are also reported in the same table. In this study, the 90th percentile of the data distribution was chosen as a drug-dependent guidance value from pharmacies with good hygiene practice.
. | Hospitals . | Pharmacies . | OpCUs . | ||||||
---|---|---|---|---|---|---|---|---|---|
. | N . | >LOD . | % . | N . | >LOD . | % . | N . | >LOD . | % . |
A | 99 | 10 | 10 | 72 | 6 | 8 | 27 | 4 | 15 |
B | 215 | 43 | 20 | 116 | 18 | 16 | 99 | 25 | 26 |
C | 38 | 11 | 29 | 18 | 8 | 44 | 20 | 3 | 15 |
D | 47 | 11 | 23 | 23 | 6 | 26 | 24 | 5 | 20 |
E | 76 | 18 | 24 | — | — | — | 76 | 18 | 24 |
F | 32 | 8 | 25 | — | — | — | 32 | 8 | 25 |
G | 156 | 47 | 30 | 36 | 6 | 17 | 120 | 41 | 34 |
H | 108 | 45 | 42 | 84 | 39 | 46 | 24 | 6 | 25 |
Total | 771 | 193 | 25 | 349 | 83 | 24 | 422 | 110 | 26 |
. | Hospitals . | Pharmacies . | OpCUs . | ||||||
---|---|---|---|---|---|---|---|---|---|
. | N . | >LOD . | % . | N . | >LOD . | % . | N . | >LOD . | % . |
A | 99 | 10 | 10 | 72 | 6 | 8 | 27 | 4 | 15 |
B | 215 | 43 | 20 | 116 | 18 | 16 | 99 | 25 | 26 |
C | 38 | 11 | 29 | 18 | 8 | 44 | 20 | 3 | 15 |
D | 47 | 11 | 23 | 23 | 6 | 26 | 24 | 5 | 20 |
E | 76 | 18 | 24 | — | — | — | 76 | 18 | 24 |
F | 32 | 8 | 25 | — | — | — | 32 | 8 | 25 |
G | 156 | 47 | 30 | 36 | 6 | 17 | 120 | 41 | 34 |
H | 108 | 45 | 42 | 84 | 39 | 46 | 24 | 6 | 25 |
Total | 771 | 193 | 25 | 349 | 83 | 24 | 422 | 110 | 26 |
. | Hospitals . | Pharmacies . | OpCUs . | ||||||
---|---|---|---|---|---|---|---|---|---|
. | N . | >LOD . | % . | N . | >LOD . | % . | N . | >LOD . | % . |
A | 99 | 10 | 10 | 72 | 6 | 8 | 27 | 4 | 15 |
B | 215 | 43 | 20 | 116 | 18 | 16 | 99 | 25 | 26 |
C | 38 | 11 | 29 | 18 | 8 | 44 | 20 | 3 | 15 |
D | 47 | 11 | 23 | 23 | 6 | 26 | 24 | 5 | 20 |
E | 76 | 18 | 24 | — | — | — | 76 | 18 | 24 |
F | 32 | 8 | 25 | — | — | — | 32 | 8 | 25 |
G | 156 | 47 | 30 | 36 | 6 | 17 | 120 | 41 | 34 |
H | 108 | 45 | 42 | 84 | 39 | 46 | 24 | 6 | 25 |
Total | 771 | 193 | 25 | 349 | 83 | 24 | 422 | 110 | 26 |
. | Hospitals . | Pharmacies . | OpCUs . | ||||||
---|---|---|---|---|---|---|---|---|---|
. | N . | >LOD . | % . | N . | >LOD . | % . | N . | >LOD . | % . |
A | 99 | 10 | 10 | 72 | 6 | 8 | 27 | 4 | 15 |
B | 215 | 43 | 20 | 116 | 18 | 16 | 99 | 25 | 26 |
C | 38 | 11 | 29 | 18 | 8 | 44 | 20 | 3 | 15 |
D | 47 | 11 | 23 | 23 | 6 | 26 | 24 | 5 | 20 |
E | 76 | 18 | 24 | — | — | — | 76 | 18 | 24 |
F | 32 | 8 | 25 | — | — | — | 32 | 8 | 25 |
G | 156 | 47 | 30 | 36 | 6 | 17 | 120 | 41 | 34 |
H | 108 | 45 | 42 | 84 | 39 | 46 | 24 | 6 | 25 |
Total | 771 | 193 | 25 | 349 | 83 | 24 | 422 | 110 | 26 |
Substances . | Areas . | . | Percentage of positive results . | Area contamination levels (ng cm−2) . | . | ||
---|---|---|---|---|---|---|---|
. | . | n/N . | >LOD (%) . | 50th . | 75th . | 90th . | max . |
CP | Pharmacies | 27/84 | 32 | <LOD | 0.3 | 3.6 | 51.3 |
BSC | 10/30 | 37 | <LOD | 0.4 | 3.6 | 45.5 | |
Work area | 14/36 | 42 | <LOD | 0.5 | 4.6 | 51.3 | |
Logistic room | 3/18 | 17 | <LOD | <LOD | 0.4 | 1.4 | |
OpCU | 29/100 | 29 | <LOD | 0.3 | 0.7 | 45.5 | |
Pt | Pharmacies | 34/99 | 34 | <LOD | 0.05 | 0.5 | 89.1 |
BSC | 21/48 | 44 | <LOD | 0.3 | 1.3 | 89.1 | |
Work area | 10/30 | 33 | <LOD | 0.06 | 0.2 | 0.9 | |
Logistic room | 3/21 | 14 | <LOD | <LOD | 0.02 | 0.3 | |
OpCU | 61/119 | 51 | 0.02 | 0.2 | 0.8 | 121.7 | |
5-FU | Pharmacies | 13/98 | 13 | <LOD | <LOD | 1.0 | 390.0 |
BSC | 8/38 | 21 | <LOD | <LOD | 1.5 | 390.0 | |
Work area | 5/46 | 11 | <LOD | <LOD | 1.0 | 4.0 | |
Logistic room | 0/14 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 13/103 | 13 | <LOD | <LOD | 1.2 | 4.0 | |
GEM | Pharmacies | 9/59 | 15 | <LOD | <LOD | 0.9 | 183.1 |
BSC | 4/13 | 31 | <LOD | 1.6 | 8.2 | 183.1 | |
Work area | 5/32 | 16 | <LOD | <LOD | <LOD | 1.4 | |
Logistic room | 0/14 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 5/44 | 11 | <LOD | <LOD | 0.1 | 1.2 | |
Epi-Doxo | Pharmacies | 0/9 | — | <LOD | <LOD | <LOD | <LOD |
BSC | 0/5 | — | <LOD | <LOD | <LOD | <LOD | |
Work area | 0/2 | — | <LOD | <LOD | <LOD | <LOD | |
Logistic room | 0/2 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 2/56 | 4 | <LOD | <LOD | <LOD | 1.1 |
Substances . | Areas . | . | Percentage of positive results . | Area contamination levels (ng cm−2) . | . | ||
---|---|---|---|---|---|---|---|
. | . | n/N . | >LOD (%) . | 50th . | 75th . | 90th . | max . |
CP | Pharmacies | 27/84 | 32 | <LOD | 0.3 | 3.6 | 51.3 |
BSC | 10/30 | 37 | <LOD | 0.4 | 3.6 | 45.5 | |
Work area | 14/36 | 42 | <LOD | 0.5 | 4.6 | 51.3 | |
Logistic room | 3/18 | 17 | <LOD | <LOD | 0.4 | 1.4 | |
OpCU | 29/100 | 29 | <LOD | 0.3 | 0.7 | 45.5 | |
Pt | Pharmacies | 34/99 | 34 | <LOD | 0.05 | 0.5 | 89.1 |
BSC | 21/48 | 44 | <LOD | 0.3 | 1.3 | 89.1 | |
Work area | 10/30 | 33 | <LOD | 0.06 | 0.2 | 0.9 | |
Logistic room | 3/21 | 14 | <LOD | <LOD | 0.02 | 0.3 | |
OpCU | 61/119 | 51 | 0.02 | 0.2 | 0.8 | 121.7 | |
5-FU | Pharmacies | 13/98 | 13 | <LOD | <LOD | 1.0 | 390.0 |
BSC | 8/38 | 21 | <LOD | <LOD | 1.5 | 390.0 | |
Work area | 5/46 | 11 | <LOD | <LOD | 1.0 | 4.0 | |
Logistic room | 0/14 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 13/103 | 13 | <LOD | <LOD | 1.2 | 4.0 | |
GEM | Pharmacies | 9/59 | 15 | <LOD | <LOD | 0.9 | 183.1 |
BSC | 4/13 | 31 | <LOD | 1.6 | 8.2 | 183.1 | |
Work area | 5/32 | 16 | <LOD | <LOD | <LOD | 1.4 | |
Logistic room | 0/14 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 5/44 | 11 | <LOD | <LOD | 0.1 | 1.2 | |
Epi-Doxo | Pharmacies | 0/9 | — | <LOD | <LOD | <LOD | <LOD |
BSC | 0/5 | — | <LOD | <LOD | <LOD | <LOD | |
Work area | 0/2 | — | <LOD | <LOD | <LOD | <LOD | |
Logistic room | 0/2 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 2/56 | 4 | <LOD | <LOD | <LOD | 1.1 |
Surface contamination percentiles (50th, 75th, 90th) and maximum value (max) for each antineoplastic drug in Pharmacies (Pharm) and Outpatients Care Units (OpCUs). The 90th percentile is the suggested HGV in pharmacies.
Substances . | Areas . | . | Percentage of positive results . | Area contamination levels (ng cm−2) . | . | ||
---|---|---|---|---|---|---|---|
. | . | n/N . | >LOD (%) . | 50th . | 75th . | 90th . | max . |
CP | Pharmacies | 27/84 | 32 | <LOD | 0.3 | 3.6 | 51.3 |
BSC | 10/30 | 37 | <LOD | 0.4 | 3.6 | 45.5 | |
Work area | 14/36 | 42 | <LOD | 0.5 | 4.6 | 51.3 | |
Logistic room | 3/18 | 17 | <LOD | <LOD | 0.4 | 1.4 | |
OpCU | 29/100 | 29 | <LOD | 0.3 | 0.7 | 45.5 | |
Pt | Pharmacies | 34/99 | 34 | <LOD | 0.05 | 0.5 | 89.1 |
BSC | 21/48 | 44 | <LOD | 0.3 | 1.3 | 89.1 | |
Work area | 10/30 | 33 | <LOD | 0.06 | 0.2 | 0.9 | |
Logistic room | 3/21 | 14 | <LOD | <LOD | 0.02 | 0.3 | |
OpCU | 61/119 | 51 | 0.02 | 0.2 | 0.8 | 121.7 | |
5-FU | Pharmacies | 13/98 | 13 | <LOD | <LOD | 1.0 | 390.0 |
BSC | 8/38 | 21 | <LOD | <LOD | 1.5 | 390.0 | |
Work area | 5/46 | 11 | <LOD | <LOD | 1.0 | 4.0 | |
Logistic room | 0/14 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 13/103 | 13 | <LOD | <LOD | 1.2 | 4.0 | |
GEM | Pharmacies | 9/59 | 15 | <LOD | <LOD | 0.9 | 183.1 |
BSC | 4/13 | 31 | <LOD | 1.6 | 8.2 | 183.1 | |
Work area | 5/32 | 16 | <LOD | <LOD | <LOD | 1.4 | |
Logistic room | 0/14 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 5/44 | 11 | <LOD | <LOD | 0.1 | 1.2 | |
Epi-Doxo | Pharmacies | 0/9 | — | <LOD | <LOD | <LOD | <LOD |
BSC | 0/5 | — | <LOD | <LOD | <LOD | <LOD | |
Work area | 0/2 | — | <LOD | <LOD | <LOD | <LOD | |
Logistic room | 0/2 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 2/56 | 4 | <LOD | <LOD | <LOD | 1.1 |
Substances . | Areas . | . | Percentage of positive results . | Area contamination levels (ng cm−2) . | . | ||
---|---|---|---|---|---|---|---|
. | . | n/N . | >LOD (%) . | 50th . | 75th . | 90th . | max . |
CP | Pharmacies | 27/84 | 32 | <LOD | 0.3 | 3.6 | 51.3 |
BSC | 10/30 | 37 | <LOD | 0.4 | 3.6 | 45.5 | |
Work area | 14/36 | 42 | <LOD | 0.5 | 4.6 | 51.3 | |
Logistic room | 3/18 | 17 | <LOD | <LOD | 0.4 | 1.4 | |
OpCU | 29/100 | 29 | <LOD | 0.3 | 0.7 | 45.5 | |
Pt | Pharmacies | 34/99 | 34 | <LOD | 0.05 | 0.5 | 89.1 |
BSC | 21/48 | 44 | <LOD | 0.3 | 1.3 | 89.1 | |
Work area | 10/30 | 33 | <LOD | 0.06 | 0.2 | 0.9 | |
Logistic room | 3/21 | 14 | <LOD | <LOD | 0.02 | 0.3 | |
OpCU | 61/119 | 51 | 0.02 | 0.2 | 0.8 | 121.7 | |
5-FU | Pharmacies | 13/98 | 13 | <LOD | <LOD | 1.0 | 390.0 |
BSC | 8/38 | 21 | <LOD | <LOD | 1.5 | 390.0 | |
Work area | 5/46 | 11 | <LOD | <LOD | 1.0 | 4.0 | |
Logistic room | 0/14 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 13/103 | 13 | <LOD | <LOD | 1.2 | 4.0 | |
GEM | Pharmacies | 9/59 | 15 | <LOD | <LOD | 0.9 | 183.1 |
BSC | 4/13 | 31 | <LOD | 1.6 | 8.2 | 183.1 | |
Work area | 5/32 | 16 | <LOD | <LOD | <LOD | 1.4 | |
Logistic room | 0/14 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 5/44 | 11 | <LOD | <LOD | 0.1 | 1.2 | |
Epi-Doxo | Pharmacies | 0/9 | — | <LOD | <LOD | <LOD | <LOD |
BSC | 0/5 | — | <LOD | <LOD | <LOD | <LOD | |
Work area | 0/2 | — | <LOD | <LOD | <LOD | <LOD | |
Logistic room | 0/2 | — | <LOD | <LOD | <LOD | <LOD | |
OpCU | 2/56 | 4 | <LOD | <LOD | <LOD | 1.1 |
Surface contamination percentiles (50th, 75th, 90th) and maximum value (max) for each antineoplastic drug in Pharmacies (Pharm) and Outpatients Care Units (OpCUs). The 90th percentile is the suggested HGV in pharmacies.
Hospitals
As shown in Table 3, the biggest hospital (hospital B) had 215 determinations corresponding to 28% of the total dataset, followed by site G with 156 data (20%) while for the other sites analyses ranged between 32 and 99 (Table 3), the overall percentages proving to be statistically significant (χ2 = 33.2, P < 0.001). Along with Bonferroni’s rule, hospital H had a number of positive samples that was three times higher than that of A with a P-value <0.0001 and 4-fold greater than that of B (P < 0.0001). Great concern arose from the pharmacy located in hospital H. For this workplace, the highest percentage of positive determinations (46%) was obtained from surfaces that had been sampled from hospital cleanroom H. These results were statistically significant (P < 0.0001) with respect to those from hospitals A, B, and G. Statistical analyses served to support the hypothesis that the pharmacy personnel at hospital H was at risk of occupational exposure. Hospital G exhibited a number of samples above the LODs that was four times higher than that of hospital A (P < 0.0001). Also, the cleanroom located in hospital C showed a percentage of positive samples higher than A but with a grade of less importance (P = 0.001). Interestingly, a further statistical investigation emphasized that in comparison with OpCUs, pharmacies had a number of positive values (at least one substance had values greater than LOD) that was statistically significant (χ2 = 42.9, P < 0.001), data among OpCUs being heterogeneous and not relevant (χ2 = 7.8, P = 0.35).
Pharmacies (BSC, Work area, and Logistic rooms)
In pharmacies, the mean and the median contamination levels were 2.7 and 0.25 ng cm−2, respectively (range 0.005–390.0 ng cm−2). Platinum compounds (34%) and CP (32%) had the highest percentages of positive values. In Pharmacies, 83 out of 349 results (24%) from wipe samples provided evidence of positive findings (Table 3). For all the five investigated drugs, the 50th percentile was always below the LODs either in BSCs or in work/logistic areas (Table 4). The 75th percentile of area contamination in pharmacies was 0.3 and 0.05 ng cm−2 for CP and Pt, respectively. For all the five markers, CP, Pt, 5-FU, GEM, and EPI, the 90th percentile of area contamination levels was 3.6, 0.5, 1.0, 0.9 ng cm−2, and <LOD, respectively. In regard to EPI, there were not enough positive results because this drug always had centiles below the LODs.
Many contaminations were found inside BSCs (43/134) and work areas (34/146), but few were in logistic rooms (6/69). Particularly, BSCs showed to be more contaminated than work areas and logistic rooms. For Pt, according to Bonferroni’s correction for multiple comparisons, an excess of positive (10 times higher) contamination levels was found inside the BSCs with respect to the logistic rooms (P = 0.009). Overall area contamination in pharmacies showed that the worktops of BSCs had the highest value. A single measure of GEM (183 ng cm−2) was obtained by sampling the worktop of the BSC at the pharmacy of hospital G equipped with a class II biosafety cabinet. This value was ascribed to a particular case of spillage which occurred during the handling of the final product of GEM. This occasional concentration value was considered irrelevant to describe data distribution at pharmacy G. Differently from hospital G, hospital H showed a mean contamination level of 5-FU of 4.4 ng cm−2 and a maximum value of 390 ng cm−2.
Out Patient Care Units
For all the drugs, the mean and the median contamination levels were 1.62 ng cm−2 and 0.10 ng cm−2, respectively (range 0.005–121.7 ng cm−2). The percentage of positive values for the most contaminated surface areas in the OpCUs was always >51%. Consistently, high contamination from Pt (121.7 ng cm−2) was found on the floor of the restroom in patient care unit B. The drug was administered via IV infusion bag. Similarly, considerable amounts of CP were detected on the floor around an infusion pole (5.5 ng cm−2) that was located in the administration area of patient care unit B. In outpatient facilities, the minimum level of drug manipulation (putting up lines, drips, etc.) was thought to have caused the spreading of the contamination into the background although the rate of contamination was not statistically significant. The highest levels of CP, 5-FU, GEM, and Pt were specially found on floor surfaces. The highest levels of CP, 5-FU, GEM, and Pt were especially found on floor surfaces. These therapy areas had detectable levels with a mean value that was 2.2 ng cm−2 (range: <LOD–12.2 ng cm−2) on floors inside the bathrooms, and 0.25 ng cm−2 (range: <LOD–5.5 ng cm−2) on floors next to the poles. The other two sampling positions i.e. outer surfaces of drug vials and door handles were found to be slightly contaminated with CP, 5-FU, GEM, and Pt. The vial surface had the highest value of 5-FU (0.9 ng cm−2) and ranged from <LOD to 0.9 ng cm−2. For all the drugs, the median values were always lower than the LOD values, whereas the 75th percentile showed surface contamination for CP (0.3 ng cm−2) and Pt (0.05 ng cm−2). At the 90th percentile of the area contamination levels, CP, platinum compounds, 5-fluorouracil (5-FU), gemcitabine (GEM), and epi-doxorubicin had values of 0.7, 0.8, 1.2, 0.1 ng cm−2, and <LOD, respectively.
Discussion
This study demonstrated that ADs will continue representing a potential risk for personnel involved in the handling of these compounds and great concerns have been raised by the presence of ADs in many investigated workplaces.
In the literature, guidance values for environmental monitoring programs of ADs were often set at the 90th percentile of data distribution and these guidance values were not health based (Schierl et al., 2009; Hedmer et al., 2012; Kiffmeyer et al., 2013). In our study, HGVs are being developed based on the 90th percentile of surface monitoring results therefore higher concentration values than those based on the 75th percentile were obtained. For instance, CP had a value of 3.6 ng cm−2 at the 90th percentile of data distribution and all other ADs had values ranging from 0.5 to 1.0 ng cm−2. Due to a clear need to control exposure within pharmacies, these guidance values were implemented and deemed as a simple and practical way of providing a benchmark for the assessment of exposure risks. We therefore agreed with the approach of proposing performance-based guidance values by using the 90th percentile of results in order to obtain HGVs. In 24 Swiss hospitals, the evaluation of chemical contamination of surfaces was conducted by means of the 50th, 75th, and 90th percentile (Fleury-Souverain et al., 2015). For cytarabine, ifosfamide, GEM, and CP, the 90th percentile ranged between 0.1 and 446 pg cm−2. They decided not to set guidance values at any of these percentiles however protective measures were suggested to improve situations where technicians might be at risk of exposure. In Sweden, the 90th percentiles were used for the first time at the administering facilities of five hospitals. For CP, the 90th percentile ranged from 2.8 to 370 pg cm−2 and for ifosfamide it was between 1.2 and 500 pg cm−2 (Hedmer et al. 2012) in relation to the selected locations. In German hospitals, Schierl et al. proposed guidance values for Pt and 5-FU in pharmacies at the 75th percentile of the data distribution because low levels of detection limits were found. The authors estimated that the 75th percentile from all the studied locations was 4 pg cm−2 for Pt and 30 pg cm−2 for 5-FU. Recently, Kiffmeyer et al. have suggested that surface contamination levels should be <0.1 ng cm−2 (1.0 µg m−2) and provided this value as a drug-independent guidance. In this study, HGVs were obtained and determined for each single substance since each AD shows different chemical properties and usages (e.g. CP is a highly toxic and resistant drug, with high skin permeability, Sessink et al., 2011). Overall contamination levels of the five studied ADs were detected in 85 per cent of the pharmacies and in 93 per cent of the OpCUs.
Pharmacies
Pharmacies and administering areas were regularly wipe-sampled yielding a figure of positive data that was very close between the two hospital units. The pharmacies showed statistically significant contamination levels differently from the work areas where chemotherapy treatments were nursed. The relatively high number of samples collected at the pharmacy of hospital H led to identifying possible contamination sources with respect to 5-FU. Although the internal safe handling guidelines had been reported and deeply explained to the personnel, adherence to safe regulations was omitted. Thus, 5-FU was detected on the gloves of a technician involved in the compounding of this drug. Contamination was found elsewhere particularly on the door handle of the cleanroom and in other adjacent areas such as the countertop where chemotherapy treatments were nursed. Gloves were stressed as a possible source of exposure and the described accidental contact with other furniture had possibly spread the contamination to the handles of doors leading into the cleanroom. The statistical study supported these findings because the pharmacy of hospital H had contamination rates significantly higher (odds ratio = 6.96; P < 0.0001) than all the other hospital pharmacies. Based on these outcomes, the working procedures were corrected in order to lower the risk of exposure. The use of over-gloves with a high frequency of changing was recommended in order to limit the transfer of chemical contamination to other surfaces. The permeability characteristics of the gloves in use at pharmacy H was also evaluated in relation to 5-FU that was supplied as a liquid substance. It was thought to reinforce this practice especially in pharmacy hospital H because the technicians compounding 5-FU were seen to wear medical gloves instead of chemotherapy gloves. Kopp et al. (2013) reported that the use of alcohols during the compounding of ADs was controlled in order to limit permeation through this material.
In Italian pharmacies, the 90th percentiles of the area contamination levels were found to be 3.6, 1.0, 0.9, 0.5 ng cm−2, and <LOD for CP, 5-FU, GEM, Pt, and EPI, respectively. By comparison, the results and the 90th percentile of the area contamination levels in Italian pharmacies were higher than those of other European workplaces although many factors such as different recovery rates and the analytical limits of determinations are thought to have affected the results. The 90th percentile (3.6 ng cm−2) for CP contamination meant that strict action was applied and measures were implemented to reduce the detected surface loads of this compound classified as a carcinogenic substance. Exceeding values at the 90th percentile that were included between 1.0 and 0.5 ng cm−2 likewise raised the need for applying corrective measures to control exposure to 5-FU, GEM, and platinum compounds. In these workplaces, a decrease in contamination was observed upon revision of the work procedures such as the frequency of changing gloves during the compounding of ADs inside the BSC. Furthermore, in many facilities, the introduction of closed system devices (Yoshida et al., 2009; Clark et al., 2013) was suggested as a practical and urgent improvement in order to control contamination inside the BSC. Moreover, setting the HGVs at the 90th percentile was aimed to begin a series of actions in order to control exposure on the 10% of results that were above the 90th percentile. The HGVs were considered a suitable tool to control occupational exposure since no PELs for surface contamination from ADs exist.
Outpatient care units
In this study, nurses working at outpatient clinics or oncology wards were potentially exposed to ADs. During the monitoring surveys carried out at the patient care units, the working procedures were reported as being continuously revised. Regulations that were in use at administering facilities were poorly standardized. Rather, practical arrangements were adopted to organize hospital wards successfully. As a result, data determined by wipe sampling monitoring programs performed according to the studied strategy showed generalized contamination with sporadic and high measurable levels of the cytotoxic compounds. It was thought that floor surface contamination with CP, 5-FU, GEM, and Pt was mainly due to a poor efficacy of decontamination and cleaning procedures. In this fragmentary work environment, the obtained HGVs were not applicable. The aim of this study was therefore to develop a wipe sampling strategy with the additional purpose of standardizing the sampling locations in the differently organized Italian OpCUs by means of repeated monitoring programs.
Conclusions
The study provides the first summary of environmental contamination detected in pharmacies and OpCUs of Italian hospitals. The overall picture described by this work was evidence that no zero levels of ADs are achievable. Therefore, the establishment of a technical performance-based limit has become an irretrievable step before performing risk assessment studies for occupational exposure to ADs. In pharmacies as a whole, the evaluation of efficient and safe hospital operations was aided by knowledge of the residual surface contamination. Based on these findings, we propose a guidance value based on the 90th percentile of actual data distribution. We therefore suggest HGVs of 3.6 ng cm−2 (CP), 1.0 ng cm−2 (5-FU), 0.9 ng cm−2 (GEM), 0.5 ng cm−2 (platinum compounds), and <LOD (epi-doxorubicin). We propose this applicable benchmark as a powerful tool to evaluate how exposure was controlled in the studied workplaces since the levels of antineoplastic agents have to be kept as low as possible and no official threshold limits are yet available. At the moment, the approach of developing a dataset with a large amount of wipe sampling data is regarded as a suitable and promising way of obtaining a recommended guidance that can be further improved to that of 0.1 ng cm−2 as reported in the literature.
Declaration
The authors declare no conflict of interest relating to the material presented in this Article. Its contents, including any opinions and/or conclusions expressed, are solely those of the authors.
References