Society for Radiological Protection J. Radiol. Prot. 34 (2014) 811–823
Journal of Radiological Protection doi:10.1088/0952-4746/34/4/811
Derivation and application of dose reduction factors for protective eyewear worn in interventional radiology and cardiology Jill S Magee1, Colin J Martin1,2, Viktor Sandblom3,4, Matthew J Carter1, Anja Almén3, Åke Cederblad3, Pernilla Jonasson3 and Charlotta Lundh3 1
Health Physics, Gartnavel Royal Hospital, Glasgow, G12 0XH, Scotland, UK Department of Clinical Physics and Bio-Engineering, University of Glasgow, Glasgow, Scotland, UK 3 Department of Medical Physics and Biomedical Engineering, Sahlgrenska University Hospital, SE-413 45 Gothenburg, Sweden 4 Department of Radiation Physics, University of Gothenburg, SE-413 45 Gothenburg, Sweden 2
E-mail: [email protected] Received 20 January 2014, revised 29 August 2014 Accepted for publication 19 September 2014 Published 21 October 2014 Abstract
Doses to the eyes of interventional radiologists and cardiologists could exceed the annual limit of 20 mSv proposed by the International Commission on Radiological Protection. Lead glasses of various designs are available to provide protection, but standard eye dosemeters will not take account of the protection they provide. The aim of this study has been to derive dose reduction factors (DRFs) equal to the ratio of the dose with no eyewear, divided by that when lead glasses are worn. Thirty sets of protective eyewear have been tested in x-ray fields using anthropomorphic phantoms to simulate the patient and clinician in two centres. The experiments performed have determined DRFs from simulations of interventional procedures by measuring doses to the eyes of the phantom representing the clinician, using TLDs in Glasgow, Scotland and with an electronic dosemeter in Gothenburg, Sweden. During interventional procedures scattered x-rays arising from the patient will be incident on the head of the clinician from below and to the side. DRFs for x-rays incident on the front of lead glasses vary from 5.2 to 7.6, while values for orientations similar to those used in the majority of clinical practice are between 1.4 and 5.2. Specialised designs with lead glass side shields or of a wraparound style with angled lenses performed better than lead glasses based on the design of standard spectacles. 0952-4746/14/040811+13$33.00 © 2014 IOP Publishing Ltd Printed in the UK
811
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Results suggest that application of a DRF of 2 would provide a conservative factor that could be applied to personal dosemeter measurements to account for the dose reduction provided by any type of lead glasses provided certain criteria relating to design and consistency of use are applied. Keywords: interventional radiology, protective eyewear, eye dose, interventional cardiology (Some figures may appear in colour only in the online journal) 1. Introduction The International Commission on Radiological Protection (ICRP) have proposed that the occupational dose limit for the lens of the eye is reduced from 150 mSv to 20 mSv, averaged over 5 years with no single year exceeding 50 mSv (ICRP 2011). The reduction has been instigated following a review of epidemiological evidence indicating that the threshold lens dose for radiation induced cataracts is 0.5 Gy instead of the previously suggested 5 Gy (Ainsbury et al 2009, Ciraf-Bjelac et al 2010, ICRP 2012). This change in the dose limit has been incorporated into the revised European and International Basic Safety Standards (BSSs) (European Commission 2013, International Atomic Energy Agency 2014). The public dose limit to the lens of the eye will remain at 15 mSv per year and, according to the European BSS, radiation workers who are liable to receive an equivalent dose greater than 15 mSv per year for the lens of the eye shall be classified in category A. Medical staff working in interventional radiology and cardiology have relatively high exposures to radiation compared to other occupational groups involved with x-rays (Kim et al 2008, Martin 2009, Koukorava et al 2011a, Kim et al 2012, Vanhavere et al 2012, ICRP 2013, Jacob et al 2013). Radiation doses to the lens of the eye have the potential to exceed the dose limit for interventional clinicians with high workloads, unless appropriate radiation protection measures are put in place (Martin 2011, Martin and Magee 2013). This requires training in personal protection techniques, including adequate use of ceiling suspended shields and the wearing of protective eyewear. Effective monitoring to assess the dose to the eye is important in order to identify where there are protection issues. Eye dosemeters available currently will not take account of protection provided by lead glasses, since these dosemeters are placed where they are not shielded by the lead glasses. Since lead glasses can provide an important component of the protection if used effectively, a methodology needs to be developed to enable personal dosemeter results to be adjusted to take the protection provided into account. The aim of the present study was to recommend a reasonable and conservative dose reduction factor for lead glasses that can be applied to dosemeter results when measurements for individual lead glasses are not available. 2. Materials and methods In this project lead glasses of varying design have been tested using anthropomorphic phantoms to simulate the patient and clinician in order to determine dose reduction factors (DRFs) equal to the ratio of the dose with no eyewear, divided by that when lead glasses are worn. Separate studies have been performed in Glasgow, Scotland, UK and Gothenburg, Sweden on lead glasses available in the two countries. Results have been pooled to provide more extensive coverage of the protective eyewear available. The designs of eyewear have been separated 812
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Figure 1. Computed tracings showing generic examples of the seven categories of lead glasses tested.
into a number of categories, which are listed below, and figure 1 displays generic computer generated drawings of the various categories. All but one had lead glass front lenses with protection equivalent to 0.75 mm of lead and protection in any side-shield was equivalent to 0.5 mm of lead. Table 1 gives details of the models and the design categories which are listed below. (a) Glasses with large flat lead glass lenses, thick frames and side shields (b) Glasses of similar design to category 1, but with smaller flat lenses and protective side shields (c) Glasses with small flat lenses and thin frames, similar to conventional spectacles, but with lead glass side shields extending for 30 mm along the side arms and following the contour of the lower part of the lens (d) Glasses held in place by a head band with flat lenses, thin frames and side shields to the full height of the lens extending for 20 mm to the side (e) Wraparound style glasses with lenses angled so as to provide more protection for irradiation from the side (f) Fitover glasses, similar in design to category 1, but arranged to fit over conventional spectacles (g) Face mask of 0.1 mm lead equivalence, held in place by a headband. The protection provided by lead glasses depends on the angle at which scatter from the patient is incident on the head (McVey et al 2013, Van Rooijen et al 2013). For the majority of the time that an interventional radiologist or cardiologist is carrying out a procedure, he/she will be viewing the resulting images rather than looking towards the patient when x-rays are 813
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Table 1. List of lead glasses tested together with the category and DRFs measured.
Glasses No.
Model
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Front + Side Front Front (0.5 mm Pb) Maxx 10 Maxx 30 Polycarbonate Adjustable Metalflex Metalite Metalite Alumilite Aviator Ultralite Wraparound Wraparound Wraparound Fitover Fitover Facemask
Supplier
Design category
Glasgow Study A. Somerville A. Somerville Rothband A. Somerville A. Somerville Protecx APC Cardiovascular Protecx Rothband Protecx A. Somerville APC Cardiovascular APC Cardiovascular Protecx Rothband A. Somerville Protecx Bartec Technologies
1 2 2 2 2 2 3 3 3 5 5 5 5 5 5 6 6 7
Gothenburg Study 19 20 21 22 23 24 25 26 27 28 29 30
58 Jarrod APX Ray BX04 APX Ray TX01 APX Ray TX07 APX Ray AX06 APX Ray AX10 9935 Ultralite Small 9935 Ultralite Large 9941 Ultralite 99 Ultralite Vista 89 Fitover
Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Mediel Scanflex Medical
4 5 5 5 5 5 5 5 5 5 6 6
DRF
DRF
(30°)
(60°)
4.5 4.5 3.3 4.7 5.0 4.6 3.2 4.0 3.8 4.7 4.7 4.6 4.4 5.0 5.0 4.2 4.0 4.0
4.3 4.2 4.2 4.1 3.2 4.1 3.5 3.9 5.0 4.1 -
(0°)
(60°)
7.2 6.7 6.7 6.0 6.0 5.8 5.2 5.5 6.7 6.0 7.2 7.6
1.6 5.2 4.8 3.0 4.1 4.9 1.9 2.9 3.5 3.0 2.2 1.4
being emitted. Therefore the DRF required must take account of x-ray beams incident from the side and below the level of the head. During a procedure the eyes of the operator will typically be about 500 mm above the level of the patient and scatter will be incident at an angle of about 45° in the vertical plane (figure 2(a)). A standard postero–anterior (PA) projection during an interventional radiology procedure was simulated in both studies using the trunk of an anthropomorphic phantom. The PA projection was chosen as the most frequent in clinical practice, although right and left anterior oblique projections with the x-ray beam directed upwards through the patient on either side of the vertical are also common. The operator usually stands adjacent to the patient couch to the right of the x-ray tube / image receptor gantry. If the clinician is looking at the monitor during the x-ray exposures, the angle in the horizontal plane between the direction the head is facing and the angle at which the x-ray beam is incident will typically be 30°–60° (figure 2(b)). 814
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Figure 2. Experimental arrangement showing the orientation of x-ray scatter beams
used for measurements of DRF for lead glasses in the simulated procedures, (a) in the vertical plane and (b) showing the two angled positions of measurement in the horizontal plane.
Methodologies used to test the lead glasses in the two centres were similar but not identical and so will be described separately. In Glasgow, 18 pairs of lead glasses, of varying styles were obtained on loan from a number of UK based companies. Experiments were performed using an anthropomorphic Rando phantom and doses measured with LiF: Mg, Cu, P thermoluminescent dosemeters (TLDs). The TLDs were calibrated in a 70 kV, 4 mm filtration x-ray beam against air kerma measured with an ionisation chamber with a calibration traceable to a national standard. Each chip was allocated an individual calibration factor. No adjustments were made for Hp(3) since results are presented in the form of ratios with and without protection, so the measurements represent the dose within the 1 mm depth of the TLD. This should adequately represent Hp(3) behaviour for x-ray photons in the diagnostic energy range (Behrens 2012). The measurements were made using a Gulmay superficial x-ray treatment unit as the x-ray source, operating at 70 kV, with a filtration equivalent to 4 mm of aluminium. Tube potentials commonly used for interventional radiology and cardiology are in the range 60 kV–80 kV, being maintained at a relatively low level to allow visualisation of the iodine contrast medium in small vessels. Use of the Gulmay unit allowed repeatable exposures to be made at a high dose level of 1000 mu equating to 10 Gy incident on the phantom to improve the accuracy of the TLD measurements. Care was taken to ensure that the x-ray beam, which was directed towards the heart region, was entirely intercepted by the phantom. The head of the phantom was separated from the body and secured on a tripod to represent the operator. It was positioned to represent a configuration in which scattered x-rays were directed upwards at an angle of about 45° in the vertical plane, and at an angle of 30° (Glasgow 30°) and in some cases at 60° (Glasgow 60°) in the horizontal plane with respect to the direction the head was facing (figure 2). The dose received by the eyes was determined using TLDs placed on each of the phantom’s eyes, one placed on the midpoint between the eyes and one positioned to the side of each eye behind the side shield (figure 3). The dose ranges recorded by the TLDs were 0.9–2.1 mGy and 4.1–6.9 mGy with and without protection respectively, and the mean results 1.3 mGy and 5.4 mGy with and without protection. Doses measured on the two eyes were within the stated uncertainty with a standard deviation for sets of TLDs of ±3%. When calculating the DRF, the dose received by the eyes was taken as the mean of the TLD readings. The overall uncertainty is considered to be ±7% when the accuracy of positioning on the phantom is taken into account. Some assessments were also made in Glasgow using Gafchromic XR-QA film, which was scanned on a flat bed scanner and converted to dose using the method described in Martin et al (2011). Strips about 10 cm wide were placed between the glasses and 815
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Figure 3. Positions of TLDs used for measurements on the head of the Rando phantom.
Table 2. Mean DRFs and ranges for each category of lead glasses (table 1).
Category
Lead equiv. (mm)
1 2 2 3 4 5 6 7
0.75 0.75 0.5 0.75 0.75 0.75 0.75 0.1
Gothenburg DRF 0°
7.2 6.1 (5.2–6.7) 7.4 (7.2–7.6)
Glasgow DRF 30°
Glasgow DRF 60°
4.5 4.7 (4.5–5.0) 3.3 3.7 (3.2–4.0)
4.3 4.2 (4.1–4.3)
4.7 (4.4–5.0) 4.1 (4.0–4.2) 4
4.1 (3.5–5.0) 4.1 4
Gothenburg DRF 60°
3.2 1.6 3.7 (1.9–5.2) 1.8 (1.4–2.2)
the phantom to allow the surface dose distribution to be viewed in order to give an indication of the level of protection provided by the frames. Dosimetry records kept by Greater Glasgow Health Board for interventional radiology staff which included details of the number and types of procedure and the doses for each operator were consulted to assess the potential importance of protective eyewear in limiting dose. In Gothenburg, 12 pairs of lead glasses, provided by Sweden based companies, were evaluated (table 1, figure 2). Anthropomorphic phantoms were used to simulate the patient and the operator. An electronic dosemeter (EDD-30, RaySafe, Billdal, Sweden) with a sensor measuring 22 × 11 × 6 mm3 was placed on the left eye of the head phantom representing the operator (position C in figure 3). The shape of the head phantom allowed the sensor to be placed on the eye without affecting the position of the lead glasses. EDD-30 was calibrated with an N80 spectrum on an ICRU body phantom, as defined in ISO 4037 to measure Hp(0.07). This provides an adequate representation of Hp(3) for x-ray photons (Behrens 2012). The x-ray unit used was a Philips BV 300 mobile C-arm, operating at 82 kV, with a filtration equivalent to 8 mm of aluminium. A standard postero–anterior projection was employed. Measurements of the dose from scattered radiation were made with and without lead glasses for the position of 60° in the horizontal plane in figure 2(b). Two different head angles were used; one when the head phantom was angled straight forward, simulating an operator looking at the monitor with radiation incident from the left (Gothenburg 60°) (figure 2), and a second with the head phantom angled obliquely down and to the left, simulating an operator looking directly at the 816
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Figure 4. Histogram showing the distribution of DRF measurements for lead glasses
tested.
irradiated part of the patient (Gothenburg 0°). The first position was similar to the 60° angle experiment carried out in Glasgow. Lead glasses number 29 and 30 were placed on the head phantom with an additional space to the skin, simulating the wearing of a pair of regular spectacles underneath. A similar adjustment was not made for the Glasgow tests. For each pair of glasses the dose was measured for 60 s fluoroscopy time (82 kV, 2.82 mA), leading to doses of 1.3–7.4 µSv, depending on incident angle and the pair of glasses, corresponding to dose rates of 78–444 µSv h−1. These were within the specified ranges for dose of 10 nSv–9999 Sv, and dose rate of 30 µSv–2 Sv h−1. The human factor involved in placement of the glasses on the phantom over the dosemeter is more significant in determining the uncertainty in the measured doses, than the uncertainties in the measured dose. As a test, measurements were repeated ten times for one pair of glasses, which were taken off the phantom and repositioned between measurements. The DRF range was 6.3–7.4, with a mean of 6.7 and the standard deviation was 0.36. The net uncertainty in the measurements including both the human factor and the uncertainty in the measured dose is ±6%. To investigate the DRF from ceiling suspended shields as an alternative to lead glasses, additional measurements of collar dose with an electronic personal dosimeter (DoseAware, RaySafe, Billdal, Sweden) were made for three interventional radiologists while performing similar procedures in Glasgow. The radiologists were observed throughout the measurements and the doses recorded while they were using and not using a ceiling suspended screen. 3. Results The DRFs for all glasses tested are shown in table 1. Means and ranges of DRFs for groupings of different categories of lead glasses are given in table 2. The DRFs from all the measurements are plotted in the form of histograms in figure 4. The mean DRF of all lead glasses and geometric set ups was 4.5. All lead glasses tested with the head angled towards the patient (Gothenburg DRF 0°) showed good protection capacities, with a mean DRF of 6.4 and 817
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Table 3. The effect of wearing lead glasses on an interventional clinician’s highest
possible workload. Protection No Lead Glasses Mean DRF for all lead glasses (4.5) Application of a DRF of 2.0
Collar dose for 315 procedures (mSv)
No. of procedures to reach 20 mSv collar dose
No. of procedures to reach 20 mSv projected eye dose
30 6.7
210 945
280 1260
15
420
560
standard deviation (SD) of 0.7. These results only showed minor differences in DRF between the different pairs, although the pair with the smallest lenses (25) had the lowest DRF. Moving through the columns in table 2 from left to right shows a steady reduction in DRF for all categories of glasses as the angle of incidence in the horizontal plane increased. When the source of scatter was below the head and the angle in the horizontal plane was 30° (Glasgow DRF 30°) the mean DRF was 4.3 (excluding face shield data), and for the 60° angulation (combined data DRF 60°) this fell to 3.6. In Gothenburg, the lowest DRF observed was 1.4 and the highest was 7.6. In the Glasgow results the lowest DRF observed for a lead equivalence of 0.75 mm was 3.2 and the highest was 5.0. The face mask, with a lead equivalence of 0.1 mm, had a DRF of 4.0 and the 0.50 mm lead equivalence category 2 glasses had a DRF of 3.3. Pictures obtained using Gafchromic XR-QA2 film were used to give a qualitative assessment of the protection afforded by the fitover glasses (category 6) and revealed that protection in the frames was much less than that in the lenses. The records kept by Greater Glasgow Health Board showed that the highest annual dose measured at the collar for one interventional radiologist at Glasgow was about 30 mSv, which equates to an eye dose of about 22.5 mSv (×0.75) (Martin 2011). This radiologist performed 315 interventional procedures in 2012, but did not at that time wear lead glasses or use the ceiling suspended shield consistently. It should be noted that since the doses for some months were between 4 mSv and 5 mSv, there is a potential for annual doses to be substantially higher than 30 mSv. Table 3 displays the effect that wearing a pair of lead glasses could have on such an interventional radiologist’s doses and the workloads that could be undertaken before exceeding the proposed dose limit. The results show that with no protection, the radiologist workload would need to be restricted to ensure that he did not reach the 20 mSv limit, but wearing lead glasses with a DRF of 4.5 could potentially increase this number to 945 (table 3) before 20 mSv was reached on the collar dosemeter. Results from other centres have shown that annual doses to the eyes of some interventional clinicians may be 50 mSv–100 mSv (Koukorava et al 2011a, Jacob et al 2013, Martin and Magee 2013), requiring substantial additional improvement in training and implementation of radiological protection (ICRP 2013). Measurements made on Glasgow interventional radiologists indicated that during periods when a ceiling suspended screen was positioned effectively, a DRF of 5 could be achieved using the screen. 4. Discussion 4.1. Variation in DRF with design of glasses
In this study assessments of DRFs have been made for protective eye wear in head orientations likely to be encountered in routine clinical practice (figure 2). Tables 1 and 2 show the 818
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
variation in DRF for different designs of lead glasses and exposure situations. Results for each group will be discussed in turn. Glasses of categories 1 and 2 provided reasonable protection for the eyes from scattered radiation with DRFs between 4.1 and 5.0. One pair of category 2 glasses with 0.50 mm lead equivalent lenses had a DRF at 30° of 3.3. The effectiveness in protecting the eyes from radiation which is angled upwards depends on the closeness of the fit to the facial contours. One pair with opaque metal side-shields had an adjustable frame, which allowed optimal fitting of the glasses and ensured no large gaps were present. Glasgow radiologists favoured this design feature as they felt it allowed them to achieve the best fit possible and, unlike some designs, the glasses did not ‘steam up’. The design of category 3 glasses was based on an adaptation of standard spectacles with side shielding extending below the side of the lens to provide a closer fit to the face. These had DRFs between 3.2 and 4.0. The pair of category 4 lead glasses tested only achieved a DRF of 1.6 at an angle of 60°. When these glasses were placed on the head phantom a relatively large gap was created between the phantom surface and the glasses due to the poor design, allowing scattered radiation to hit the dosemeter placed on the eye directly without penetrating the lead glass. Category 5 glasses had front lenses angled so as to provide more protection for irradiation from the side and eliminated the need for side shields and the gaps between the frames and the head were smaller. DRFs in the forward and 30° positions were between 4.4 and 6.7, while most at 60° were between 2.9 and 5.2. However, one pair, number 25, had smaller lenses and a DRF at 60° of 1.9. The lower value for the DRF for this pair is thought to be caused by the proximity of irradiated tissue resulting from the restricted coverage by the smaller lenses, as the dosemeter had not affected the positioning of the glasses. When pairs of category 5 glasses were trialled by radiologists in Glasgow, it was reported that some of these glasses had a tendency to ‘steam up’ and therefore reduce vision. The fitover glasses (category 6) were designed to be worn over prescription spectacles. These glasses were bulky, not a secure fit and had large gaps underneath to allow wearing of conventional spectacles. The front lenses were the largest of all and provided the best protection when the head was facing the patient (Gothenburg 0°). However, when the head was angled the DRF fell dramatically from 7.4 (Gothenburg 0°) to about 1.8 (Gothenburg 60°) for experiments in which a gap was left to replicate fitting over conventional spectacles. The Glasgow DRF for this category, where no additional gap was left during the tests, was significantly higher, 4.1 at both 30° and 60°. The low DRF values were due mainly to the larger space left between the glasses and the head for the prescription spectacles. It is difficult to allow for the gap precisely, because of the rigid nature of the phantom surfaces, but results highlight the need for caution where this geometry is used. Gafchromic XR-QA2 film analysis highlighted that protection in the frames of the fitover lead glasses appeared to be virtually non-existent, presumably to keep the weight down. If the operator’s head is angled towards the monitor (figure 2(b)), which is likely to be the case for the majority of the time, then the scattered radiation is able to pass through gaps behind the lenses and through parts of the frame that are not protected to irradiate the eyes directly. Glasgow radiologists found the fitover glasses to be less comfortable, possibly due to the greater weight, and so purchase of prescription lead glasses provided a better option for several reasons. The face mask (category 7) had a DRF of 4.0, despite being of a lower lead equivalence. The mask covered the whole of the face and so also reduced the exposure of regions of the head surrounding the eyes that would make a significant contribution to the eye dose from backscatter. The dose reduction in this study were similar to that reported by Galster et al (2013), who observed that eye dose could be increased by a face mask when the operator was facing away from the patient due to scatter. However the face mask was not favoured by 819
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
radiologists due to its inconvenient size and some reported reduced vision when testing it, because of the tendency to steam up. 4.2. Comparison with results of other studies
DRFs from the Gothenburg measurements with the operator facing the patient were between 5.2 and 7.6. Values for DRFs between 5 and 10 have been reported for a variety of lead glasses when protecting against x-rays incident from the front in the same horizontal plane as the eyes from experimental measurements (Moore et al 1980, Marshall et al 1992, Thornton et al 2010, Galster et al 2013, McVey et al 2013, Van Rooijen et al 2013) and Monte Carlo simulations (Koukorava et al 2014). When the head was at an angle to the direction of irradiation, the DRF was lower, usually between 1.4 and 5.2. McVey et al (2013) reported that DRFs for a pair of category 5 lead glasses varied between 5 and 7 with the angle of incidence for scattered x-ray beams incident in the same horizontal plane as the head. Results from the angled measurements in the present study are at the lower end of the range reported in other studies. This is likely to be due to the use of scattered radiation incident from below the level of the eyes in an arrangement closer to that occurring in clinical practice. Van Rooijen et al (2013) found that in an experiment simulating clinical practice with the head above the level of the phantom and rotated by an angle of 45° to the direction of the x-ray scatter in the horizontal plane, the mean DRF for lead glasses tested was 2.1, compared to 7.9–10.0 for frontal exposures. 4.3. Factors determining DRFs for lead glasses
Other studies of lead glasses have concluded that the dose to the eyes when protective eyewear is worn results primarily from radiation scattered from surrounding tissues (Moore et al 1980, Day and Forster 1981, Cousin et al 1987, McVey et al 2013). This conclusion is supported by results from the face mask (category 7), which had a relatively high DRF of 4.0, despite being of a lower lead equivalence (0.1 mm), because of the reduction in the contribution from backscatter. This result demonstrates that the proximity to the eyes of tissue on which unattenuated x-rays are incident is a major factor in determining the dose to the eye lens. For exposures from the front, differences between various categories of glasses appeared to relate to the sizes of the lenses, and so the proximity of unprotected and therefore irradiated tissue. For example the pair of glasses number 25 which had smaller lenses had a DRF of 5.2 when irradiated from the front, compared to an average of 6.5 for other models. With exposures from the side, the eye dose depends on the extent of the protection from the side. When the scattered radiation is incident from a below, radiation may enter not only from the side, but also through the gap underneath the glass lenses. The wraparound design tended to give a closer fit, but model 25, which had smaller lenses had a lower DRF when irradiated from the side, presumably relating to the proximity of irradiated tissue to the eye itself. The PA projection employed for the measurements is more frequently used and the results should be representative of most exposure scenarios where the x-ray beam passes through the patient. Monte Carlo simulations have shown that the DRF may be lower for left lateral projections where the x-ray tube is adjacent to the operator and x-rays are scattered towards the operator from the surface of the body (Koukorava et al 2014), but this arrangement is used infrequently for the majority of procedures and not recommended because of the protection implications. For larger patients, the tube potential can rise towards 100 kV, but Monte Carlo simulations have shown that this will change the level of protection by less than 5% (Koukorava et al 2014). 820
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
One pair of category 2 glasses with 0.50 mm lead equivalent lenses had a DRF of 3.3. Thus 0.50 mm lead equivalence lenses could provide sufficient reduction to keep the eye dose well below the proposed dose limit, as suggested by Koukorava et al (2011b), particularly if the lenses were large. A lighter pair of lead plastic goggles equivalent to between 0.3 mm and 0.5 mm of lead, providing a close fit at the base and offering more extensive protection to the side could be an alternative design option. Lead glasses should be individually tried out by the user prior to purchase to minimise the gap between the skin and the lead glasses and thereby maximising the DRF. 4.4. Recommendations on protective eyewear
Measurements of the protection offered by lead glasses such as those described in this paper can provide useful data based on which adjustments can be made to dose values recorded by unshielded eye dosemeters to derive a dose representing that to the lens of the eye. However, DRFs are critically dependent on the experimental arrangement and a straight measure of attenuation of x-rays incident from the front will significantly overestimate the DRF in clinical practice. Moreover, many departments will not have the facility to make such measurements. Therefore it would be beneficial to have a factor that could be applied to dosemeter results for any interventional clinicians for whom it could be guaranteed that they wore the protective eyewear consistently. The mean DRFs from the measurements in the studies carried out are combined in a histogram in figure 4. Only for the least favourable orientation for the fitover lead glasses is the DRF less than 2.0, and for most situations the DRF is between 3 and 6. Based on these results, division by a DRF of 2 would appear to be both a reasonable and conservative approach to account for the protection offered by all lead glasses. The majority will have a DRF over 3 and, if an arrangement for approval of designs for protective eyewear could be established, a higher value could be used or approved designs. However any calculations assume that the lead glasses are worn for every procedure. Therefore, in order for this or any other factor to be applied, certain quality assurance measures must be in place. Firstly, the eyewear should be of appropriate design, with either side shields or of a wraparound design. Secondly, quality assurance procedures should be in place, such as regular documented checks to confirm that the interventional clinician concerned always wears the protective eyewear. Reports on dose reductions to the eyes achieved through use of ceiling suspended screens are varied. The DRF of 5 recorded for use of a ceiling suspended screen in the present study corresponds to periods where the interventional radiologist carefully positioned the screen throughout the procedure. A large scale report of clinical measurements for interventional procedures gave DRFs between 1.3 and 7 (Vanhavere et al 2012), a review comparing doses from groups at different centres performing similar procedures gave DRFs of 0.7–2.5 (Jacob et al 2013), while a study at one cardiology centre reported a DRF of 19 (Maeder et al 2006). Monte Carlo simulations of the reduction in collar dose from the use of ceiling suspended screens gave a range of DRFs from 1.03 to 33 for different projections, with a quartile range from 1.1 to 8, and a DRF calculated from the mean of the dose ratios of 2.0 (Koukorava et al 2014). Although the screens in principle give good protection, their effective use for the range of projections throughout clinical procedures is likely to limit the overall level of protection that can be achieved, but reductions by a factors of 2 should be achievable. Nevertheless, use of protective eyewear is an important component of the protection armoury. Calculations showed that wearing lead glasses should minimise any restrictions that would need to be placed on an interventional radiologist’s workload (table 3). Application of a DRF of 2 for those who consistently wear protective eyewear coupled with training in effective use of 821
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
ceiling suspended screens should ensure that few interventional clinicians exceed an annual dose constraint of 15 mSv. 5. Conclusions Measurements of DRFs performed with a variety of lead glasses, using head orientations similar to those encountered in clinical practice, have shown that reductions in the dose to the eyes is typically by factors of 3–5, in addition to the protection offered by ceiling suspended shields. However, glasses designed to fit over prescription spectacles and designs with a poorer fit or smaller lenses gave a DRF less than 2 when irradiated at 60°. The consistent wearing of lead glasses and effective use of ceiling suspended shields could eliminate the need for restrictions on workload for the majority of interventional radiologists and cardiologists. Lead glasses of wraparound design (category 5) or with large front lenses and side shielding (categories 1 and 2) provide a reasonable level of protection. The closeness of the fit to the facial contours, particularly beneath the eyes, and the size of the lenses are both important. When assessments of the dose to the eyes for operators wearing lead glasses are made, it is necessary to take the protection provided by the lead glasses into account. If local measurements to determine the DRF for specific lead glasses are not possible, hospitals should apply a DRF of 2 for clinicians wearing lead glasses consistently. Acknowledgments The authors wish to thank APC Cardiovascular Ltd, A. Somerville Ltd, Bartec Technologies, Protecx, Rothband Ltd, Scanflex Medical and Mediel for the loan of protective eyewear. The glasses tested were a selection of models and did not cover all models available. References Ainsbury E A, Bouffler S D, Dörr W, Graw J, Muirhead C R, Edwards A A and Cooper J 2009 Radiation cataractogenesis: a review of recent studies Radiat. Res. 172 1–9 Behrens R 2012 On the operational quantity Hp(3) for eye lens dosimetry J. Radiol. Prot. 32 455–64 Ciraj-Bjelac O, Rehani M M, Sim K H, Liew H B, Vano E and Kleiman N J 2010 Risk for radiationinduced cataract for staff in interventional cardiology: is there reason for concern? Catheter. Cardiovasc. Interv. 76 826–34 Cousin A J, Lawdahl R B, Chakraborty D P and Koehler R E 1987 The case for radioprotective eyewear / facewear. Practical implications and suggestions Invest. Radiol. 22 688–92 Day M J and Forster E 1981 Protective effect of spectacles Br. J. Radiol. 54 137–8 European Commission 2013 Council Directive (Euratom 2013/59) laying down basic safety standards for protection against the dangers arising from exposure to ionising radiation Official J. Eur. Commun. 2014 L13 (http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2014:013:0001:0073: EN: PDF) Galster M, Guhl C, Uder M and Adamus R 2013 Exposition of the operator’s eye lens and efficacy of radiation shielding in fluoroscopically guided interventions Fortschr. Röntgenstr. 185 474–81 ICRP 2011 Statement on tissue reactions ICRP 4825-3093-1464 April 2011 ICRP 2012 ICRP Statement on tissue reactions / early and late effects of radiation in normal tissues and organs—threshold doses for tissue reactions in a radiation protection context ICRP Publication 118 Ann. ICRP 41 1–2 ICRP 2013 Radiological protection in cardiology, ICRP Publication 120 Ann ICRP 42 International Atomic Energy Agency 2014 Radiation Protection and Safety of Radiation Sources: International Basic Safety Standards (General Safety Requirements Part 3) (Vienna: IAEA) (www-pub.iaea.org/MTCD/Publications/PDF/Pub1578_web-57265295.pdf) 822
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Jacob S, Donadille L, Maccia C, Bar O, Boveda S, Laurier D and Bernier M-O 2013 Eye lens radiation exposure to interventional cardiologists: a retrospective assessment of cumulative doses Rad. Prot. Dosim. 153 282–93 Kim K P, Miller D L, Balter S, Kleinerman R A, Linet M S, Kwon D and Simon S L 2008 Occupational radiation doses to operators performing cardiac catheterization procedures Health Phys. 94 211–27 Kim K P, Miller D L, Berrington de Gonzalez A, Balter S, Kleinerman R A, Ostroumova E, Simon S L and Linet M S. 2012 Occupational radiation doses to operators performing fluoroscopically-guided procedures Health Phys. 103 80–99 Koukorava C, Carinou E, Ferrari P, Krim S, Struelens L, International Workshop on Optimization of Radiation Protection of Medical Staff and ORAMED 2011b Study of the parameters affecting operator doses in interventional radiology using Monte Carlo simulations ORAMED 46 1216–22 Koukorava C, Carinou E, Simantirakis G, Vrachliotis T G, Archontakis E, Tierris C and Dimitriou P 2011a Doses to moperators during interventional radiology procedures: focus on eye lens and extremity dosimeters Rad. Prot. Dosim. 144 482–6 Koukorava C, Farah J, Struelens L, Clairand I, Donadille L, Vanhavere F and Dimitriou P 2014 Efficiency of radiation protection equipment in interventional radiology: a systematic Monte Carlo study of eye lens and whole body doses J. Radiol. Prot. 34 509–28 Maeder M, Brunner-La Rocca H P, Wolber T, Ammann P, Roelli H, Rohner F and Rickli H 2006 Impact of a lead glass screen on scatter radiation to eyes and hands in interventional cardiologists Catheter. Cardiovasc. Interv. 67 18–23 Marshall N W, Faulkner K and Clarke P 1992 An investigation into the effect of protective devices on the dose to radiosensitive organs in the head and neck Br. J. Radiol. 65 799–802 Martin C J 2009 A review of radiology staff doses and dose monitoring requirements Radiat. Prot. Dosim. 136 140–57 Martin C J 2011 What are the implications of the proposed revision of the eye dose limit for interventional operators? Br. J. Radiol. 84 961–2 Martin C J and Magee J S 2013 Assessment of eye and body dose for interventional radiologists, cardiologists, and other interventional staff J. Radiol. Prot. 33 445–60 Martin C J, Gentle D J, Sookpeng S and Loveland J 2011 Application of Gafchromic film in the study of dosimetry methods in CT phantoms J. Radiol. Prot. 31 389–409 McVey S, Sandison A and Sutton D G 2013 An assessment of lead eyewear in interventional radiology J. Radiol. Prot. 33 647–59 Moore W E, Ferguson G and Rohrmann C 1980 Physical factors determining the utility of radiation safety glasses Med. Phys. 7 8–12 Thornton R, Dauer L T, Altamirano J P, Alvarado K J, St. Germain J and Solomon S B 2010 Comparing strategies for operator eye protection in the interventional suite J. Vasc. Int. Radiol. 21 1703–7 Van Rooijen B D, de Hann M W, Das M, Arnoldussen C W K P, de Graaf R, van Zwam W H, Backes W H and Jeukens C R L P N 2013 Efficacy of radiation safety glasses in interventional radiology Cardiovasc. Intervent. Radiol. 37 1149–55 Vanhavere F et al 2012 ORAMED: optimisation of radiation protection for medical staff 7th EURADOS Report 2012-02
823
Copyright of Journal of Radiological Protection is the property of IOP Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Journal of Radiological Protection doi:10.1088/0952-4746/34/4/811
Derivation and application of dose reduction factors for protective eyewear worn in interventional radiology and cardiology Jill S Magee1, Colin J Martin1,2, Viktor Sandblom3,4, Matthew J Carter1, Anja Almén3, Åke Cederblad3, Pernilla Jonasson3 and Charlotta Lundh3 1
Health Physics, Gartnavel Royal Hospital, Glasgow, G12 0XH, Scotland, UK Department of Clinical Physics and Bio-Engineering, University of Glasgow, Glasgow, Scotland, UK 3 Department of Medical Physics and Biomedical Engineering, Sahlgrenska University Hospital, SE-413 45 Gothenburg, Sweden 4 Department of Radiation Physics, University of Gothenburg, SE-413 45 Gothenburg, Sweden 2
E-mail: [email protected] Received 20 January 2014, revised 29 August 2014 Accepted for publication 19 September 2014 Published 21 October 2014 Abstract
Doses to the eyes of interventional radiologists and cardiologists could exceed the annual limit of 20 mSv proposed by the International Commission on Radiological Protection. Lead glasses of various designs are available to provide protection, but standard eye dosemeters will not take account of the protection they provide. The aim of this study has been to derive dose reduction factors (DRFs) equal to the ratio of the dose with no eyewear, divided by that when lead glasses are worn. Thirty sets of protective eyewear have been tested in x-ray fields using anthropomorphic phantoms to simulate the patient and clinician in two centres. The experiments performed have determined DRFs from simulations of interventional procedures by measuring doses to the eyes of the phantom representing the clinician, using TLDs in Glasgow, Scotland and with an electronic dosemeter in Gothenburg, Sweden. During interventional procedures scattered x-rays arising from the patient will be incident on the head of the clinician from below and to the side. DRFs for x-rays incident on the front of lead glasses vary from 5.2 to 7.6, while values for orientations similar to those used in the majority of clinical practice are between 1.4 and 5.2. Specialised designs with lead glass side shields or of a wraparound style with angled lenses performed better than lead glasses based on the design of standard spectacles. 0952-4746/14/040811+13$33.00 © 2014 IOP Publishing Ltd Printed in the UK
811
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Results suggest that application of a DRF of 2 would provide a conservative factor that could be applied to personal dosemeter measurements to account for the dose reduction provided by any type of lead glasses provided certain criteria relating to design and consistency of use are applied. Keywords: interventional radiology, protective eyewear, eye dose, interventional cardiology (Some figures may appear in colour only in the online journal) 1. Introduction The International Commission on Radiological Protection (ICRP) have proposed that the occupational dose limit for the lens of the eye is reduced from 150 mSv to 20 mSv, averaged over 5 years with no single year exceeding 50 mSv (ICRP 2011). The reduction has been instigated following a review of epidemiological evidence indicating that the threshold lens dose for radiation induced cataracts is 0.5 Gy instead of the previously suggested 5 Gy (Ainsbury et al 2009, Ciraf-Bjelac et al 2010, ICRP 2012). This change in the dose limit has been incorporated into the revised European and International Basic Safety Standards (BSSs) (European Commission 2013, International Atomic Energy Agency 2014). The public dose limit to the lens of the eye will remain at 15 mSv per year and, according to the European BSS, radiation workers who are liable to receive an equivalent dose greater than 15 mSv per year for the lens of the eye shall be classified in category A. Medical staff working in interventional radiology and cardiology have relatively high exposures to radiation compared to other occupational groups involved with x-rays (Kim et al 2008, Martin 2009, Koukorava et al 2011a, Kim et al 2012, Vanhavere et al 2012, ICRP 2013, Jacob et al 2013). Radiation doses to the lens of the eye have the potential to exceed the dose limit for interventional clinicians with high workloads, unless appropriate radiation protection measures are put in place (Martin 2011, Martin and Magee 2013). This requires training in personal protection techniques, including adequate use of ceiling suspended shields and the wearing of protective eyewear. Effective monitoring to assess the dose to the eye is important in order to identify where there are protection issues. Eye dosemeters available currently will not take account of protection provided by lead glasses, since these dosemeters are placed where they are not shielded by the lead glasses. Since lead glasses can provide an important component of the protection if used effectively, a methodology needs to be developed to enable personal dosemeter results to be adjusted to take the protection provided into account. The aim of the present study was to recommend a reasonable and conservative dose reduction factor for lead glasses that can be applied to dosemeter results when measurements for individual lead glasses are not available. 2. Materials and methods In this project lead glasses of varying design have been tested using anthropomorphic phantoms to simulate the patient and clinician in order to determine dose reduction factors (DRFs) equal to the ratio of the dose with no eyewear, divided by that when lead glasses are worn. Separate studies have been performed in Glasgow, Scotland, UK and Gothenburg, Sweden on lead glasses available in the two countries. Results have been pooled to provide more extensive coverage of the protective eyewear available. The designs of eyewear have been separated 812
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Figure 1. Computed tracings showing generic examples of the seven categories of lead glasses tested.
into a number of categories, which are listed below, and figure 1 displays generic computer generated drawings of the various categories. All but one had lead glass front lenses with protection equivalent to 0.75 mm of lead and protection in any side-shield was equivalent to 0.5 mm of lead. Table 1 gives details of the models and the design categories which are listed below. (a) Glasses with large flat lead glass lenses, thick frames and side shields (b) Glasses of similar design to category 1, but with smaller flat lenses and protective side shields (c) Glasses with small flat lenses and thin frames, similar to conventional spectacles, but with lead glass side shields extending for 30 mm along the side arms and following the contour of the lower part of the lens (d) Glasses held in place by a head band with flat lenses, thin frames and side shields to the full height of the lens extending for 20 mm to the side (e) Wraparound style glasses with lenses angled so as to provide more protection for irradiation from the side (f) Fitover glasses, similar in design to category 1, but arranged to fit over conventional spectacles (g) Face mask of 0.1 mm lead equivalence, held in place by a headband. The protection provided by lead glasses depends on the angle at which scatter from the patient is incident on the head (McVey et al 2013, Van Rooijen et al 2013). For the majority of the time that an interventional radiologist or cardiologist is carrying out a procedure, he/she will be viewing the resulting images rather than looking towards the patient when x-rays are 813
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Table 1. List of lead glasses tested together with the category and DRFs measured.
Glasses No.
Model
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Front + Side Front Front (0.5 mm Pb) Maxx 10 Maxx 30 Polycarbonate Adjustable Metalflex Metalite Metalite Alumilite Aviator Ultralite Wraparound Wraparound Wraparound Fitover Fitover Facemask
Supplier
Design category
Glasgow Study A. Somerville A. Somerville Rothband A. Somerville A. Somerville Protecx APC Cardiovascular Protecx Rothband Protecx A. Somerville APC Cardiovascular APC Cardiovascular Protecx Rothband A. Somerville Protecx Bartec Technologies
1 2 2 2 2 2 3 3 3 5 5 5 5 5 5 6 6 7
Gothenburg Study 19 20 21 22 23 24 25 26 27 28 29 30
58 Jarrod APX Ray BX04 APX Ray TX01 APX Ray TX07 APX Ray AX06 APX Ray AX10 9935 Ultralite Small 9935 Ultralite Large 9941 Ultralite 99 Ultralite Vista 89 Fitover
Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Scanflex Medical Mediel Scanflex Medical
4 5 5 5 5 5 5 5 5 5 6 6
DRF
DRF
(30°)
(60°)
4.5 4.5 3.3 4.7 5.0 4.6 3.2 4.0 3.8 4.7 4.7 4.6 4.4 5.0 5.0 4.2 4.0 4.0
4.3 4.2 4.2 4.1 3.2 4.1 3.5 3.9 5.0 4.1 -
(0°)
(60°)
7.2 6.7 6.7 6.0 6.0 5.8 5.2 5.5 6.7 6.0 7.2 7.6
1.6 5.2 4.8 3.0 4.1 4.9 1.9 2.9 3.5 3.0 2.2 1.4
being emitted. Therefore the DRF required must take account of x-ray beams incident from the side and below the level of the head. During a procedure the eyes of the operator will typically be about 500 mm above the level of the patient and scatter will be incident at an angle of about 45° in the vertical plane (figure 2(a)). A standard postero–anterior (PA) projection during an interventional radiology procedure was simulated in both studies using the trunk of an anthropomorphic phantom. The PA projection was chosen as the most frequent in clinical practice, although right and left anterior oblique projections with the x-ray beam directed upwards through the patient on either side of the vertical are also common. The operator usually stands adjacent to the patient couch to the right of the x-ray tube / image receptor gantry. If the clinician is looking at the monitor during the x-ray exposures, the angle in the horizontal plane between the direction the head is facing and the angle at which the x-ray beam is incident will typically be 30°–60° (figure 2(b)). 814
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Figure 2. Experimental arrangement showing the orientation of x-ray scatter beams
used for measurements of DRF for lead glasses in the simulated procedures, (a) in the vertical plane and (b) showing the two angled positions of measurement in the horizontal plane.
Methodologies used to test the lead glasses in the two centres were similar but not identical and so will be described separately. In Glasgow, 18 pairs of lead glasses, of varying styles were obtained on loan from a number of UK based companies. Experiments were performed using an anthropomorphic Rando phantom and doses measured with LiF: Mg, Cu, P thermoluminescent dosemeters (TLDs). The TLDs were calibrated in a 70 kV, 4 mm filtration x-ray beam against air kerma measured with an ionisation chamber with a calibration traceable to a national standard. Each chip was allocated an individual calibration factor. No adjustments were made for Hp(3) since results are presented in the form of ratios with and without protection, so the measurements represent the dose within the 1 mm depth of the TLD. This should adequately represent Hp(3) behaviour for x-ray photons in the diagnostic energy range (Behrens 2012). The measurements were made using a Gulmay superficial x-ray treatment unit as the x-ray source, operating at 70 kV, with a filtration equivalent to 4 mm of aluminium. Tube potentials commonly used for interventional radiology and cardiology are in the range 60 kV–80 kV, being maintained at a relatively low level to allow visualisation of the iodine contrast medium in small vessels. Use of the Gulmay unit allowed repeatable exposures to be made at a high dose level of 1000 mu equating to 10 Gy incident on the phantom to improve the accuracy of the TLD measurements. Care was taken to ensure that the x-ray beam, which was directed towards the heart region, was entirely intercepted by the phantom. The head of the phantom was separated from the body and secured on a tripod to represent the operator. It was positioned to represent a configuration in which scattered x-rays were directed upwards at an angle of about 45° in the vertical plane, and at an angle of 30° (Glasgow 30°) and in some cases at 60° (Glasgow 60°) in the horizontal plane with respect to the direction the head was facing (figure 2). The dose received by the eyes was determined using TLDs placed on each of the phantom’s eyes, one placed on the midpoint between the eyes and one positioned to the side of each eye behind the side shield (figure 3). The dose ranges recorded by the TLDs were 0.9–2.1 mGy and 4.1–6.9 mGy with and without protection respectively, and the mean results 1.3 mGy and 5.4 mGy with and without protection. Doses measured on the two eyes were within the stated uncertainty with a standard deviation for sets of TLDs of ±3%. When calculating the DRF, the dose received by the eyes was taken as the mean of the TLD readings. The overall uncertainty is considered to be ±7% when the accuracy of positioning on the phantom is taken into account. Some assessments were also made in Glasgow using Gafchromic XR-QA film, which was scanned on a flat bed scanner and converted to dose using the method described in Martin et al (2011). Strips about 10 cm wide were placed between the glasses and 815
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Figure 3. Positions of TLDs used for measurements on the head of the Rando phantom.
Table 2. Mean DRFs and ranges for each category of lead glasses (table 1).
Category
Lead equiv. (mm)
1 2 2 3 4 5 6 7
0.75 0.75 0.5 0.75 0.75 0.75 0.75 0.1
Gothenburg DRF 0°
7.2 6.1 (5.2–6.7) 7.4 (7.2–7.6)
Glasgow DRF 30°
Glasgow DRF 60°
4.5 4.7 (4.5–5.0) 3.3 3.7 (3.2–4.0)
4.3 4.2 (4.1–4.3)
4.7 (4.4–5.0) 4.1 (4.0–4.2) 4
4.1 (3.5–5.0) 4.1 4
Gothenburg DRF 60°
3.2 1.6 3.7 (1.9–5.2) 1.8 (1.4–2.2)
the phantom to allow the surface dose distribution to be viewed in order to give an indication of the level of protection provided by the frames. Dosimetry records kept by Greater Glasgow Health Board for interventional radiology staff which included details of the number and types of procedure and the doses for each operator were consulted to assess the potential importance of protective eyewear in limiting dose. In Gothenburg, 12 pairs of lead glasses, provided by Sweden based companies, were evaluated (table 1, figure 2). Anthropomorphic phantoms were used to simulate the patient and the operator. An electronic dosemeter (EDD-30, RaySafe, Billdal, Sweden) with a sensor measuring 22 × 11 × 6 mm3 was placed on the left eye of the head phantom representing the operator (position C in figure 3). The shape of the head phantom allowed the sensor to be placed on the eye without affecting the position of the lead glasses. EDD-30 was calibrated with an N80 spectrum on an ICRU body phantom, as defined in ISO 4037 to measure Hp(0.07). This provides an adequate representation of Hp(3) for x-ray photons (Behrens 2012). The x-ray unit used was a Philips BV 300 mobile C-arm, operating at 82 kV, with a filtration equivalent to 8 mm of aluminium. A standard postero–anterior projection was employed. Measurements of the dose from scattered radiation were made with and without lead glasses for the position of 60° in the horizontal plane in figure 2(b). Two different head angles were used; one when the head phantom was angled straight forward, simulating an operator looking at the monitor with radiation incident from the left (Gothenburg 60°) (figure 2), and a second with the head phantom angled obliquely down and to the left, simulating an operator looking directly at the 816
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Figure 4. Histogram showing the distribution of DRF measurements for lead glasses
tested.
irradiated part of the patient (Gothenburg 0°). The first position was similar to the 60° angle experiment carried out in Glasgow. Lead glasses number 29 and 30 were placed on the head phantom with an additional space to the skin, simulating the wearing of a pair of regular spectacles underneath. A similar adjustment was not made for the Glasgow tests. For each pair of glasses the dose was measured for 60 s fluoroscopy time (82 kV, 2.82 mA), leading to doses of 1.3–7.4 µSv, depending on incident angle and the pair of glasses, corresponding to dose rates of 78–444 µSv h−1. These were within the specified ranges for dose of 10 nSv–9999 Sv, and dose rate of 30 µSv–2 Sv h−1. The human factor involved in placement of the glasses on the phantom over the dosemeter is more significant in determining the uncertainty in the measured doses, than the uncertainties in the measured dose. As a test, measurements were repeated ten times for one pair of glasses, which were taken off the phantom and repositioned between measurements. The DRF range was 6.3–7.4, with a mean of 6.7 and the standard deviation was 0.36. The net uncertainty in the measurements including both the human factor and the uncertainty in the measured dose is ±6%. To investigate the DRF from ceiling suspended shields as an alternative to lead glasses, additional measurements of collar dose with an electronic personal dosimeter (DoseAware, RaySafe, Billdal, Sweden) were made for three interventional radiologists while performing similar procedures in Glasgow. The radiologists were observed throughout the measurements and the doses recorded while they were using and not using a ceiling suspended screen. 3. Results The DRFs for all glasses tested are shown in table 1. Means and ranges of DRFs for groupings of different categories of lead glasses are given in table 2. The DRFs from all the measurements are plotted in the form of histograms in figure 4. The mean DRF of all lead glasses and geometric set ups was 4.5. All lead glasses tested with the head angled towards the patient (Gothenburg DRF 0°) showed good protection capacities, with a mean DRF of 6.4 and 817
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Table 3. The effect of wearing lead glasses on an interventional clinician’s highest
possible workload. Protection No Lead Glasses Mean DRF for all lead glasses (4.5) Application of a DRF of 2.0
Collar dose for 315 procedures (mSv)
No. of procedures to reach 20 mSv collar dose
No. of procedures to reach 20 mSv projected eye dose
30 6.7
210 945
280 1260
15
420
560
standard deviation (SD) of 0.7. These results only showed minor differences in DRF between the different pairs, although the pair with the smallest lenses (25) had the lowest DRF. Moving through the columns in table 2 from left to right shows a steady reduction in DRF for all categories of glasses as the angle of incidence in the horizontal plane increased. When the source of scatter was below the head and the angle in the horizontal plane was 30° (Glasgow DRF 30°) the mean DRF was 4.3 (excluding face shield data), and for the 60° angulation (combined data DRF 60°) this fell to 3.6. In Gothenburg, the lowest DRF observed was 1.4 and the highest was 7.6. In the Glasgow results the lowest DRF observed for a lead equivalence of 0.75 mm was 3.2 and the highest was 5.0. The face mask, with a lead equivalence of 0.1 mm, had a DRF of 4.0 and the 0.50 mm lead equivalence category 2 glasses had a DRF of 3.3. Pictures obtained using Gafchromic XR-QA2 film were used to give a qualitative assessment of the protection afforded by the fitover glasses (category 6) and revealed that protection in the frames was much less than that in the lenses. The records kept by Greater Glasgow Health Board showed that the highest annual dose measured at the collar for one interventional radiologist at Glasgow was about 30 mSv, which equates to an eye dose of about 22.5 mSv (×0.75) (Martin 2011). This radiologist performed 315 interventional procedures in 2012, but did not at that time wear lead glasses or use the ceiling suspended shield consistently. It should be noted that since the doses for some months were between 4 mSv and 5 mSv, there is a potential for annual doses to be substantially higher than 30 mSv. Table 3 displays the effect that wearing a pair of lead glasses could have on such an interventional radiologist’s doses and the workloads that could be undertaken before exceeding the proposed dose limit. The results show that with no protection, the radiologist workload would need to be restricted to ensure that he did not reach the 20 mSv limit, but wearing lead glasses with a DRF of 4.5 could potentially increase this number to 945 (table 3) before 20 mSv was reached on the collar dosemeter. Results from other centres have shown that annual doses to the eyes of some interventional clinicians may be 50 mSv–100 mSv (Koukorava et al 2011a, Jacob et al 2013, Martin and Magee 2013), requiring substantial additional improvement in training and implementation of radiological protection (ICRP 2013). Measurements made on Glasgow interventional radiologists indicated that during periods when a ceiling suspended screen was positioned effectively, a DRF of 5 could be achieved using the screen. 4. Discussion 4.1. Variation in DRF with design of glasses
In this study assessments of DRFs have been made for protective eye wear in head orientations likely to be encountered in routine clinical practice (figure 2). Tables 1 and 2 show the 818
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
variation in DRF for different designs of lead glasses and exposure situations. Results for each group will be discussed in turn. Glasses of categories 1 and 2 provided reasonable protection for the eyes from scattered radiation with DRFs between 4.1 and 5.0. One pair of category 2 glasses with 0.50 mm lead equivalent lenses had a DRF at 30° of 3.3. The effectiveness in protecting the eyes from radiation which is angled upwards depends on the closeness of the fit to the facial contours. One pair with opaque metal side-shields had an adjustable frame, which allowed optimal fitting of the glasses and ensured no large gaps were present. Glasgow radiologists favoured this design feature as they felt it allowed them to achieve the best fit possible and, unlike some designs, the glasses did not ‘steam up’. The design of category 3 glasses was based on an adaptation of standard spectacles with side shielding extending below the side of the lens to provide a closer fit to the face. These had DRFs between 3.2 and 4.0. The pair of category 4 lead glasses tested only achieved a DRF of 1.6 at an angle of 60°. When these glasses were placed on the head phantom a relatively large gap was created between the phantom surface and the glasses due to the poor design, allowing scattered radiation to hit the dosemeter placed on the eye directly without penetrating the lead glass. Category 5 glasses had front lenses angled so as to provide more protection for irradiation from the side and eliminated the need for side shields and the gaps between the frames and the head were smaller. DRFs in the forward and 30° positions were between 4.4 and 6.7, while most at 60° were between 2.9 and 5.2. However, one pair, number 25, had smaller lenses and a DRF at 60° of 1.9. The lower value for the DRF for this pair is thought to be caused by the proximity of irradiated tissue resulting from the restricted coverage by the smaller lenses, as the dosemeter had not affected the positioning of the glasses. When pairs of category 5 glasses were trialled by radiologists in Glasgow, it was reported that some of these glasses had a tendency to ‘steam up’ and therefore reduce vision. The fitover glasses (category 6) were designed to be worn over prescription spectacles. These glasses were bulky, not a secure fit and had large gaps underneath to allow wearing of conventional spectacles. The front lenses were the largest of all and provided the best protection when the head was facing the patient (Gothenburg 0°). However, when the head was angled the DRF fell dramatically from 7.4 (Gothenburg 0°) to about 1.8 (Gothenburg 60°) for experiments in which a gap was left to replicate fitting over conventional spectacles. The Glasgow DRF for this category, where no additional gap was left during the tests, was significantly higher, 4.1 at both 30° and 60°. The low DRF values were due mainly to the larger space left between the glasses and the head for the prescription spectacles. It is difficult to allow for the gap precisely, because of the rigid nature of the phantom surfaces, but results highlight the need for caution where this geometry is used. Gafchromic XR-QA2 film analysis highlighted that protection in the frames of the fitover lead glasses appeared to be virtually non-existent, presumably to keep the weight down. If the operator’s head is angled towards the monitor (figure 2(b)), which is likely to be the case for the majority of the time, then the scattered radiation is able to pass through gaps behind the lenses and through parts of the frame that are not protected to irradiate the eyes directly. Glasgow radiologists found the fitover glasses to be less comfortable, possibly due to the greater weight, and so purchase of prescription lead glasses provided a better option for several reasons. The face mask (category 7) had a DRF of 4.0, despite being of a lower lead equivalence. The mask covered the whole of the face and so also reduced the exposure of regions of the head surrounding the eyes that would make a significant contribution to the eye dose from backscatter. The dose reduction in this study were similar to that reported by Galster et al (2013), who observed that eye dose could be increased by a face mask when the operator was facing away from the patient due to scatter. However the face mask was not favoured by 819
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
radiologists due to its inconvenient size and some reported reduced vision when testing it, because of the tendency to steam up. 4.2. Comparison with results of other studies
DRFs from the Gothenburg measurements with the operator facing the patient were between 5.2 and 7.6. Values for DRFs between 5 and 10 have been reported for a variety of lead glasses when protecting against x-rays incident from the front in the same horizontal plane as the eyes from experimental measurements (Moore et al 1980, Marshall et al 1992, Thornton et al 2010, Galster et al 2013, McVey et al 2013, Van Rooijen et al 2013) and Monte Carlo simulations (Koukorava et al 2014). When the head was at an angle to the direction of irradiation, the DRF was lower, usually between 1.4 and 5.2. McVey et al (2013) reported that DRFs for a pair of category 5 lead glasses varied between 5 and 7 with the angle of incidence for scattered x-ray beams incident in the same horizontal plane as the head. Results from the angled measurements in the present study are at the lower end of the range reported in other studies. This is likely to be due to the use of scattered radiation incident from below the level of the eyes in an arrangement closer to that occurring in clinical practice. Van Rooijen et al (2013) found that in an experiment simulating clinical practice with the head above the level of the phantom and rotated by an angle of 45° to the direction of the x-ray scatter in the horizontal plane, the mean DRF for lead glasses tested was 2.1, compared to 7.9–10.0 for frontal exposures. 4.3. Factors determining DRFs for lead glasses
Other studies of lead glasses have concluded that the dose to the eyes when protective eyewear is worn results primarily from radiation scattered from surrounding tissues (Moore et al 1980, Day and Forster 1981, Cousin et al 1987, McVey et al 2013). This conclusion is supported by results from the face mask (category 7), which had a relatively high DRF of 4.0, despite being of a lower lead equivalence (0.1 mm), because of the reduction in the contribution from backscatter. This result demonstrates that the proximity to the eyes of tissue on which unattenuated x-rays are incident is a major factor in determining the dose to the eye lens. For exposures from the front, differences between various categories of glasses appeared to relate to the sizes of the lenses, and so the proximity of unprotected and therefore irradiated tissue. For example the pair of glasses number 25 which had smaller lenses had a DRF of 5.2 when irradiated from the front, compared to an average of 6.5 for other models. With exposures from the side, the eye dose depends on the extent of the protection from the side. When the scattered radiation is incident from a below, radiation may enter not only from the side, but also through the gap underneath the glass lenses. The wraparound design tended to give a closer fit, but model 25, which had smaller lenses had a lower DRF when irradiated from the side, presumably relating to the proximity of irradiated tissue to the eye itself. The PA projection employed for the measurements is more frequently used and the results should be representative of most exposure scenarios where the x-ray beam passes through the patient. Monte Carlo simulations have shown that the DRF may be lower for left lateral projections where the x-ray tube is adjacent to the operator and x-rays are scattered towards the operator from the surface of the body (Koukorava et al 2014), but this arrangement is used infrequently for the majority of procedures and not recommended because of the protection implications. For larger patients, the tube potential can rise towards 100 kV, but Monte Carlo simulations have shown that this will change the level of protection by less than 5% (Koukorava et al 2014). 820
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
One pair of category 2 glasses with 0.50 mm lead equivalent lenses had a DRF of 3.3. Thus 0.50 mm lead equivalence lenses could provide sufficient reduction to keep the eye dose well below the proposed dose limit, as suggested by Koukorava et al (2011b), particularly if the lenses were large. A lighter pair of lead plastic goggles equivalent to between 0.3 mm and 0.5 mm of lead, providing a close fit at the base and offering more extensive protection to the side could be an alternative design option. Lead glasses should be individually tried out by the user prior to purchase to minimise the gap between the skin and the lead glasses and thereby maximising the DRF. 4.4. Recommendations on protective eyewear
Measurements of the protection offered by lead glasses such as those described in this paper can provide useful data based on which adjustments can be made to dose values recorded by unshielded eye dosemeters to derive a dose representing that to the lens of the eye. However, DRFs are critically dependent on the experimental arrangement and a straight measure of attenuation of x-rays incident from the front will significantly overestimate the DRF in clinical practice. Moreover, many departments will not have the facility to make such measurements. Therefore it would be beneficial to have a factor that could be applied to dosemeter results for any interventional clinicians for whom it could be guaranteed that they wore the protective eyewear consistently. The mean DRFs from the measurements in the studies carried out are combined in a histogram in figure 4. Only for the least favourable orientation for the fitover lead glasses is the DRF less than 2.0, and for most situations the DRF is between 3 and 6. Based on these results, division by a DRF of 2 would appear to be both a reasonable and conservative approach to account for the protection offered by all lead glasses. The majority will have a DRF over 3 and, if an arrangement for approval of designs for protective eyewear could be established, a higher value could be used or approved designs. However any calculations assume that the lead glasses are worn for every procedure. Therefore, in order for this or any other factor to be applied, certain quality assurance measures must be in place. Firstly, the eyewear should be of appropriate design, with either side shields or of a wraparound design. Secondly, quality assurance procedures should be in place, such as regular documented checks to confirm that the interventional clinician concerned always wears the protective eyewear. Reports on dose reductions to the eyes achieved through use of ceiling suspended screens are varied. The DRF of 5 recorded for use of a ceiling suspended screen in the present study corresponds to periods where the interventional radiologist carefully positioned the screen throughout the procedure. A large scale report of clinical measurements for interventional procedures gave DRFs between 1.3 and 7 (Vanhavere et al 2012), a review comparing doses from groups at different centres performing similar procedures gave DRFs of 0.7–2.5 (Jacob et al 2013), while a study at one cardiology centre reported a DRF of 19 (Maeder et al 2006). Monte Carlo simulations of the reduction in collar dose from the use of ceiling suspended screens gave a range of DRFs from 1.03 to 33 for different projections, with a quartile range from 1.1 to 8, and a DRF calculated from the mean of the dose ratios of 2.0 (Koukorava et al 2014). Although the screens in principle give good protection, their effective use for the range of projections throughout clinical procedures is likely to limit the overall level of protection that can be achieved, but reductions by a factors of 2 should be achievable. Nevertheless, use of protective eyewear is an important component of the protection armoury. Calculations showed that wearing lead glasses should minimise any restrictions that would need to be placed on an interventional radiologist’s workload (table 3). Application of a DRF of 2 for those who consistently wear protective eyewear coupled with training in effective use of 821
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
ceiling suspended screens should ensure that few interventional clinicians exceed an annual dose constraint of 15 mSv. 5. Conclusions Measurements of DRFs performed with a variety of lead glasses, using head orientations similar to those encountered in clinical practice, have shown that reductions in the dose to the eyes is typically by factors of 3–5, in addition to the protection offered by ceiling suspended shields. However, glasses designed to fit over prescription spectacles and designs with a poorer fit or smaller lenses gave a DRF less than 2 when irradiated at 60°. The consistent wearing of lead glasses and effective use of ceiling suspended shields could eliminate the need for restrictions on workload for the majority of interventional radiologists and cardiologists. Lead glasses of wraparound design (category 5) or with large front lenses and side shielding (categories 1 and 2) provide a reasonable level of protection. The closeness of the fit to the facial contours, particularly beneath the eyes, and the size of the lenses are both important. When assessments of the dose to the eyes for operators wearing lead glasses are made, it is necessary to take the protection provided by the lead glasses into account. If local measurements to determine the DRF for specific lead glasses are not possible, hospitals should apply a DRF of 2 for clinicians wearing lead glasses consistently. Acknowledgments The authors wish to thank APC Cardiovascular Ltd, A. Somerville Ltd, Bartec Technologies, Protecx, Rothband Ltd, Scanflex Medical and Mediel for the loan of protective eyewear. The glasses tested were a selection of models and did not cover all models available. References Ainsbury E A, Bouffler S D, Dörr W, Graw J, Muirhead C R, Edwards A A and Cooper J 2009 Radiation cataractogenesis: a review of recent studies Radiat. Res. 172 1–9 Behrens R 2012 On the operational quantity Hp(3) for eye lens dosimetry J. Radiol. Prot. 32 455–64 Ciraj-Bjelac O, Rehani M M, Sim K H, Liew H B, Vano E and Kleiman N J 2010 Risk for radiationinduced cataract for staff in interventional cardiology: is there reason for concern? Catheter. Cardiovasc. Interv. 76 826–34 Cousin A J, Lawdahl R B, Chakraborty D P and Koehler R E 1987 The case for radioprotective eyewear / facewear. Practical implications and suggestions Invest. Radiol. 22 688–92 Day M J and Forster E 1981 Protective effect of spectacles Br. J. Radiol. 54 137–8 European Commission 2013 Council Directive (Euratom 2013/59) laying down basic safety standards for protection against the dangers arising from exposure to ionising radiation Official J. Eur. Commun. 2014 L13 (http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2014:013:0001:0073: EN: PDF) Galster M, Guhl C, Uder M and Adamus R 2013 Exposition of the operator’s eye lens and efficacy of radiation shielding in fluoroscopically guided interventions Fortschr. Röntgenstr. 185 474–81 ICRP 2011 Statement on tissue reactions ICRP 4825-3093-1464 April 2011 ICRP 2012 ICRP Statement on tissue reactions / early and late effects of radiation in normal tissues and organs—threshold doses for tissue reactions in a radiation protection context ICRP Publication 118 Ann. ICRP 41 1–2 ICRP 2013 Radiological protection in cardiology, ICRP Publication 120 Ann ICRP 42 International Atomic Energy Agency 2014 Radiation Protection and Safety of Radiation Sources: International Basic Safety Standards (General Safety Requirements Part 3) (Vienna: IAEA) (www-pub.iaea.org/MTCD/Publications/PDF/Pub1578_web-57265295.pdf) 822
J S Magee et al
J. Radiol. Prot. 34 (2014) 811
Jacob S, Donadille L, Maccia C, Bar O, Boveda S, Laurier D and Bernier M-O 2013 Eye lens radiation exposure to interventional cardiologists: a retrospective assessment of cumulative doses Rad. Prot. Dosim. 153 282–93 Kim K P, Miller D L, Balter S, Kleinerman R A, Linet M S, Kwon D and Simon S L 2008 Occupational radiation doses to operators performing cardiac catheterization procedures Health Phys. 94 211–27 Kim K P, Miller D L, Berrington de Gonzalez A, Balter S, Kleinerman R A, Ostroumova E, Simon S L and Linet M S. 2012 Occupational radiation doses to operators performing fluoroscopically-guided procedures Health Phys. 103 80–99 Koukorava C, Carinou E, Ferrari P, Krim S, Struelens L, International Workshop on Optimization of Radiation Protection of Medical Staff and ORAMED 2011b Study of the parameters affecting operator doses in interventional radiology using Monte Carlo simulations ORAMED 46 1216–22 Koukorava C, Carinou E, Simantirakis G, Vrachliotis T G, Archontakis E, Tierris C and Dimitriou P 2011a Doses to moperators during interventional radiology procedures: focus on eye lens and extremity dosimeters Rad. Prot. Dosim. 144 482–6 Koukorava C, Farah J, Struelens L, Clairand I, Donadille L, Vanhavere F and Dimitriou P 2014 Efficiency of radiation protection equipment in interventional radiology: a systematic Monte Carlo study of eye lens and whole body doses J. Radiol. Prot. 34 509–28 Maeder M, Brunner-La Rocca H P, Wolber T, Ammann P, Roelli H, Rohner F and Rickli H 2006 Impact of a lead glass screen on scatter radiation to eyes and hands in interventional cardiologists Catheter. Cardiovasc. Interv. 67 18–23 Marshall N W, Faulkner K and Clarke P 1992 An investigation into the effect of protective devices on the dose to radiosensitive organs in the head and neck Br. J. Radiol. 65 799–802 Martin C J 2009 A review of radiology staff doses and dose monitoring requirements Radiat. Prot. Dosim. 136 140–57 Martin C J 2011 What are the implications of the proposed revision of the eye dose limit for interventional operators? Br. J. Radiol. 84 961–2 Martin C J and Magee J S 2013 Assessment of eye and body dose for interventional radiologists, cardiologists, and other interventional staff J. Radiol. Prot. 33 445–60 Martin C J, Gentle D J, Sookpeng S and Loveland J 2011 Application of Gafchromic film in the study of dosimetry methods in CT phantoms J. Radiol. Prot. 31 389–409 McVey S, Sandison A and Sutton D G 2013 An assessment of lead eyewear in interventional radiology J. Radiol. Prot. 33 647–59 Moore W E, Ferguson G and Rohrmann C 1980 Physical factors determining the utility of radiation safety glasses Med. Phys. 7 8–12 Thornton R, Dauer L T, Altamirano J P, Alvarado K J, St. Germain J and Solomon S B 2010 Comparing strategies for operator eye protection in the interventional suite J. Vasc. Int. Radiol. 21 1703–7 Van Rooijen B D, de Hann M W, Das M, Arnoldussen C W K P, de Graaf R, van Zwam W H, Backes W H and Jeukens C R L P N 2013 Efficacy of radiation safety glasses in interventional radiology Cardiovasc. Intervent. Radiol. 37 1149–55 Vanhavere F et al 2012 ORAMED: optimisation of radiation protection for medical staff 7th EURADOS Report 2012-02
823
Copyright of Journal of Radiological Protection is the property of IOP Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
