Occupational radiation exposure in doctors: an analysis of exposure rates over 25 years

Abstract

Objectives:

Healthcare professionals’ occupational exposure to ionising radiation may be increasing due to increasing use of imaging and image-guided intervention. This study aims to assess the occupational exposure of doctors over a 25-year period at an NHS teaching hospital.

Methods:

Dosemeter measurements were collected prospectively from 1995 to 2019. Two retrospective analyses were performed over time (first including all measurements, second excluding “zero-dose” measurements), and by speciality. Group comparisons were undertaken using multilevel linear regression; a p-value <0.05 was deemed significant.

Results:

8,892 measurements (3,983 body, 1,514 collar, 649 eye, 2,846 hand), of which 3,350 were non-zero measurements (1,541 body, 883 collar, 155 eye, 771 hand), were included. Whole dataset analysis found a significant decrease in exposure for radiologists and cardiologists, as measured by body, hand and collar dosemeters over the last 25 years (p < 0.01 for all). The non-zero readings reflect the whole cohort analysis except in the case of eye dosemeters, which showed a significant decrease in exposure for cardiologists (p < 0.01), but a significant increase for radiologists and surgeons/anaesthetists (p < 0.01 for both).

Conclusions:

Whilst ionising radiation remains an occupational risk for doctors, the overall decreasing trend in occupational exposure is reassuring. However, a significant rise in eye dose for radiologists, surgeons and anaesthetists is concerning, and close monitoring is required to prevent future issues.

Advances in knowledge:

This paper is one of few evaluating the occupational radiation exposure to doctors over a 25-year period, showing that although most dosemeter measurements reflect decreasing exposure, the increase in eye exposure warrants caution.

Background

The use of ionising radiation is vital in modern medicine, for both diagnostic and interventional procedures. In the UK, the Diagnostic Imaging Dataset,¹ which collects data on the diagnostic imaging tests performed on NHS patients in England, shows imaging using ionising radiation has markedly increased over the last 10 years. While the use of ionising radiation is always justified on a per patient basis when the indication is appropriate and the dose is optimised, the occupational exposure to the operator is increasingly considered.²

The Multi Specialty Occupational Health Group, comprised of professional societies whose members perform invasive procedures, identified health concerns related to a career working with ionising radiation,³ and there is an abundance of evidence of describing adverse health effects due to exposure to ionising radiation. The International Nuclear Workers Study (INWORKS),⁴ a multinational cohort study, demonstrated a significant association between red bone marrow dose and the risk of leukaemia, and between colon dose and the risk of solid cancers. The recently updated analysis of the National Registry for Radiation Workers⁵ showed mortality and incidence risks were significantly raised for malignant neoplasms. Relative risk for both mortality and incidence of cancer remained significantly raised when information relating to cumulative doses above 100  mSv was excluded. These studies have assembled strong evidence to strengthen the scientific basis for the protection of adults from low dose, low-dose rate, exposures to ionising radiation. The typical exposure within a healthcare setting mimics this chronic low-dose exposure; however, there are relatively few studies examining adverse effects of occupational exposure in healthcare workers.^6,7 Medical radiation workers now represent the largest group of workers who are occupationally exposed to radiation, and their number is rapidly increasing worldwide.^8,9

Due to these potential risks, the use of ionising radiation within healthcare settings is highly regulated. The Health and Safety Executive Ionising Radiation Regulations¹⁰ delineate the approved code of practice and guidance for working with ionising radiation and set annual limits for occupational radiation exposure. These UK dose limits can be traced back to recommendations made by International Commission on Radiological Protection (ICRP), which have been adopted in many other countries.^11,12 Healthcare staff working with ionising radiation are routinely supplied with personal dosemeters, which measure operational dose quantities. Typically operational dose quantities will overestimate effective dose unless a conversion factor is applied. Effective dose represents a radiation dose quantity created by the ICRP¹³ specifically for the assessment of radiation risk. Several personal dosemeters can be worn simultaneously by those thought to be exposed to high levels of ionising radiation; commonly a body dosemeter, collar dosemeter, eye dosemeter, and left- and right-hand dosemeters. The purpose of monitoring individual staff doses is not confined only to demonstrating compliance with dose limits. There is a regulatory requirement to ensure doses are as low as reasonably practicable (ALARP). In practice, measuring doses to individuals is a key step in the optimisation process required to keep them as low as reasonably practicable. When dose investigation levels are exceeded within our organisation, individuals are notified, and asked to consider the reasons and discuss them with local radiation protection supervisors and the medical physics team.

The increase in imaging and interventional work reliant on ionising radiation has led to thought leaders in this area expressing concerns regarding the potentially increasing rate of occupational exposure to ionising radiation.¹⁴ As per the Ionising Radiation Regulations,¹⁰ data from dosemeters worn by healthcare professionals are collected prospectively at our institution. We performed a retrospective analysis of dosemeter data, with the aim of assessing how doctors’ occupational exposure to ionising radiation has changed over 25 years at our tertiary institution.

Methods

Our hospital trust is a tertiary referral centre for interventional radiology, performing a wide range of interventional procedures in five cardiac catheter laboratories, three interventional radiology theatres, and a hybrid X-ray theatre. Staff working with ionising radiation wear appropriate personal protective equipment (PPE), usually in the form of wrap-around lead aprons (0.25–0.5  mm lead equivalence) and a thyroid collar (0.5  mm equivalence), plus lead glasses when deemed appropriate. Staff who work with ionising radiation may be issued with LANDAUER luxel +personal dosemeters (body and collar), LANDAUER VISION (eye) personal dosemeters and LANDAUER Saturn ring dosemeters (hand). Those who are thought to be exposed to low levels (and unlikely to accrue high eye or extremity doses) are generally issued with one (body), or two (body and collar) dosemeters, and those thought to be exposed to higher levels are issued with up to five dosemeters (body, collar, eye, left and right hand) as their exposure is not considered to be homogeneously received, with some parts of the body more affected, for example, extremities. Current UK practice dictates that the body dosemeter is worn inside the lead apron, whereas all the other dosemeters are worn on the outside of the PPE. Practice has changed within the study data collection period, and for the initial years of this study, the collar badge was worn under the thyroid shield. Data from these dosemeters are collected prospectively by the Medical Physics department and has been electronically stored since 1995.

Institutional approval was granted for this study, which retrospectively analysed the data collected from dosemeters from 1995 to 2019 (25 year period).

Inclusion criteria:

• Doctors who were issued a dosemeter during any of the period from 1995 to 2019.

Exclusion criteria:

• Doctors who work primarily at another institution

• All non-doctor staff members.

The doctors who were issued dosemeters fell under four specialities: radiology, cardiology, surgery and anaesthesiology.

Statistical methods

Each staff member was allocated a participant number in order to pseudoanonymise the data set. Separate statistical analyses were performed for each of three groups: radiologists, cardiologists, and surgeons and anaesthetists (due to the low numbers of surgeons and anaesthetists, these were grouped together for the purpose of statistical analysis).

Data were collected and analysed at the year level, with one outcome measurement representing the radiation exposure for a single doctor over a one-year period. Doctors contributed data over different periods of time, with some contributing data for just a single year, whilst others contributed data in multiple years. As a result, all analyses were performed using multilevel linear regression. Two-level models were used with individual measurements nested within individual staff members. To account for skew, the outcome was log-transformed prior to analysis.

Two analyses were performed: one for the dataset in its entirety and a second excluding those returning an annual dose of ‘below measurement threshold’, interpretated as a 0 mSv or ‘zero’ dose. For the first analysis, a small constant was added to all values before transformation so that those staff members who were issued dosemeters but had a dosemeter reading of 0 mSv could be included in the analysis. For the second, dosemeter readings of 0 mSv (‘non-zero’) were excluded, and subsequently there was no requirement for a constant to be added.

The analyses were performed with radiation dose as the outcome variable, and time as the explanatory (independent) variable. To allow a flexible nature to the relationship between the outcome and time, higher order terms (squared and cubic terms) were considered in the analyses. If significant, these were retained, whilst these were omitted if not significantly adding to the fit of the model.

Results

Summary of study data

8,892 yearly measurements (typically each composed of the sum of 6 bimonthly readings) were collected between 1995 and 2019 (3,983 body dosemeter measurements, 1,514 collar dosemeter measurements, 649 eye dosemeter measurements, and 2,846 hand dosemeter measurements). Of these, 3,350 measurements were non-zero measurements (1541 body dosemeter measurements, 883 collar dosemeter measurements, 155 eye dosemeter measurements, and 771 hand dosemeter measurements). As per the methodology, two analyses were performed, the first including all available data, and a second analysis excluding ‘0 mSv’ readings; these are presented in the below tables as “Analysis 1” and ‘Analysis 2”. ’The results of the data collected are summarised in Tables 1 and 2. Table 1 highlights the figures of the number of individual yearly measurements, and also the number of doctors from which these measurements have been derived. The final column of Table 1 shows the data collection period for each speciality, as data were not collected from all groups from the whole study period.

Table 1.

Summary of analysed data. Overall numbers presented first, with number of non-zero readings presented in parentheses.

Dosemeter type	Speciality	N. measurements	N. staff members	Data collection period
Body	Radiology	2243 (900)	435 (254)	1995–2019
	Cardiology	1297 (475)	203 (137)	1995–2019
	Surgeon/Anaesthetist	443 (166)	57 (37)	1995–2019
Collar	Radiology	270 (106)	72 (39)	1995–2019
	Cardiology	1130 (736)	182 (151)	1995–2019
	Surgeon/Anaesthetist	114 (41)	21(10)	2001–2019
Eye	Radiology	166 (44)	46 (19)	2001–2019
	Cardiology	401(105)	90 (38)	1995–2019
	Surgeon/Anaesthetist	82 (6)	21(4)	2011–2019
Left hand	Radiology	459 (108)	90 (44)	1995–2019
	Cardiology	842 (273)	158 (90)	1999–2019
	Surgeon/Anaesthetist	103 (19)	24 (5)	2002–2019
Right hand	Radiology	476 (120)	92 (49)	1996–2019
	Cardiology	842 (228)	158 (79)	1999–2019
	Surgeon/Anaesthetist	124 (23)	28 (5)	2002–2019

Table 2.

Describes the mean dose, median dose and interquartile range over the last 25 years, broken down by dosemeter type and separated by speciality.

Analysis 1
Dosemeter type	Speciality	Mean (mSv)	Median (mSv)	Interquartile range
Body	Radiology	0.23	0.00	0.12
	Cardiology	0.19	0.00	0.08
	Surgeon/Anaesthetist	0.31	0.00	0.11
Collar	Radiology	0.37	0.00	0.11
	Cardiology	2.21	0.21	1.67
	Surgeon/Anaesthetist	0.59	0.00	0.21
Eye	Radiology	1.36	0.00	0.30
	Cardiology	1.14	0.00	0.30
	Surgeon/Anaesthetist	0.28	0.00	0.30
Left hand	Radiology	1.36	0.00	0.30
	Cardiology	1.14	0.00	0.30
	Surgeon/Anaesthetist	0.28	0.00	0.30
Right hand	Radiology	1.17	0.00	0.07
	Cardiology	0.72	0.00	0.30
	Surgeon/Anaesthetist	0.59	0.00	0.00
Analysis 2
Dosemeter type	Speciality	Mean	Median	Interquartile range
Body	Radiology	0.57	0.18	0.50
	Cardiology	0.53	0.23	0.45
	Surgeon/Anaesthetist	0.83	0.28	0.68
Collar	Radiology	0.94	1.01	1.10
	Cardiology	3.40	0.28	3.20
	Surgeon/Anaesthetist	0.63	0.21	0.45
Eye	Radiology	5.12	2.45	5.10
	Cardiology	4.37	2.10	3.70
	Surgeon/Anaesthetist	3.82	2.55	4.25
Left hand	Radiology	5.12	2.45	3.65
	Cardiology	4.37	2.10	2.80
	Surgeon/Anaesthetist	3.82	2.55	3.48
Right hand	Radiology	4.67	1.10	2.51
	Cardiology	2.67	1.20	2.30
	Surgeon/Anaesthetist	3.17	2.50	3.90

Table 2 highlights the overall descriptive statistics of the dataset over the 25-year study period, again with analysis one including all data, and analysis two excluding ‘0 mSv’ readings.

Body dosemeter analysis

There were 3,983 body dosemeter measurements collected between 1995 and 2019, of which 1,541 measurements recorded a non-zero dose, representing the largest group by dosemeter type. The body dosemeter analysis is summarised in Table 3, showing that for all three groups, across both analyses, there was a significant decrease in radiation dose over the course of the study period (p < 0.01 for all). In the first analysis, for both radiologists and surgeons/anaesthetists, there was a slight increase in radiation from 1995 to 2003, after which the dose declined, and by 2019 doses were at a similar level to those observed in 1995. Cardiologists’ doses declined slightly over time to a similar level observed in the other two speciality groups. In the second analysis, there was a general decrease in radiation dose over time for all three cohorts. These changes over time are represented graphically in Figure 1.

Table 3.

Change in body dosemeter measurements over time by speciality.

Speciality	Year term	Ratio (95% CI)a		P -value
		Analysis 1 (entire dataset)	Analysis 2 (excluding 0 mSv readings)	Analysis 1 (entire dataset)	Analysis 2 (excluding 0 mSv readings)
Radiology	Linear term	0.88 (0.86, 0.91)	0.69 (0.62, 0.77)	<0.01	<0.001
	Squared term	0.98 (0.97, 0.99)	0.93 (0.87, 0.99)
	Cubic term	1.02 (1.01, 1.03)	-
Cardiology	Linear term	0.91 (0.88, 0.94)	0.53 (0.43, 0.66)	<0.01	<0.001
	Squared term	1.02 (1.00, 1.03)	1.27 (1.13, 1.42)
	Cubic term	1.01 (1.00, 1.02)	1.08 (1.02, 1.15)
Surgeon/ Anaesthetist	Linear term	0.84 (0.78, 0.91)	0.60 (0.46, 0.79)	<0.01	0.001
	Squared term	0.97 (0.97, 1.00)	0.84 (0.70, 1.00)
	Cubic term	1.02 (1.00, 1.04)	-

^aAs the analysis was performed on a log scale, the regression coefficients have been converted to ratios, which are presented along with corresponding confidence intervals and represent the relative change in radiation per 5 year increase in time.

Figure 1.

Fitted regression curves for body, collar and hand data, comparing entire dataset (Analysis 1), and excluding 0mSv doses (Analysis 2).

Collar dosemeter analysis

There were 1,514 collar dosemeter measurements collected between 1995 and 2019, of which 883 measurements recorded a non-zero dose. The collar dosemeter analysis is summarised in Table 4, showing there was a significant decrease in radiation dose over time for both radiologists and cardiologists across both analyses. There was no change for the surgeon/anaesthetist group in either analysis; however, this group was relatively much smaller, representing only 114 measurements from 21 doctors, and only 42 measurements from 10 doctors when the non- zero measurements are removed. Of note, these measurements were over a shorter period (2001–2019 vs 1995–2019), and as shown by Figure 1, the highest measurements of radiologists and cardiologists were prior to 2000. There was a sharp decrease in radiation dose from 1995 to 2005 in cardiologists, with a lower rate of decline up to 2019. There was a similar trend for radiologists over this time period; however, this group started from a lower base level, and so there was a much smaller decline in radiation dose for this group. All three groups had similar exposure levels in the latter years of the study. These changes over time are represented graphically in Figure 1.

Table 4.

Change in collar dosemeter measurements over time by speciality.

Speciality	Year term	Ratio (95% CI)a		P -value
		Analysis 1 (entire dataset)	Analysis 2 (excluding 0 mSv readings)	Analysis 1 (entire dataset)	Analysis 2 (excluding 0 mSv readings)
Radiology	Linear term	0.94 (0.88, 1.01)	0.75 (0.55, 1.01)	<0.01	0.02
	Squared term	1.12 (1.06, 1.17)	1.34 (1.06, 1.68)
	Cubic term	-	-
Cardiology	Linear term	0.78 (0.71, 0.85)	0.54 (0.49, 0.61)	<0.01	<0.001
	Squared term	1.11 (1.07, 1.15)	1.24 (1.15, 1.33)
	Cubic term	0.96 (0.93, 0.98)	-
Surgeon/ Anaesthetist	Linear term	1.0 (0.91, 1.10)	1.11 (0.59, 2.08)	0.96	0.76
	Squared term	-	-
	Cubic term	-	-

^aAs the analysis was performed on a log scale, the regression coefficients have been converted to ratios, which are presented along with corresponding confidence intervals and represent the relative change in radiation per 5 year increase in time.

Eye dosemeter analysis

There were 649 eye dosemeter measurements collected between 1995 and 2019, of which 155 measurements recorded a non-zero dose. The eye dosemeter analysis is summarised in Table 5, showing that a significant decrease in radiation dose over time was observed for cardiologists across both analyses, as shown in Figure 2. Cardiologists were issued eye dosemeters during the time period 1995–2019, meaning data were recorded over a longer period of time for this group, with the highest readings seen prior to the year 2000. Conversely, for radiologists, for whom data were collected from 2001 to 2019, when ‘below measurement threshold’ doses were excluded, there was a significant rise in radiation exposure over the studied time period (p < 0.01). In the second analysis for the surgeon/anaesthetist group, the numbers were very low, with six measurements from four doctors, meaning these results are difficult to interpret. The fitted relationship is shown graphically in Figure 2, illustrating that, for cardiologists, there was a sharp decline in dose from 1995 to 2005 seen in both analyses; a period in which there were few measurements from the radiologists’ group and none from the surgeons/anaesthetists group.

Table 5.

Change in eye dosemeter measurements over time by speciality.

Speciality	Year term	Ratio (95% CI)a		P -value
		Analysis 1 (entire dataset)	Analysis 2 (excluding 0 mSv readings)	Analysis 1 (entire dataset)	Analysis 2 (excluding 0 mSv readings)
Radiology	Linear term	1.02 (0.90, 1.15)	4.68 (2.16, 10.1)	0.80	<0.001
	Squared term	-	1.58 (0.95, 2.61)		-
	Cubic term	-	0.54 (0.38, 0.77)
Cardiology	Linear term	0.92 (0.83, 1.01)	0.99 (0.72, 1.36)	<0.01	<0.001
	Squared term	1.06 (1.03, 1.10)	0.83 (0.72, 0.95)
	Cubic term	0.96 (0.94, 0.99)	-
Surgeon/ Anaesthetist	Linear term	1.15 (0.94, 1.40)	0.87 (0.81, 0.94) 15.6 (2.77, 87.5)	0.18	0.002
	Squared term	-	15.6 (2.77, 87.5)
	Cubic term	-

^aRatios reported for a 5-year increase in time.

Figure 2.

Change in eye dosemeter measurements over time by speciality.

Hand dosemeter analysis

There were 2,846 (1,404 left hand, 1442 right hand) hand dosemeter measurements collected between 1995 and 2019, of which 771 (400 left hand, 371 right hand) represented non-zero measurements. The left and right hands were analysed separately as it is considered that the dominant operating hand may be more exposed to ionising radiation that the less dominant hand.¹⁵ The hand dosemeter analysis is summarised in Table 6. For both hands, there was a significant decrease in radiation dose for cardiologists and surgeons/anaesthetists over the study period, seen in both analyses (p≤0.05 for all). There was a significant decrease in the left-hand dose for the radiologist’s group in both analyses, but no significant change in dose of right-hand measurements when non- zero doses were excluded (p = 0.93). These findings are represented graphically in Figure 1.

Table 6.

Change in hand dosemeter measurements over time by speciality.

	Speciality	Year term	Ratio (95% CI)a		P-value
			Analysis 1 (entire dataset)	Analysis 2 (excluding 0 mSv readings)	Analysis 1 (entire dataset)	Analysis 2 (excluding 0 mSv readings)
Left Hand	Radiology	Linear term	0.90 (0.84, 0.95)	1.18 (0.75, 1.86)	<0.01	0.05
		Squared term	1.08 (1.04, 1.13)	1.02 (0.86, 1.20)
	Cardiology	Cubic term	-	0.89 (0.79, 1.00)
		Linear term	0.84 (0.80, 0.87)	0.63 (0.55, 0.71)	<0.01	<0.001
	Surgeon/  Anaesthetist	Squared term	0.96 (0.93, 1.00)	0.88 (0.79, 0.98)
		Linear term	0.87 (0.75, 1.00)	0.51 (0.30, 0.86)	0.05	0.01
Right Hand	Radiology	Linear term	0.89 (0.84, 0.94)	1.01 (0.82, 1.24)	<0.01	0.93
		Squared term	1.08 (1.04, 1.12)	-
	Cardiology	Linear term	0.80 (0.75, 0.86)	0.73 (0.63, 0.84)	<0.01	<0.001
		Squared term	1.00 (0.97, 1.02)	-
	Surgeon/  Anaesthetist	Cubic term	1.04 (1.01, 1.07)	-
		Linear term	0.81 (0.72, 0.92)	0.51 (0.32, 0.83)	0.01	<0.007

^aRatios reported for a 5-year increase in time.

Discussion

Occupational exposure to ionising radiation requires close and careful monitoring to ensure occupational limits are not breached. These data show that in our centre, the occupational exposure of radiologists and cardiologists, as measured by body and collar dosemeters over the last 25 years, has decreased (p < 0.01 for both). There has also been a decrease in ionising radiation exposure to the eye in the cardiologists’ cohort (p < 0.01), but an increase in ionising radiation exposure to the eye in the radiologists’ cohort. Due to lower numbers, plus a reduced monitoring period, definitive trends in the surgeon/anaesthetists’ group were more difficult to establish, although a significant decrease in exposure was demonstrated by body and hand dosemeter measurements, whereas there was an apparent increase in eye exposure. It is difficult to explain the increase in eye dose when there is no corresponding increase in collar dose, as collar dose is intended as a surrogate for eye dose. There are multiple potential reasons for this discrepancy. There may be poor compliance with the wearing of collar dosemeters, as there has been a change in practice over the study period, whereas originally they were worn under the thyroid shield and are now worn above, and some staff members may still be wearing this badge under the thyroid shield, resulting in artificially low levels of exposure recorded. There may also have been a concurrent improvement in compliance with eye dosemeters; perhaps due to increased awareness regarding the risk of eye issues such as cataracts, or perhaps due to a recent reduction in the equivalent dose limit for the lens of the eye.¹⁰ Another explanation may be that as more staff members in a wider range of roles are issued collar badges, whereas relatively fewer staff members, generally with higher occupational exposure levels, are issued eye badges, overall low exposure of the large number of staff with collar badges may be decreasing the average exposure, whereas a few people with high eye dose exposure may be increasing the average within this smaller pool.

Our study shows cardiologists have had the highest occupational exposure over the last 25 years when compared to the other specialties. This supports previous literature, for example, a study published by Picano and Vano,¹⁶ stating that cardiologists have an exposure per person that is two to three times higher than the exposure of radiologists. However, our results show this is predominantly in the earlier years of the study period, with exposure falling into line with radiologists more recently. This may be because eye and hand dosemeters appear to have been issued to cardiologists prior to the other groups, suggesting cardiologists were monitored more closely in the earlier part of the study period. There were substantially fewer surgeons and anaesthetists issued with dosemeters, despite the fact they are often exposed to intraoperative radiation, and we will likely see increased awareness of the risk among this group in the future, with several recent papers published to this effect.^17–19

Overall, these results suggest that doctors’ occupational exposure to ionising radiation is decreasing despite a perceived increase in imaging and interventional work reliant on the use of ionising radiation. There are several reasons why this may be the case. The European Working Time Directive has limited the working hours of doctors, meaning that the increase in overall workload may be shared between more doctors, resulting in spread of the ionising radiation dose. The decrease in exposure may also be due to the increasing awareness of the risks of ionising radiation due to the changing legal framework. In 2011, the International Commission on Radiological Protection (ICRP) recommended a substantial reduction in the equivalent dose limit for the lens of the eye, in line with a reduced threshold of absorbed dose for radiation-induced cataracts,¹⁰ which has likely increased awareness and improved monitoring due to punitive results of exceeding the eye dose limit. In Europe, the Euratom 2013/59 directive²⁰ established uniform safety standards for the protection of the health of individuals subject to occupational exposures, stating operational protection of exposed workers shall be achieved through prior evaluation, optimisation of radiation protection, classification of exposed workers, control measures and monitoring, medical surveillance and education and training. As such, the undertaking of radiation protection courses and certification is now routine in many institutions, as supported by a number of specific recommendations on radiation protection published by major cardiological societies, such as the American Heart Association,²¹ European Society of Cardiology,²² and the American College of Cardiology.²³ A further influencing factor in dose reduction may be the greater availability and usage of ceiling suspended lead screens and lead glasses, and improvements in imaging equipment technology and optimisation, such as dynamic filtration, pulsed fluoroscopy and flat panel detectors. A randomised controlled trial that has been carried out evaluating patient and staff exposure with state-of-the-art X‐ray technology in cardiac catheterisation,²⁴ found that the technology significantly reduced patient dose in coronary angiographies and percutaneous coronary interventions, and in general, reduced scatter dose.

Despite this decrease, the risks of exposure to ionising radiation remain significant. The figures presented in this study show average values of radiation exposure, however, included in this study are several staff that have been ‘classified’,¹⁰ meaning they are likely to receive an effective dose in excess of 6 mSv per year or an equivalent dose in excess of 3/10ths of any relevant dose limit. Academic literature has highlighted the numerous adverse effects of ionising radiation on health,²⁵ even in when this exposure of a repeated low-dose nature.^4,5 Although this risk has not been as fully characterized in medicine compared with other industries in which staff are occupationally exposed, there is some evidence to support this heightened risk in a medical cohort. A recent study considering all diagnostic medical radiation workers in South Korea over a 15-year period²⁶ looked at potential cancer risk from occupational radiation exposure and found that although the radiation-related risk was small in most cases, it varied widely by gender and occupational group.

Due to this perceived individual risk, scientific interest has moved to individual’s genetic susceptibility to ionising radiation exposure. Gaetani et al²⁷ assessed the effect of ionising radiation exposure via DNA damage response markers in circulating lymphocytes of medical personnel occupationally exposed to radiation, finding increased DNA repair activity in ionising radiation-exposed groups, and subjects exposed to high ionising radiation doses accumulated DNA damage in their circulating cells. The authors concluded that “results indicate that chronic exposure to a low dose of ionising radiation in occupational settings induces DNA damage response in exposed subjects and may be mutagenic in workers with family history of cancer, suggesting that periodic surveillance might be advisable, along with exposure monitoring”. Similarly, El-Sayed et al²⁸ studied blood samples from clinicians performing endovascular aortic repairs, looking for specific biomarkers for DNA damage/repair and DNA damage response. They found these biomarkers were elevated immediately after radiation exposure and returned to normal by 24 h, reflecting DNA damage and repair after exposure. Interestingly, they found huge individual variation in the levels of DNA damage sustained following a single endovascular procedure, suggesting individual susceptibility to the cellular effects of ionising radiation. Borghini et al²⁹ assessed genetic markers in blood samples obtained from interventional cardiologists exposed to ionizing radiation, finding deregulation of the brain-specific microRNA-134, which is associated with certain forms of epilepsy, Alzheimer’s disease, and brain malignancies. These findings of differences in individual susceptibility to ionising radiation emphasise the importance of closely monitoring occupational exposure, even in a scenario where the average occupational exposure is decreasing over time.

Limitations

The major limitation of the study is non-compliance with dosemeter usage limiting the interpretability of results. We found high numbers of dosemeters returning an annual dose of 0 mSv, and it is very difficult to determine if these measurements are representative of zero radiation exposure, or due to the participants not wearing their dosemeters, despite being legally mandated to do so. The dosemeters used by our institution are designed in a way that requires the staff to remove a ‘tab’ in order to attach the dosemeter to its holder. When the dosemeters are collected and sent back to the manufacturer for processing, only the badges without the tab are processed, implying at least a proportion of those with 0 mSv annual reading have mounted their badge on its holder, inferring at least the intention to comply with dose monitoring. The reasons for non-compliance are complex to analyse- some staff may find it burdensome to wear multiple dosemeters, some may forget to wear their dosemeters or wear them sporadically, or some may not be wearing their dosemeters correctly. To try to counteract some of this inbuilt limitation, two analyses were performed, one including the whole cohort and one excluding ‘zero’ doses, to analyse trends within the cohort that routinely return dosemeters with positive radiation doses. The true readings are likely somewhere in between the values found in analysis 1 and 2, and a wide margin of error must be applied to this analysis. None the less, the fact that the majority of the first and second analyses give the same result is reassuring. However, in the case of eye dose, the results are concerning, and further work is required to establish a more watertight picture.

Non-compliance with dosemeters has been described as an issue in other studies, for example O’Connor et al,²⁹ who found that when issuing interventional radiologists with a second dosemeter to be worn at the neck or shoulder, the dose recorded on the majority of these second dosemeters was extremely low or zero, despite a higher recording on their first dosemeter, indicating that it is likely these staff were not complying with dose monitoring. Future advances in radiation monitoring need to determine a more robust way of ensuring compliance with dosemeters, and ideally include GPS tracking of devices linked to the user’s mobile phone.

Conclusion

The overall occupational exposure to ionising radiation of doctors appears to be decreasing, despite increasing use of imaging using ionising radiation. However, the increase in eye dose exposure is concerning, and ongoing close monitoring is required to ensure this upward trajectory is not maintained. Maintaining low levels of occupational exposure to ionising radiation is extremely important, due to the adverse health risks and likely individual susceptibility to developing adverse outcomes.