Diagnostic reference levels for fluoroscopically guided procedures in a South African tertiary hospital

Abstract

Background

The burgeoning usage and complexity of fluoroscopically guided procedures (FGPs) contribute to extended examination times and increased risk of adverse radiation effects. Diagnostic reference levels (DRLs) play a pivotal role in dose optimization. There are limited DRL data for FGPs in low- and middle-income countries (LMICs).

Purpose

To determine local DRLs (LDRLs) for common FGPs in the South African (SA) context and compare these with published international data.

Material and Methods

A three-year, retrospective study of the 15 most frequently performed FGPs at a SA institution. For each procedure, the 50th and 75th percentiles of kerma area product (KAP), reference point air kerma (K_a,r), and fluoroscopy time data were derived. Published international FGP DRL data were collated and compared with the 75th percentiles of local institutional dosage parameters.

Results

The commonest FGPs were aorto-bifemoral diagnostic angiography (n = 590), aorto-bifemoral interventional angiography (n = 287), nephrostomy (n = 265), and bronchial arterial embolization (BAE) (n = 208). Selective abdominal vessel interventional angiography (KAP = 170 Gy . cm²; K_a,r = 877 mGy) recorded the highest LDRL dosages; BAE was the longest procedure (LDRL = 38 min). Nephrostomies achieved the lowest LDRLs across all parameters (KAP = 10 Gy . cm²; K_a,r = 63 mGy, fluoroscopy time = 4.3 min). All Tygerberg Hospital LDRLs with comprehensive comparable data were within or below published ranges.

Conclusion

This study advances international radiation protection initiatives, addresses the paucity of LMIC DRL data, demonstrates broad alignment of Tygerberg Hospital FGP practice with international norms and highlights areas for optimization of institutional practice.

Keywords

Diagnostic reference levels fluoroscopy interventional radiology radiation protection dose optimization

Introduction

The demand for fluoroscopically guided procedures (FGPs) has increased substantially in recent decades, as a result of its established contribution to reduced patient morbidity, shorter hospital stays, and curtailed medical expenses. In some instances, FGPs are the only feasible treatment option, due to patient debility (1). However, the burgeoning usage and complexity of procedures have contributed to extended examination times, with increased risk of radiation-effects. FGPs are implicated in the majority of fluoroscopy-related skin injuries (2).

The International Commission on Radiological Protection (ICRP) introduced the term “Diagnostic Reference Level” (DRL) in 1996 (3) and provided practical guidance on implementation in 2001 (4). In 2002, the International Atomic Energy Agency (IAEA) adopted an “International Action Plan for Radiological Protection of Patients,” resolving to advocate widespread use of DRLs in the radiological domain (5).

According to updated ICRP terminology (6), the DRL is a notional value representing the 75th percentile of all collected median values (DRL quantities) for specific procedures, performed under similar circumstances, at specified locations, over a defined period. DRLs collected from facilities in a part of a country generate a local DRL (LDRL), those from a representative sample of a country’s healthcare facilities a national DRL, and data from multiple countries in the same geographical region, a regional DRL (6).

A range of radiation dose metrics are utilized in diagnostic imaging. For estimation of energy delivered to the patient, and as a reasonable indicator of stochastic risk, the ICRP recommends air kerma area product (KAP/P_KA). This is measured in mGy . cm². KAP is the preferred acronym, defined as the product of air kerma and the X-ray beam area, in the absence of backscatter. Dose area product (DAP) is an older, equivalent term (6).

For assessment of skin dose and the risk of skin injury, air kerma at the patient entrance reference point (K_a,r) is used. Defined as cumulative air kerma at a fixed distance from the focal spot, this is measured in Gy, with K_a,r the ICRP’s preferred acronym. Synonymous terms are “reference air kerma” and “cumulative air kerma” (6).

Multiple dose parameters are recommended for FGP DRLs, including KAP, K_a,r, fluoroscopy time, and the number of acquired images. Such an approach facilitates causal analysis when addressing “consistently exceeded” DRL values per survey (6 –10).

Setting DRLs for interventional fluoroscopy is challenging, in view of the potential for wide institutional inter-patient dose variation for the same investigation (10,11). Such variation is multifactorial and may result from differential fluoroscopy times, technique, case complexity, radiologist experience, patient size, and the number of acquired images (8,12).

A recent literature review (13) showed that less than one-quarter of the world’s LMICs have any published DRL data, with just 13 sub-Saharan African (SSA) publications, of which only four documented DRLs for FGPs. Of these, three addressed cardiology investigations (14 –16) while a single manuscript included limited data for diagnostic radiology (17).

The aim of the present study was therefore to determine DRLs for common FGPs in the South African (SA) context and to compare these with published international data.

Material and Methods

This was a retrospective study from 1 October 2015 to 30 September 2018 at Tygerberg Hospital (TBH) in Cape Town, South Africa. TBH is a 1386-bed tertiary-level public-sector facility and the main teaching hospital affiliated to the Faculty of Medicine and Health Sciences of Stellenbosch University. The TBH vascular and interventional suite is equipped with a Philips Allura Clarity fluoroscopy unit, commissioned in 2015. The radiation dose was automatically calculated at the conclusion of each procedure, with a resultant radiation structured dose report (RSDR).

An initial, comprehensive, customized search of the institutional picture archiving and communication system (PACS) was conducted for all FGPs performed at TBH in the review period. Data were collated and the 15 most frequently performed TBH FGPs were identified, stratified as diagnostic or interventional procedures, and included in this study. Less commonly performed procedures were excluded from analysis. The KAP, K_a,r, and fluoroscopy time for each included procedure were captured and the 50th and 75th percentiles of the collected data were derived, the latter defining the respective LDRLs for each procedure. Published international FGP DRL data were collated and compared with the TBH LDRLs. Lower and upper range values were considered necessary for meaningful comparison to international DRL data.

The Health Research Ethics Committee of the Faculty of Medicine and Health Sciences of Stellenbosch University approved this study (HREC reference no. S19/07/118).

Results

A total of 2446 FGPs were performed at TBH in the survey period.

The 15 most frequently performed TBH FGPs, representing a total of 2292 studies, or 94% of all procedures, with corresponding 50th and 75th dosage percentiles, are reflected in Table 1.

Table 1.

Dosimetric data for Tygerberg Hospital.

Procedure	Cases (n)	Fluoroscopy time (min)			Reference point air kerma (mGy)			Kerma air product (Gy . cm²)
Procedure	Total	Median	75th percentile	75th:50th	Median	75th percentile	75th:50th	Median	75th percentile	75th:50th
	2292
Diagnostic
Leg – aorto-bifemoral angiogram	590	4	6	1.50	84	150	1.79	31	51	1.65
Leg – lower limb angiogram (trauma)	70	4	8	2.00	79	160	2.03	18	42	2.33
Diagnostic cerebral angiogram	61	8	14	1.75	147	289	1.97	27	55	2.04
Interventional
Leg – aorto-bifemoral intervention	287	13	18	1.34	145	249	1.72	47	73	1.55
Nephrostomy (unilateral)	265	1	4	4.00	19	63	3.32	3	10	3.33
Bronchial artery embolization	208	28	38	1.36	139	259	1.86	41	73.	1.78
PTC drainage	173	12	20	1.67	161	227	1.41	23	46	2.00
Nephrostomy (bilateral)	147	2	4	2.00	21	57	2.71	4	9	2.25
Nephrostomy and stent (unilateral)	122	8	12	1.50	100	196	1.96	21	39	1.86
Selective abdominal vessels – interventional angiogram	77	16	29	1.81	381	877	2.30	102	170	1.67
Selective upper extremity – interventional angiogram (trauma)	73	6	13	2.17	48	94	1.95	14	30	2.14
Selective neck vessel – interventional angiogram (trauma)	65	15	27	1.80	404	587	1.45	45	75	1.67
Interventional cerebral angiogram	55	14	25	1.79	394	505	1.28	47	63	1.34
PTC stent ± dilatation	54	13	19	1.46	285	443	1.55	37	80	2.16
Nephrostomy and stent (bilateral)	45	17	24	1.41	118	342	2.90	22	54	2.45

PTC, percutaneous transhepatic cholangiogram.

The commonest FGPs were aorto-bifemoral diagnostic angiography (n = 590), aorto-bifemoral interventional angiography (n = 287), unilateral nephrostomy (n = 265), bronchial arterial embolization (BAE) (n = 208), and percutaneous transhepatic cholangiogram (PTC) drainage (n = 173), together accounting for two-thirds of the TBH FGP workload (1523/2292; 66%) in the review period. The highest radiation dosages were recorded for selective abdominal interventional angiography (KAP = 170 Gy.cm²; K_a,r =877 mGy); BAE was the longest procedure (LDRL =38 min). Nephrostomies achieved the lowest LDRLs across all parameters (KAP = 10 Gy.cm²; K_a,r = 63 mGy, fluoroscopy time = 4 min).

Unilateral nephrostomies showed the widest variation in dose metrics, with the 75th percentile being more than three times the 50th centile across all parameters. The narrowest range of dose metrics was achieved in interventional cerebral angiography, with the 75th percentile being, on average, 1.45 times the 50th percentile across the parameters.

TBH LDRLs are compared with published international data in Tables 2 –4.

Table 2.

Comparison of TBH with published KAP DRL data.

Procedure	KAP/DAP 75th percentile (Gy.cm²)
	TBH, 2019	International range	Rizk et al., 2019 (21)	Etard et al., 2016 (27)	Heilmaier et al., 2016 (24)	Ruiz-Cruces et al., 2016 (29)	Erskine et al., 2014 (9)	Korir et al., 2014 (17)	Zotova et al., 2012 (30)	D’Ercole et al., 2010 (28)	Miller et al., 2009 (8)	Hart et al., 2009 (25)	Trueb et al., 2009 (26)	Bleeser et al., 2008 (31)	Vano et al., 2008 (22)	Aroua et al., 2007 (12)	Verdun et al., 2005 (20)	Aroua et al., 2004 (18)	McParland, 1998 (23)
	South Africa		Lebanon	France	Switzerland	Spain	Australia	Kenya	Bulgaria	Italy	USA	UK	Switzerland	Belgium	Spain	Switzerland	Switzerland	Switzerland	Saudi Arabia
Diagnostic
Diagnostic cerebral angiogram	55	41–180.4	71	87.5			82.6		41	180.4			125	71	107	125	124	50	82.5
Leg – aorto-femoral angiogram	42	N/A					75.1
Leg – aorto-bifemoral angiogram	51	47–226	47	169		78	105.8	112	47				210	74.6	68	210	226	90	103
Interventional
Interventional cerebral angiogram	63	152.9–440	197	186.5			152.9			487.4	341					440	352
Selective neck vessel – interventional angiogram	75	N/A
Selective upper extremity – interventional angiogram	30	N/A
Bronchial artery embolization	73	131.4–170		131.4							170
Leg – aorto-bifemoral intervention	73	42–460	83		52		134.6		42				460			460
Selective abdominal vessels – interventional angiogram	170	314.1–478			314.1													478
PTC drainage	46	30–215	145	26.8	60.5	30		56	42		94	50						215	107
PTC stent ± dilatation	80	42.1–352	352	43.1												240	312		42.1
Nephrostomy (unilateral)	10	10.8–62.7	40		12.6		10.8	47			32	14			18				62.7
Nephrostomy (bilateral)	9	N/A
Nephrostomy and stent (unilateral)	39	N/A					24.1
Nephrostomy and stent (bilateral)	54	N/A

DAP, dose area product; DRL, diagnostic reference level; KAP, kerma air product; N/A, not applicable; PTC, percutaneous transhepatic cholangiogram; TBH, Tygerberg Hospital.

Table 3.

Comparison of TBH with published fluoroscopy time DRL data.

Procedure		Fluoroscopic time 75th Percentile (min)
	TBH, 2019	International range	Rizk et al., 2019 (21)	Etard et al., 2016 (27)	Ruiz-Cruces et al., 2016 (29)	Erskine et al., 2014 (9)	Korir et al., 2014 (17)	Zotova et al., 2012 (30)	D’Ercole et al., 2010 (28)	Miller et al., 2009 (8)	Hart et al., 2009 (25)	Trueb et al., 2009 (26)	Vano et al., 2008 (22)	Aroua et al., 2007 (12)	McParland, 1998 (23)
	South Africa		Lebanon	France	Spain	Australia	Kenya	Bulgaria	Italy	USA	USA	Switzerland	Spain	Switzerland	Saudi Arabia
Diagnostic
Diagnostic cerebral angiogram	14	8–15	8	10.3		6.25		12.2	12.3			15	12	15	14.2
Leg – aorto-femoral angiogram	8	N/A				3.5
Leg – aorto-bifemoral angiogram	6	2.9–12.5	4	7.2	4	5.25	9	2.9				8	3.8	8	12.5
Interventional
Interventional cerebral angiogram	25	28–90	28	58		32			46.3	90				50
Selective neck vessel – interventional angiogram	27	N/A
Selective upper extremity – interventional angiogram	13	N/A
Bronchial artery embolization	38	37.4–50		37.4						50
Leg – aorto-bifemoral intervention	18	11.25–25	16			11.25		15				25
Selective abdominal vessels – interventional angiogram	29	N/A
PTC drainage	20	9.2–30	20	14.2	17.3		23	9.2		30	15				16.6
PTC stent ± dilatation	19	9.8–31	31	17.5										25	9.8
Nephrostomy (unilateral)	4	3.5–18	11			3.5	18			15	5.1		15		10.5
Nephrostomy (bilateral)	4	N/A
Nephrostomy and stent (unilateral)	12	N/A				9
Nephrostomy and stent (bilateral)	24	N/A

DRL, diagnostic reference level; N/A, not applicable; PTC, percutaneous transhepatic cholangiogram; TBH, Tygerberg Hospital.

Table 4.

Comparison of TBH with published K_a,r DRL data.

Procedure
	TBH, 2019	International range	Rizk et al.,2019 (21)	Etard et al., 2016 (27)	Heilmaier et al., 2016 (24)	Korir et al., 2014 (17)	M iller et al., 2009 (8)
	South Africa		Lebanon	France	Switzerland	Kenya	USA
Diagnostic
Diagnostic cerebral angiogram	289	596–628	596	628
Leg – aorto-femoral angiogram	160	N/A
Leg – aorto-bifemoral angiogram	150	142–300	260	142		300
Interventional
Interventional cerebral angiogram	505	2583–4750	2583	2763			4750
Selective neck vessel – interventional angiogram	587	N/A
Selective upper extremity – interventional angiogram	94	N/A
Bronchial artery embolization	259	827–2000		827			2000
Leg – aorto-bifemoral intervention	249	260–503	503		260
Selective abdominal vessels – interventional angiogram	877	N/A			3710
PTC drainage	227	253–1406	1406	253	580	320	1400
PTC stent ± dilatation	443	305–2418	2418	305
Nephrostomy (unilateral)	63	80–412	412		80	245	400
Nephrostomy (bilateral)	56	N/A
Nephrostomy and stent (unilateral)	196	N/A
Nephrostomy and stent (bilateral)	342	N/A

DRL, diagnostic reference level; K_a.r, reference point air kerma; N/A, not applicable; PTC, percutaneous transhepatic cholangiogram; TBH, Tygerberg Hospital.

Comprehensive comparable international DRL data are available for just over half the TBH FGPs (8/15, 53%), namely diagnostic and interventional cerebral angiography, diagnostic and interventional aorto-bifemoral angiography, bronchial artery embolization, PTC drainage, PTC dilatation with stent placement, and unilateral nephrostomy. All three TBH LDRLs for these procedures are within, or below, published international ranges. Of note, TBH’s five commonest FGPs have comprehensive comparable published data.

Conversely, almost half the frequently performed TBH FGPs (7/15; 47%) have no, or limited, comparable international DRL data. Included in this group are single-limb diagnostic aorto-femoral angiography, selective interventional angiographic procedures of the neck, arm and abdomen, bilateral nephrostomy, as well as both unilateral and bilateral nephrostomy with stent placement.

The TBH FGPs with the most comprehensive comparative data available across all three DRL parameters included diagnostic aorto-bifemoral angiography (n = 26 values), diagnostic cerebral angiography (n = 24 values), and PTC drainage (n = 23 values). KAP was the most widely available comparable DRL parameter across international publications, with 13, 12, and 10 prior DRL studies for diagnostic aorto-bifemoral angiography, diagnostic cerebral angiography, and PTC drainage, respectively.

Internationally, the highest KAP DRL value has been recorded for selective interventional abdominal angiography (478 Gy.cm²) while interventional cerebral angiography has the highest documented DRLs for both K_a,r (4750 mGy) and procedure time (90 min) (8,18). Of note, TBH’s highest KAP and K_a,r LDRLs were recorded for selective abdominal interventional angiography. Although the limited available international data for this examination preclude definitive comparison, TBH DRL parameters are nonetheless substantially lower than values published to date. Nephrostomies achieved the lowest published DRLs across all parameters (KAP = 10.8 Gy.cm²; K_a,r = 80 mGy, fluoroscopy time = 3.5 min) and were very closely aligned with TBH data.

Discussion

A key finding of the present study is that all TBH procedures with comprehensive comparable international DRL data have DRL parameters within, or below, published norms. This suggests that TBH FGP practice is broadly aligned with global best practice.

The present study also provides seminal insights into FGP practice in the SA public sector. Three investigations, namely interventional angiography of the neck (n = 65), upper extremity (n = 73), and lower limb (n = 70), have no comparable international data and are thus unique to this manuscript. They are typically performed for Cape Town’s highly prevalent, gang-related gunshot injuries (19).

While higher KAP values tended to be associated with higher K_a,r measurements, these were not consistently associated with prolonged procedure times. This disjunction is reflected in the international literature, highlighting the multiplicity of factors contributing to the overall FGP radiation dose. It supports the ICRP recommendation (6) that a range of DRL parameters be invoked to facilitate FGP quality assurance and dose optimization.

The present study underscores the truism that low DRL parameters are not necessarily fully optimized. Unilateral nephrostomies are a case in point, recording some of the lowest institutional DRLs, but the widest range in recorded values across all parameters, as reflected in high 75th:50th percentile ratios (KAP = 3.33, K_a,r = 3.32, and fluoroscopy time = 4 min). There is thus potential for improvement through greater consistency. The wide parameter range is likely the result of inconsistent technique by junior registrars, who typically perform the study as the “entry-level” interventional procedure. Verdun et al. (20) highlighted the role of junior staff in broadening the range of recorded parameter values for a particular procedure. In contrast, TBH neuro-FGI procedures achieved both low DRLs and a narrow range of values across all parameters. This is indicative of striking consistency in technique and likely results from these procedures being the exclusive domain of a limited number of experienced personnel.

The present study also highlights the need for standardized global FGP terminology, as exemplified by nephrostomies. Our analysis stratified nephrostomy drainage as unilateral or bilateral, with or without JJ-stent placement, in line with Erskine et al.’s 2014 publication (9). Accordingly, the TBH and Erskine nephrostomy DRLs are lower than previously recorded for the undifferentiated “nephrostomy” grouping (8,21 –25). Future TBH work will focus on enhancing the specificity of abdominal angiographic descriptors, to include stratification into splenic, visceral, and renal artery embolization, in line with the majority of international publications (8,9,21,24,26,27). This principle will also be applied to diagnostic cerebral angiography, in line with the work of Etard and Erskine et al. (9,27).

A strength of the present study is its identification of the typical spectrum of FGPs currently performed in a public-sector academic hospital in an upper-middle-income country. It also highlights those procedures not yet performed in our setting, such as transjugular intrahepatic portosystemic shunt (TIPS), transarterial chemoembolization (TACE), percutaneous gastrostomy placement, and prostate artery embolization. It is anticipated that these studies will be introduced in the medium term, as radiologists complete the necessary training, and clinicians increasingly embrace minimally invasive treatment options.

The ICRP recommends the analysis of as many parameters as possible in the quest for optimization (6). The retrospective design of the present study limited analysis of a number of factors that potentially impact dosage, including patient weight and girth, procedural complexity, operator experience, and the quantum of images. Of note, Miller et al. (8) suggest that DRLs uncorrected for weight are of substantial benefit. Previous studies, notably those of D’Ercole et al. (28) and Ruiz-Cruces et al. (29) incorporated notional complexity factors for cardiology and radiology procedures.

To the best of our knowledge, this work presents the most comprehensive SSA FGP LDRL data in the radiological domain to date. As such, it makes an important contribution to ICRP initiatives for the radiation protection of patients. It also addresses the known paucity of LMIC DRL data. The single previous manuscript in this domain, the 2014 Kenyan study by Korir et al. (17), included interventional radiology DRLs for four-vessel cerebral angiography, lower-limb arteriography, nephrostomy, biliary drainage, vasogram, uterine tube catheterization, renal angiography, arterial chemoembolization, and IVC filter placement. However, the study was limited by small cohort numbers, with only three investigations (coronary angiography, four-vessel angiography, and lower limb arteriography) meeting the threshold of 30 studies as recommended by Miller et al. (8) for the calculation of DRLs with reasonable 95% confidence intervals.

In conclusion, the present study advances international radiation protection initiatives, addresses the paucity of LMIC DRL data, demonstrates broad alignment of TBH FG practice with international norms, and highlights areas for potential optimization of institutional practice.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Leon Malan

References

The Royal College of Radiologists. Investing in the clinical radiology workforce-the quality and efficiency case. London: RCR, 2012. Available at: https://www.rcr.ac.uk/sites/default/files/RCR_CRWorkforce_June2012.pdf (accessed 19 July 2018).

International Atomic Energy Agency. Establishing guidance levels in x ray guided medical interventional procedures: a pilot study. Vienna: IAEA, 2009. Available at: https://www-pub.iaea.org/MTCD/Publication/PDF/Pub1346_web.pdf (accessed 30 June 2019).

ICRP. Radiological protection and safety in medicine. ICRP Publication 73. Ann ICRP 1996;26:1–47. Available at: http://www.icrp.org/publication.asp?id=ICRP%20Publication%20135 (accessed 28 May 2018).

ICRP. 2001 Diagnostic reference levels in medical imaging: review and additional advice. Ann ICRP 2001;31:33–52. Available at: http://www.icrp.org/publication. asp?id=ICRP%20Supporting%20Guidance%202 (accessed 11 March 2018).

International Atomic Energy Agency, Pan American Health Organization, World Health Organization. Radiological protection for medical exposure to ionizing radiation. Vienna: IAEA, 2002. Available at: www.iaea.org/ns/coordinit (accessed 24 July 2018).

ICRP. Diagnostic reference levels in medical imaging. ICRP Publication 135. Ann ICRP 2017;46:7–143. Available at: http://www.icrp.org/publication.asp?id=ICRP%20 Publication%20135 (accessed 11 March 2018).

ICRP. The 2007 Recommendations of the International Commission on Radiological Protection. ICRP Publication 103. Ann ICRP 2007;37:3–34. Available at: https://journals. sagepub.com/doi/pdf/10.1177/ANIB_37_2-4 (accessed 16 June 2018).

Miller

Kwon

Bonavia

GH.

Reference levels for patient radiation doses in interventional radiology: proposed initial values for U.S. practice. Radiology 2009; 253:753–764.

Erskine

Brady

Marshall

EM.

Local diagnostic reference levels for angiographic and fluoroscopic procedures: Australian practice. Australas Phys Eng Sci Med 2014; 37:75–82.

10.

Miller

Hilohi

Spelic

, et al. Patient radiation doses in interventional cardiology in the U.S.: advisory data sets and possible initial values for U.S. reference levels. Med Phys 2012; 39:6276–6286.

11.

Vogiatzi

Kipouros

Chobis

, et al. Guidance on diagnostic reference levels (DRLs) for medical exposures. Radiat Prot Dosimetry 1998; 80:1–26.

12.

Aroua

Rickli

Stauffer

J-C

, et al. How to set up and apply reference levels in fluoroscopy at a national level. Eur Radiol 2007; 17:1621–1633.

13.

Meyer

Groenewald

Pitcher

RD.

Diagnostic reference levels in low- and middle-income countries: early “ALARAm” bells?

Acta Radiol 2017; 58:442–448.

14.

De Vos

Radiation dose optimization in interventional radiology and cardiology using diagnostic reference levels. University of Cape Town 2016. Available at: https://open.uct.ac.za/bitstream/handle/11427/20928/thesis_hsf_2016_de_vos_hendrik_johannes.pdf?sequence=1 (accessed 14 July 2018).

15.

Makosa

Conradie

Diagnostic reference levels for the vascular, screening and adult cardiology units at Universitas Academic Hospital. Phys Medica 2015; 31:S7.

16.

Jonas

Trauernicht

Kotze

A survey of initial patient diagnostic reference levels from selected fluoroscopically guided interventional procedures at Groote Schuur Hospital. Phys Medica 2016; 32:161.

17.

Korir

Wambani

Yuko-Jowi

, et al. Establishing diagnostic reference levels for interventional procedures in Kenya. Radiography 2014; 20:148–152.

18.

Aroua

Besancon

Buchillier-Decka

, et al. Adult reference levels in diagnostic and interventional radiology for temporary use in Switzerland. Radiation Prot Dosimetry 2004; 111:289–295.

19.

Statistics South Africa. Victims of Crime Survey 2017/2018. Available at: http://www.statssa.gov.za/publications/P0341/P03412018.pdf (accessed 18 August 2019).

20.

Verdun

Aroua

Trueb

, et al. Diagnostic and interventional radiology: a strategy to introduce reference dose level taking into account the national practice. Radiat Prot Dosimetry 2005; 114:188–191.

21.

Rizk

Farah

Vanhavere

, et al. National diagnostic reference levels in interventional radiology suites in Lebanon: a multicenter survey. Radiat Prot Dosimetry 2019; 187:50–60.

22.

Vano

Järvinen

Kosunen

, et al. Patient dose in interventional radiology: a European survey. Radiat Prot Dosimetry 2008; 129:39–45.

23.

McParland

BJ.

A study of patient radiation doses in interventional radiological procedures. Br J Radiol 1998; 71:175–185.

24.

Heilmaier

Zuber

Berthold

, et al. Establishing local diagnostic reference levels in IR procedures with dose management software. J Vasc Interv Radiol 2017; 28:429–441.

25.

Hart

Hillier

Wall Bf.

National reference doses for common radiographic, fluoroscopic and dental x-ray examinations in the UK. Br J Radiol 2009; 82:1–12.

26.

Trueb

Aroua

Stuessi

, et al. Diagnostic reference levels in cardiology and interventional radiology. IFMBE Proceedings 2009; 25:136–139.

27.

Etard

Bigand

Salvat

, et al. Patient dose in interventional radiology: a multicentre study of the most frequent procedures in France. Eur Radiol 2017; 27:4281–4290.

28.

D’Ercole

Thyrion

Bocchiola

, et al. Proposed local diagnostic reference levels in angiography and interventional neuroradiology and a preliminary analysis according to the complexity of the procedures. Phys Medica 2012; 28:61–70.

29.

Ruiz-Cruces

Vano

Carrera-Magarino

, et al. Diagnostic reference levels and complexity indices in interventional radiology: a national programme. Eur Radiol 2016; 26:4268–4276.

30.

Zotova

Vassileva

Hristova

, et al. A national patient dose survey and setting of reference levels for interventional radiology in Bulgaria. Eur Radiol 2012; 22:1240–1249.

31.

Bleeser

Hoornaert

Smans

, et al. Diagnostic reference levels in angiography and interventional radiology: a Belgian multi-centre study. Radiat Prot Dosimetry 2008; 129:50–55.