Article Text

Download PDFPDF

Invasive imaging
Radiation exposure in invasive cardiology
  1. Eberhard Kuon
  1. Priv. Doz. Dr Eberhard Kuon, Klinik Fraenkische Schweiz, Feuersteinstr. 2, D–91320 Ebermannstadt, Germany; eberhard.kuon{at}

Statistics from

Mean patient dose area products (DAP) in invasive cardiology are high and for specific interventional procedures vary greatly. For this reason patients may face serious radiation injuries and an increased risk of cancer in the future. Such increased risk is due less to inappropriate equipment than to inadequate operational technique and a lack of awareness of the potential for injury by individual cardiac interventionists.

This article focuses on the “as low as reasonably achievable” (ALARA) principle in invasive cardiology from the viewpoint of an experienced cardiologist, analysing the ranking and interdependency of those factors which influence patient radiation exposure—adequate length and number of radiographic runs, consistent collimation, adequate instead of best possible image quality, and basic knowledge about less irradiating angulations. Training in techniques to reduce radiation exposure enables patient doses to be achieved that are far below those used in non-invasive coronary imaging such as multislice computed tomography (MSCT). In addition, improved lead shielding is effective in maintaining operator radiation exposure below 1% of typical levels in advanced catheterisation laboratories. Traditional radiological convictions, misinterpreted and misleading since the early times of fluoroscopy guided cardiac interventions, will be disclosed and discussed.


As early as 1925, FM Groedel, founder of both the German Cardiac Society and the American College of Cardiology, identified the following factors as playing a role in the adverse effects of x rays: excessively long fluoroscopic examinations, short focus skin distances, insufficient filtering of the x ray beam, and, not least, excessive numbers of radiographs made by insufficiently trained operators and assistants.1

In invasive cardiology, however, up to now these problems have been underappreciated: directives from the International Commission of Radiation Protection (ICRP) point out that “Unfortunately there is growing evidence that many interventionists have a less than ideal understanding of the risks of radiation-induced injuries from X-rays and…are not aware of the simple methods for decreasing their incidence utilising dose control strategies”.2 This statement is bolstered by results of a UK survey, which noted that most doctors generally underestimate the radiation doses received from commonly requested radiological investigations. However, there is no dose of radiation that can be considered safe or harmless.

Currently, reported mean DAP to patients from coronary angiography and coronary angioplasty are high and typically vary widely—that is, within 16–106 Gy × cm2 and 34–109 Gy × cm2 up to mean levels of 191 Gy × cm2 for combined interventions.3 Serious skin lesions in the patients, including chronic radiodermatitis and musculocutaneous ulceration, may result.4 Radiation exposure to cardiologists may also be high and it has been suggested that hazards may include radiation induced cancer,2 particularly brain cancer. For this reason international directives from both the ICRP and the EURATOM council have established radiation protection training courses for cardiac interventionists over and above that required by the radiology community2 5 6; however, these official programmes have yet to prove that they have resulted in less radiation exposure to patients in daily routine. According to the recommended ALARA principle, “All medical exposure for radiodiagnostic purposes…shall be kept as low as reasonably achievable”,2 5 even below accepted reference values. Recently validated efficient dose reduction techniques3 7 have delivered mean (SD) patient DAP far below typical current dose levels: 4.2 (1.6) Gy × cm2 for elective coronary angiography and 6.7 (5.2) Gy × cm2 for elective one vessel percutaneous coronary intervention (PCI), equivalent to mean effective doses of 0.8 and 1.3 mSv, respectively.8 9 On the basis of this experience the “Encourage to less irradiating cardiologic interventional techniques” (ELICIT) initiative, an educational 90 min course, reduced patient dose in elective diagnostic coronary angiography by 38%.10

This article deals with the delivery of the ALARA principle in daily routine from a specific, pragmatic viewpoint of an experienced cardiologist. Improvements will not occur by repeating well known recommendations and warnings about radiotoxic effects alone. A careful analysis of the ranking and interdependency of the various factors which cause radiation exposure is required, together with education on radiation reducing techniques. How can we minimise occupational radiation risk for cardiac operators? What about patient radiation exposure from alternative methods, such as non-invasive MSCT of the heart? Could it be that well established but misinterpreted radiological recommendations, dogmatic educational traditions and even some technical improvements, insufficiently reflected in the cardiology community, are unproductive?


Roentgen (R) and rad have been replaced by the SI radiation unit Gray (Gy  =  joule/kg) for air kerma (kinetic energy released in matter) and for absorbed dose, as well as by the Sievert (Sv), applied for dose equivalent. For x rays the biological quality factor is 1 for conversion from absorbed dose to dose equivalent. Entrance skin air kerma (ESAK) is the dose to the air in the entrance plane of the patient without backscatter. Entrance skin dose includes backscatter from the upper skin layers and quantifies most appropriately radiation induced skin lesions. DAP is the product of the dose in air in a given plane by the area of this irradiating beam. It is independent of the distance from the x ray source and is best suited for interventions with varying angulations. The effective dose (ED) is the sum of all weighted dose equivalents subjected to the organs in the body and characterises stochastic cancer risks. DAP to ED conversion factors for cardiac interventions under conditions of undercouch tube position have been calculated to be approximately 0.20 mSv/Gy × cm2.11


Deterministic risks

In clinical practice faint erythema due to local exposure of 2–3 Gy, eliciting an activation of histamine-like substances, typically goes unnoticed because of its brief presence. Main erythema and epilation following threshold values of 6–8 Gy are caused by an inflammation subsequent to the destruction of basal cells in the epidermis; 10–15 Gy may cause teleangiectasis and chronic radioderma, and 16–18 Gy may cause skin ulcerations. The higher the acute radiation dose, the greater the likelihood skin damage will develop, usually within 4–16 weeks after exposure.2

A DAP of 100 Gy × cm2, applied in one constant tube angulation from a 13 cm intensifier magnification, results in exposure of an area of the patient’s skin of approximately 40 cm2, which will induce a calculated skin entrance air kerma of 2.5 Gy and—including backscatter—a skin entrance dose of approximately 3.0 Gy. For complex interventions with estimated skin doses of ⩾2 Gy, patients should be counselled on the future possibility of erythema or more serious skin damage and followed up for several weeks after exposure.2

Stochastic risks

Although there are clear benefits from the use of diagnostic x rays, it is generally acknowledged that their use involves some risk of cancer. The additional individual lifetime cancer mortality risk of an effective dose of 1 Sv over the course of 20–40 years has been calculated at 7–11% for high level and 5–10% for low level (<0.1 Sv/h) radiation intensity, and that of an intervention generating an effective dose of 40 mSv (equivalent to approximately 200 Gy × cm2) is approximately 0.2–1.0%.12 13 For patients over the age of 50 years, the risk is considerably less than the latter figure.13 In an international survey of the Cancer Research UK Epidemiology Unit, the average number of cases of radiation induced cancer per million coronary angiographies was recently estimated to be 280.14


The patient’s physical state, such as his or her weight, as well as the complexity of intervention, are factors which influence patient dose. For percutaneous coronary interventions (PCI) in clinical routine, consistent collimation could prove to be the most efficient radiation reducing factor. For diagnostic coronary angiographies, restriction to the essential number of radiographic frames, consistent collimation, and realisation of an adequate instead of best possible image quality, are each more efficient than minimising fluoroscopy time.15

Minimisation of radiographic beam time

This is a very effective step toward dose reduction, for radiographic documentation creates 12- to 20-fold higher dose intensities than fluoroscopy mode. Radiography therefore generates 60–70% of total DAP during coronary angiography. Ensuring a safe, ostial catheter tip position during fluoroscopy will enable the experienced interventionist to inject contrast medium into the coronary tree just during non-visual automatic dose control regulation of the tube immediately before radiographic documentation. After clarification of coronary flow and exclusion of collateral pathways, digital acquisition of short radiographic runs of one or two heart cycle lengths, repeated as often as necessary, will provide adequate visual impressions without any diagnostic or procedural impact.3 8 10

Optimal beam collimation to the region of interest

Using partly closed collimator leaves, optimal beam collimation to the region of interest will deliver far less DAP than collimation from an image intensifier entrance field that is wide open. In practice collimation to the sinus of Valsalva during intubation of the respective coronary orifice (“buttonhole technique”) will be achieved efficiently by short taps on the foot switch, taking advantage of the “last image hold” function.3 8 Modern x ray systems with virtual control of the collimator and semi-transparent diaphragm indeed enable collimation as desired without any fluoroscopy.

Adequate instead of best possible image quality

Achieving an adequate image, as opposed to the highest quality image that is possible, results in slightly higher contrast and an insignificantly course mottled background.16 As often as possible, image intensifier low level modes should be used. If the interventional communities accept a radiographic frame rate of 12.5/s for an adequate documentation of cardiac structures in motion, an identical fluoroscopic pulse rate of no more than 12.5/s, supported by the commonly available technical advancement of “gap filling”, should consequently be judged sufficient to guarantee diagnostic and interventional safety and accuracy.8 Moreover, the distance between the x ray tube and the patient should be practicably maximised: keeping the image intensifier as close to the patient as possible minimises the source to image distance (SID), which decreases blurring of the image and allows the image intensifier to serve as a barrier between patient and operator. Use of the lowest degree of image magnification, required for accurate interpretation, substantially reduces skin dose—dose increments due to magnification generally being higher for conventional image intensifiers systems than for flat panel detector systems.13 Again consistent collimation, plus full inspiration during radiography, and x ray equipment in good repair and calibration, are preconditions for minimising patient entrance skin air kermas (ESAK) during both fluoroscopy and radiography, below 20 mGy × cm2 per second and frame, respectively.

Less irradiating angulations

For the chest region of an anthropomorphic Rando–Alderson phantom, the mean fluoroscopic patient DAP/s were lowest in right anterior oblique (RAO) 20°/0° tube angulation. Patient DAP/s for all posteroanterior (PA) projections, cranialised and caudalised together, rose significantly: 3.7 times the PA 0° baseline values toward both, 100° left (LAO) and 100° right (RAO) anterior oblique angulation. Patient DAP/s for all PA projections, angulated to the right and left, increased approximately 2.5 times toward 30° craniocaudal angulations. Favouring craniocaudal PA view over typical LAO angulations will reduce radiation exposure in the course of diagnostic evaluation of the left coronary artery by 50%, because skeletal structures now will run just beside the collimated beam (fig 1): the caudal PA view documents the left coronary main stem, proximal and distal left anterior descending artery (LAD), and proximal and mid circumflex segments. The cranial PA view, however, is suitable for the left coronary orifice, circumflex periphery, all LAD segments including the bifurcations into its diagonal (D) branches, and collateral pathways towards the right coronary artery (RCA) (fig 2). Left coronary artery (LCA) standard angiography is completed by lateral 90°/0° LAO angulation. The 50°/0° LAO angulation visualises the proximal and mid RCA. The more proximal the bifurcation, the more the second standard cranial PA view for RCA should vary from cranial PA view toward the cranial 30°/30° RAO and finally the 30°/0° RAO view. Due to its anatomic proximity to the spine, visualisation of the RCA, however, should be shortened as far as possible.3 17 Specific diagnostic features optionally require further projections which, however, should consistently be collimated to the region of interest.3

Figure 1 Time adjusted fluoroscopic radiation exposure (mGy × cm2/s) to an anthropomorphic Rando–Alderson phantom as a function of tube angulation: 50% reduction of patient dose in the course of left coronary arteriography is achieved by favouring radiation reducing angulations (blue positions) over typical angulations (red positions). Broken line indicates additional angulations for details of the circumflex artery, if necessary.
Figure 2 30° caudal posteroanterior (PA) (left panel) and 30° cranial PA (right panel) views enable comparable left coronary visualisation, but significantly lower radiation exposures to patients and operators than the respective typical craniocaudal left anterior oblique views.

Well rested operators

Radiation exposure to patients resulting from PCI procedures is influenced by operator fatigue. Radiation exposure significantly rose by 28%—due to more and longer radiographic runs—after the cardiologists have been working for more than 6 h.18

Keep DAP within sight and mind

The DAP dose should be displayed in the catheterisation laboratory near the monitors, and not in the control centre, in order for the operator to maintain awareness of radiation safety.


For non-invasive coronary imaging by 64 slice computed tomography (CT), sensitivities ranging from 86–99% and specificities from 93–97% have been claimed. However, up to 12% of the included coronary segments were not interpretable in the respective studies. The negative predictive values ranged from 95–100%, partly due to the low prevalence of significant coronary stenoses in the study populations. MSCT tends to overestimate the extent of coronary artery disease, especially if picture quality is poor.19 The positive predictive values were lower in a wide range between 50–100%. Mean (SD) effective doses from MSCT, however, even with use of dose reducing acquisition techniques, are reported to vary from 9.5 (3.4) to 13.6 (2.8) mSv,20 21 which is substantially higher than 0.8 (0.3) mSv for coronary angiography, when consistently applied under the conditions of radiation reducing interventional techniques mentioned above.8 Moreover, the usefulness of 16-slice CT in daily routine in a community hospital outpatient setting was substantially restricted by limited diagnostic accuracy (sensitivity, specificity, positive and negative predictive values: 82%, 75%, 75%, and 81%, respectively) and non-interpretable scans (21%).22 For these reasons, particularly to exclude coronary artery disease in younger individuals with chest pain, cardiologists should maintain the undeniably valuable high resolution gold standard of invasive coronary angiography with justifiable future risks, a negligible risk of interventional complications, and the possibility of primary treatment.


The most effective technique for reducing the operator’s occupational dose entails diminishing the patient’s DAP. PA and right anterior angulations (fig 2), which enable significantly lower scatter radiation intensities (fig 3), should be favoured over left anterior angulations whenever possible. Moreover, the operator’s occupational dose depends on case load and his or her care over an adequate use of lead protection devices. Doubling the source to operator distance (SOD) will decrease operator dose to approximately one quarter.13 Interventionists should remember to step back from the patient’s isocentre diagonal to the tube, and not simply backwards in order not to leave the safe area, ensured by the overcoach lead glass screen.7

Figure 3 Dose corrected operator exposure (μSv/mGy × cm) during fluoroscopy at an anthropomorphic Rando–Alderson phantom as a function of tube angulation: significant reduction of radiation intensity during left arteriography by favouring craniocaudal, posteroanterior views over typical craniocaudal left anterior oblique views and 90° right lateral over 90° left lateral angulation, respectively.

At an anthropomorphic Rando–Alderson phantom, sufficiently characterising in vivo conditions, absolute (μSv/h) and DAP corrected (μSv/Gy × cm2) fluoroscopic scatter radiation, dependent on angulation, may be reduced by a typical 0.5 mm and improved 1.0 mm overcouch and undercouch lead shielding to 9–20% and to 2–12% of baseline, respectively. This moderate reduction in scatter radiation is due to significant radiation leakage at a height of 80–105 cm between the overcouch and undercouch shielding.7 For all angulations, significant reduction of that radiation leakage was achieved by an additional 1.0 mm lead equivalent undercouch top and overcouch flap, which—as depicted for the caudal LAO angulation (fig 4)—resulted in mean S-ESAK-O below 1% of baseline; below the personal lead protection garments (glasses, collar and apron of 0.5 mm lead equivalency and footswitch shield of 1.0 mm lead equivalency)—even for the scatter intensive “spider view”— the local S-ESAK-O levels during fluoroscopy amount to 100–200 nSv/h, which is below natural radiation exposure of 2–2.5 mSv/year. However, operators “using protective overcouch screen appropriate only occasionally”, as recently disclosed in daily routine, and working without glasses, may be exposed to a 5–10 000-fold scatter radiation exposure to the lens of their eye. Recently a 0.5 mm lead equivalent cap and a 1.0 mm lead equivalent cover around the patient’s thighs proved effective in reducing occupational exposure to the operator’s head and hands towards levels below 1% and 16%, respectively.23

Figure 4 Scatter entrance skin air kerma in operator’s position (μSv/h) during fluoroscopy, for 0–180 cm height in 20° caudal 60° left anterior oblique angulation: without lead shielding, with typical 0.5 mm and with improved 1.0 mm lead equivalent shielding, with leakage closure by undercouch (UC) top and overcouch (OC) flap, as well as with additional personal body protection.

Optimised interventional techniques, restriction to adequate instead of “best possible” image quality, elimination of essential radiation leakage by improved lead shielding, and consistent use of lead protection devices enabled individual scatter ESAK to the operator <10 nSv per applied Gy cm2 and <60 nSv per coronary angiography—far below typically published operator skin doses of <10 μSv per coronary angiography, respectively.7

DAP standardised dose parameters, because they are not appreciably influenced by the equipment age and type, or by the image intensifier entrance dose rate of the respective catheterisation system, are best suited for evaluation and comparison of various radiation protection devices.


Misinterpretation and limits of short fluoroscopy times

According to international radiological recommendations and legislation, fluoroscopy time is widely considered to be a valid criterion for the operator’s efforts to reduce patient DAP. This conviction goes back to the early days of fluoroscopy guided contrast passages of the alimentary canal in gastroenterology, which required merely a few radiographic frames for diagnostic documentation. Conclusively, fluoroscopy time—without considering collimation efforts—was a useable predictor of total DAP.

However, a 10 year long term follow up of 3115 elective coronary angiographies revealed that consistent reduction of mean fluoroscopy times to 50% resulted in an overall DAP reduction of only 20%—a logical consequence of the fact that 60–70% of total DAP for coronary angiography is due to radiographic acquisition. Minimisation of radiographic frames towards essential numbers, restriction to adequate instead of best possible image quality, as well as consistent collimation were much more efficient and enabled DAP reductions of approximately 50% each. During PCI, collimation reduced DAP by a remarkable 65%. In consequence, educational efforts to reduce radiation during routine clinical invasive cardiology should—against widely held opinion—focus less exclusively towards a reduction of fluoroscopy time but more towards reduction of radiographic frames, restriction to an adequate image quality, and consistent collimation to the region of interest, even though primarily longer fluoroscopy times may result.15

Digital acquisition is not realised in the cardiologist’s mind

The most effective step towards dose reduction is avoidance of lengthy filming. In the former era of radiographic films, acquisition runs below 4 s were not possible for a reliable visual real time diagnosis. However, the cardiac community should not adhere to historic traditions of lengthy cinegraphic filming: digital radiographic documentation consisting of short radiographic loops, one heart cycle in length, created for repeated use in clinical routine has been shown to provide satisfactory visual impressions of the coronary tree. Bypass grafts, late filling collateral pathways, and slow flow phenomena should be adequately documented by longer series. Furthermore, for a given data volume of 0.26 MB per radiographic frame with a matrix of 5122 pixels, the data volume for complete coronary angiography and PCI is approximately 25 MB and 50 MB, respectively, which in bail-out situations will be easily transferred within 100 s to the next tertiary cardiac referral centre for further online discussion.

Craniocaudal PA angulations

Craniocaudal PA views for coronary visualisation have been typically judged to be rather radiation intensive for the patient, as the radiation beam crosses the spine and sternum. This applies to the RCA due to its anatomical proximity to the thoracic spine. In the course of diagnostic evaluation of the LCA, however, craniocaudal PA views should be favoured over typical LAO angulations because skeletal structures will now run just beside a well collimated radiation beam (figs 1–3).

Reducing radiation exposure in invasive cardiology: key points

  • Reduce radiographic runs towards adequate length and number.

  • Collimate consistently to the region of interest both during fluoroscopy as well as for radiographic acquisition. Do not worry about slightly longer fluoroscopy times.

  • Optimise image intensifier entrance dose towards an adequate instead of best possible image quality.

  • Improve your knowledge about less irradiating angulations and implement them in daily practice.

  • Try to reduce fluoroscopy time, but not to the detriment of collimation or the number of radiographic runs.

Three dimensional angulation: progress backwards?

In the midst of enthusiasm concerning new coronary views now made possible by three dimensional angulation, neither cardiac interventionists nor physicists have realised the potential adverse effects with respect to the ALARA principle. Typical former catheterisation systems consisted of fixed, installed undercouch tubes and overcouch image intensifier systems and a rotatable couch to realise lateral projections. To achieve a strictly lateral panoramic view of the LCA, the patient, while fastened in his couch belts, was rotated towards the operator and his cardiac colleague, resulting in a left coronary view identical to today’s LAO 90°/0° projection. A contralateral rotation from the operator’s position—corresponding to the actual RAO 90°/0° projection—for obvious reasons was obsolete for it increased the risk of patient injury from falling out of the couch. However, taking over familiar coronary steep left lateral 90°/0° views in daily routine by the new technical device of three dimensional tube angulation, without reflecting basic principles of backscatter geometry, resulted in an approximately threefold occupational operator exposure compared to the contralateral right lateral 90°/0° angulation (fig 5).

Figure 5 Backscatter geometry from the patient towards the operator during fluoroscopy before (left panel) and after the introduction of three dimensional tube angulation for the steep lateral 90° left anterior oblique tube angulation (right panel).

You can get CPD/CME credits for Education in Heart

Education in Heart articles are accredited by both the UK Royal College of Physicians (London) and the European Board for Accreditation in Cardiology—you need to answer the accompanying multiple choice questions (MCQs). To access the questions, click on BMJ Learning: Take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to:

Please note: The MCQs are hosted on BMJ Learning—the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into Heart with their journal’s username and password. All users must also complete a one-time registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.


Based on the 75th centiles of six European reference centres (n = 100 patients, in each centre) in Greece, Spain, Italy, England, Ireland, and Finland, the European DIMOND (Measures for Optimising Radiological Information and Dose in Digital Imaging and Interventional Radiology) research cardiology group recently proposed the following reference levels for PCI: 75 Gy × cm2 for DAP, 17 min for TF, and 1300 for radiographic frames.24

Our experience of 542 elective PCIs, performed in clinical routine over the course of 3 years, demonstrated significantly lower 75th centiles—that is, 10 Gy × cm2 for DAP, 9 min for TF, and 180 for radiographic frames, respectively. This is a result of consistent and continued training of the personnel involved in PCI according to our guidelines toward reduction of radiation exposure, mentioned above.

In conclusion, reference levels should be oriented to new, promising state of the art techniques for reducing radiation exposure. Consistent realisation of radiation reducing principles in invasive cardiology and individual self surveillance in daily routine enabled 75th centiles for elective one, two and three vessel PCI, for recanalisations of chronic coronary occlusions, and for emergency PCI in acute myocardial infarction of 9, 14, 26, 21, and 24 Gy × cm2 for DAP, of 8, 14, 24, 16, and 16 min for TF, and of 153, 253, 490, 449 and 407 for the number of cinegraphic frames, respectively.9


  1. 1.
  2. 2.
  3. This detailed ICRP report emphasises patient skin injuries as a result of mainly poor operational technique and a lack of awareness of the potential for injury from cardiac interventions. It provides well illustrated educational case reports, detailed information about radiopathology of the skin and eye, but most importantly an overview of simple methods for decreasing radiation injuries utilising dose control strategies.

  4. 3.
  5. 4.
  6. 5.
  7. 6.
  8. 7.
  9. Closure of radiation leakage and optimisation of both table attached lead shielding and personal lead garments, as illustrated and proven for daily practice, enabled a reduction of occupational radiation exposure down to 0.8% of typical operator entrance skin air kerma levels in advanced catheterisation laboratories.

  10. 8.
  11. 9.
  12. The cardiology community needs reference levels considering the complexity of PCIs, carried through by consistent use of radiation reducing techniques: this study presents mean radiographic and fluoroscopic DAP, number of radiographic frames and runs, fluoroscopy times dependent on PCI target vessel, one, two and three vessel PCI, chronic occlusions, acute myocardial infarction, number of implanted stents, and excimer laser angioplasty of diffuse in-stent restenoses.

  13. 10.
  14. 11.
  15. 12.
  16. 13.
  17. This statement deals in great detail with the physics of radiation, principles of x ray image formation, functionalities of an x ray cinefluorographic unit, and determinants of patient exposure and their potential effects, considers the clinical risk: benefit ratio for typical interventions, and presents a recommended radiation safety curriculum for cardiologic interventionists.

  18. 14.
  19. 15.
  20. 16.
  21. 17.
  22. Representative and detailed data tool of both time adjusted fluoroscopic DAP to the patient and DAP corrected mean local scatter dose in the operator’s position, mapped at an anthropomorphic Rando–Alderson phantom for each conceivable tube angulation. It enables the interventionist to check the individual radiation risk resulting from favoured coronary views and to find less radiation intensive angulations.

  23. 18.
  24. 19.
  25. 20.
  26. 21.
  27. 22.
  28. 23.
  29. 24.
View Abstract


  • Competing interests: In compliance with EBAC/EACCME guidelines, all authors participating in Education in Heart have disclosed potential conflicts of interest that might cause a bias in the article. The author has no competing interests.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.