Author + information
- Received April 17, 1998
- Revision received September 4, 1998
- Accepted October 22, 1998
- Published online February 1, 1999.
- Jack T Cusma, PhDa,* (, )
- Malcolm R Bell, MBBS, FRACP, FACCa,
- Merrill A Wondrowa,
- Jerome P Taubela and
- David R Holmes Jr., MD, FACCa
- ↵*Reprint requests and correspondence: Jack T. Cusma, Cardiac Catheterization Laboratory, Mayo Clinic, 200 First St. SW, Rochester, MN 55905
The aim of this study was to accurately assess the radiation exposure received by patients during cardiac catheterization in a large sample representative of the current state of practice in cardiac angiography.
Radiation exposure to patients and laboratory staff has been recognized as a necessary hazard in coronary angiography. The effects on x-ray exposure of the increased complexity of coronary angiographic procedures and, in particular, the increasing use of coronary artery stenting, have not been adequately addressed in previous studies.
X-ray exposure measurements were performed on a consecutive series of 972 patients undergoing 992 diagnostic and interventional studies in the Mayo Clinic catheterization laboratory within an eight week period in late 1997. Data were acquired from 706 diagnostic procedures and 286 interventional procedures using a real-time exposure measurement system to continuously calculate and record the exposure rate and total exposure, reflecting all parameters relevant to the specific patient and procedure situation.
The median exposure for all 992 procedures was 41.8 mC/kg (162.1 R); the corresponding values for diagnostic and interventional procedures were 34.9 and 95.6 mC/kg, respectively (135.3 vs. 370.5 R). There were significant differences in the fluoroscopy exposure time between diagnostic and interventional procedures: 4.7 min vs. 21.0 min. Heavier patients (>83 kg) received x-ray exposures at a significantly higher rate than did lighter patients (<83 kg) during both fluoroscopy and cine; 44.9 mC/kg/min (173.9 R/min) vs. 27.9 mC/kg/min (108.3 R/min) for cine exposure rate and 2.3 mC/kg/min (8.8 R/min) vs. 1.5 mC/kg/min (5.8 R/min) for fluoroscopy exposure rate.
Changes in practice have led to higher values for patient x-ray radiation exposures during cardiac catheterization procedures. The real-time display and recording of x-ray exposure facilitates the reduction of exposure in the catheterization laboratory.
Diagnostic coronary angiography and percutaneous coronary interventions are being performed in increasing numbers. Treatment of multivessel disease, intervention on increasingly complex stenoses and repeat procedures necessitated by restenosis have all led to a significant increase in the amount of fluoroscopic guidance required with an accompanying increase in radiation exposure. Radiation exposure to the attending physician and other personnel are recognized as potential occupational hazards (1–3); although monitoring of the exposure over time to catheterization laboratory personnel is mandatory, there has been no universal attempt to monitor radiation doses received by the patient. Despite numerous studies (4–13), there remains some uncertainty as to what radiation risks these patients are exposed, but skin reactions after coronary angioplasty have been reported (14–16).
Measurements of radiation dose delivered to patients during coronary angiography and interventional procedures such as percutaneous transluminal coronary angioplasty (PTCA) have been reported previously (6,10–13,17–20), with the highest radiation doses documented with PTCA, particularly multivessel procedures and complex procedures (17–20). However, in few of these studies was radiation dose to the patient measured directly (or indirectly) while the procedure was being performed (6,12–13,17,18). There are few data available describing radiation exposure in the current era of widespread use of new interventional devices (17,21), particularly stents which are now used in 40 to 70% of all interventional cases. Implementation of ionization chamber dosimeters or thermoluminescent detectors (TLD’s) for the routine monitoring of radiation exposure to patients has not been practical. Such techniques are also difficult to implement in an “on-line” manner, i.e., in such a way that operators can be aware of the radiation exposure to the patient as the procedure is being performed. Recently, an on-line radiation dose measurement system has been developed which can be used to provide real-time measurements of radiation exposure to the patient in the catheterization laboratory. The measurement system can be networked throughout a catheterization laboratory and stores acquired data, making it available for cumulative measurements. In this study, we document the results obtained with its use in a consecutive series of 972 patients undergoing diagnostic and interventional coronary procedures at Mayo Clinic.
A consecutive series of 972 patients undergoing 992 diagnostic and interventional cardiac procedures was studied during an eight week period in 1997. These procedures represent all patients who underwent procedures in five adult cardiac catheterization suites but do not include an additional 76 procedures which were performed in one of the rooms over a two week period. These 76 procedures were not captured from the single laboratory due to an equipment malfunction but were representative of the overall mix in the total sample. Table 1lists the number of procedures in several major categories along with patient characteristics within those categories.
Data were recorded from 706 diagnostic cases where no coronary intervention was performed. The analysis of diagnostic procedures was performed on the majority of these procedures, after exclusion of: right heart catheterization procedures (independent of coronary angiography), stand-alone endomyocardial biopsies and valvuloplasties, intracoronary ultrasound studies, and Doppler coronary flow measurements. This resulted in a total of 597 diagnostic procedures in which coronary angiography was performed; in 207 of these, selective coronary angiography was performed without ventriculography while, in another 233 cases, left ventriculography was also performed. The remaining diagnostic procedures included, in combination with coronary angiography: visualization of coronary artery bypass grafts (aorto-saphenous veins and mammary arteries) − 81; right heart catheterization − 56; miscellaneous procedures such as biopsy and aortography − 20.
Coronary interventional procedures were performed during a total of 286 examinations: 84 as a “stand alone” procedure while 202 were combined with a diagnostic coronary angiogram. Two hundred eighty four of the procedures included coronary angioplasty (PTCA). There were 235 stent procedures, 17 rotational atherectomy, 1 laser and 2 directional atherectomy procedures. The types of interventions were not mutually exclusive and the majority of procedures involved combined interventions.
The Mayo Clinic Cardiac Catheterization Laboratory currently includes six procedure rooms which can be used for adult cardiac angiography, each of which contains a General Electric MPX-100/LUC biplane x-ray acquisition system (General Electric Medical Systems, Waukesha, Wisconsin) with a C arm on each plane. One of the procedure rooms is primarily used for pediatric patients and occasionally for adult procedures—exposure records from this room were not included in the analyzed data set. All imaging chains are identical and include a triple mode image intensifier (11/15/23 cm modes). Entrance exposures at the image intensifier (I.I.) in all rooms are monitored routinely and set for the 23 cm I.I. mode to values of 0.34 nC/kg/frame (1.3 μR/frame) for fluoroscopy and 3.9 nC/kg/frame (15 μR/frame) during cineangiography; corresponding entrance exposures for the 15 cm I.I. mode typically employed during coronary angiography are 0.62 nC/kg/frame (2.4 μR/frame) and 7.7 nC/kg/frame (30 μR/frame) for fluoroscopy and cine, respectively. The total filtration for each x-ray tube is adjusted at installation to an equivalent of 3.5 mm A1 at 80 kVp. The x-ray system employs pulsed progressive scanning of the video camera during fluoroscopy (22)and cineangiography—all imaging is performed at 30 frames/s. All rooms have parallel cine film and digital acquisition capability as well as operating modes where the system is operated in digital-only acquisition or “digital cine.” Entrance exposure rates during the parallel modes are equivalent to the standard cine exposure rate; during digital cine acquisition, the exposure rate ranges from cine-equivalent down to one-fourth of the cine rate. On these systems, the digital cine options are used—at one-fourth of the cine exposure—in a manner similar to high level control (HLC) fluoroscopy, when increased contrast and lower noise relative to standard fluoroscopy are required, e.g., during attempts to visualize intracoronary stents of low radio-opacity.
Radiation monitoring system
Radiation measurements are performed in our laboratory using a Patient Exposure Monitoring Network (PEMNET) System (Clinical Microsystems, Inc, Arlington, Virginia). During each x-ray exposure, the PEMNET system calculates and displays the real-time exposure rate in R/min based upon the peak kilo Volts (kVp), milliAmpere (mA), exposure pulse width, and the geometry of the table and positioners. This provides a measurement of exposure rate both during fluoroscopy and cine/digital acquisitions reflecting changes in patient tissue thickness, projection angles, table height, etc. It should be noted that, for the purposes of the exposure measurements, “digital cine” is recorded as a “cine” exposure because the acquistions are controlled by the cine control circuits of the x-ray generator. The total cumulative exposure to the patient is displayed following the end of any single exposure sequence. Data acquisition modules are interfaced to every x-ray acquisition plane; at the conclusion of a procedure, the data are transferred from each procedure room to a standard personal computer from which data are available for export and analysis.
The PEMNET system is routinely monitored and calibrated to ensure that the patient Entrance Skin Exposure (ESE) calculated from the acquisition parameters are within a predetermined tolerance with direct measurements acquired with an ionization chamber (Radcal Corp., Monrovia, California). The calibration is performed under conditions which include the effects of back-scattered radiation. Another set of measurements is also made with the table and pad placed between the dosimeter ion chamber and the x-ray source. The attenuation characteristics of the table and pad at different energies are determined using these two sets of calibration measurements and these are compensated for during patient exposure measurements. The PEMNET software uses statistical analysis of the calibration data to determine the equations to use in calculating the patient ESE. Quality assurance procedures in our laboratory include monthly monitoring of the agreement between PEMNET and ionization chamber measurements; discrepancies greater than 15% result in a recalibration of the measurement system.
In addition to the routine calibration procedures, exposure measurements were performed using phantoms to simulate the range of patient thickness and projection angles representative of clinical routine in our laboratory. A quality control phantom consisting of 20 cm of lucite and 4 mm of aluminum was used in combination with up to 12.5 cm of lucite to simulate a range of patient thicknesses. This served as a validation of the PEMNET calculations with phantom thickness and reflects the relative increase in the x-ray factors one would find during clinical procedures with patients of different size. In this set of measurements, the phantom thickness was varied for a fixed projection angle—straight AP—and a calibrated ionization chamber measured exposure rates at the entrance plane to the phantom. The range of thicknesses simulated patient thickness of approximately 20 to 30 cm. In a second set of measurements, an anthropomorphic chest phantom was used to demonstrate the increase in radiation exposure which would result from typically employed angulation and correspondingly longer path lengths through patients. This phantom, consisting of tissue and bone within a housing which resembles a human torso, provides attenuation of x-rays similar to what one might expect in an average patient. Using this phantom, the PEMNET measurements were recorded as the projection angle was varied.
Numerical data are presented as median with interquartile range unless otherwise stated. Comparisons of continuous data were performed using the Student’s t-test. A p value <0.05 was considered to be significant.
The results of the PEMNET comparison with dosimeter measurements are reported in Table 2where the exposure rates during fluoroscopy and cineangiography are shown for a range of phantom thicknesses. The average deviation between the two measurements is −5.4% for the fluoroscopic rates and −4.9% for cine rates. It should also be noted that, in the straight antero-posterior (AP) projection, where the effect of projection angle on the effective patient thickness is minimized, the exposure rates for our mean patient thickness of 23 cm were 1.82 mC/kg/min (7.07 R/min) and 21.39 mC/kg/min (82.9 R/min) for fluoroscopy and cineangiography, respectively. The effect of projection angle is illustrated in Table 3where the fluoroscopy and cine exposure rate measurements are shown as a function of the projection angle. As is evident from these results, the left anterior oblique (LAO) Cranial projection requires exposure rates over five times those which are sufficient in the standard right anterior oblique (RAO) projection.
Radiation exposure measurements along with measurements of the fluoroscopic and cineangiographic time components are listed in Tables 4 and 5. ⇓⇓Fluoroscopy comprised 89% of imaging time during diagnostic cases and 94% during interventional cases. Fluoroscopic imaging duration was markedly longer during interventional cases relative to diagnostic cases (median of 21.0 versus 4.7 minutes, respectively) and the variability was wide as reflected by an interquartile range of 13.7 to 31.3 min for all interventional cases combined into a single group. Cineangiographic acquisition times were relatively short for all procedures compared with the fluoroscopy time. A breakdown for several subgroups of interventional procedures is shown in Table 5. The only significant difference among the measurements was for the cine time measurements for stent and nonstent procedures.
One of the determinants of radiation dose received during cardiac angiographic procedures is the size of the patient. A frequency histogram of the study population’s weight is shown in Figure 1. For the purpose of analysis, patients were divided into two groups based on their weight relative to the overall mean weight (<83 kg and ≥83 kg) and average exposure rate calculated for fluoroscopy and cineangiography from all procedures. Results are shown in Figure 2. Patients weighing more than 83 kg were exposed to significantly more radiation per unit time during fluoroscopy than those less than 83 kg: 2.3 versus 1.5 mC/kg/min (8.8 vs. 5.8 Roentgen (R)/min), p < 0.001. Similarly the exposure per unit time during cineangiography was greater for heavier patients than for lighter patients: 44.9 mC/kg/min vs. 27.9 mC/kg/min (173.9 vs. 108.3 R/min). The relative durations of fluoroscopy and cineangiography did not differ significantly among patients of different size. For example, the median value for fluoroscopy time per case was 7.5 and 7.7 minutes for patients <83 kg and ≥83 kg, respectively. The median cine exposure time was 0.8 minutes for both groups. Alternatively, the patients were grouped by chest thickness—chest thickness ≤23 cm and >23 cm: the results for fluoro rates are 1.5 mC/kg/min vs. 2.2 mC/kg/min (6.0 vs. 8.7 R/min); for cine rates—28.6 mC/kg/min versus 44.6 mC/kg/min (110.7 vs. 172.7 R/min). All differences were significant (p < 0.001) except for the differences in the fluoro time and cine time measurements. A further breakdown of the measurements by both chest thickness and procedure type—diagnostic versus intervention—is shown in Table 6. Here, it can be seen that the major differences in exposure rates occur as a function of patient size; for all rate measurements within a type of procedure, the differences between thin and thick patients were significant. The effect of procedure type on exposure rates for the same patient size group was significant except for cine rate for thin patients (p = 0.38 between diagnostic and interventional procedures). The differences in cine time for all pairings and for fluoroscopy time across procedure types were significant but there was no significant difference in fluoroscopy time within the same type of procedure. The total measured exposure is also significantly different when comparisons are made across both procedure type and patient size.
Comparison with previous studies
Relatively high values of radiation exposure have been considered a necessary consequence of cardiac angiographic procedures. With increasing complexity of the procedures, there has been growing concern regarding the magnitude of the exposure to operators and patients (1,12,13,17,23). In prior studies, a complete assessment has been hampered by the fact that little direct data available on the magnitude of the range of exposures in currently accepted angiographic procedures.
Early studies that investigated radiation exposure during cardiac angiography in relatively large numbers of patients were limited to approximating the value of the x-ray exposure using indirect measures, namely, the fluoroscopic times during the procedure and the number of cineframes that were recorded for the procedure. These approaches do not take into account variations in patient size, position of the patient relative to the x-ray tube and detector and the angulation of the X-ray image intensifier relative to the patient. As demonstrated in the phantom measurements reported in this study, such variation can increase exposure rates by as much as a factor of ten compared to the exposure rate delivered to a relatively thin patient in the AP projection.
The results presented here are significantly higher than those from previous studies that have investigated large numbers of procedures retrospectively. Without dividing procedures into diagnostic and interventional procedures, Johnson et al. (1)reported that, from a survey of catheterization laboratories, a typical exposure to the patient of 15.5 to 20.6 mC/kg (60 to 80 R) could be expected. In the report by Pattee et al. (12), an average skin ESE of 32.0 mC/kg (124 R) for a typical angioplasty procedure was calculated for over 1800 PTCA procedures. The calculations in their study assumed a value for fluoroscopic and cine exposure rate which is often exceeded with steep angulations and thicker patients. As shown in our data, their values of 0.52 mC/kg/min (2.0 R/min) during fluoroscopy and 13.9 mC/kg/min (54 R/min) during cine are characteristic of an average patient in the RAO view but the rates go up significantly with steeper LAO angulations. A survey of practice in the U.K. reported by Coulden et al. (11)resulted in patient exposure estimates from 5.2–16.8 mC/kg (20– 65 R) but no corrections were made for varying projections and their effect on effective patient thickness, nor was there any differentiation made between diagnostic and interventional procedures. The patient exposure estimates for all but five of the 36 responding sites were also made using fluoroscopy time and length of cine film. Huyskens et al. (10)applied representative “kerma-exposure product” measurements from a small number of procedures to two years worth of procedures in a single large laboratory in the Netherlands. The actual calculations were derived from fluoroscopy time and cine film length for over 3000 procedures. A previous report from our laboratory (21), using an average estimation approach from fluoroscopy and cineangiographic times, calculated patient exposures for interventional procedures ranging from 12.9–30.9 mC/kg (50– 120 R). There was no correction made for varying projection angles; the authors emphasized that such measurements should be considered the minimum expected exposure.
There have been previous direct measurements or calculations of exposure to operators and patients on a case-by-case basis. Dash et al. (9)reported one of the first investigations of the variation with PTCA using film badges and TLD’s to measure cumulative exposure to operators during a total of 58 angiographic procedures. Although they did not measure exposure to the patient, they did find that angioplasty resulted in 93% greater exposure to the operator than did diagnostic procedures. No quantitative measurements were made but the authors did note the increased use of cranial angulation under fluoroscopy during angioplasty compared with routine angiography. Cascade et al. (18)used TLD measurements in approximately 110 patients to compare PTCA to coronary angiography. Their technique was designed to take into account the effect of projection angle and they found that angioplasty procedures produced nearly four times as much exposure to patients as did angiography—17.8 versus 5.2 mC/kg (69 vs. 20 R). They also found a factor of two increase in exposure for procedures where two interventions were performed compared to one. A dosimetric study on 58 procedures by Zorzetto et al. (13)looked at dose to operators using TLD’s and to the patient using a meter indicating dose-area-product (DAP). They found that the DAP was about 70% higher for PTCA than for diagnostic angiography. The results of the DAP measurements were mean values of 55.9 and 91.8 Gy-cm2for diagnostic and interventional procedures, respectively. A recent study by Bakalyar et al. (17)collected Area-Exposure-Product (AEP) data for 510 consecutive procedures, resulting in AEPs of 144.0–227.0 Gy-cm2(16,500–26,000 R-cm2) for interventional procedures of various types compared with 94.3 Gy-cm2(10,800 R-cm2) for diagnostic procedures. Assuming an entrance area of 100 cm2, this corresponds to a range of 42.6–67.0 mC/kg (165–260 R) and 27.9 mC/kg (108 R), respectively. They also differentiated among different types of interventions as well with, for example, stent procedures being 60% higher exposure than PTCA alone.
Advantages of real-time exposure measurement
Among the questions prompted by the results in this study is whether they may be attributable to unique aspects of our laboratory or the type of procedures performed or whether, instead, they are representative of general changes in angiographic practice over time. A major difference in the methodology of this study is that the exposure to the patient is calculated in real-time and reflects all factors which affect the radiation exposure during an angiographic procedure. As such, these results cannot be directly compared with other experiences because other reported methods do not measure parameters under these conditions. The increase in exposure with patient size, for example, is demonstrated by the phantom simulation in this study and agrees with results reported by Boone et al. (24)who looked at changes in scatter radiation to personnel as patient thickness increases. Similar issues with respect to the effects on exposure to the operator of modifying views during coronary angioplasty were addressed by Pitney et al. (25). From these results, one can see that moving from straight AP to shallow RAO to steep LAO cranial angulation can increase the exposure by a factor of three or more. The results in our study are also consistent with computer simulations and measurements by Stern et al. (26).
Another factor that cannot be accounted for in “average” approaches is the increasing use of “high level control” (HLC) fluoroscopy as a result of an increase in the proportion of interventional procedures that require the deployment of stents in the coronary arteries. Recognized as a contributor to higher radiation exposure with the advent of PTCA (27), HLC fluoroscopy can result in exposure rates as high as that of cineangiography—on the order of ten times normal fluoro—and, when used during positioning of stents and balloons, can lead to significantly greater exposure times than during cine. Although the digital cine mode used in our laboratory is not strictly an HLC fluoroscopy mode, it is utilized in a similar manner during interventional procedures and is representative of practice in a number of laboratories. The longer total cine exposure times for interventional procedures reported in our data can most likely be attributed to the contribution from the digital cine mode of acquisition. Other laboratories’ use of HLC may be logged as fluoroscopic exposure despite the fact that the exposure rate may in fact be closer to that of cine acquisition. The significant difference in our data for cine time measurement between stent and nonstent procedures would support this hypothesis.
Potential for radiation-induced skin injury in coronary angiography
Another question that needs to be addressed is whether, considering the magnitude of the total procedure exposures, one should expect to report a high frequency of skin effects. There have been no such reports in our catheterization laboratory and this can most likely be attributed to the fact that the totals reported in this study are, in fact, integrated over all projection angles and do not represent the total exposure to any single area of skin. As illustrated in the comprehensive analysis by Stern et al. (26), the distribution of x-ray exposure during a typical catheterization procedure varies greatly over the range of typical views. If, instead, all of the radiation had been delivered over the same area of skin, exposures at the higher end of the values reported in our measurements—77.4 mC/kg (300 R) and greater—might be considered at the threshold of such effects.
There is some uncertainty in the literature related to the appropriate degree of concern regarding the amount of x-ray exposure to the patient (6,11,12,17). Although all agree that the highest quality diagnostic images are the primary consideration during angiography, any precautions that can reduce unnecessary exposure should be taken as well. Careful monitoring of the performance of equipment—especially x-ray and I.I. tubes—should be performed to detect increases in the amount of x-rays required to form an adequate image. Reduction of exposure can be achieved through the use of reduced frame rate fluoroscopy and cineangiography; the increasing use of digital angiography without cinefilm recording should make the latter more feasible, and attention should be paid to patient and I.I. position. The use of HLC fluoroscopy should be considered only when necessary and, if available, the frame rate should be no higher than necessary: 15 frames/s results in a straightforward reduction in exposure by a factor of two.
Patient x-ray radiation exposures during cardiac catheterization procedures have increased as a result of the increased complexity of the angiographic procedures performed in current clinical practice. The increasing prevalence of coronary artery stent deployment has led to significant increases in the exposure to patients during fluoroscopic and cineangiographic imaging. Monitoring of x-ray exposure in the catheterization laboratory is facilitated through the use of real-time exposure measurements that display and record exposure as parameters are changed continuously during a procedure.
We gratefully acknowledge the technical assistance of Tom Vrieze during the acquisition of x-ray exposure measurements and calibration of the radiation measurement system.
- area exposure product
- coulomb (also mC-milliCoulomb, nC-nanoCoulomb, mC/kg, mC/kg/min)
- dose area product
- entrance skin exposure
- high level control
- image intensifier
- left anterior oblique
- peak kilo Volts
- Roentgen (also R/min and μR-microRoentgen)
- percutaneous transluminal coronary angioplasty
- right anterior oblique
- thermoluminescent detector
- Received April 17, 1998.
- Revision received September 4, 1998.
- Accepted October 22, 1998.
- American College of Cardiology
- Brinker J.A,
- Pepine C.J,
- Block P.C,
- et al.
- Reuter F.G
- Faulkner K,
- Love H.G,
- Sweeney J.K,
- Bardsley R.A
- McParland B.J,
- Nosil J,
- Burry B
- Balter S,
- Sones F.M Jr.,
- Brancato R
- Dash H,
- Leaman D.M
- Huyskens C.J,
- Hummel W.A
- Coulden R.A,
- Readman L.P
- Pattee P.L,
- Johns P.C,
- Chambers R.J
- Sovik E,
- Klow N.E,
- Hellesnes J,
- Lykke J
- Federman J,
- Bell M.R,
- Wondrow M.A,
- Grill D.E,
- Holmes D.R Jr.
- Pitney M.R,
- Allan R.M,
- Giles R.W,
- et al.
- ↵Stern SH, Rosenstein M, Renaud L, Zankl M. Handbook of selected tissue doses for fluoroscopic and cineangiographic examination of the coronary arteries. Health and Human Services Publication FDA 95-8288. 1995.