Abstract
Objectives
This study sought to identify clinical and procedural predictors of elevated radiation dose received by patients with acute myocardial infarction (AMI) undergoing percutaneous coronary intervention (PCI) and to determine if elevated radiation dose was predictive of mortality in this population.
Background
Little data exist regarding the impact of excessive radiation burden on clinical outcomes in patients undergoing PCI.
Methods
The study population included 1,039 patients who underwent PCI for an AMI between January 1, 2007 and December 31, 2008 at an academic tertiary care teaching hospital. Cumulative skin dose (measured in milligray [mGy]) was selected as a measurement of patient radiation burden. Clinical and procedural variables were analyzed in multiple logistic and linear regression models to determine predictors of higher skin dose, and its impact was evaluated on all-cause intermediate-term mortality at two years.
Results
Median skin dose was 2120 mGy (IQR 1379–3190 mGy) in the overall population, of which 153 (20.8%) patients received an elevated skin dose (defined as a skin dose > 4,000 mGy). Independent predictors of elevated skin dose included male gender, obesity, multivessel intervention, and presentation with a non-ST-elevation MI (NSTEMI) versus an ST-elevation MI (STEMI). Increased skin dose was not predictive of intermediate-term mortality by multivariate analysis in the overall population or in either subgroup of STEMI and NSTEMI.
Conclusions
In this contemporary observational study examining patients with AMI undergoing PCI, male gender, obesity, multivessel intervention, and presentation with a NSTEMI were associated with increased radiation exposure.
1
Introduction
Reducing radiation dose during percutaneous coronary intervention (PCI) and other fluoroscopic procedures helps to minimize the risk of both stochastic (e.g. malignancy) and deterministic (e.g. skin injury) effects for patients and physicians, and therefore efforts to optimize radiation safety in the cardiac catheterization lab are strongly encouraged . A comprehensive understanding of factors contributing to elevated radiation dose and its effects is vital to reducing radiation burden to the patient, operator, and laboratory staff. Prior studies have examined clinical and procedural variables associated with radiation dose in patients presenting for coronary angiography and PCI . Previously identified predictors have included male gender, higher body mass index (BMI), and prior coronary artery bypass graft (CABG) surgery .
However, most of these data have focused on access route (femoral versus radial) and elective versus non-emergent procedures . Moreover, no studies to date have examined whether there is a relationship between elevated radiation dose and clinical outcomes. The aim of the present study was to identify the clinical determinants of higher radiation dose during PCI in patients presenting with an acute myocardial infarction (AMI) and to determine if higher radiation dose is associated with worse clinical outcomes, namely all-cause mortality.
2
Methods
2.1
Patient population and clinical definitions
Data were prospectively collected from all patients presenting with an AMI undergoing PCI between January 1, 2007 and December 31, 2008 at an academic tertiary care teaching hospital. This study was approved by the institutional review board. PCI procedures were performed as per current guidelines by one of fifteen interventional cardiologists and an interventional or general cardiology fellow-in-training. Fellow involvement ranged from assisting during the procedure to performing the PCI under direct supervision of an attending interventionalist. Procedures were confined to the first PCI performed within the index hospitalization. If any procedures were performed as staged procedures whether within the index hospitalization or subsequent to hospital discharge, these procedures were not included in the analysis.
Data elements in the registry included patient demographic information, baseline medical history, clinical presentation, angiographic and procedural characteristics, and clinical outcomes. The population consisted of 1,087 consecutive patients presenting with an AMI and undergoing emergent/urgent PCI. A total of 48 patients were excluded from the analysis due to absent data regarding radiation skin dosage. Of the remaining 1,039 patients, the following demographic and medical history data were collected: age, gender, BMI, prior CABG, prior PCI, prior MI, diabetes mellitus, hypertension, chronic kidney disease, chronic obstructive pulmonary disease, and atrial fibrillation. Presentation and procedural characteristics recorded included AMI type [ST-elevation MI (STEMI) vs. non-ST-elevation MI (NSTEMI)], operator experience (high volume vs. low volume), left ventricular ejection fraction (LVEF), extent of coronary disease, culprit coronary artery treated, number of coronary arteries intervened, contrast amount, and radiation dosage, which included cumulative skin dose, total body dose, and fluoroscopy time. A high volume operator was defined as an interventional cardiologist who had performed at least 100 PCI cases in AMI patients during the study period. Coronary arteries (and their respective branches) were grouped as left main, left anterior descending, left circumflex, right coronary artery, and bypass (saphenous vein or arterial) graft. Clinical outcomes, including mortality (both in-hospital and at two years post hospital discharge), hospital length of stay, and discharge location, were also recorded. Mortality was determined from the patient’s electronic medical record and use of the Social Security Death Index. Over 95% of the PCI procedures were performed via the femoral approach, and therefore access site was not included in this analysis.
2.2
Radiation dose
PCI procedures were performed using one of three Siemens Axion Artis Systems (Siemens Healthcare, Erlangen, Germany). The X-ray systems are configured to use frame rates of 15 s − 1 for both fluoroscopy and cine acquisition imaging. Any changes made to this default setting were not specifically recorded. Operators selected the appropriate angiographic views. All physicians, nurses, and radiation technicians were fully trained in the principles of radiation safety in accordance with mandatory regulations and wore lead aprons and thyroid collars during the procedure. A transparent 1.0-cm lead screen and table shield were utilized by operators throughout the entirety of the procedure. Radiation doses are reviewed monthly by an institutional cardiovascular radiation safety committee.
The X-ray systems provided a real-time display of radiation exposure on the monitor of the systems. The total body dose [microgray per meter squared (μGy/m 2 )], cumulative skin dose [milligray (mGy)], and fluoroscopy time (minutes) were displayed on the monitors throughout the procedure. Fluoroscopy time is an indicator of procedural duration. While it has some influence on radiation dose delivered, it does not account for the dose rate or intensity and thus is not useful for risk estimation of deterministic or stochastic effects . Cumulative skin dose, which is typically measured at the interventional reference point (the point on the x-ray beam axis lying 15 centimeters from the imaging system’s isocenter on the x-ray tube side), represents an accurate measure of the risk of fluoroscopy-associated skin injury and has been validated as an effective indicator of maximum radiation dose to the skin during PCI . Chida et al. noted that cumulative skin dose is approximately twice the maximum skin dose for each target vessel during PCI . Thus, it can be deduced that the threshold for early transient erythema (maximum skin dose of 2,000 mGy) has been crossed if the total skin dose is approximately 4,000 mGy. For the purpose of the present study, an elevated skin dose was therefore defined as a cumulative skin dose > 4,000 mGy, and the patient population was divided into 2 groups based on the total PCI cumulative skin dose above or below the 4,000 mGy at the end of the procedure.
2.3
Statistical analysis
Data were summarized by descriptive statistics. Univariate analyses were performed to compare characteristics of patients receiving higher versus lower radiation doses. Chi-squared test (or Fisher’s exact test, when applicable) was used to compare differences in categorical variables, and student’s t test was used for continuous variables. Multivariable logistic and linear regression was utilized to evaluate independent clinical determinants of elevated skin dose in the population while controlling for demographic characteristics, medical history, clinical presentation, and procedural characteristics. Multivariate logistic regression was also performed to identify independent predictors of all-cause mortality. Predictors for multivariate analyses were selected based on statistical significance in the univariate analysis (p < 0.1). An ROC curve and plot of residuals were constructed to determine fit of logistic and linear regression models, respectively. A Kaplan–Meier survival curve was formulated to determine the impact of elevated skin dose on mortality at follow up in the overall population as well as in the STEMI and NSTEMI sub-groups. SPSS version 18.0 (SPSS, Inc. Chicago, IL) was used for data analysis and a two-tailed p-value of < 0.05 was regarded as statistically significant.
No extramural funding was used to support this work.
2
Methods
2.1
Patient population and clinical definitions
Data were prospectively collected from all patients presenting with an AMI undergoing PCI between January 1, 2007 and December 31, 2008 at an academic tertiary care teaching hospital. This study was approved by the institutional review board. PCI procedures were performed as per current guidelines by one of fifteen interventional cardiologists and an interventional or general cardiology fellow-in-training. Fellow involvement ranged from assisting during the procedure to performing the PCI under direct supervision of an attending interventionalist. Procedures were confined to the first PCI performed within the index hospitalization. If any procedures were performed as staged procedures whether within the index hospitalization or subsequent to hospital discharge, these procedures were not included in the analysis.
Data elements in the registry included patient demographic information, baseline medical history, clinical presentation, angiographic and procedural characteristics, and clinical outcomes. The population consisted of 1,087 consecutive patients presenting with an AMI and undergoing emergent/urgent PCI. A total of 48 patients were excluded from the analysis due to absent data regarding radiation skin dosage. Of the remaining 1,039 patients, the following demographic and medical history data were collected: age, gender, BMI, prior CABG, prior PCI, prior MI, diabetes mellitus, hypertension, chronic kidney disease, chronic obstructive pulmonary disease, and atrial fibrillation. Presentation and procedural characteristics recorded included AMI type [ST-elevation MI (STEMI) vs. non-ST-elevation MI (NSTEMI)], operator experience (high volume vs. low volume), left ventricular ejection fraction (LVEF), extent of coronary disease, culprit coronary artery treated, number of coronary arteries intervened, contrast amount, and radiation dosage, which included cumulative skin dose, total body dose, and fluoroscopy time. A high volume operator was defined as an interventional cardiologist who had performed at least 100 PCI cases in AMI patients during the study period. Coronary arteries (and their respective branches) were grouped as left main, left anterior descending, left circumflex, right coronary artery, and bypass (saphenous vein or arterial) graft. Clinical outcomes, including mortality (both in-hospital and at two years post hospital discharge), hospital length of stay, and discharge location, were also recorded. Mortality was determined from the patient’s electronic medical record and use of the Social Security Death Index. Over 95% of the PCI procedures were performed via the femoral approach, and therefore access site was not included in this analysis.
2.2
Radiation dose
PCI procedures were performed using one of three Siemens Axion Artis Systems (Siemens Healthcare, Erlangen, Germany). The X-ray systems are configured to use frame rates of 15 s − 1 for both fluoroscopy and cine acquisition imaging. Any changes made to this default setting were not specifically recorded. Operators selected the appropriate angiographic views. All physicians, nurses, and radiation technicians were fully trained in the principles of radiation safety in accordance with mandatory regulations and wore lead aprons and thyroid collars during the procedure. A transparent 1.0-cm lead screen and table shield were utilized by operators throughout the entirety of the procedure. Radiation doses are reviewed monthly by an institutional cardiovascular radiation safety committee.
The X-ray systems provided a real-time display of radiation exposure on the monitor of the systems. The total body dose [microgray per meter squared (μGy/m 2 )], cumulative skin dose [milligray (mGy)], and fluoroscopy time (minutes) were displayed on the monitors throughout the procedure. Fluoroscopy time is an indicator of procedural duration. While it has some influence on radiation dose delivered, it does not account for the dose rate or intensity and thus is not useful for risk estimation of deterministic or stochastic effects . Cumulative skin dose, which is typically measured at the interventional reference point (the point on the x-ray beam axis lying 15 centimeters from the imaging system’s isocenter on the x-ray tube side), represents an accurate measure of the risk of fluoroscopy-associated skin injury and has been validated as an effective indicator of maximum radiation dose to the skin during PCI . Chida et al. noted that cumulative skin dose is approximately twice the maximum skin dose for each target vessel during PCI . Thus, it can be deduced that the threshold for early transient erythema (maximum skin dose of 2,000 mGy) has been crossed if the total skin dose is approximately 4,000 mGy. For the purpose of the present study, an elevated skin dose was therefore defined as a cumulative skin dose > 4,000 mGy, and the patient population was divided into 2 groups based on the total PCI cumulative skin dose above or below the 4,000 mGy at the end of the procedure.
2.3
Statistical analysis
Data were summarized by descriptive statistics. Univariate analyses were performed to compare characteristics of patients receiving higher versus lower radiation doses. Chi-squared test (or Fisher’s exact test, when applicable) was used to compare differences in categorical variables, and student’s t test was used for continuous variables. Multivariable logistic and linear regression was utilized to evaluate independent clinical determinants of elevated skin dose in the population while controlling for demographic characteristics, medical history, clinical presentation, and procedural characteristics. Multivariate logistic regression was also performed to identify independent predictors of all-cause mortality. Predictors for multivariate analyses were selected based on statistical significance in the univariate analysis (p < 0.1). An ROC curve and plot of residuals were constructed to determine fit of logistic and linear regression models, respectively. A Kaplan–Meier survival curve was formulated to determine the impact of elevated skin dose on mortality at follow up in the overall population as well as in the STEMI and NSTEMI sub-groups. SPSS version 18.0 (SPSS, Inc. Chicago, IL) was used for data analysis and a two-tailed p-value of < 0.05 was regarded as statistically significant.
No extramural funding was used to support this work.
3
Results
3.1
Patient demographics and procedural data
Of the 1,039 patients studied, a total of 153 (20.8%) patients received elevated radiation doses while 886 (79.2%) did not. A histogram of skin dose values is provided in Fig. 1 . Median skin dose was 2120 mGy (IQR 1379–3190 mGy) in the overall population, 1906 mGy (IQR 1295–2813 mGy) in patients with STEMI and 2341 mGy (IQR 1444–3507 mGy) in the NSTEMI cohort. Patient demographic and baseline clinical characteristics are summarized in Table 1 . Patients receiving elevated radiation doses were more likely to be male, had higher rates of obesity, prior CABG, and diabetes mellitus. Clinical presentation, procedural and angiographic characteristics are summarized in Table 2 . Patients receiving elevated radiation doses were more likely to present with an NSTEMI rather than a STEMI, had more extensive coronary disease and a greater number of intervened vessels. Increasing radiation dose directly correlated with greater contrast administration ( Fig. 2 ). Performance of PCI by a high volume vs. low volume operator was not significantly different between both groups. Patients with elevated skin dose demonstrated a trend towards having lower LVEF, more left main interventions and fewer right coronary artery interventions. Total body dose (26,705 vs. 10,735 μGy/m 2 , p < 0.001), skin dose (5,070 vs. 1,976 mGy, p < 0.001), and fluoroscopy time (30.4 vs. 14.5 minutes, p < 0.001) also differed significantly between both groups.
| Increased radiation dose | p value | ||
|---|---|---|---|
| No (n = 886) | Yes (n = 153) | ||
| Age (years + SD) | 63 + 13 | 63 + 12 | 0.919 |
| Male | 611 (69.0%) | 123 (80.4%) | 0.004 |
| BMI (kg/m2) | 29.1 (6.1) | 33.3 (7.6) | < 0.001 |
| BMI group | |||
| Normal/Underweight | 211 (24.0%) | 12 (7.9%) | < 0.001 |
| Overweight | 337 (38.3%) | 43 (28.5%) | |
| Obese | 333 (37.8%) | 96 (63.6%) | |
| Prior coronary artery bypass surgery | 100 (11.3%) | 30 (19.6%) | 0.004 |
| Prior percutaneous coronary intervention | 184 (20.8%) | 36 (23.5%) | 0.440 |
| Prior myocardial infarction | 100 (11.3%) | 20 (13.1%) | 0.527 |
| Diabetes mellitus | 212 (23.9%) | 66 (43.1%) | < 0.001 |
| Hypertension | 499 (56.3%) | 96 (62.7%) | 0.138 |
| Chronic kidney disease | 63 (7.1%) | 12 (7.8%) | 0.746 |
| Atrial fibrillation | 53 (6.0%) | 8 (5.2%) | 0.714 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
