In the United States, there has been a dramatic rise in the number of complex diagnostic and therapeutic procedures performed in the cardiac catheterization laboratory exceeding 4 million cases per year. These procedures deliver doses of ionizing radiation from 300-3000 times that of a standard chest x-ray and are often associated with prolonged fluoroscopy times. As a result, patients are exposed to high radiation doses, and this has contributed to an increase in radiation-induced skin injury. Furthermore, exposure to ionizing radiation also increases the patient's risk for the development of solid cancers and leukemias.
Accordingly, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) recently added a new provision to the list of events that are reviewable under its Sentinel Event Policy to include prolonged fluoroscopy with a cumulative dose of >15000 mSv to a single field. Yet, these procedures generally occur without radiation dose monitoring protocols to warn operators of high radiation doses to the patients and avoiding potential injury. Furthermore, there are many dose reduction techniques that can easily be implemented to mitigate the risk of immediate and long term radiation morbidity to the patient.
As such, the Food and Drug Administration and VA Central Office have identified radiation safety to be an area of high priority. The objectives of this RRP are to define the distribution of fluoroscopy time and radiation dose for procedures performed in the cardiac catheterization laboratory and to identify patient, provider and hospital factors associated with high radiation exposure. These initial aims are very achievable using the fluoroscopy and radiation dose data on over 50,000 patients in the Cardiovascular Assessment Reporting and Tracking for Cardiac Catheterization Laboratories data repository (CART-CL). Informed by the results of the initial aims, we will pilot a radiation safety tool-kit at the Denver and Ann Arbor VAMCs (education course in radiation reducing technologies, an in-lab monitoring protocol and data feedback to operators) focused on physicians and catheterization lab staff to increase radiation awareness and decrease patient radiation exposure in the catheterization lab. We will compare fluoroscopy and radiation doses before and after the pilot intervention. A highly qualified team of investigators experienced with using the Cardiovascular Assessment Reporting and Tracking for Cardiac Catheterization Laboratories data repository (CART-CL) and implementation projects within IHD-QUERI will complete the aims of this proposal.
1.Define the distribution of fluoroscopy time and radiation dose by procedure (e.g. coronary angiography, peripheral intervention, etc) performed in the VA cardiac catheterization laboratories
2.Assess patient, provider and hospital factors associated with the highest quartile of fluoroscopy time and radiation dose
3.Informed by specific aim 1 and 2, pilot a radiation safety tool-kit at the Denver and Ann Arbor VAMCs (1. provide focused education on radiation dose-reduction techniques, 2. provide an in-lab radiation monitoring protocol, and 3. provide monthly radiation physician report cards). We will then compare fluoroscopy time and radiation doses before (3 months) and after the intervention.
Data Analysis Specific Aim 1: We will define the distribution of fluoroscopy time and effective radiation dose (mean, standard deviation, median, interquartile range, range, 25th and 75th percentiles) for the procedure categories listed below.
1.Coronary angiography without bypass grafts
2.Coronary angiography with bypass grafts
3.Percutaneous coronary intervention
Stratified variables- Patients will be further stratified within peripheral procedures into diagnostic versus therapeutic procedure and inflow (iliac) vs. outflow (common and superficial femoral and popliteal artery) versus runoff (infrapopliteal) procedures. Electrophysiology procedures will be further stratified into dual versus single lead pacemaker implantation, dual versus single lead defibrillator implantation, biventricular lead placement, and ablation procedures.
Data Analysis Specific Aim 2: Assess patient, provider, and hospital factors associated with the highest quartile of fluoroscopy time and radiation dose
Statistical Analysis: For each of the procedure categories listed above, we used a hierarchical linear regression model to evaluate the association of fluoroscopy and effective radiation dose with patient, provider and hospital factors. This multi-level model allowed us to adjust for clustering at the provider and hospital levels and to determine the relative contribution of patient, provider, and hospital characteristics to variation in fluoroscopy time and effective radiation dose. Baseline characteristics were included in the model. We also used hierarchical logistic regression models to assess the likelihood that these factors are associated with fluoroscopy time or effective radiation dose within the highest quintile of radiation exposure for each procedure group.
Patient level variables: Age, gender, height, weight, hypertension, Dyslipidemia, prior MI, prior CHF, family h/o premature CAD, prior PCI, Prior CABG, ESRD, cerebrovascular disease, peripheral arterial disease, chronic lung disease, diabetes mellitus, presenting signs and symptoms
Patient/Procedure level variables: arterial access site (femoral, brachial, radial, other), intra-aortic balloon pump, mechanical ventricular support devices, coronary anatomy (left main, LAD, LCx, RCA), lesion characteristics (previously treated, graft lesion, high complexity, thrombus present, bifurcation lesion), procedure complications (MI, Cardiogenic shock, CVA/Stroke, Tamponade, vascular complication requiring treatment)
Provider level variables: Years in practice, board certification in cardiology, board certification in interventional cardiology
Hospital level variables: Teaching hospital, >250 beds, academic affiliation, VISN
Data analysis specific aim 3: After a 3 month run-in period, operators and cath lab staff will undergo the 3 tiered intervention. Radiation dosing was measured following the intervention at montly intervals over a 3 month period. Cases were compared in aggregate and stratifed by procedure (coronary angiography +/- ventriculography and PCI). The Mann-Whitney-U test was used to assess for differences between pre and post-intervention radiation dosing.
Aim 1 and 2
We examined 87,658 CA, BGA and PCI procedures performed by 362 operators at 58 VA facilities from 2007 through 2010. There is wide variation in fluoroscopy time by operator regardless of procedure performed. Median fluoroscopy time (IQR) for coronary angiogram was 4.7 (3.0-8.0) min, 12.0 (7.8-18.0) for coronary angiography with bypass grafts and 16.3 (10.2-25.9). Multiple patient, provider and hospital factors were associated with fluoroscopy times in the highest quartile. Patient factors included age and a history of peripheral arterial disease; provider factors included operator volume and experience; and hospital factors included whether or not they were associated with a teaching program.
Patient, provider and hospital factors associated with the highest quartile of fluoroscopy time after multivariable adjustment for coronary angiography were as follows [Odds Ratio (95% CI)]
1. Age 1.03 (1.02-1.03)
2. Secondary Access1.71 (1.60-1.82)
3. H/o Peripheral arterial disease1.11 (1.08-1.15)
4. Teaching hospital (v. non)1.44 (1.35-1.53)
5. Operator Volume
--2nd quartile v. 1st quartile 1.22 (1.14-1.30)
--3rd quartile v. 1st quartile 0.85 (0.80-0.89)
--4th quartile v. 1st quartile 0.76 (0.72-0.80)
6. Each additional 5 years of operator experience 1.0 (0.96-1.03)
Therefore in a large VA system capturing data on all cardiac catheterization procedures, fluoroscopy time is highly variable and dependent on multiple factors relating to patient, provider and hospital domains.
We examined interventional cardiology procedures prior to and following implementation of the multi-modal intervention. Site 1 had 624 cases prior to the intervention and 502 cases following the intervention. Site 2 had 258 cases pre-intervention and 208 cases post-intervention. Nationally there were 4,684 cases pre-intervention and 3,603 cases post-intervention.
Site 1 did not differ significantly in median Radiation Dose (mRD) following intervention (p = 0.34)
Pre: 71.9 Gy*cm2 [IQR 48.0-114.0]
Post: 79.5 Gy*cm2 [50.0-124.8]
Site 2: significant decrease in mRD following intervention (p = 0.004)
Pre: 118.72 Gy*cm2 [IQR 73.6-190.0]
Post: 92.8 Gy*cm2 [56.6-158.3]
National: mRD did not show a significant change (p=0.47)
Pre: 91.53 Gy*cm2 [58.0-145.4] vs. Post: 90.0 Gy*cm2 [56.3-142.0]
Secondary analysis excluding bypass graft cases, stratified by PCI vs. angiography:
Site 1: PCI and coronary angiography mRD post-intervention did not differ significantly. PCI trended towards increase in MRD.
--Coronary Angiography (p=1.0)
Pre: 60.0 Gy*cm2 [IQR 42.0-89.0]
Post: 60.0 Gy*cm2 [ 42.0-95.25]
Pre: 100.0 Gy*cm2 [60.50-144.80]
Post: 121.0 Gy*cm2 [78.25-179.50]
Site 2: PCI and coronary angiography mRD post-intervention did not differ significantly. Both trended towards mRD reduction.
--Coronary Angiography (p=0.28)
Pre: 85.81 Gy*cm2 [51.80-121.39]
Post: 69.45 Gy*cm2 [50.60-98.13]
Pre: 205.96 Gy*cm2 [146.0-288.8]
Post: 170.82 Gy*cm2 [117.4-282.8]
Therefore, a three-tiered multi-modal intervention to reduce radiation was associated with decreases in radiation exposure in a high radiation utilization laboratory at baseline. These findings suggest that an approach to radiation reduction interventions targeted at higher radiation use centers may result in meaningful decreases in patient radiation exposure.
Aim 1 established highly variable radiation doses by procedure, site and physician while Aim 2 showed that patient, provider and hospital factors were independent predictors of high radiation utilization. Aim 3 developed a multi-modal radiation reduction intervention at two sites and suggested a decrease in radiation exposure in a high radiation utilization laboratory without significant differences in a low baseline radiation utilization laboratory.
Using this pilot data, we are submitting for a SDP to implement the multifaceted intervention to reduce radiation exposure to a total of 6 high radiation utilization sites. We proposed to roll this intervention into 3 phases, randomizing 2 sites to the intervention during each roll-out. During each roll-out, we will use semi-structured interviews and process measures to assess barriers and facilitators to implementation fidelity and then incorporate lessons learned from each roll-out phase into subsequent intervention phases.
If the intervention is demonstrated to be effective, we will work with our operational partners to disseminate the intervention to the rest of the VA Cath Labs and other specialites that perform procedures requiring radiation.
External Links for this Project
- Valle JA, Petrich M, Carey E, Bradley SM, Gurm HS, Grossman PM, Maddox TM, Duvernoy CS, Nallamothu BK, Rumsfeld JS, Ho M, Tsai TT. A Multimodal Radiation Reduction Intervention for Intra-procedural Radiation Exposure in Patients Undergoing Cardiac Catheterization in Veterans Affairs Hospitals. Poster session presented at: American Heart Association Quality of Care and Outcomes Research in Cardiovascular Disease and Stroke Annual Scientific Sessions; 2013 May 16; Baltimore, MD. [view]
- Tsai TT. Improving Care During and After PCI: Optimal Use of Technology and Periprocedural Complications. Paper presented at: American Heart Association Annual Scientific Sessions; 2011 Nov 15; Orlando, FL. [view]