| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Health Policy and Practice |
1 From the Department of Radiology, Harborview Medical Center, 325 Ninth Ave, Box 359728, Seattle, WA 98104-8560 (C.C.B.); the Departments of Radiology (C.C.B.) and Health Policy and Administration (W.N.Z.), University of North Carolina, Chapel Hill; and the Department of Management Services, University of North Carolina Health Care System, Chapel Hill (N.D.G.). Received August 25, 2000; revision requested October 5; revision received March 9, 2001; accepted March 16. C.C.B. received funding for this study as a GERRAF Fellow. Address correspondence to C.C.B (e-mail: craige@u.washington.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Direct measurement was made of technologist labor and supply costs of a cohort of 409 consecutive patients with trauma who underwent cervical spine radiography. Probability of cervical spine injury was determined by reviewing emergency department medical records. An animated simulation model was used to combine cost and injury probability estimates to determine resource costs. Sensitivity analysis explored factors that determined costs and estimated uncertainty in model estimations. Comparison was made with other cost estimates.
RESULTS: The average technical resource cost for cervical spine radiography was $49.60. Both direct labor ($19.60 vs $13.33; P < .005) and film ($8.39 vs $6.76; P < .005) costs were greater in patients with high probability of injury than in those with low probability of injury. Overall costs in patients with high probability of injury exceeded those in patients with low probability of injury by 33%. Resource costs exceeded Medicare resource-based relative value unit reimbursements for all patients with trauma.
CONCLUSION: Resource costs of the technical components of cervical spine radiography varied with patient probability of injury and were higher than Medicare reimbursements.
Index terms: Cost-effectiveness • Economics, medical • Spine, injuries • Spine, radiography
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Estimates of cost can be made depending on the perspective of the analysis (eg, societal, institutional, payer) (1–3), the type of cost (eg, variable, indirect, fixed) (1,4,5), and the definition of cost (resource cost, charge, reimbursement) (6,7). Different types of costs may be relevant to address different issues. This article will consider resource costs, the resources that are consumed by the performance of an imaging study. In addition to any institutional financial use, resource costs are relevant because they are the basis for resource-based relative value scale (RBRVS) reimbursement systems (8,9) and are the most appropriate when making cost comparisons among providers. Also, resource costs are an estimate of the opportunity cost of the procedure and therefore represent the appropriate cost for use in the cost-effectiveness analysis (4,10).
In the setting of emergency radiologic examination, how to optimize screening imaging of the cervical spine remains a contentious issue (11). Who should undergo screening and with what imaging modality is an ongoing controversy in the radiology literature. A key component of this debate is the cost of radiography (12), which previously has not been systematically evaluated and reported in the medical literature, to our knowledge.
The primary objective of this study was to apply an activity-based microcosting approach to estimate the resource cost to perform cervical spine radiography in an acute trauma setting. Secondary objectives were to determine the factors, including patient risk of fracture, that drive the cost of cervical spine radiography, to provide a model for the resource cost estimation, and to compare resource costs with other types of cost estimation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Patient Selection
To estimate resource costs of cervical spine imaging, from September 1998 through January 1999 we prospectively enrolled 500 consecutive patients who were examined at our level 1 trauma center or urgent care center and who underwent emergent radiographic evaluation of the cervical spine to exclude injury. The age of patients varied from less than 1 year to more than 90 years of age. Patients were excluded if they did not have trauma or did not undergo radiographic evaluation consisting of at least three views. In addition, for patients seen multiple times, only the initial cervical spine radiography was considered. After exclusions, 409 patients (age range, 0–99 years; mean age, 36 years) were eligible for the study. All examinations were performed with radiographic units (MultiX UH; Siemens Medical System, Erlangen, Germany) with digital acquisition of images and a printout of film hard-copy images for interpretation and image archiving.
Costs
Direct costs consisted of labor, supplies, and equipment. Labor costs consisted of radiologic technologist costs, film management clerk costs, and patient transportation personnel costs. Radiologist costs and other costs of interpreting the images were beyond the scope of this analysis. Technologist time was determined as the total time for the radiographic examination, as recorded by the technologists in the computerized radiologic information system (IDXrad; IDX Systems, Burlington, Vt).
The technologists used a bar code to start and stop each radiologic examination, with the time between the bar codes representing the time devoted by the technologist to that examination. When bar code times were unavailable (scheduled and unscheduled information system downtime and technologist noncompliance), we estimated the technologist’s time from the exposure times recorded on the film hard-copy images, plus we added 5 minutes for setup and patient transfer. Correlation between the two methods of time measurement (radiologic information system and times recorded on film hard-copy images) was evaluated with the Pearson correlation coefficient in 40 (10%) of 409 patients who were selected by using a random number generator and who underwent both methods of measurement.
Often, an overall imaging examination will consist of several imaging procedures, not only radiography of the cervical spine but also of other body parts. For example, a single examination might include radiography of the cervical spine as well as of the lumbar spine and pelvis. To determine the proportion of the total examination time that could be allocated to the cervical spine portion of the examination, we divided the total time by the number of exposures in the entire examination, to determine a time per exposure. The time for the cervical spine radiography consisted of the time per exposure multiplied by the number of exposures in the cervical spine portion of the overall examination. This method assumes that the time to acquire an exposure is approximately the same regardless of the body part being imaged.
According to protocol, at our institution, all trauma evaluations of the cervical spine consist of either three views—anteroposterior, lateral, and open-mouth—or five views—anteroposterior, lateral, open-mouth, and obliques. In practice, both examinations usually also include a lateral swimmer’s view and an additional Fuch view of the odontoid. Multiple attempts may be made for each view to achieve an adequate examination.
To estimate the clerical and transportation time required per examination, we prospectively measured the time spent by the relevant personnel on a series of 104 radiographic evaluations. Student research assistants observed 16 four-hour day and evening shifts and recorded clerical and transportation personnel times for all emergent radiographic procedures. The shifts were selected randomly, by frequency matching to the expected volumes for each 4-hour block during the upcoming month. We assumed that cervical spine examinations would require similar clerical and transportation services as did the other examinations. We did not include the cost of nurses or other emergency department personnel in this assessment of radiographic costs, making the assumption that emergency department staffing will be approximately the same whether the patient is in the emergency department or in the radiography suite.
Supply costs consisted of films, with film storage jackets, and labels. Film jacket and label costs were assumed to be $0.50 per examination on the basis of retail costs. Film use (total number of exposures for the entire examination and for the cervical spine portion of the examination) was recorded prospectively by the radiologic technologists for each of the 409 cervical spine examinations. Cost per sheet of film was determined from the hospital wholesale cost of film and the cost of the film-processing supplies.
Equipment costs consisted of the radiography room purchasing and installation costs, represented by the range of costs paid by our institution recently for suitable equipment. We also included the annual service contracts for maintenance of the machinery, with all costs amortized over the expected 5-year lifetime of the equipment. To determine the maximum number of examinations that could be performed in a given radiography room (total capacity), we measured the daily volume for a year. The average volume in the 5% of days that were busiest was used as an estimate of the maximum attainable volume. The 5% limit was chosen arbitrarily to represent a volume that was attainable but that had face validity as indicative of the system at full capacity.
Overhead Costs
All of our cost measurements reflect the direct resource consumption in the emergency radiology section. However, any hospital or imaging center has some overhead costs not encompassed by a single procedure. Such overhead costs by their nature defy accurate assessment and attribution to a particular procedure and therefore must be estimated. Such overhead costs include the cost of administrative personnel, housekeeping and maintenance staff, physical plant (excluding the radiography room), utilities, and taxes (if applicable). We included an estimate of institutional overhead of 33% in the base case analysis and tested a range of 28%–35% in the sensitivity analysis. These estimates are obtained from published national benchmarks for large academic and nonacademic medical centers (13).
Capacity Costs
The measurements just described of technologist time and cost will only reflect the resource costs if the technologists can work on some task at all times, that is, when the system is functioning at full capacity. However, it is not possible to always function at full capacity in the emergency setting, because patient arrivals and imaging needs are unpredictable. Some reserve or idle time is essential to ensure adequate reserves for times of crisis or high demand. However, like overhead costs, capacity costs cannot be reasonably measured and therefore must be estimated.
To encompass the inefficiency caused by obligatory reserve capacity of the emergency radiology setting, we estimated resource costs on the basis of the average actual capacity at which our institution functioned. Because they included inefficiencies, the average capacity cost estimates per procedure were higher than full-capacity costs. To determine the average capacity at which the emergency radiology section functions, we first determined the daily number of procedures for a calendar year. The number of examinations completed on the 5% of the days with highest volumes was used as an estimate of the maximum capacity. The added cost from this obligatory reserve was described as the capacity costs in the results section. Also, we tested a range of 50%–100% of maximum capacity in the sensitivity analysis; average capacity at our institution in 1998 was 69%.
Subject Risk Stratification
For each of the 409 patients in the cohort, we also determined the probability of cervical spine injury and developed separate cost estimates for three injury probability strata: high, moderate, and low risk (Table 1). Probability of cervical spine injury was determined by means of a retrospective review of the patient’s emergency department medical records, and it represents an estimate of injury probability prior to cervical spine imaging.
|
statistic in a random sample of 39 patients who were examined by both reviewers.
Modeling Strategy and Sensitivity Analysis
To estimate activity-based costs from the multiple cost component measurements, we used an animated simulation modeling software program (MEDMODEL 4.1; ProModel, Orem, Utah). The program operates by running a series of Monte Carlo simulations. For each simulation, values of the component costs and other variables were determined by obtaining samples from the distributions of those variables, as determined from the actual data measurement. The software program allowed for the incorporation into the analysis of the uncertainty inherent in each of the variables that were measured. Our goal was to estimate resource costs at our institution and to derive resource cost estimates that might apply to other emergency radiography settings. Accordingly, we used the software program to perform broad sensitivity analyses that include the uncertainty in our estimates and also vary the estimates of labor and equipment costs to increase generalizability. This sensitivity analysis was designed to encompass the range of equipment costs from other vendors and labor costs from other geographic regions within the United States.
In addition, the modeling system was based on a visual representation of the emergency radiology section that was developed from departmental blueprints. This visual format facilitated assessment of face validity of the cost estimation. All of the steps in the process of acquiring radiographs were included in the model, with appropriate variables for the costs and times for each component. All of the potential pathways that a patient might follow, with associated probabilities, were encompassed. Distributions were entered for each relevant variable, and broad sensitivity analyses were performed. Estimates were derived from the model of resource costs of cervical spine radiography, with differing conditions and scenarios. The details of the model have been reported separately (15,16).
We also compared our resource costs with other methods of estimating costs, including charges, cost-to-charge ratios, and reimbursements. Both charges and radiology department–specific cost-to-charge ratios were obtained from our institutional accounting department. Reimbursements were obtained from the 1999 Medicare RBRVS payment schedules (17). Comparisons of costs between different risk strata were made by using the t test with STATA software (Stata, College Station, Tex).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The source of data was not a significant predictor of cervical spine imaging time in multiple regression analysis performed to assess for confounding by data source and did not contribute appreciably to the model, which indicated that residual confounding by data source was unlikely. Agreement between the two chart abstractors who determined the levels of injury probability was assessed in 38 (9%) of the 409 patients. For simple agreement, the value was 0.86, and for weighted agreement, the value was 0.88.
Cost Estimation
The average resource cost of the technical component of cervical spine radiography was $49.60 and included overhead and capacity costs, with the direct costs (labor, equipment, and supplies) accounting for $29.08. The major components of the direct cost of cervical spine radiography were technologist labor (55%) and film (26%), with minor contributions from nontechnologist labor (9%), equipment (8%), and supplies other than film (2%). Overhead and capacity costs were additional major components of the total cost. Capacity costs ranged from 22.8% ($13.15/$57.77) in patients with a high probability of injury (high risk) to 21.6% ($9.40/$43.51) in patients with a low probability of injury (low risk). Overhead was 33% in the base-case estimates.
The resource cost of cervical spine radiography varied with the level of probability of injury. Resource cost was $43.51 in patients with a low probability of injury and $57.77 in patients with a high probability of injury, a difference of 33% (P < .005). Table 2 illustrates the cost components, which are separated into three risk strata.
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
In earlier work, McNeil et al (18) and Trisolini et al (19) demonstrated that technologist time and therefore radiographic costs increase with the number of views. Our study findings show that the patient characteristic of fracture probability predicts cost. Patients who are involved in high-energy causes of injury, such as high-speed motor vehicle collisions, and patients who have associated injuries, such as head injury, are at higher risk of cervical spine fracture (14). These same patients are also intrinsically more difficult to image because they are placed on spine immobilization boards and in cervical collars. Finally, the presence of other injuries may limit the ability to position high-risk patients for adequate imaging. As a result, more inadequate examinations occur in this group, resulting in a greater need for additional imaging such as computed tomography (CT) (20).
Overall, the greater time and number of radiographs used for patients at high risk of injury lead to greater cost. Our mean cervical spine imaging time for all risk groups combined was also higher than that reported by Trisolini et al (19). This difference is likely due to the fact that all patients in our study had trauma and therefore were more difficult to image. Estimation of cervical spine radiographic time and cost for patients without trauma is beyond the scope of this analysis.
The results of this study of radiographic costs are relevant to the ongoing debate regarding the role of CT as an initial screening strategy in the cervical spine. A number of recent articles (21,22) have demonstrated that the accuracy of CT as an initial cervical spine screening imaging modality exceeds that of radiography. In addition, a Medicare reimbursements-based cost-effectiveness analysis of competing cervical spine screening strategies published previously (12) demonstrated that CT was more cost-effective in high-risk patients. Our current study findings suggest that these Medicare reimbursements underestimated cervical spine radiographic costs, particularly in higher-risk patients. Applying the current higher-resource cost estimates for radiography to the previous cost-effectiveness analysis of CT versus radiography would further support CT as an initial screening strategy in high-risk patients. Of course, whether Medicare reimbursements accurately reflect resource costs of CT screening remains unknown.
Since the early 1990s, the Health Care Financing Administration has been basing reimbursements for outpatient medical procedures on estimates of the resource costs of those procedures. This RBRVS system produced a standardized method of reimbursement nationwide (17,23). In parallel, the American College of Radiology developed an RBRVS for radiologic procedures that was eventually adopted by the Health Care Financing Administration (9). An assumption of the RBRVS reimbursement system is that the resources consumed for a procedure will, on average, be the same at different institutions (assuming efficiency), regardless of the case mix at those institutions. Our results demonstrate that resource costs exceed RBRVS reimbursements for cervical spine radiography in a trauma setting. Further, there was a 33% difference between resource costs of patients with a high probability of injury compared with those with a low probability of injury.
In this study, we examined only a single procedure; therefore the results should not be extrapolated to all of trauma radiology. However, our work does indicate that further research is necessary to investigate the appropriateness of the RBRVS system for major trauma centers. If the injury severity–related cost differences that we observed for cervical spine radiography are also present in other radiologic interventions, then level 1 trauma centers that see higher-risk patients are potentially being reimbursed at a rate that is below that of the actual resource costs.
The differences that we observed between high- and low-risk patients also have implications for cost-effectiveness analysis. One common and accepted method for estimating costs in a cost-effectiveness analysis is to use Medicare reimbursements as a proxy for cost (4,24). However, as the cervical spine example demonstrates, reimbursements may not accurately reflect resource costs, particularly if different patients demand different resources. In the case of cervical spine imaging, any cost-effectiveness analysis must take into consideration the case mix of the included patients. Obviously, data collected from an outpatient clinic that examines patients at low risk of injury would produce different conclusions from those from data collected from a trauma center that sees patients with more severe injuries and therefore has higher costs of imaging.
Collecting precise cost data can be a time-consuming and expensive task, and such efforts may not be necessary for all economic analyses. In many cases, reimbursements may be a suitable and less expensive proxy. However, to justify the use of reimbursements as a proxy for costs, it is incumbent on the investigator to demonstrate that the average resource costs that the RBRVS system is designed to estimate are, in fact, relevant to the population under consideration (1,5).
An additional factor influencing resource cost is capacity. In an ideal situation, there would be no idle time or unfilled capacity in a radiology department. In such a theoretic occurrence, the measured resource costs would approximate the costs allocated from management accounting systems. However, particularly in the emergency or trauma setting, it is essential to have excess capacity to allow for reserves in time of crisis or high demand. Rhea and St Germain (25) demonstrated that waiting time increases exponentially as the amount of excess capacity is decreased. In fact, even at 50% of capacity, emergency radiologic waiting times may exceed 60 minutes. Many institutions address this issue of excess capacity in times of low demand by shifting technologists and patients to and from other sections of the radiology department. However, some excess capacity remains essential.
At our institution, by using the 5% of days with greatest volume as an estimate of maximum capacity, we function on average at approximately 69% of capacity. We can be certain that our maximum capacity for both technologists and equipment is not below this value because we were able to achieve this level consistently. However, the maximum capacity may in fact be higher than our estimate because the 5% limit is arbitrary. We would argue that having the capacity to handle demand that occurs on less than 5% of the days may be inefficient and therefore not a component of resource costs. The use of a higher maximum-capacity estimate would in fact have elevated our cost estimates further. The 69% estimate that we use is somewhat higher than the ideal capacity estimate of Rhea and St Germain (25), a difference that may be explained with our method of determining achievable maximum capacity and our ability to shift inpatients to the emergency radiology section for imaging during times of lower demand.
Ideally, in a resource-based reimbursement system, institutions would be reimbursed on the basis of efficient resource costs. Because institutional allocated costs approach resource costs only at high capacity, there would be a strong incentive to function efficiently (at 100% of capacity), reduce unit costs, and therefore have reimbursements exceed costs. Our analysis indicates that even at high efficiency (100% of capacity), resource costs of cervical spine imaging exceed reimbursements for patients at all risk levels for injury. However, as explained earlier, 100% of capacity is not attainable in the emergency setting. At the more realistic 69% capacity, resource costs are even more in excess of reimbursements.
This activity-based microcosting analysis differs from other published cost analyses that are based on management accounting systems. In management accounting systems (Transition I; Transition Systems, Boston, Mass), all direct costs are totaled and divided by the number of procedures performed, usually weighted by the labor times and/or supply costs of the procedures (26). The key component here is that procedure costs are derived by dividing total costs by volume. Accordingly, the derived costs may be largely dependent on volume (27). In comparison, in our activity-based microcosting analysis, resource consumption is measured directly, with appropriate overhead and capacity costs added. In our approach, resource consumption is determined independently of volume or capacity. Either approach may be more appropriate, depending on the question at hand.
We acknowledge several limitations to our analysis. We relied on data from several sources, including technologist-completed data forms, the radiologic information system, emergency department records, and cervical spine radiographs. There is some potential for error from each of these sources, although efforts were made to ensure the completeness of the data. Because we used two data sources for the technologist times, there is potential for confounding by data source. However, comparison of both methods in a random sample of 10% of the patients revealed no significant differences.
In addition, we explored our data by using a variable for the source of time data and found no significant effects. Finally, we were unable to assess the effect of injury severity on equipment and nontechnologist labor time. However, as equipment and nontechnologist labor reflect only 8% and 9%, respectively, of the total direct cost, the effect on the final cost estimates should be relatively small.
Not all costs can be measured accurately. Factors such as administration, housekeeping, taxes, and utilities are costs to maintain a facility or program, but the fraction of these costs that are due to each department and to each radiologic procedure cannot be accurately assessed. All cost analyses are dependent on the estimation of this overhead or indirect expense. In this article, we estimate the magnitude of the overhead expense by using published national benchmarks applied to direct costs at full capacity. An alternate and less conservative method of estimating overhead costs would be to make calculations on the basis of total direct costs inclusive of excess capacity. This would increase our cost estimates to $62.11, $52.92, and $45.37 for high-, moderate-, and low-risk patients, respectively. To maximize the validity of our overhead estimates, we also provide sensitivity analyses. Even on the basis of the range of overhead costs reported nationally, the precise overhead figure that we use has only mild effect on our results.
Finally, we emphasize that we only estimated costs for the performance of cervical radiography. Patients with trauma may undergo several other cervical spine procedures that were not within the scope of this analysis. Specifically, many severely injured patients undergo lateral cervical spine radiography with a portable x-ray machine used at the site of resuscitation prior to the performance of a full cervical radiography study in the radiology department. Such radiography is performed in addition to the standard cervical spine evaluation and therefore is not included in our study. In addition, questionable findings on radiographs may lead to CT scanning in as many as 22% of the patients in some high-risk groups who do not have a fracture. These induced imaging costs were also outside the scope of this analysis (20).
This investigation was concerned only with patients with trauma who were undergoing radiographic screening. We specifically excluded patients who underwent cervical spine imaging for reasons other than trauma or those who underwent imaging only with CT. In addition, we considered only the technical component of the radiographic imaging. We did not attempt to estimate the cost to interpret the radiographs, but instead we focused on the resource cost to produce the images.
In conclusion, the resource cost of the technical component of cervical spine trauma radiography is dependent on the patient probability of injury, and it exceeds current Medicare RBRVS reimbursement. The higher cost of cervical spine radiography in patients with a high probability of fracture adds further support in favor of CT as an initial screening strategy. These results indicate that further research may be necessary to assess the appropriateness of the current reimbursements for trauma radiologic procedures.
|
FOOTNOTES |
|---|
Abbreviation: RBRVS = resource-based relative value scale
Author contributions: Guarantor of integrity of entire study, C.C.B.; study concepts and design, C.C.B., W.N.Z., N.D.G.; literature research, C.C.B., W.N.Z., N.D.G.; clinical studies, C.C.B., W.N.Z., N.D.G.; data acquisition, C.C.B.; data analysis/interpretation, C.C.B., W.N.Z., N.D.G.; statistical analysis, C.C.B., N.D.G.; manuscript preparation, definition of intellectual content, editing, revision/review, and final version approval, C.C.B., W.N.Z., N.D.G.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  K. Muchantef and H. P. Forman Cost Accounting in Radiology: New Directions and Importance for Policy Am. J. Roentgenol., December 1, 2005; 185(6): 1404 - 1407. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. C. Blackmore Clinical Prediction Rules in Trauma Imaging: Who, How, and Why? Radiology, May 1, 2005; 235(2): 371 - 374. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. D. Bub, C. C. Blackmore, F. A. Mann, and F. M. Lomoschitz Cervical Spine Fractures in Patients 65 Years and Older: A Clinical Prediction Rule for Blunt Trauma Radiology, January 1, 2005; 234(1): 143 - 149. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Tigges, D. L. Roberts, K. H. Vydareny, and D. A. Schulman Routine Chest Radiography in a Primary Care Setting Radiology, November 1, 2004; 233(2): 575 - 578. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. T. Gray, W. Hollingworth, C. C. Blackmore, M. A. Alotis, B. I. Martin, S. D. Sullivan, R. A. Deyo, and J. G. Jarvik Conventional Radiography, Rapid MR Imaging, and Conventional MR Imaging for Low Back Pain: Activity-based Costs and Reimbursement Radiology, June 1, 2003; 227(3): 669 - 680. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Cohen Determining Costs of Imaging Services Radiology, September 1, 2001; 220(3): 563 - 565. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
