Assessing Radiologist Assistant Supervision Levels in Florida

Transcription

1 ... PEER REVIEW Assessing Radiologist Assistant Supervision Levels in Florida REBECCA LUDWIG, PhD, R.T.(R)(QM) BETH A TORSIELLO, MIS, R.R.A., R.T.(R) WILLIAM BLAINE B NORTON, MIS, R.R.A., R.T.(R), CNMT, RSO Background As radiologist assistants (RAs) become more prevalent in the radiology work force, it is important to re-evaluate how they practice and are supervised. Purpose To analyze the impact of RA supervision levels on clinical practice. Methods A survey was conducted to evaluate whether a random sample of radiologists in Florida would recommend changes to the current levels of supervision required for advanced-practice technologists. The survey form solicited radiologists opinions concerning practice activities an advanced-practice radiographer would perform and the appropriate supervision levels for each of those activities. Results Reduced levels of supervision were recommended by the respondents for 62.5% of the changes to procedures overall, although the responses varied considerably across the procedures. Recommendations for change were more likely when the radiologist had experience working with a physician extender and for supervision of certain types of procedures. Discussion The study revealed that as more experience is gained working with RAs, radiologists increasingly are supportive of lowering supervision levels. The radiologist assistant (RA) and the radiology practitioner assistant (RPA) function similar to physician extenders in other specialties and are supervised by radiologists according to supervision levels designated by the Centers for Medicare and Medicaid Services (CMS). The levels of supervision are termed personal, direct and general. For personal supervision, the radiologist is required to be present in the room during the critical part of the exam, while direct supervision requires the radiologist to be present in the facility and readily available to provide assistance if needed. For general supervision, the radiologist provides oversight, but is not necessarily required to be physically present. These prescriptive levels of supervision may be viewed by some as limiting the potential benefit of RAs in clinical practice because they allow radiologists no autonomy in providing oversight based on the individual technologist s ability, the patient s condition or the complexity of the particular procedure to be performed. 1 The purpose of this research project was to evaluate whether radiologists support changing the current levels of supervision for procedures performed by RAs. The original survey used as the foundation for developing the RA concept as an advanced-practice model was conducted by the American Registry of Radiologic Technologists (ARRT) in January The findings of the survey translated into the initial role delineation model for advanced practice, with a list of procedures to be performed by the RA and specific levels of supervision to be provided by the radiologist. 2 The original survey was modified for this study of actual clinical practice by removing individual procedural components (eg, starting an intravenous [IV] line, explaining the procedure, inserting the spinal needle, administering the contrast once the needle is placed). Each procedure was listed as a whole (eg, myelogram or upper gastrointestinal tract study [UGI]) and the radiologist was asked to identify the level of supervision required for the RA or RPA performing the procedure. An additional change was that the survey for this study was sent to a sample of radiologists in Florida, whereas the original ARRT study was national in scope. Although these adaptations potentially reduce generalizability of the study, the modifications enabled the investigators to specifically target radiologist preferences in a limited sample as a potential indicator of change since the implementation of the RA model. RADIOLOGIC TECHNOLOGY March/April 2009, Vol. 80/No

2 ASSESSING RA SUPERVISION LEVELS Revisiting the appropriateness of supervision levels is crucial because allowing greater autonomy for the RA in caring for patients is expected to improve the continuity of care, expand access to care in overloaded practices and enhance productivity within the practice. As continuity of care improves, patient satisfaction will likely increase and medical errors will likely decrease. In the words of Bowen et al, RAs can expand their ever-increasing contributions into the department of radiology and can improve the quality of service to patients as a result. 3 Furthermore, if RAs can perform more functions with less radiologist supervision, radiologists would have more time for image interpretation, and as productivity increases, the practice revenues also should improve. As the number of RAs in the work force increases, radiologists would appear to benefit from reducing some of the supervision levels to optimize patient management. Evaluating the impact of nonphysician providers in an academic radiology setting for 5 years showed that physician extenders fill an important niche by increasing the consistency and continuity of care while improving the quality of service patients experience. 3 An important step toward maximizing the acceptance and integration of the RA into clinical practice is evaluating current radiologist opinion regarding modifying the supervision levels that were assigned prior to inception of the RA. RAs can improve the lives of patients and radiologists; this study suggests needed changes to maximize the benefits derived from RAs. Background Previous studies have shown that lower levels of supervision for physician extenders have not adversely affected patient care. One study of patients undergoing stress tests in echocardiography evaluated the complication rates for procedures performed by registered nurses with advanced training. The findings demonstrated comparable complication rates to previously reported studies for physicians conducting the procedure. 4 Another study compared the complication rates for cardiac catheterization with coronary angioplasty performed by physician assistants (PAs) in 929 cases and cardiology fellows in 4521 cases under similar supervision. The 2 groups had comparable complication rates (0.54% vs 0.58%), although the PAs took less time to complete the procedure and exposed the patients to less radiation. 5 A literature review by VanValkenburg et al in 2000 comparing R.T.s with advanced training to radiology residents found the advanced-practice technologists were consistently rated as high or better than the residents in image quality, efficiency and patient safety across the studies. 6 More recently, complication rates in more than 2000 venous access procedures performed by interventional radiology staff, fellows, residents and an RPA were analyzed retrospectively. Although the complication rates for all operators were below the acceptable threshold, the complication rate for the RPA was lower than for the fellows or the residents. 7 All of these studies suggest that midlevel providers with specialized training perform procedures safely and with minimal complications. Using physician extenders with specific expertise also can lead to higher patient satisfaction ratings. Evaluation of collaboration between nurse practitioners and physicians suggests that successful practices garner high satisfaction ratings from patients and their families through good communication, respect for one another s expertise and strong feelings of respect and trust. 8 These types of findings also were reported for pediatric nurse practitioners in an emergency department. Increased continuity of care efficiently delivered by these trained professionals produced higher satisfaction for patients, physicians and administrators. 9 A review of PAs in interventional radiology prior to the introduction of RAs reported that procedure-specific training is an excellent way to improve patient care, although PAs role in imaging may be restricted by regulations related to administering ionizing radiation. 10 Given that RAs have strong technical expertise in imaging combined with advanced patient care and procedural skills taught as a part of the educational process for advanced practice, similar high levels of patient satisfaction are anticipated in practices with RAs. Historically, patients readily accept the physician extender in performing procedures because these providers typically have more time than physicians to respond to patients concerns and provide patient education. 11 Other midlevel providers, such as PAs and nurse practitioners, are supervised according to a standard comparable to the general-level supervision that applies to RAs. Reducing advanced-practice technologists supervision to this level for all procedures would reduce unnecessary restrictions in clinical practice, allow RAs to perform more procedures and thus help maximize patient satisfaction. Reducing supervision levels is expected to further ease the burden on radiologists while increasing practice productivity. A nonexperimental descriptive study of RA students in one program found the potential for improved patient satisfaction and higher productivity. 12 Advantages previously identified for hiring RAs include cost-effectiveness, increased productivity and improved quality of 310 March/April 2009, Vol. 80/No. 4 RADIOLOGIC TECHNOLOGY

3 LUDWIG, TORSIELLO, NORTON patient care. 1,3,7-9,11-13 The American College of Radiology (ACR) endorsed the development of the RA as a strategy to alleviate the shortage of radiologists while maintaining a high degree of imaging quality and patient care. 14 Methods Following internal review board approval, a slightly modified version of a survey originally developed by the ARRT was mailed to 500 randomly selected ACR-registered radiologists in Florida. No consent forms were used; however, all responses were kept confidential and contained no individual identifying information. The survey categorized 36 delineated procedures and activities, including 5 personal supervision procedures, 15 direct supervision procedures, 15 general supervision procedures and 9 additional procedures, performed by RAs according to the current supervision level for each procedure. The responding radiologist indicated whether he or she thought the supervision level should remain the same or should be changed. The survey also asked the respondent to indicate which level of supervision (ie, general, direct or personal) the respondent believed would be appropriate if the level was changed. This represents a slightly different format from the survey conducted by the ARRT in January 2004 when no levels of supervision had been previously assigned. Although the original ARRT survey also listed specific procedural tasks such as acquiring IV access, administering contrast agents and others, this survey listed only the procedures. Other modifications to the survey included asking respondents to identify which physician extender they believed was best suited for their practice and their perceptions regarding how informed they were about the RA concept. A final modification was providing an opportunity at the end of the survey to add other procedures or comments Table 1 Responder Demographics Missing or Multiple Responses Number Percent Number Percent Gender Male Female Years of Experience Range: 2-41 Mean: Specialty General radiology Interventional radiology Neuroradiology Other Practice Type Academic 5 7 Private Government Hospital regarding RA roles and responsibilities. The mailings to radiologists included a survey, contact information in case the respondents had questions and a letter of intent with instructions, an expression of appreciation and a deadline for returning the survey. No further correspondence in the form of reminders or thankyou letters was sent to the sample radiologists. All data received by the deadline were tabulated and analyzed for this project; data received after the deadline were omitted. Results Ninety-five questionnaires were returned from 500 surveys mailed, yielding a response rate of 19%. Seven of the returned surveys were missing pages. The missing items were coded as missing data where applicable. The data were analyzed using basic frequencies, cross-tabulation, Chi-square analysis and multiple regression analysis methods. The P value was set at <0.05 and the confidence interval level was 95%. The typical responder was a male radiologist in a private general practice with 15 years of experience (see Table 1). RADIOLOGIC TECHNOLOGY March/April 2009, Vol. 80/No

4 ASSESSING RA SUPERVISION LEVELS Table 2 Selected Supervision Level by Practice Setting a Academic Private Government Hospital ARRT Delineation (additional procedure Supervision Level not included) No. (%) Personal 20 (13.9) 235 (13.5) 21 (36.2) 44 (12.4) 16.7% Direct 67 (46.5) 712 (41.0) 22 (37.9) 145 (41.0) 41.7% General 57 (39.6) 790 (45.5) 15 (25.9) 165 (46.6) 41.7% Mean b Number Who Rated One Or More Procedures a Number and percentage of procedures for which respondents indicated a preference for each supervision level, summed across all ARRT-delineated procedures. b Personal = 3, direct = 2, general = 1. None of the differences among practice settings in mean supervision level was statistically significant at even the.05 level. Some radiologists responded that current supervision levels are appropriate but others believed that the required supervision level should be changed. For the procedures currently at personal level, about one-half of respondents (57.6%, averaged across all 6 personallevel procedures) selected to keep the same level of supervision. The remaining 42.4% of respondents indicated that those procedures should be changed to either direct (32.8%) or general (9.7%) supervision. For procedures currently listed at the direct level, 78.2% of the radiologists responded that the current level should be maintained, while 13.6% suggested lowering the supervision level to general and 8.2% suggested changing to personal. This percentage varied considerably across the 15 direct supervision procedures, with 5 of the 15 procedures garnering more recommendations for switching to personal supervision than general supervision. For procedures currently listed as general supervision, fewer than 10% of respondents recommended higher levels of supervision for those procedures; of that 10%, most (85.3%) of these respondents selected direct supervision. Overall, 62.5% of the suggestions for change in supervision level recommended less strict supervision. Practice Setting Some differences in the sample results were evident among different types of practice settings, ie, government, academic, private or hospital-based practices. However, given the small number of respondents from academic settings and the low number of personal supervision recommendations, none of these differences were statistically significant for an individual procedure. A dichotomous variable was calculated, but no significant variation was identified between radiologists in private practice compared to nonprivate practice for the preferred levels of supervision except for the additional procedure hysterosalpingogram cannulation. Of the 59 respondents indicating their preferred level of supervision for this procedure, 10 (71.4%) of the 14 respondents practicing in nonprivate settings vs 14 (31.1%) of those in private settings believed that this procedure should be performed under personal (rather than direct or general) supervision. The difference between private and nonprivate settings was not statistically significant for any of the overall indices (ie, for the average of all 45 procedures, all 36 delineated procedures, the 5 personal procedures, all 15 direct procedures, the 15 general procedures or all 9 additional procedures.) Overall, the respondents from government settings appeared to recommend higher levels of supervision, averaged across all procedures for which the ARRT Role Delineation document prescribes a level of supervision (see Table 2). However, this difference was not statistically significant because the difference is based on a comparison among settings. In the academic setting, there were only 2 respondents. Moreover, 1 of these 2 respondents rated only 22 of the 36 delineated procedures; if this respondent s ratings are ignored, the mean level of supervision recommended by the remaining respondent is very close to but slightly lower 312 March/April 2009, Vol. 80/No. 4 RADIOLOGIC TECHNOLOGY

5 LUDWIG, TORSIELLO, NORTON than the means for the other 3 practice settings. Similarly, radiologists with specialized practices supported slightly lower levels of supervision overall (mean averaged across all delineated procedures and all respondents who rated 27 or more of those procedures = 1.577) than the radiologists in general practices (mean = 1.638), but this difference was not statistically significant (see Table 3). Table 3 Selected Supervision Level by Type of Practice a Supervision Level General Radiology Specialized Personal 10.2% 8.6% Direct 43.4% 40.5% General 46.4% 50.9% Mean b Number Who Rated One or More Procedures a Number and percentage of procedures for which respondents indicated a preference for each supervision level, summed across all ARRT-delineated procedures. b Personal = 3, direct = 2, general = 1. None of the differences among practice settings in mean supervision level was statistically significant at even the.05 level. Type of Physician Extender A couple of surprising sample differences emerged in the study. More responders preferred the RA for their practice environments than preferred any of the other 3 types of physician extenders (see Table 4). Furthermore, when the responses of radiologists currently working with RPAs were compared with radiologists working with RAs, radiologists working with RAs appeared to support reducing supervision levels more frequently than radiologists working with RPAs. Further investigation with a higher sample number might be needed to strengthen the validity of this finding. Participants were asked whether they were confused about RAs and if so, in what regard (see Table 5). The responses were broken into 4 parts, but can be combined into 2 categories: 28.4% were uncertain about the scope of practice for RAs and 75% were unclear about billing, reimbursement and supervision levels, which are interrelated factors. The prevalence of these responses may explain why some radiologists are hesitant to employ and use RAs at this time. Table 4 Which Physician Extender Do You Feel Is Best Suited for Your Practice? Frequency Percent Valid Percent Valid RA RPA PA NP Total Invalid Missing Multiple response Total Total RA = radiologist assistant; RPA = radiology practitioner assistant; PA = physician assistant; NP = nurse practitioner Radiologist Characteristics Multiple regression analysis indicated that radiologists with fewer years of practice tended to prefer hiring midlevel providers more often than radiologists with more years of experience. The analysis was performed with 6 predictor variables: whether advanced-practice technologist(s) were on staff, whether there was confusion about the role of the RA, type of practice dichotomized as general or specialized, size of practice, years in radiology and practice type dichotomized as private or nonprivate. The overall numeric index variable for all procedures was the dependent variable. This model did not prove to be very productive; only 4% of the variability in the RADIOLOGIC TECHNOLOGY March/April 2009, Vol. 80/No

6 ASSESSING RA SUPERVISION LEVELS Table 5 Are You Confused About the Role of the RA? Frequency Percent Valid Percent Valid No Yes Total Missing Missing Total If yes, how are you confused? Responses N Percent N Billing % 78.1% Supervision Levels % 75.0% Reimbursement % 75.0% Scope of Practice % 90.6% Total % dependent variable was accounted for by the predictors and the multiple correlation of.209 was not statistically significant. Nevertheless, the model confirmed that, after correcting for the additive effects of the other predictors, respondents who had more years of experience selected higher levels of supervision and those whose practices were specialized preferred lower levels of supervision compared with radiologists practicing general radiology (see Table 6). Although the majority of respondents indicated current supervision levels should remain the same, further analyses suggested that those radiologists who had direct experience with physician extenders supported lowering supervision levels, while radiologists lacking direct experience supported no change to supervision levels. For those procedures with personal level supervision, most radiologists without physician extenders preferred maintaining the current level of supervision, while most respondents with physician extenders preferred lowering the supervision level to direct or general. Similarly, the majority of radiologists without physician extenders indicated that procedures currently listed as direct supervision should be increased to personal supervision, while radiologists with physician extenders were equally likely to prefer maintaining the direct supervision level or reducing the level to general supervision. Among those lacking experience with physician extenders, there was a desire to raise some of the procedures currently listed as general supervision to a personal level (a total of 15 such recommendations across the 15 general procedures and 71 raters). In comparison, there was no desire for increasing supervision level to personal among those radiologists who had experience with physician extenders (0 recommendations Percent of Cases across the 15 general procedures and 13 raters). A majority of radiologists without physician extenders suggested supervision at the personal level (55.4% of the 504 recommendations) for the procedures with no level of supervision currently specified. In contrast, most radiologists with physician extenders suggested direct supervision for these procedures (52.2% of the 113 recommendations) and a higher proportion of employers with physician extenders (12.4%) than of nonemployers (7.5%) preferred general supervision. Radiologists working with physician extenders who participated in this study consistently supported reducing the supervision levels for most of the procedures performed by RAs. For the 6 procedures currently delineated as requiring personal supervision, 41 of the 78 recommendations made by the 13 physician-extender employers were to lower the level of supervision to direct or general. For the 15 procedures currently delineated as requiring direct supervision, more of the 13 employers recommended moving to general than recommended moving to personal supervision for 8 of the procedures, while there were more recommendations for moving to personal than to general supervision in only 5 of the cases. The data also were analyzed to compare the responses from radiologists in one area of practice to radiologists in another area of practice. For example, those whose specialty was women s health were more likely to prefer maintaining the current level of direct supervision for a small bowel examination than the general radiologists. Conversely, general radiologists were more likely to prefer maintaining the current level of supervision for a breast biopsy, whereas those specializing in women s health were 314 March/April 2009, Vol. 80/No. 4 RADIOLOGIC TECHNOLOGY

7 LUDWIG, TORSIELLO, NORTON Table 6 Regression Analysis: Overall Mean Level of Supervision Predicted From Respondent and Practice Characteristics Coefficients(a) Model Unstandardized Coefficients Standardized Coefficients t Significance B Standard Error B Standard Error (Constant = overall mean) Focus Dichotomized Practice Dichotomized (Dichotomous) Do you have an advanced-practice technologist (RPA or RA) working in your facility? How many radiologists are employed in your practice? How long have you been in radiology practice? Are you confused about the role of the RA? 1.15e Dependent variable: Overall mean for all procedures if at least 34 rated. A positive B or thus indicates that those with a high score on the predictor variable preferred a higher mean supervision level than did those with a low score on the predictor. Multiple R =.209; R2 =.044, not statistically significantly different from 0. much more likely to suggest lowering the supervision level for breast biopsy from personal to general. Following the same pattern, the majority of interventional radiologists were more interested in lowering supervision levels for invasive procedures and less supportive of reducing supervision levels for gastrointestinal and genitourinary (GI/ GU) studies, while general radiologists were much more likely to support reducing the supervision levels for GI/ GU procedures than for procedures routinely performed in interventional radiology. Not surprisingly, radiologists employing any type of physician extender opted for lower supervision levels more frequently (25 of the 45 total procedures, 3 of the 6 personal procedures, 2 of the 15 direct procedures, 11 of the 15 general procedures and all 9 of the additional procedures) than radiologists who did not have a physician extender in their practice, although none of these differences was statistically significant. The radiologists employing RPAs consistently preferred stricter levels of supervision than those who employed RAs across most procedures. Mean level of supervision recommended was higher for employers of RPAs than for employers of RAs for 34 of the 45 procedures (5 of the 6 personal procedures, 14 of 15 direct procedures, 13 of 15 general procedures and 2 of 9 additional procedures). However, the sample size was small and none of these differences was statistically significant. Of the 48 respondents who indicated which of the 4 types of radiologist extender (RA, PA, RPA or NP) was best for their practice, a plurality (41.7%) chose the RA. This was statistically significantly greater than the 25% who might have been expected to designate the RA if their choice was random; similarly, the 10.4% who designated the RPA as best for their practice was significantly less than would have been expected by chance. Unsurprisingly, the preference for the RA was stronger (7 of 9 = 78% designating the RA as best) among the respondents who employed some type of radiologist extender than among those who did not (13 of 39 = 33%); this difference is statistically significant (1 = 5.943, P =.015). Surprisingly, the RA vs RPA comparison did not differ among respondents employing RAs, RPAs or both; none of these 8 respondents (including none of the 3 RPA employers) chose the RPA as best for their practice; 7 of them chose the RA as best; and 1 (an RPA employer) chose the PA as best (as did the 1 respondent who RADIOLOGIC TECHNOLOGY March/April 2009, Vol. 80/No

8 ASSESSING RA SUPERVISION LEVELS employed an advanced technologist of unspecified type). One procedure, peripherally inserted central catheter (PICC) placement, was placed in 2 different categories. The ARRT lists PICC placement as direct/general supervision and adds a note that level of supervision [is] dependent upon complexity of procedure. 2 When PICC placement was listed as a direct supervision level procedure in the current survey, 68.6% of respondents indicated the supervision should stay the same, 15.1% preferred a general level and 7.0% suggested it should be personal level. However, when PICC placement was later listed as a general supervision level procedure, 65.1% kept the supervision the same, 24.4% selected direct supervision and only 2.3% restricted it to personal supervision. This suggests that a substantial percentage of respondents (between 44% and 50% of those responding to the PICC placement questions) would adopt the ARRT recommendations as the appropriate level of supervision. For instance, 47% of those responding to either of the PICC placement questions simply could have chosen what was listed as the ARRT-delineated level of supervision and 43% of the respondents to each question could have listed their own opinion. If so, the results for these 2 questions could be accounted for by 34.6%, 49.4%, and 16.0% of the respondents who expressed their own opinions preferring general, direct or personal supervision, respectively when PICC placement was listed as a direct procedure vs 40.4%, 54.5%, and 5.1% when it was listed as a general procedure. In other words, the respondents appeared to have a stronger preference for general supervision of PICC placements. Many comments included on the survey by the respondents indicated that physician extenders are much needed and commonly used in radiology practices. While some radiologists listed needs that were best addressed by specific types of physician extenders, most of the comments were more general in nature. Respondents frequently expressed that the most beneficial physician extender is one who is strong in the technical aspects of imaging. Discussion At the time the ARRT conducted its original survey, no RAs were practicing and a low percentage of radiologists had experience working with RPAs. Given the many unknowns with this type of physician extender, a preference among radiologists for greater restrictions on the practice of RAs seemed a logical expectation. However, a preference for easing the supervision levels over time to facilitate more efficient and effective practice was anticipated as radiologists experienced working with RAs and their comfort levels increased. 1 Our analysis concludes that although the data reflect that nearly one-half of radiologists want to lower supervision levels overall, radiologists who have worked with physician extenders and subspecialists respond differently. There is overwhelming evidence in the survey to show that radiologists who have experience with physician extenders, and in particular those who supervise procedures that radiologists routinely perform and are most used to performing, are becoming very comfortable with physician extenders performing procedures and assuming responsibilities previously carried out solely by radiologists. The perspectives of radiologists are changing as more radiologists become familiar with the role of the RA; one-third of the respondents suggested easing supervision levels for RAs while only about 1% indicated supervision levels should be increased. Therefore, the investigators hypothesis that the opinions of radiologists would indicate the potential need for change is supported by the finding of a preference for reducing the current levels of supervision. To address the study s weaknesses, the authors recommend future surveys be conducted blindly so that the current levels of supervision are not provided, and respondents simply are asked to indicate what level of supervision should be assigned to the procedure. This approach may more accurately reflect radiologists opinions without introducing bias based on the current levels of supervision. In addition, the findings would be strengthened by a larger study of radiologists across the nation to eliminate the geographic restrictions and limitations of a small sample size inherent in this study. A more comprehensive investigation is needed; this study may provide useful information for increasing validity in the research design for future studies. The results of this investigation indicate a positive shift in acceptance of and confidence in RAs. Radiologists appear likely to continue placing more responsibility on RAs and giving RAs more autonomy. 3,11-13 The hard work and strong efforts of the pioneers in this new field already have demonstrated that RAs are reliable and provide significant benefits to radiology; this research indicates the need for further research in support of an expanded role for RAs. 316 March/April 2009, Vol. 80/No. 4 RADIOLOGIC TECHNOLOGY

9 LUDWIG, TORSIELLO, NORTON References 1. Daniels C, Lung C. Radiologist assistants: new developments in a growing profession. RBMA Bulletin. November 2008: American Registry of Radiologic Technologists. Registered radiologist assistant role delineation Accessed February 21, Bowen MA, Torres WE, Small WC. Nonphysician providers in radiology: The Emory University experience. Radiology. 2007;245(1): Kane GC, Hepinstall MJ, Kidd GM, et al. Safety of stress echocardiography supervised by registered nurses: results of a 2-year audit of 15,404 patients. J Am Soc Echocardiogr. 2008;21(4): Krasuski RA, Wang A, Ross C, et al. Trained and supervised physician assistants can safely perform diagnostic cardiac catherization with coronary angioplasty. Catheter Cardiovasc Interv. 2003;58: Van Valkenburg J, Ralph B, Lopatofsky L, Campbell M, Brown D. The role of the physician extender in radiology. Radiol Technol. 2000;72(1): Benham JR, Culp WC, Wright LB, McCowan TC. Complication rate of venous access procedures performed by a radiology practitioner assistant compared with interventional radiology physicians and supervised trainees. J Vasc Interv Radiol. 2007:18(8); Resnick B, Bonner A. Collaboration: foundation for a successful practice. J Am Med Dir Assoc. 2003;4(6): Silvestri A, McDaniel-Yakscoe N. The expanded role of the nurse practitioner in a pediatric emergency department extended care unit. Pediatr Emerg Care. 2005;21(3): Stecker MS, Armenoff D, Johnson MS. Physician assistants in interventional radiology practice. J Vasc Interv Radiol. 2004;15(3): Wright DL, Killion JB, Johnston J, et al. RAs increase productivity. Radiol Technol. 2008;79(4): Smith WL, Applegate KE. The likely effects of radiologist extenders on radiology training. J Am Coll Radiol. 2004;1(6): Bennett SE. Advantages of hiring a radiologist assistant. Radiol Technol. 2006;78(1): American College of Radiology. White paper on the radiologist assistant. /quality_safety/radiologistassistant/whitepaperonthe RadiologistAssistantDoc13.aspx. Accessed February 21, at the university. Beth A Torsiello, MIS, R.R.A., R.T.(R), is a graduate of the University of Arkansas for Medical Sciences master s program for radiologist assistants. She works as an RA at Florida Hospital in Orlando. William Blaine B Norton, MIS, R.R.A., R.T.(R), CNMT, RSO, is a graduate of the University of Arkansas for Medical Sciences master s program for radiologist assistants; he works at Fairfield Memorial Hospital in Fairfield, Illinois. The authors thank the American College of Radiology (ACR) for providing the subject sample and Florida Hospital for their cooperation in conducting the study. The authors also appreciate the leadership and staff of the American Society of Radiologic Technologists (ASRT) and the American Registry of Radiologic Technologists (ARRT) for their ongoing support with implementation of the RA concept and with related research projects. Their commitment, expertise and resources enable advancement of the profession and facilitate the research efforts of others in this area, including our project. Reprint requests may be sent to the American Society of Radiologic Technologists, Communications Department, Central Ave SE, Albuquerque, NM , or communications@asrt.org by the American Society of Radiologic Technologists. Rebecca Ludwig, PhD,R.T.(R)(QM), is associate professor and chairman of the department of imaging and radiation sciences of the University of Arkansas for Medical Sciences in Little Rock. She also is director of the radiologist assistant program RADIOLOGIC TECHNOLOGY March/April 2009, Vol. 80/No

10