Diagnostic reference levels from the ACR CT Accreditation Program.

PURPOSE: The aim of this study was to assess the distribution of CT dose index (CTDI) values reported by sites undergoing ACR CT accreditation between 2002 and 2004. METHODS: Weighted CTDI (CTDI(w)) values were measured and reported by sites applying for ACR CT accreditation, and the percentage of scanners with values above the 2002 ACR diagnostic reference levels (DRLs) was determined. Acquisition parameters for a site's adult head, adult abdominal, and pediatric abdominal examinations were used to calculate volume CTDI (CTDI(vol)), and the average and standard deviation were calculated by year. Histogram analysis was performed to determine 75th and 90th percentiles of CTDI(vol). RESULTS: Between September 2002 and December 2004, 829 scanners underwent the accreditation process. Volume CTDI values (average ± SD) for 2002, 2003, 2004, and 2002 to 2004, respectively, were 66.7 ± 23.5, 58.5 ± 17.5, 55.8 ± 15.7, and 59.1 ± 18.6 mGy for adult head examinations; 18.7 ± 8.0, 19.2 ± 8.6, 17.0 ± 7.6, and 18.4 ± 8.3 for adult abdominal examinations; and 17.2 ± 9.7, 15.9 ± 8.6, 14.0 ± 7.0, and 15.5 ± 8.4 for pediatric abdominal examinations. For 2004 data, 23.8%, 2.3%, and 6.9% of sites reported doses above the 2002 CTDI(w) reference levels, compared with 49.6%, 4.7%, and 15% for 2002 data for adult head, adult abdominal, and pediatric abdominal examinations, respectively. Seventy-fifth percentiles of CTDI(vol) were 76.8 mGy (adult head, 2002 only), 22.2 mGy (adult abdominal), and 20.0 mGy (pediatric abdominal). CONCLUSIONS: From 2002 to 2004, average CTDI(vol) values decreased by 10.9, 1.7, and 3.2 mGy for adult head, adult abdominal, and pediatric abdominal examinations. Effective January 1, 2008, the ACR program implemented United States-specific diagnostic reference levels of 75, 25, and 20 mGy, respectively, for the CTDI(vol) of routine adult head, adult abdominal, and pediatric abdominal CT scans.

Clarifying the relationship between nonradiologists' financial interest in imaging and their utilization of imaging.

OBJECTIVE: Several limitations and deficiencies have been identified in existing studies of physician financial interest in imaging that show financial interest is associated with more imaging. We conducted extensive quantitative analysis of seven deficiencies that have been identified. MATERIALS AND METHODS: Using Symmetry's Episode Grouper, we created episodes of care from all the 2004-2007 health care claims for a random 5% sample of Medicare fee-for-service beneficiaries. We compared utilization of imaging in nonhospital episodes having a nonradiologist physician who had a financial interest in imaging with utilization in episodes with no such physician. We studied 23 combinations of medical conditions with imaging modalities commonly used for these conditions. RESULTS: Across four different definitions of financial interest and the 23 combinations, the relative probability (risk ratio) of imaging was uniformly higher for episodes of physicians with a financial interest, predominantly at p < 0.001. The mean relative probability was 1.87. This mean was little affected by the definition of financial interest used or the definition of the physician deemed responsible for the imaging. Controlling for patient characteristics, illness severity, and physician specialty likewise had little effect. Physicians who had acquired a financial interest averaged a 49% increase in the odds of imaging relative to physicians who had not. Physicians with a financial interest in an imaging modality used other modalities more than did physicians without a financial interest in the index modality. The Deficit Reduction Act's 2007 payment reductions had little effect. CONCLUSION: A financial interest in imaging is associated with higher utilization, probably causally. Limiting nonradiologists' financial interest in imaging may be desirable.

Physician self-referral for imaging and the cost of chronic care for Medicare beneficiaries.

BACKGROUND: As the cost of both chronic care and diagnostic imaging continue to rise, it is important to consider methods of cost containment in these areas. Therefore, it seems important to study the relationship between self-referral for imaging and the cost of care of chronic illnesses. Previous studies, mostly of acute illnesses, have found self-referral increases utilization and, thus, probably imaging costs. OBJECTIVE: To evaluate the relationship between physician self-referral for imaging and the cost of episodes of chronic care. RESEARCH DESIGN: Using Medicare's 5% Research Identifiable Files for 2004 to 2007, episodes of care were constructed for 32 broad chronic conditions using the Symmetry Episode Treatment Grouper. Using multivariate regression, we evaluated the association between whether the treating physician self-referred for imaging and total episode cost, episode imaging cost, and episode nonimaging cost. Analyses were controlled for patient characteristics (eg, age and general health status), the condition's severity, and treating physician specialty. RESULTS: Self-referral in imaging was significantly (P < 0.01) associated with total episode costs in 41 of the 76 medical condition and imaging modality (computed tomography, magnetic resonance imaging, etc.) combinations studied. Total costs were higher in 38 combinations and lower in 3. Even nonimaging costs were much more often significantly higher (in 24 combinations) with self-referral than being lower (in 4 combinations). CONCLUSIONS: We find broad evidence that physician self-referral for imaging is associated with significantly and substantially higher chronic care costs. Unless self-referral has empirically demonstrable benefits, curbing self-referral may be an appropriate route to containing chronic care costs.

Can experienced CT radiologists use technique parameters to predict excessive patient dose? An analysis of the ACR CT accreditation database.

PURPOSE: The aim of this study was to determine, for 3 basic clinical examinations, whether blinded, experienced CT radiologists participating in the ACR's CT Accreditation Program could use scan parameters such as tube current-time product (mAs), tube voltage (kVp), and pitch to predict scanner output settings, expressed as weighted CT dose index (CTDIw) and volume CT dose index (CTDIvol), exceeding CTDIw diagnostic reference levels (DRLs) set by the ACR in 2002 and CTDIvol DRLs adopted by the ACR in 2008. METHODS: CT sites with 829 scanners submitted examinations to the ACR between 2002 and 2004, yielding 518 eligible examinations for analysis (138 adult head CT scans, 333 adult abdominal CT scans, and 47 pediatric abdominal CT scans). The sites' measured CTDIw values for each type of examination were compared with the ACR's CTDIw DRLs in effect from 2002 to 2004 and compared with comments regarding excessively high mAs or kVp made by radiologist clinical reviewers to determine if excessively high-dose index measurements could be predicted. The same analysis was repeated using CTDIvol DRLs adopted by the ACR in 2008 and compared with excessively high mAs or kVp and excessively low-pitch comments. RESULTS: Excessively high mAs or kVp comments yielded sensitivity of only 21.2% in predicting examinations above the CTDIw DRLs, with specificity of 87.6%. Using the 2008 CTDIvol DRLs, the corresponding sensitivity was 13.1% and specificity was 86.2%. Significance was not achieved for use of clinical parameters in predicting either the CTDIw or CTDIvol. CONCLUSION: Experienced CT radiologists cannot reliably use scan parameters to predict examinations that exceed CTDIw or CTDIvol DRLs.

Feasibility of economic analysis of Radiation Therapy Oncology Group (RTOG) 91-11 using Medicare data.

PURPOSE: The specific aim of this analysis was to evaluate the feasibility of performing a cost-effectiveness analysis using Medicare data from patients treated on a randomized Phase III clinical trial. METHODS AND MATERIALS: Cost data included Medicare Part A and Part B costs from all providers--inpatient, outpatient, skilled nursing facility, home health, hospice, and physicians--and were obtained from the Centers for Medicare & Medicaid Services for patients eligible for Medicare, treated on Radiation Therapy Oncology Group (RTOG) 9111 between 1992 and 1996. The 47-month expected discounted (annual discount rate of 3%) cost for each arm of the trial was calculated in 1996 dollars, with Kaplan-Meier sampling average estimates of survival probabilities for each month and mean monthly costs. Overall and disease-free survival was also discounted 3%/year. The analysis was performed from a payer's perspective. Incremental cost-effectiveness ratios were calculated comparing the chemotherapy arms to the radiation alone arm. RESULTS: Of the 547 patients entered, Medicare cost data and clinical outcomes were available for 66 patients. Reasons for exclusion included no RTOG follow-up, Medicare HMO enrollment, no Medicare claims since trial entry, and trial entry after 1996. Differences existed between groups in tumor characteristics, toxicity, and survival, all which could affect resource utilization. CONCLUSIONS: Although we were able to test the methodology of economic analysis alongside a clinical trial using Medicare data, the results may be difficult to translate to the entire trial population because of non-random missing data. Methods to improve Medicare data capture and matching to clinical trial samples are required.

The practice of imaging self-referral doesn't produce much one-stop service.

Imaging as a result of self-referral-when a physician refers patients for imaging tests at a facility owned or leased by the same physician-is widespread. The practice has come under much scrutiny because it is associated with higher volumes of imaging services. Proponents of such self-referral argue that the practice offers patients convenient same-day, one-stop service and allows treatment to start sooner. Our analysis of 2006 and 2007 Medicare data showed that self-referral provided same-day imaging for 74 percent of straightforward x-rays, but for only 15 percent of more-advanced procedures such as computed tomography and magnetic resonance imaging. Policy makers attempting to make the use of imaging more responsible should consider narrowing Medicare's special provision allowing referrals to a physician's own practice so that the provision covers x-rays only.

Imaging self-referral associated with higher costs and limited impact on duration of illness.

Self-referral for imaging services occurs when a physician sends patients to receive an imaging procedure from a device that the physician owns or leases. Advocates argue that this shortens the duration of illness and lowers costs. For twenty common combinations of medical conditions and types of imaging, we evaluated the association between self-referral, duration of illness episode, and three measures of cost. Self-referral was associated with significantly and substantially higher episode costs for most of the combinations of medical conditions and imaging that we studied. There was no decrease in the length of illness, except when doctors self-referred patients to receive x-rays for a few common conditions. These findings indicate that except for x-rays, constraining the self-referral of imaging may be appropriate.

Inpatient CT and MRI utilization: trends in the academic hospital setting.

PURPOSE: The aim of this study was to determine trends in the utilization of inpatient CT and MRI at academic medical centers. METHODS: Surveys requesting inpatient CT volumes, inpatient MRI volumes, discharges excluding newborns, and case-adjusted mix index from 2002 to 2007 were e-mailed to all 123 members of the Society of Chairmen of Academic Radiology Departments. CT and MRI studies per discharge were adjusted using the case mix index (CMI) provided by each hospital to adjust for the differences in patient mix at participating institutions. Trends in adjusted inpatient imaging utilization were compared over time and across responding institutions. RESULTS: Twenty-two of 123 chairs (17.9%) of academic radiology departments, representing all geographic regions and a wide variability in National Institutes of Health research funding ranking, provided responses to our survey. Between 2002 and 2007, there was an increase in median CMI-adjusted CT studies per discharge of 28.0% and an increase in median CMI-adjusted MRI studies per discharge of 19.8%. The largest annual percentage increase in CT utilization (20.2%) occurred from 2003 to 2004, and there was negative growth between 2006 and 2007 of -3.74%. The largest annual percentage increase in MRI utilization (13.9%) occurred from 2006 to 2007, with 3 years of negative growth from 2002 to 2003, 2004 to 2005, and 2005 to 2006. In 2007, there was a wide range in CMI-adjusted CT studies per discharge between institutions from 0.16 to 0.75, with a mean of 0.40 ± 0.18, with a corresponding wide range in CMI-adjusted MRI studies per discharge of 0.04 to 0.16, with a mean of 0.09 ± 0.03. CONCLUSION: There has been large growth in inpatient CT and MRI utilization at academic medical centers. This growth is variable over time and between institutions. Practice leaders can use this information to compare themselves with their peers and to monitor the impact of programmatic improvements on inpatient imaging utilization and in discussions with health system leaders who would like to improve system profitability by decreasing costly inpatient imaging procedures.

Effect of advanced imaging technology on how biopsies are done and who does them.

PURPOSE: To assess national levels and trends in utilization of biopsy procedures during the past decade and investigate the relative roles of biopsy approaches (open, endoscopic, and percutaneous) and physician specialties. MATERIALS AND METHODS: Institutional review board approval was not necessary because only public domain data were used. Aggregated Medicare claims data were used to determine utilization of biopsies performed in 10 anatomic regions from 1997 to 2008. Utilization levels according to biopsy approach and anatomic region were calculated. Trends in the relative utilization of percutaneous needle biopsy (PNB) and imaging-guided percutaneous biopsy (IGPB) were assessed. The relative roles of radiologists and nonradiologists in the performance of all biopsies, PNBs, and IGPBs were evaluated. RESULTS: Biopsy procedures with all approaches increased from 1380 to 1945 biopsies per 100,000 Medicare enrollees between 1997 and 2008, which represents a compound annual growth rate (CAGR) of 3%. Utilization of non-PNBs fell, while the absolute level and relative share of PNBs increased. In 2008, 67% of all biopsies were performed by using a percutaneous route. IGPB as a percentage of all PNBs increased over time in the regions for which data were available. Radiology was the leading specialty providing biopsy services. The total number of biopsies performed by radiologists increased at a CAGR of 8%, and radiologists' share of all biopsies increased from 35% to 56%. CONCLUSION: During the past decade, there was continuing substitution away from invasive approaches and non-imaging-guided percutaneous approaches in favor of PNBs and IGPBs, likely related to increasing use of advanced imaging modalities for biopsy guidance. Consequently, radiologists are performing an increasing share of biopsies across all anatomic regions.

Frequency of use of imaging tests in the diagnosis of pulmonary embolism: effects of physician specialty, patient characteristics, and region.

OBJECTIVE: The purpose of this study was to ascertain whether clinical practice in diagnosing pulmonary embolism is consistent with recommendations in the literature and to explore variations in practice across site of care (e.g., emergency department), physician and patient characteristics, and geographic location. MATERIALS AND METHODS: Medicare 5% research identifiable files were analyzed. The cases of patients with emergency department visits or inpatient stays for a diagnosis of pulmonary embolism or for symptoms related to pulmonary embolism (shortness of breath, chest pain, and syncope) were identified. We determined the number of patients who underwent each type of relevant imaging test and evaluated variations in the first non-chest-radiographic test by site of care and treating physician specialty. Using logistic regression, we studied variations in the use of common imaging tests, exploring variations associated with patient characteristics, physician specialty, site of care, and geographic location. RESULTS: For patients in whom pulmonary embolism might have been suspected, the most common tests were echocardiography (26% of the patients), CT or CT angiography of the chest (11%), cardiac perfusion study (6.9%), and duplex ultrasound (7.3%). For patients with an inpatient diagnosis of pulmonary embolism, the most common tests were chest CT or CT angiography (49%), duplex ultrasound (18%), echocardiography (10.9%), and ventilation-perfusion scintigraphy (10.9%). For patients for whom pulmonary embolism might have been suspected, many large variations were found in practice patterns among physician specialties and geographic locations. There were fewer variations among patients with the inpatient diagnosis of pulmonary embolism. CONCLUSION: Physician practice in the diagnosis of pulmonary embolism is broadly consistent with recommendations. However, variations by physician specialty and geographic location may be evidence of inappropriate imaging.

Who's underworked and who's overworked now? An update on radiologist shortage and surplus.

OBJECTIVE: The purpose of this study was to present an updated report on the radiologist surplus and shortage situation using a recently developed improved measure-namely, the extent to which radiologists desire less or more work if their income were to change by the same percentage as their workload. MATERIALS AND METHODS: Non-individually identifiable data from the American College of Radiology's (ACR's) 2007 Survey of Diagnostic Radiologists were used. Responses were weighted to be representative of all posttraining professionally active radiologists in the United States. Information is presented for all radiologists and according to such factors as type and size of practice, radiologist subspecialty, and geographic region. Multivariable regression analysis was used to identify the probable causal links between desired workload change and characteristics of radiologists and the practices where they work. Comparisons were made with ACR's similar 2003 Survey of Radiologists. RESULTS: The net average workload change sought in 2007 was an approximately 3% increase. In 2003, radiologists on average did not desire a statistically significant change in workload. Regression analysis for 2007 showed a pattern of relative shortages and surpluses that was generally, but not entirely, different from that found in 2003. CONCLUSION: The overall balance between the demand and the supply of radiologists shifted toward a surplus between 2003 and 2007. According to our measure, we judge there was a close balance in 2003, but a 3% surplus in 2007. The employment market seems generally, but not universally, to self-correct relative shortages and surpluses in individual geographic areas and subspecialties within a few years.

What factors affect the productivity and efficiency of physician practices?

BACKGROUND: Increasing the productivity and efficiency of physician practices could help relieve the rapid growth of US healthcare costs and the expected physician shortage. Radiology practices are an attractive specific focus for research on practices' productivity and efficiency because they are home to many purportedly productivity-enhancing operational technologies. This affords an opportunity to study the effect of production technology on physicians' output. As well, radiology is a leader in the general movement of physicians out of very small practices. And imaging is by the fastest-growing category of physician expenditure. RESEARCH DESIGN: Using data from 2003 to 2007 surveys of radiologists, we estimate a stochastic frontier model to study the effects of practice characteristics, such as work hours, practice size, and output mix, and technologies used in work production, on practices' productivity and efficiency. RESULTS: At the mean, the elasticities of output with respect to practice size and annual hours worked per full-time physician were 0.73 and 0.51, respectively. Some production technologies increase productivity by 15% to 20%; others generate no increase. Using "nighthawks"--ie, contracting out after-hours work to external firms that consolidate workflow--significantly increases practice efficiency. CONCLUSIONS: The general US trend toward larger practice size is unlikely to relieve cost or physician shortage pressures. The actual effect of purportedly productivity-enhancing operational technologies needs to be carefully evaluated before they are widely adopted. As the recently-developed innovations of nighthawks and hospitalists show, practices should give more attention to a possible choice to "buy," rather than "make," part of their output.

Imaging evaluation for suspected pulmonary embolism: what do emergency physicians and radiologists say?

OBJECTIVE: The utility of various imaging techniques and strategies for the diagnosis of pulmonary embolism has been studied in randomized control trials and extensively described in the literature. CT and ventilation-perfusion scintigraphy are the mainstays of diagnosis, and MRI is emerging. The purpose of this study was to assess the diagnostic approach to pulmonary embolism practiced by emergency physicians and advised by radiologists. MATERIALS AND METHODS: Questionnaires were sent to emergency physicians and radiologists in Pennsylvania. The questions covered diagnostic strategies for the detection of pulmonary embolism in the usual situations and in clinical circumstances in which the primary imaging technique is considered less desirable. RESULTS: Sixty-two radiologists and 52 emergency physicians completed the survey. Ninety percent of radiologists and 96% of emergency physicians answered that CT was their first-line choice for the diagnosis of pulmonary embolism. The use of ventilation-perfusion scintigraphy increased in the care of patients with renal failure and allergy to iodinated contrast material. MRI was chosen infrequently. CONCLUSION: CT is the overwhelmingly preferred technique for the diagnosis of pulmonary embolism. The role of ventilation-perfusion scintigraphy increases when the use of iodinated contrast material is contraindicated. MRI does not seem to have an important role in practice.

Projected cancer risks from computed tomographic scans performed in the United States in 2007.

BACKGROUND: The use of computed tomographic (CT) scans in the United States (US) has increased more than 3-fold since 1993 to approximately 70 million scans annually. Despite the great medical benefits, there is concern about the potential radiation-related cancer risk. We conducted detailed estimates of the future cancer risks from current CT scan use in the US according to age, sex, and scan type. METHODS: Risk models based on the National Research Council's "Biological Effects of Ionizing Radiation" report and organ-specific radiation doses derived from a national survey were used to estimate age-specific cancer risks for each scan type. These models were combined with age- and sex-specific scan frequencies for the US in 2007 obtained from survey and insurance claims data. We estimated the mean number of radiation-related incident cancers with 95% uncertainty limits (UL) using Monte Carlo simulations. RESULTS: Overall, we estimated that approximately 29 000 (95% UL, 15 000-45 000) future cancers could be related to CT scans performed in the US in 2007. The largest contributions were from scans of the abdomen and pelvis (n = 14 000) (95% UL, 6900-25 000), chest (n = 4100) (95% UL, 1900-8100), and head (n = 4000) (95% UL, 1100-8700), as well as from chest CT angiography (n = 2700) (95% UL, 1300-5000). One-third of the projected cancers were due to scans performed at the ages of 35 to 54 years compared with 15% due to scans performed at ages younger than 18 years, and 66% were in females. CONCLUSIONS: These detailed estimates highlight several areas of CT scan use that make large contributions to the total cancer risk, including several scan types and age groups with a high frequency of use or scans involving relatively high doses, in which risk-reduction efforts may be warranted.

Radiologic and nuclear medicine studies in the United States and worldwide: frequency, radiation dose, and comparison with other radiation sources--1950-2007.

The U.S. National Council on Radiation Protection and Measurements and United Nations Scientific Committee on Effects of Atomic Radiation each conducted respective assessments of all radiation sources in the United States and worldwide. The goal of this article is to summarize and combine the results of these two publicly available surveys and to compare the results with historical information. In the United States in 2006, about 377 million diagnostic and interventional radiologic examinations and 18 million nuclear medicine examinations were performed. The United States accounts for about 12% of radiologic procedures and about one-half of nuclear medicine procedures performed worldwide. In the United States, the frequency of diagnostic radiologic examinations has increased almost 10-fold (1950-2006). The U.S. per-capita annual effective dose from medical procedures has increased about sixfold (0.5 mSv [1980] to 3.0 mSv [2006]). Worldwide estimates for 2000-2007 indicate that 3.6 billion medical procedures with ionizing radiation (3.1 billion diagnostic radiologic, 0.5 billion dental, and 37 million nuclear medicine examinations) are performed annually. Worldwide, the average annual per-capita effective dose from medicine (about 0.6 mSv of the total 3.0 mSv received from all sources) has approximately doubled in the past 10-15 years.

The actual role of CT and ventilation-perfusion scanning in workup for suspected pulmonary embolism: evidence from hospitals.

OBJECTIVE: Over the past two decades, CT has been found valuable in the diagnosis of pulmonary embolism (PE). We sought to ascertain the relative roles of CT and ventilation-perfusion (V/Q) scanning, the previously preferred technique, in the diagnosis of PE in recent practice and whether there is variation among hospital types. MATERIALS AND METHODS: Using the Medicare anonymized 5% of beneficiaries complete claims file for 2005, we studied the use of relevant CT and V/Q scanning in the evaluation of patients with a diagnosis of PE and of patients with symptoms that might have been due to PE (chest pain, syncope, difficulty breathing). In 2008, we surveyed the radiology departments of Pennsylvania hospitals about the use of CT and V/Q scanning for PE, service availability hours, and what equipment was used. RESULTS: In all data, we found that CT was used approximately six times as frequently as V/Q scanning. In the Medicare data, only small differences in frequency of use of CT and V/Q scanning were associated with hospital characteristics. Academic hospitals did not differ in a major way from other hospitals, nor did small or rural hospitals. In the survey, 97% of radiology departments reported that CT was available for evaluation of PE 24 hours a day 7 days a week. Ninety-three percent of departments reported V/Q scanning was available at some times; 77% reported V/Q available at all times. CONCLUSION: CT was a fully disseminated and dominant technique for the diagnosis of PE by 2005, and it was readily available at small and rural hospitals. The lack of availability of off-hours V/Q scanning at a substantial fraction of hospitals may be a problem for patients with contraindications to CT.

Radiology practices' use of external off-hours teleradiology services in 2007 and changes since 2003.

OBJECTIVE: Our objective is to report patterns of utilization of external off-hours teleradiology services (EOTSs) in 2007 and changes since 2003. MATERIALS AND METHODS: We analyzed non-individually identified data from the American College of Radiology's 2007 Survey of Member Radiologists and its 2003 Survey of Radiologists. Responses were weighted to be nationally representative of individual radiologists and radiology practices. We present descriptive statistics and multivariable regression analysis results on the use of EOTSs in 2007 and comparisons with 2003. RESULTS: Overall, 44% of all radiology practices in the United States reported using EOTSs in 2007. These practices included 45% of all U.S. radiologists. Out-of-practice teleradiology had been used by 15% of practices in 2003. Regression analysis indicates that, other practice characteristics being equal, in 2007, primarily academic practices had lower odds of using EOTSs than private radiology practices. Also, large practices (>or= 30 radiologists) had lower odds of using EOTSs than practices with 15-29 radiologists. Small practices (1-10 radiologists) had high odds, but nonmetropolitan practices did not. There were no significant differences by geographic region of the United States. CONCLUSION: Use of EOTSs was widespread by 2007, and it had been increasing rapidly in the preceding few years. Patterns of use were generally as might be expected except that nonmetropolitan practices did not have high odds of using EOTSs.

Workload of radiologists in United States in 2006-2007 and trends since 1991-1992.

PURPOSE: To determine radiologists' workloads in 2006-2007, as measured by both procedures per full-time equivalent (FTE) radiologist and relative value units (RVUs) per FTE radiologist, and to discover trends since 1991-1992. MATERIALS AND METHODS: Non-individually identifiable data from the American College of Radiology (ACR) 2007 Survey of Radiologists were compared with data from previous ACR surveys; all surveys were weighted to make them nationally representative. Under National Institutes of Health rules for protection of human subjects, studies based on anonymized surveys do not require approval by an institutional review board. Workload according to individual practice characteristics, such as type (eg, academic, private, multispecialty) and setting, was tested for statistically significant differences from the average for all radiologists. Time trends and the independent effect on workload of practice characteristics were measured with regression analyses. Changes in average procedure complexity were calculated in physician work RVUs per Medicare procedure. RESULTS: In 2006-2007, the average annual workload per FTE radiologist was 14,900 procedures, an increase of 7% since 2002-2003 and 34.0% since 1991-1992. Annual RVUs per FTE radiologist were 10 200, an increase of 10% since 2002-2003 and 70.3% since 1991-1992. Academic practices performed about one-third fewer procedures per FTE radiologist than others. In most types of practice, radiologists in a 75th-percentile practice performed at least 65% more procedures annually than radiologists in a 25th-percentile practice. Regression analysis showed that practices that used external off-hours teleradiology services performed 27% more procedures than otherwise similar practices that did not use these services. CONCLUSION: Radiologists' workload continued to increase in recent years. Because there is much unexplained variation, averages or medians should not be used as norms. However, such statistics can help practices to understand how they compare with other, similar practices. SUPPLEMENTAL MATERIAL: http://radiology.rsnajnls.org/cgi/content/full/2522081895/DC1.

The effect of teleradiology on time to interpretation for CT pulmonary angiographic studies.

PURPOSE: The aim of this study was to evaluate the impact of a teleradiology service on the time to interpretation for computed tomographic (CT) pulmonary angiographic studies. METHODS: A survey of clinical and imaging physicians was performed to develop achievable goals for the interpretation of CT pulmonary angiographic studies. Percentages of studies given preliminary written reports within these thresholds were compared for 485 CT pulmonary angiographic studies completed 3 months before teleradiology was implemented and 617 studies completed 3 months afterward. A total of 1,638 CT brain studies completed over identical periods were used for comparison. Statistical significance (P < .05) was evaluated with 2-tailed t-tests. RESULTS: The median of the optimal time to the preliminary written interpretation of a CT pulmonary angiographic study reported by radiology chairs was 60 minutes, compared with 20 minutes for emergency medicine physicians, who also reported a 40-minute limit for an acceptable time to interpretation. There were statistically significant improvements in the percentages of these studies interpreted within the 60-minute (51% to 62%; P < .01) and 20-minute (9% to 13%; P < .05) optimal time thresholds, as well as within the 40-minute acceptable time threshold (34% to 43%; P < .01). No statistically significant improvements occurred for CT brain studies. CONCLUSIONS: The use of teleradiology to interpret off-hours inpatient imaging serves as an important process improvement tool in decreasing the time to preliminary written reports for CT pulmonary angiographic studies. By establishing agreed-on time standards for reporting such examinations, radiologists and clinicians can collaborate to ensure the prompt diagnosis and treatment of potentially lethal illnesses, such as pulmonary embolism.

Diagnostic radiologists' subspecialization and the new final board examination.

OBJECTIVE: Recognizing that subspecialization can consist of concentration in multiple fields as well as in a single main field, we conducted this study to profile in detail the subspecialization of diagnostic radiologists in the United States in ways that illuminate issues related to the American Board of Radiology plan for a new final examination. MATERIALS AND METHODS: We tabulated nonindividually identified data from the American College of Radiology 2003 Survey of Radiologists, a stratified random-sample mail survey with 1,924 responses and a 63% response rate. Respondents were guaranteed confidentiality. Responses were weighted to make them representative of all radiologists in the United States. RESULTS: Sixty-nine percent of respondents reported specializing at least to a small extent. If concentration in a field is defined as spending 10% or more of clinical work time in the field, 51% of radiologists concentrated in one or two fields, 24% in three or four fields, and 21% in more than four fields. An examination covering a radiologist's four most time-intensive fields would cover 80% of the clinical work of the median radiologist. However, the one fourth of radiologists whose work is most varied would have 40% or more not covered by the examination, but the one fourth with the most concentrated work would have 100% covered. CONCLUSION: Most radiologists concentrate in a few fields, making the American Board of Radiology plan for an examination that covers four fields--or fewer, at an examinee's discretion--a major step forward in recognizing the nature of current practice. Four fields, however, are too many for the practice patterns of many radiologists but too few for the practice patterns of a substantial minority. We offer for consideration more far-reaching reforms.