HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Conoley, P. M. Search for Related Content PubMed PubMed Citation Articles by Conoley, P. M. Social Bookmarking                      What's this?

Productivity of Radiologists in 1997

Estimates Based on Analysis of Resource-Based Relative Value Units

¹ Kelsey-Seybold Clinic, P.A., 4527 Nenana Dr., Houston, TX 77035-3627.

Received December 16, 1999; accepted after revision February 16, 2000.

 
Address correspondence to P. M. Conoley

Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
OBJECTIVE. Productivity of radiologists was quantified using the resource-based relative value scale for examining trends in workload.

MATERIALS AND METHODS. Staffing and workload data for 1997 were collected in a survey of radiology departments in multispecialty clinics. Workload ratios were calculated and were compared with prior surveys of the same clinics and with published data.

RESULTS. Fifteen clinics reported 3,234,730 examinations and 1,860,729 resource-based relative value units (RBRVUs) performed by 284 radiologists and 28 fellows serving 6305 providers. Productivity ratios were as follows: physician index, 19 physicians per radiologist; provider index, 23 providers per radiologist; availability index, 0.78; difficulty index, 0.54 RBRVUs per examination; examination index, 11,559 examinations per year per radiologist; RBRVU index, 6090 RBRVUs per year per radiologist. Each index had roughly a twofold range of variation from lowest to highest ratio observed. Among diagnostic and interventional procedures, 37% of the supervision and interpretation RBRVUs were in general radiography, 41% in sectional imaging, and 22% in special procedures. Since 1973, the percentages of sectional imaging and special procedure examinations and RBRVUs have increased, and the difficulty index has increased. The physician index has been relatively stable. Non-supervision and interpretation codes constitute approximately 18% of the reported RBRVUs. RBRVU valuation of total radiology services has held steady or slightly increased between 1993 and 1997.

CONCLUSION. RBRVU workload of radiologists in the clinics appears to be increasing primarily because of an increase in the percentages of highly valued sectional imaging and interventional and angiographic studies, which constituted 63% of the diagnostic imaging RBRVU workload. The ranges of the indexes among the clinics varies greatly.

Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
In 1989, the American College of Radiology (ACR) and the Health Care Financing Administration (HCFA) quantified radiology services for purposes of reimbursement using relative value units (RVUs) [1]. In the assignment of the professional component of radiology services, consideration was given to such factors as the training, knowledge, skill, stress, and time required of the radiologist in performing each procedure. The ACR relative value scale was used as a measure of radiologist productivity for 1989 [2]. In 1992, the ACR relative value scale was supplanted by the resource-based relative value scale (RBRVS), which has been revised annually [3, 4]. For purposes of reimbursement, each procedure code of Current Procedural Terminology (CPT) [5] was assigned various technical and professional RBRVU fractions for work, practice expense, and professional liability. This study was undertaken to update the method reported for 1989 [2] by applying the RBRVS to quantify the productivity of radiologists in multi-specialty group practices, ignoring the expense and liability fractions and using only the work fraction as the measure of productivity. The work fraction was used as published without correction by geographic practice cost indexes.

Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
The study population consisted of 21 large multi-specialty clinics across the United States that are members of the American Society of Clinic Radiologists, an ACR-affiliated organization. The clinics that are members of the society have been unchanged for the last 15 years. The clinics were asked to provide a comprehensive list of the volume of procedures performed during a recent 12-month period (calendar or fiscal) part of which overlapped calendar year 1997, for each individual code in the diagnostic radiology, sonography, nuclear imaging, radiotherapy, vascular, surgery, and general sections of the CPT [5]. For procedures for which RBRVUs had not been assigned by HCFA, the investigator assigned RBRVUs of similar techniques so that all reported examinations would be included in the RBRVU analysis. Approximately 3.8% of the reported workload was estimated in this manner, predominantly in screening mammography (CPT 76092), which has no RBRVU. (For screening mammography, the value 0.35 was used; that value was assigned to it by HCFA in the 1993 RBRVS. Doppler sonographic imaging studies were categorized as sonography even through they have CPT codes in the 90000 series.) No audit was performed to determine the accuracy of the use of the coding system by the clinics or to ascertain whether there was overreporting or underreporting of examinations. This study assumes that differences among clinics in the use of the CPT system have introduced only random error into the data.

An RBRVU workload was calculated by multiplying the number of times each procedure was performed by its corresponding professional (work component) RBRVU. The sum of these individual CPT workloads gave the entire professional RBRVU workload for each clinic. The total RBRVU workload and the total number of examinations were numerators for the productivity calculations. Data were also obtained regarding the number of radiologists performing the work of each clinic's departments, and these data served as denominators for the workload ratios. Several of the clinics had fellowship programs. In the workload calculations, fellows were counted as two thirds of a full-time equivalent, consistent with the method of the 1989 report [2].

Four indexes were calculated to describe operational characteristics of the radiology practices. A physician index (PI) was calculated as the ratio of the total number of physicians (medical doctors or doctors of osteopathy) in the multispecialty practice to the total number of radiologists in the practice as follows:

(1)

A provider index (PrI) was calculated as the ratio of the sum of physicians plus other providers (e.g., nurse practitioners and physician's assistants) to the total number of radiologists:

(2)

An availability index (AI) was calculated as the ratio of the number of radiologists available to do clinical work at any given time to the total number of radiologists in the group:

(3)

A difficulty index (DI) was calculated by dividing the total RBRVU workload by the total number of examinations to express, in RBRVUs per examination, the weighted level of difficulty of the procedures performed in the practice:

(4)

Finally, two workload indexes were calculated. An examination index (EI), the total number of examinations per year per capita radiologist:

(5)

and an RBRVU index (RI), the number of RBRVUs per year per capita radiologist:

(6)

These indexes are similar to the indexes defined by Conoley and Vernon [2]. An important difference is that the examination and RBRVU indexes in this study were calculated per capita radiologist rather than per available radiologist, as previously defined.

The indexes were calculated on each clinic's data with the clinic—rather than the individual radiologist—being the unit of analysis. The indexes then were statistically described. The data were also analyzed for the entire population of workers and workload in aggregate (i.e., as if all radiologists and all workload were in one department).

After exclusion of codes for radiotherapy, nuclear medicine, and office visits, each of the remaining diagnostic and interventional CPT codes was categorized to analyze the workload distribution among examination techniques. The categories were general radiography and fluoroscopy (head and neck, chest, extremity, and spinal conventional radiography; genitourinary, gastrointestinal, musculoskeletal, and intrathecal contrast-enhanced studies; and mammography); sectional imaging (sonography, MR imaging, and CT of the head and body); and special procedures (angiography, neuroangiography, and interventional radiology).

This same manpower survey was performed previously for RBRVU productivity in the years 1993 and 1995 (Conoley P, unpublished data [American Society of Clinic Radiologists Manpower Reports, 1993, 1995]). The current 1997 data are presented here in depth, along with highlights of the surveys from 1993 and 1995 to display trends. Clinics that reported in both 1993 and 1997 provided a set of matched data. For these clinics, the Student's t test was performed on the differences of the indexes between 1993 and 1997 (standard for significance, p = 0.05).

Johnson and Abernathy [6] reported a breakdown by imaging technique of projected national radiology procedures in the United States for 1973 and 1980 that was used as a baseline for comparing the data from 1989 [7]. The 1973 and 1980 data from Johnson and Abernathy have been recategorized, and RBRVUs by technique were applied to the categories by the method of Conoley et al. [7], allowing evaluation of trends in RBRVU distribution from 1973 through 1997. Nuclear and radiation therapy procedures were excluded in the longitudinal analysis because not all surveyed departments offered those services, and thus they would be underestimated in comparison with historic baselines.

Finally, since its implementation in 1992, the RBRVS has been evolving with additions and deletions of CPT codes and modifications of the RBRVUs assigned to existing codes. More important, in implementing federal policy HCFA has systematically reweighted the RBRVS to transfer health care reimbursement dollars away from specialty physicians (including radiologists) to primary care physicians. To assess the deflationary effect of these changes, the RBRVU total was calculated on the 1997 examination counts using the 1993 RBRVU values for comparison with the 1997 RBRVU total.

Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Of the 21 clinics, 17 responded to the questionnaire for 1997; however, only 15 provided a comprehensive list of CPT code volumes for inclusion in the RBRVU analysis. Of these 15 clinics, seven were located in the midwestern United States, two in the South, four in the Northeast, and two in the West and Southwest. Data were submitted for 12-month periods with ending dates from as early as July 1997 to as recently as June 1998.

Altogether, these 15 practices reported 3,234,730 examinations corresponding to 1,860,729 RBRVUs. These examinations were performed by 284 radiologists and 28 fellows supporting 6305 physicians or providers in the clinics. Eleven clinics provided information regarding the percentage of clinical work that was under prospective payment plans. The weighted prepaid fraction of the examinations in these clinics was 28%. Only three of the surveyed departments performed radiation therapy services, and only eight performed nuclear imaging.

The productivity indexes calculated on individual clinic data and aggregated data are presented in Table 1. The table includes counts and RBRVU totals for all reported imaging techniques, including nuclear medicine and radiotherapy examinations, office visits, support codes, and surgery codes. There was roughly a twofold variation in the range (highest observed to lowest observed values) of each of the indexes. On the basis of aggregated data, the 1997 examination index was 10,746 examinations per year per capita radiologist, and the RBRVU index was 6144 RBRVUs per year per capita radiologist. The physician index was 19 physicians per capita radiologist, the availability index was 0.78, and the difficulty index was 0.57.

Table 2 shows the productivity indexes and percentage of staffing in the 1993, 1995, and 1997 American Society of Clinic Radiologists surveys. The clinics were at about 98% of the approved staffing level between 1993 and 1997. The data indicate an increase in RBRVU workload caused by an increase in difficulty index during the period. Some clinics that reported in earlier surveys did not participate in the 1997 survey, and conversely, some clinics participated in 1997 but not in some earlier year. To eliminate noise in the data caused by changing subjects in a small population, the data were also analyzed restricted to the 15 clinics that participated in both 1997 and 1993. Table 2 also displays this modified analysis, once again showing the trend to increasing RBRVU workload. The 1997-1993 difference in the difficulty index is significant at p = 0.01 (Student's t test).

Table 3 presents the breakdown of the reported 1997 examination population and the distribution of RBRVUs by technique category. Surgery, visit, and support codes outside the 70000 range of CPT codes constituted approximately 3.0% of the examinations and 17.8% of the RBRVUs.

Table 4 and Figure 1A,1B compare the current workload distributions for the American Society of Clinic Radiologists RBRVU surveys from 1993, 1995, and 1997 with the previous RVU survey from 1989, and with the data of Johnson and Abernathy [6] gathered for 1973 and 1980. In this analysis, nuclear medicine and radiation therapy procedures, visits, surgery, and support codes are excluded. The analysis shows a clear trend in distribution of diagnostic imaging (nontherapy, nonnuclear) workload among techniques—namely, the percentages of sectional imaging and special procedures have increased. Because sectional imaging and special procedures have higher difficulty indexes than general radiography, the percentage of increase of RBRVU workload is even greater in these techniques. Accordingly, the overall difficulty index for all diagnostic procedures has risen from 0.32 in 1973 to 0.57 in 1997.

View larger version (43K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1A. —Changes in distribution among techniques between 1973 and 1997. Bar charts show percentage of diagnostic examinations (A) and resource-based relative value units (B) by category: gray areas indicate general imaging, white indicates sectional imaging, and black indicates special imaging. Codes for surgery, nuclear medicine, radiation therapy, support, and visits are excluded from these charts. Data for 1973 and 1980 are reprinted with permission from [6]; data for 1989 are reprinted from [2]; and data for 1993, 1995, and 1997 are from American Society of Clinic Radiologists manpower surveys (Conoley PM, unpublished data).

View larger version (41K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1B. —Changes in distribution among techniques between 1973 and 1997. Bar charts show percentage of diagnostic examinations (A) and resource-based relative value units (B) by category: gray areas indicate general imaging, white indicates sectional imaging, and black indicates special imaging. Codes for surgery, nuclear medicine, radiation therapy, support, and visits are excluded from these charts. Data for 1973 and 1980 are reprinted with permission from [6]; data for 1989 are reprinted from [2]; and data for 1993, 1995, and 1997 are from American Society of Clinic Radiologists manpower surveys (Conoley PM, unpublished data).

Table 5 shows the RBRVU deflation analysis applying the 1993 and 1997 RBRVSs to the 1997 examination volumes. The general radiography, fluoroscopy, CT, sonography, angiography and interventional radiology, nuclear medicine, and radiation therapy subtotals did systematically decrease as programmed. These decreases were offset by radiology's participation in the programmed increases in surgery and visit code valuations, a small increase in MR imaging RBRVUs, and a marked increase in mammography RBRVUs. Altogether, the 1997 RBRVU total is 2.6% higher than the total using the 1993 scale.

Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
This study used management information made available by radiology practices in a select group of specific multispecialty clinics. The clinics do not constitute a random sample of multispecialty clinics, nor of national radiology practices in general. Almost two thirds of the radiologists in this study are involved with radiology training programs, which is a higher proportion than in the general population of radiologists in the United States. Nevertheless, the multispecialty clinic model entails a fairly integrated practice with a relatively defined base of physicians and other providers referring a relatively defined population of patients to a captive radiologist workforce. This is useful in studying the proportion of radiologists in a practice community. Overall, for the responding clinics, this proportion has been stable between 1989 and 1997.

Even these clinics, however, are dynamic organizations, some exhibiting significant staffing and operational changes over the period studied. Specifically, it is quite likely that learning about the structure of the coding system accounts for some of the observed increases. Certainly, feeling the pressure of declining reimbursement, departments benefit from becoming more proficient at capturing more of the billable RBRVUs in their practices. Such improved clerical efficiency could cause an apparent increase in the RBRVU measure of work without any real change in the activity of the radiologists. The extent to which the increases reported in this paper are apparent, rather than real, was not investigated.

The study by Johnson and Abernathy [6] was designed to project overall national volumes of examinations and the distribution of examinations among techniques. That study shows actual examinations performed by a subset of practitioners in selected multispecialty clinics. Although extrapolation of this data to estimate the 1997 national distribution is beyond the scope of our study, the dramatic shift of RBRVU workload from general radiology to advanced imaging techniques observed in Table 5 would seem to exemplify a national trend.

The increase in mammography RBRVUs between 1993 and 1997 was caused by the changes in unilateral diagnostic mammography (CPT 76090), which increased 132% from 0.25 to 0.58 RBRVUs, and in bilateral diagnostic mammography (CPT 76091), which increased 68% from 0.41 to 0.69 RBRVUs. No increase over 1993 was imputed by the investigator for screening mammography (CPT 76092); consequently, this report may underestimate the screening mammography contribution to workload.

The ACR Research Department was consulted to provide an external reference of comparison with the data. According to the ACR (Busheé GR, personal communication), the examinations in this study represent approximately 1% of all radiology examinations performed by radiologists in the United States in 1997, and the radiologists in the study are approximately 1% of all radiologists estimated by the ACR to have been working in the United States in that year. In a recent study from the ACR, Sunshine and Burckhardt [8] reported a national annual mean of approximately 11,700 examinations and 6000 RBRVUs per radiologist in nonacademic practices and 4000 RBRVUs per radiologist in academic practices. Their figures corroborate the findings in our survey. Most likely, some of the institutions from our survey also responded to the ACR practice survey. Regarding RBRVU deflation, the 2.6% increase in valuation from 1993 to 1997 is a welcome, if surprising, result.

The CPT and RBRVU systems capture only direct (billable) patient care work; there are many other necessary and valuable activities of radiologists that are not captured by RBRVUs. These include continuous availability for consultation, call responsibilities, participation in case conferences, commuting between practice sites, continuing education, research, didactic teaching, clinical instruction of residents, and administrative duties. None of these is easily quantifiable in terms of clinical care-based measures of productivity; yet they must be thoroughly considered in the evaluation of the performance of the staff of any radiology practice.

Over decades, each of the clinics in the study has evolved a radiology department that uniquely and individually meets the needs of the clinic in its unique and individual practice setting. The presumption of the investigator is that each department is prima facie right for its setting, that the various clinics have optimized operations in their individually appropriate ways, according to the values that underlie their decision making. For an experienced observer, the differences among the clinics are more interesting than the similarities. Because the reported means are artifical constructs, they ought to be used only as rough guides that evaluate only one aspect of a practice. The "tyranny of the average" should be avoided; that is, the data must not be interpreted with the judgmental expectation or requirement that any one practice should have the same performance profile as any other practice. Even more important, no practice should be expected to perform at the maximum value of all the indexes.

In conclusion, our study attempted to quantify only patient care work that is captured by the CPT and RBRVU systems. The difficulty of examinations (as measured by RBRVUs) and the RBRVU workload appears to be increasing. The ranges of indexes among the clinics vary widely. Sectional imaging, angiography, and interventional radiology now constitute approximately 74% of the diagnostic imaging RBRVU workload. Nonsupervision codes constitute approximately 18% of the reported RBRVUs. Radiologists' work measures may not be as adversely affected as thought by recent systematic changes in the RBRVSs. Many activities of radiologists are not captured by RBRVUs but must be recognized as valuable—even if not reimbursable—productivity. In applying RBRVUs to measure a practice, great care must be taken to perceive those unique operational characteristics that explain why a given practice understandably differs from mean values or trends.

Acknowledgments
 
I thank Mike Nelson of Minnesota Health Plans and the other members of the American Society of Clinic Radiologists for their help over the years in defining and refining the survey instrument and for their diligence in providing the exhaustive operational data required for this analysis; and Jonathan Sunshine and other members of the Research Department of the American College of Radiology for their technical support and for encouraging that these data be published.

References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Medicare programs: fee schedules for radiologists services. Federal Register. March 2, 1989:8922 -9023
2. Conoley PM, Vernon SW. Productivity of radiologists: estimates based on analysis of relative value units. AJR 1991;157:1337 -1340[Abstract/Free Full Text]
3. Hsiao WC, Braun P, Becker ER, et al. A national study of resource-based relative value scales for physician services: final report to the Health Care Financing Administration (publication 17-C-98795/1-03). Cambridge, MA: Harvard School of Public Health, 1998
4. American Medical Association. Medicare RBRVS: the physician's guide. Chicago: American Medical Association, 1998
5. American Medical Association. CPT: physicians' current procedural terminology. Chicago: American Medical Association, 1998
6. Johnson JL, Abernathy DL. Diagnostic imaging procedure volume in the United States. Radiology 1983;146:851 -853[Abstract/Free Full Text]
7. Conoley PM, Vernon SW, Burtram SG. Productivity of radiologists in the United States by imaging technique: a 16-year analysis based upon relative value units. Eur J Radiol 1992;15:258 -263[Medline]
8. Sunshine JH, Burckhardt JH. Radiology groups' workload in relative value units and factor affecting it. Radiology 2000;214:815 -822[Abstract/Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				J. H. Thrall Reinventing Radiology in the Digital Age: Part I. The All-Digital Department Radiology, August 1, 2005; 236(2): 382 - 385. [Full Text] [PDF]

				S. Ondategui-Parra, J. G. Bhagwat, K. H. Zou, E. Nathanson, I. E. Gill, and P. R. Ros Use of Productivity and Financial Indicators for Monitoring Performance in Academic Radiology Departments: U.S. Nationwide Survey Radiology, July 1, 2005; 236(1): 214 - 219. [Abstract] [Full Text] [PDF]

				T. Vertinsky and B. Forster Prevalence of Eye Strain Among Radiologists: Influence of Viewing Variables on Symptoms Am. J. Roentgenol., February 1, 2005; 184(2): 681 - 686. [Abstract] [Full Text] [PDF]

				M. Bhargavan and J. H. Sunshine Workload of Radiologists in the United States in 1998-1999 and Trends Since 1995-1996 Am. J. Roentgenol., November 1, 2002; 179(5): 1123 - 1128. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Conoley, P. M. Search for Related Content PubMed PubMed Citation Articles by Conoley, P. M. Social Bookmarking                      What's this?