Printer Friendly

Accounting for the professional work of pathologists performing autopsies.

More and more, both academic pathology departments and private pathology practices have begun using professional relative value unit (RVU) workload values to measure the clinical productivity of their faculty and/or members, and to compare that productivity with national averages. This metric is being used to determine such things as salaries, bonuses, access to additional resources, etc. Those pathologists who spend a significant portion of their clinical time performing and reviewing autopsies are at a disadvantage when RVUs are used as the metric because RVUs have never been formally established for autopsy procedures. Although the American Medical Association has created Current Procedural Terminology (CPT) codes for autopsies, neither they nor the Centers for Medicare and Medicaid Services have ever established recommended RVUs for these codes.

The Autopsy Committee of the College of American Pathologists sought to assist practices in developing their own practice-specific assessment of the professional work involved in the performance of autopsy by determining proposed values for typical autopsy activities. The Autopsy Committee includes members who practice in academic and/or community hospital settings, as well as in medical examiner offices, and each has at least 10 years of experience performing autopsies.

Members of the Autopsy Committee chose to express their perspective on practice-specific autopsy professional work by relating selected professional autopsy activities to established surgical pathology CPT codes rather than attempting to assign specific RVU values to the existing, but typically not used, autopsy CPT codes. This approach was chosen for a number of reasons. Depending upon the table a particular practice may decide to use, the listed RVU values associated with CPT codes may include adjustments for geographic practice cost indices. The professional RVU component includes not only physician work values (a relative measure of the time, skill, training, and intensity needed to provide a particular service) but also practice expense adjustments and malpractice adjustments. These can vary from year to year and region to region. If the professional work associated with the performance of autopsies is expressed as multiples of established surgical pathology CPT codes, the precise RVU values do not need to be further adjusted for these factors. Finally, however, and perhaps most practically, the main reason for determining the amount of professional work associated with autopsy performance is to assist practices in recognizing this work. In many settings, individual productivity is measured by counting up the number of fee codes each pathologist has billed during a particular period of time. These numbers are then entered into a spreadsheet in the appropriate CPT code column. These spreadsheets typically have columns for the standard, established CPT codes and may not readily accommodate additional CPT codes or write-in RVU values.

It is important to recognize that many individuals have become accustomed to associating RVU values with physician reimbursement. Although the issue of appropriate professional reimbursement for autopsy services is an important one, the goal of the process discussed in this paper is simply to provide a means of recognizing the relative professional productivity associated with autopsy performance and not to propose any specific reimbursement model.


Survey of Autopsy Committee Members

The members of the College of American Pathologists Autopsy Committee, representing individuals from both academic and community practice settings, many of whom are directors of autopsy services, were asked to individually estimate the most appropriate numerical multiplier for the 88309 surgical pathology CPT code that would represent the professional work involved in performing, reviewing, and signing out a typical autopsy. Respondents were asked to separately assess a typical adult autopsy and a typical fetal/neonatal autopsy and were specifically instructed that fractional multipliers were allowed. Evaluation of the brain was to be excluded from the estimate because a different pathologist (neuropathologist) often does that professional work. Submitted values were compiled and then discussed at a face-to-face meeting.

Survey of Autopsy Pathology Education Program Subscribers

In 2009, a survey was conducted of subscribers to the College of American Pathologists Autopsy Pathology (AU) education program, a continuing medical education program based on autopsy cases. This survey asked respondents to indicate, as a free-response question, the amount of time they spent on the professional component (reviewing history, gross organs, glass slides, and writing and/or reviewing the autopsy report) of a typical adult autopsy and a typical neonatal/fetal autopsy. The exact language used was: "Approximately how many hours (on average) are spent per autopsy case for the professional component of an adult autopsy?" and "Approximately how many hours (on average) are spent per autopsy case for the professional component of a fetal/neonatal autopsy?" Those data were subsequently acquired and incorporated into the Autopsy Committee's final recommendation.


The Autopsy Committee of the College of American Pathologists is a 9-member committee of pathologists with a particular interest in autopsy pathology. Members typically serve for staggered 6-year-maximum terms. There is usually representation from academic pathology departments, community hospital pathology departments, and medical examiners' offices. Many members direct the autopsy services at their home institutions. The membership was asked to estimate the professional work involved in adult and neonatal autopsies at their institutions, to be expressed as a multiple of the 88309 surgical pathology CPT code (Table 1). Evaluation of the brain was specifically excluded, because at many centers, a different pathologist evaluates the brain. Six members provided values for adults; committee members who were medical examiners did not participate because the goal was to develop benchmark values for hospital (ie, nonforensic) autopsies. Only 4 members provided values for fetal/ neonatal autopsies because of the limited experience with that type of autopsy at some of the represented institutions. Average values of 6.0 X 88309 and 4.6 X 88309 were obtained for adult and fetal autopsies, respectively. For evaluations of the brain, the consensus opinion among the members of the Autopsy Committee, considering typical cases, was 1.5 X 88309 for an adult brain, and 0.5 X 88309 for a fetal brain. This brings the average total multipliers to 7.5 and 5.1 for complete adult and fetal autopsies.

To validate the values proposed by the Autopsy Committee members, data from a previously conducted survey of subscribers to the College of American Pathologists AU education program were analyzed. The AU program is a case-based continuing education program designed for pathologists who perform autopsies as part of their professional activities. Each year, there is a survey of subscribers to evaluate the program. In 2009, the survey was supplemented with additional questions developed to learn more about autopsy practices at the institutions represented by the program subscribers. The 2 questions relevant to the current study asked how many hours (on average) were spent per autopsy case for the professional component of an adult or a fetal/neonatal autopsy. One hundred seventy-two responses were received for adult autopsies (Figure, A), and 159 for fetal autopsies (Figure, B). In both cases, values ranged from 0 to 80 hours. Those individuals reporting zero hours apparently did not perform autopsies, and those reporting values in excess of 12 hours presumably misinterpreted the question (ie, were not providing per-autopsy values but rather aggregate values). Nonetheless, the data show that there is significant variability in the autopsy practices across institutions. Mean values were significantly higher than median values (Table 2) because of the disproportional effect of values provided by responders who most likely misinterpreted the question. Removing these outlier values (those providing values of zero or values greater than 12 hours) produces mean values very close to the median values.


Decreasing reimbursements for pathologists' professional activity have led many pathology practices to develop methods of measuring the productivity of their professional staff. Despite obvious inequities in fee-code-based calculations (eg, the amount of work required to evaluate a skin biopsy is not comparable to the amount of work required to evaluate a breast biopsy), CPT-based systems are typically being used, both because this is the most common approach in other specialties and because it is arguably the best measure readily available. Business offices obtain counts of the number of CPT codes billed by each pathologist, enter these into a spreadsheet, and then convert these to RVUs to arrive at a total measure of professional productivity. Pathologists who perform autopsies are disadvantaged by this approach because, although CPT codes do exist for autopsy performance, these codes are rarely used because the American Medical Association has never developed RVU values for these codes.


The Autopsy Committee of the College of American Pathologists sought to assist practices wishing to recognize the professional work associated with the performance of autopsies by developing proposed relative measures for this professional activity. In order to most easily allow practices to incorporate our proposal into currently used CPT-based evaluation methods, we chose to express the professional work associated with autopsy performance as a multiple of existing surgical pathology CPT codes, because the above-mentioned spreadsheets already have columns for those codes. Understanding that there is great variability in the autopsy practices across institutions, a conclusion very much supported by the survey data presented here, the Autopsy Committee still felt there was value in providing benchmark values that individual institutions could then tailor to their own specific practice environments.

The first approach used was to poll the members of the Autopsy Committee, a 9-member group of pathologists with extensive familiarity with the autopsy practices at their institutions. Values obtained were remarkably similar (Table 1) with the exception of the values from institution 1, which were higher than the others. On further discussion, it was determined that institution 1 included a detailed clinical-pathologic correlation discussion (typically 1 to 2 pages) in each autopsy report, including an in-depth discussion of the findings correlated with the clinical history, presentation, and circumstances of the death, an explicit statement as to the most likely underlying cause and sequence of events leading to the death, and a researched and referenced discussion of the major disease process(es) present in the patient. This practice was not typical of the other institutions. It was decided that this activity, estimated as being equivalent to approximately 1.5 X 88309, should be removed from the Autopsy Committee's final recommendations and included as an add-on activity. Excluding this from the values of institution 1 drops the average multipliers for adult and fetal autopsies, respectively, to 5.8 and 4.2 without inclusion of the brain, or 7.3 and 4.7 with inclusion of the brain (1.5 for an adult brain, 0.5 for a fetal brain).

To validate these values, the Autopsy Committee analyzed survey data obtained in 2009 from subscribers to the College of American Pathologists's AU education program. Although these data had the advantage of being already available, they used a different unit of measure: the Autopsy Committee members provided values as multiples of the 88309 surgical pathology CPT code, whereas the AU program subscribers provided values in hours. Thus, a conversion factor is needed in order to directly compare these numbers. Literature benchmarks from other specialties include typical values of 3.5 to 6.5 RVUs per hour for emergency medicine, (1,2) 2.5 RVUs per hour for internal medicine, (3) 2.5 to 3.0 for radiology, (4) and approximately 2.1 for primary care. (5) No values could be found in the literature for pathology. An informal poll of pathologists at the author's institution suggested that, based on the types of surgical specimens currently billed as 88309, this typically represented about 50 minutes of professional work. That would suggest a conversion factor of 1.2 X 88309 per hour, or 3.36 RVUs per hour (the 88309 CPT code currently represents 2.80 work RVUs). Using this conversion factor and the mean values obtained from the survey of AU program subscribers after excluding the outliers yields 5.2 X 1.2 = 6.24 for adult autopsies and 4.8 X 1.2 = 5.76 for fetal autopsies. These values presumably include evaluation of the brain, because that was not excluded in the survey question. Comparable values from the Autopsy Committee members were 7.3 and 4.7 (without a detailed discussion).

The existing literature provides very little information about other attempts to determine appropriate RVUs for autopsy performance in the United States. Some of the reluctance to do this seems to stem from the association of RVUs with reimbursements, and this extends into the controversy about whether or not autopsy services are appropriately reimbursed under the current model of being incorporated into Part A payments to hospitals, or whether they should be separately reimbursed. However, in the original study from Hsiao et al (6) at the Harvard School of Public Health in 1991, research that formed the basis for the RVU system adopted by the then United States Department of Health and Human Services proposed some preliminary values for the work involved in performing an autopsy. The scenario described was: "Autopsy, gross and microscopic, with brain for sudden unexpected death of a 38-year-old single male admitted 2 days prior with fever, hypertension, and cardiac arrhythmias." The work value proposed, relative to suggested values for 88309, ranged from 3.4 to 6.9 X 88309, with an average of 4.9 X 88309. This was a relatively uncomplicated autopsy scenario and is unlikely to be representative of the typical medical autopsy performed today. More recently, the Massachusetts General Hospital has developed a compensation model for academic pathologists, derived from the RVU system but modified to account for a number of other factors and to address compensation for autopsies. (7) Details of these authors' conclusions and process were presented in the form of a short course at the United States and Canadian Academy of Pathology, and are available online. (8) They concluded that an adult autopsy, without brain, was the equivalent of 7.5 X 88309. An autopsy with the brain was 9.0 X 88309. Stillborn autopsies were given a value of 5.0 X 88309. These values did include detailed clinical-pathologic discussions, and are very comparable with the values proposed by the Autopsy Committee. Brain-only autopsies were given a value of 3.0 X 88309.

After lengthy deliberation considering all of the information available, the Autopsy Committee proposed the values presented in Table 3. This recommendation credits the pathologist with 5.5 X 88309 for a typical adult autopsy, without the brain. Evaluation of the brain represents an additional 1.5 X 88309 worth of work, but should obviously be credited to the pathologist who evaluates the brain, who may be a neuropathologist rather than the pathologist who evaluates the rest of the case. The complexity of the autopsy report is also an important element to consider. Some institution's autopsy reports simply list significant findings at autopsy, but do not provide further discussion of these findings. Others provide detailed correlation of the autopsy findings with the clinical presentation and course, and include a specific statement as to the underlying cause of death and likely sequence of events leading up to the death. This discussion may be supplemented by literature citations. The Autopsy Committee suggests crediting the pathologist with up to an additional 1.5 X 88309 for the time and effort involved in preparing this type of discussion. Discussions intermediate between these 2 extremes should be appropriately scaled. For a typical fetal autopsy, a base value of 4 X 88309 is suggested for the full fetal autopsy without a brain, with 0.5 X 88309 credit to the pathologist who evaluates the fetal brain (assuming the fetal brain is evaluable) and up to 1.5 X 88309 for a detailed discussion. Additionally, if the placenta is evaluated as part of the autopsy (rather than as a separate surgical specimen), the appropriate already-established CPT code, based on the gestational age, should be used. The lower value for fetal autopsies is not in any way intended to suggest that all fetal autopsies are easier or otherwise require less professional work than adult autopsies. Rather, as with all CPT codes, the value assigned is intended to represent an average for all fetal autopsies, and many fetal autopsies (eg, anatomically normal fetuses that died from placental events) do require less professional effort.

There is clearly significant variability in autopsy practices from institution to institution, as evidenced by the wide range of responses for the average time per autopsy obtained from a survey of pathologists. These variations include the type of cases that typically come to autopsy, the detail of dissection, the number of special studies performed and evaluated, the level of technical and professional assistance available, and the detail and scope of the final autopsy report. Individual practices choosing to recognize the professional work of pathologists who perform autopsies are encouraged to take into account the details of autopsy practices at their institutions, and to adapt the values proposed here to account for those differences. Variation in the detail of the report, as discussed above, is important to consider. Additionally, the type of typical fetal autopsy encountered at one's institution may vary. Some institutions, typically those associated with children's hospitals, may routinely encounter fetal and neonatal autopsies with complex malformations and/or genetic syndromes for which professional work values closer to the adult values may be more appropriate. In contrast, other institutions may have a greater number of elective early terminations where the purpose of the autopsy is simply to confirm that the fetus is normal, and for these cases, lower values for the professional work may be more appropriate.

Individual institutions may also elect to take other factors into account when doing an accounting of the pathologist work associated with the autopsy. Certainly, if the pathologist performs elements of the technical work associated with the case, such as the evisceration or organ dissection without the assistance of a pathology resident, pathologist's assistant, or autopsy technician, that effort should be recognized. Additionally, the inherent inefficiencies associated with intermittent individual cases (ie, it is far more efficient, per case, to sit down and sign out 5 cases in succession that to sign out 1 case every few days), particularly when coordinated viewing between an attending and a resident needs to be arranged to move a case along, may be considered. Finally, some practices use structured templates for their reports, which may decrease the amount of work associated with the case.

The purpose of this study was to develop a mechanism that practices can easily adopt to recognize the professional work of pathologists who perform autopsies, and was not done with any goal toward changing the reimbursement model for autopsies. However, general adoption of mechanisms to recognize this professional activity may represent an important early step in the process of reexamining the current compensation model or in collecting the necessary data to propose alternative models.

Financial support for this work came from the College of American Pathologists, but only to the extent that it funded the committee and carried out the survey discussed in the manuscript.


(1.) Carter KA, Dawson BC, Brewer K, Lawson L. RVU ready?: Preparing emergency medicine resident physicians in documentation for an incentive-based work environment. Acad Emerg Med. 2009;16(5):423-428.

(2.) Hemphill RR, Heavrin BS, Lesnick J, Santen SA. Those who can, do and they teach too: faculty clinical productivity and teaching. West J Emerg Med. 2011;12(2):254-257.

(3.) Johnson T, Shah M, Rechner J, King G. Evaluating the effect of resident involvement on physician productivity in an academic generalinternal medicine practice. Acad Med. 2008;83(7):670-674.

(4.) Sunshine JH, Burkhardt JH. Radiology groups' workload in relative value units and factors affecting it. Radiology. 2000;214:815-822.

(5.) Fairchild DG, McLoughlin KS, Gharib S, et al. Productivity, quality, and patient satisfaction: comparison of part-time and full-time primary care physicians. J Gen Intern Med. 2001;16(10):663-667.

(6.) Hsiao WC, Frazier HS, Loconte M, et al. A National Study of a Resource-Based Relative Value Scale for Pathologists' Services: Phase II. A Report to the College of American Pathologists: 53-60. January 3, 1991.

(7.) Hynes WM, MacMillan DH. Establishing a compensation model in an academic pathology department. Am J Clin Pathol. 2005;124(suppl 1):S8-S15.

(8.) Black-Schaffer WS, Johnson RL. Management and compliance in large and academic pathology practices: USCAP short course 36 [course handout]. http:// Published February 17, 2006. Accessed March 5, 2012.

John H. Sinard, MD, PhD, for the Autopsy Committee of the College of American Pathologists

Accepted for publication March 13, 2012.

From the Department of Pathology, Yale University School of Medicine, New Haven, Connecticut.

The authors have no relevant financial interest in the products or companies described in this article.

The Autopsy Committee during the period of time covered by this project included Cassie Boggs, MD; Elizabeth Burton, MD; Richard Conran, MD, PhD, JD; Michael Bell, MD; Charles Hitchcock, MD, PhD; Jody Hooper, MD; Patrick Lento, MD; Dylan Miller, MD; Joseph Parisi, MD; Harold Sanchez, MD; and Carmella Tan, MD.

Reprints: John H. Sinard, MD, PhD, 310 Cedar St, BML B38, PO Box 208023, New Haven, CT 06520-8023 (e-mail:
Table 1. Survey of College of American Pathologists Autopsy
Committee Members--Summary (a)

                              Institution No.

                 1      2     3     4     5     6    Average

Adult autopsy   7.5    6.5   5.0   6.1   6.1   5.0     6.0
Fetal autopsy   6.0    4.3   4.0   4.1   ...   ...     4.6

(a) Estimates of each pathologist on the Autopsy Committee as
to the amount of professional work involved in performing
autopsies of the indicated type are shown, expressed as a
multiple of the 88309 surgical pathology Current Procedural
Terminology code (eg, 5.0 means 5.0 X 88309). Values in both
cases exclude evaluation of the brain, which was separately
valued, by consensus, at 1.5 X 88309 for an adult brain and
0.5 X 88309 for a fetal brain.

Table 2. Survey of Autopsy Pathology Program Subscribers--Summary (a)

Autopsy Type     Responses   Range,   Median,   Mean,     Mean
                               h         h        h     Excluding

Adult               172       0-80       5       8.2       5.2
Fetal/neonatal      159       0-80       4       6.5       4.8

(a) Participants were asked, "Approximately how many hours (on
average) are spent per autopsy case for the professional component
of a xxxxxx autopsy?" Median and mean responses are shown, as well
as a mean after excluding outlier values (those who responded 0 or
those providing values greater than 12 hours).

Table 3. Autopsy Committee Suggested Values for Measuring
the Professional Activity Associated With the Performance
of Autopsies (a)

Adult autopsy

  Full autopsy (without brain):     88309-26 X 5.5

  Additional (add-on):

    Adult brain                     88309-26 X 1.5

    Detailed clinical-pathologic    88309-26 X 1.5
    discussion (b)

Fetal/neonatal autopsy

  Full autopsy (without brain):     88309-26 X 4.0

  Additional (add-on):

    Fetal brain                     88309-26 X 0.5

    Placenta (if done with          88307-26 X 1.0 for
    autopsy)                        third trimester

                                    88305-26 X 1.0 prior
                                    to third trimester

    Detailed clinical-pathologic    88309-26 X 1.5
    discussion (b)

(a) The final multiplier for a given case is the sum of the
base multiplier for that case type and any appropriate
add-on multipliers. These values do not include any technical
component that may be performed by the pathologist.

(b) For purposes of these recommendations, a detailed
clinical-pathologic discussion is defined as an in-depth
discussion of the findings at autopsy, correlated with the
clinical history, presentation, and clinical course; an
explicit statement as to the most likely underlying cause and
sequence of events leading to the death; and a researched and
referenced discussion of the major disease process(es) present
in the patient.
COPYRIGHT 2013 College of American Pathologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Sinard, John H.
Publication:Archives of Pathology & Laboratory Medicine
Article Type:Report
Geographic Code:1USA
Date:Feb 1, 2013
Previous Article:Approach to intraoperative consultation for donor liver biopsies.
Next Article:Outside case review of surgical pathology for referred patients: the impact on patient care.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters