Regular Article

Academic Pathology
Volume 5: 1–10
Report and Recommendations ª The Author(s) 2018
Article reuse guidelines:

of the Association of Pathology Chairs’ sagepub.com/journals-permissions

DOI: 10.1177/2374289518793988
journals.sagepub.com/home/apc
Autopsy Working Group

Gregory G. Davis, MD, MSPH1 , Gayle L. Winters, MD2, Billie S. Fyfe, MD3,
Jody E. Hooper, MD4, Julia C. Iezzoni, MD5, Rebecca L. Johnson, MD6,
Priscilla S. Markwood, CAE7, Wesley Y. Naritoku, MD, PhD8,
Marcus Nashelsky, MD9, Barbara A. Sampson, MD, PhD10,
Jacob J. Steinberg, MD11, James R. Stubbs, MD12,
Charles Timmons, MD, PhD13, and Robert D. Hoffman, MD, PhD14

Abstract
Autopsy has been a foundation of pathology training for many years, but hospital autopsy rates are notoriously low. At
the 2014 meeting of the Association of Pathology Chairs, some pathologists suggested removing autopsy from the
training curriculum of pathology residents to provide additional months for training in newer disciplines, such as
molecular genetics and informatics. At the same time, the American Board of Pathology received complaints that newly
hired pathologists recently certified in anatomic pathology are unable to perform an autopsy when called upon to do so.
In response to a call to abolish autopsy from pathology training on the one hand and for more rigorous autopsy training
on the other, the Association of Pathology Chairs formed the Autopsy Working Group to examine the role of autopsy
in pathology residency training. After 2 years of research and deliberation, the Autopsy Working Group recommends
the following:
1. Autopsy should remain a component of anatomic pathology training.
2. A training program must have an autopsy service director with defined responsibilities, including accountability to the
program director to record every autopsy performed by every resident.

1
Forensic Division, Department of Pathology, University of Alabama at Birmingham, Birmingham, AL, USA
2
Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA
3
UMDNJ-Robert Wood Johnson Medical School, New Brunswick, NJ, USA
4
Department of Pathology, The Johns Hopkins University, Baltimore, MD, USA
5
Department of Pathology, University of Virginia School of Medicine, Charlottesville, VA, USA
6
American Board of Pathology, Tampa, FL, USA
7
Association of Pathology Chairs, Wilmington, DE, USA
8
Department of Pathology and Laboratory Medicine, USC/LACþUSC Medical Center, Los Angeles, CA, USA
9
Department of Pathology, University of Iowa Carver College of Medicine, Iowa City, USA
10
City of New York Office of Chief Medical Examiner, New York, NY, USA
11
Department of Pathology, Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, NY, USA
12
Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA
13
Department of Pathology and Laboratory Medicine, University of Texas Southwestern Medical Center at Dallas, Dallas, TX, USA
14
Department of Pathology, Microbiology, and Immunology, Vanderbilt University School of Medicine, Nashville, TN, USA

Corresponding Author:
Gregory G. Davis, Forensic Division, Department of Pathology, University of Alabama at Birmingham, 1515 Sixth Avenue South, Room 220, Birmingham,
AL 35233, USA.
Email: gdavis@uabmc.edu

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 License
(http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of the work without further
permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).
2 Academic Pathology

3. Specific entrustable activities should be defined that a resident must master in order to be deemed competent in autopsy
practice, as well as criteria for gaining the trust to perform the tasks without direct supervision.
4. Technical standardization of autopsy performance and reporting must be improved.
5. The current minimum number of 50 autopsies should not be reduced until the changes recommended above have been
implemented.

Keywords
autopsy, pathology training, residency, anatomic pathology, autopsy service director, entrustable activities, rapid autopsy

Received April 30, 2018. Received revised July 11, 2018. Accepted for publication July 16, 2018.

Introduction ACGME, ABP, APC, PRODS, CAP, and NAME. Following

is the list of the scope of tasks that the APC envisioned for
The autopsy has been the foundation of pathology training for
the Group.
generations, but since 1971, when The Joint Commission no
longer required autopsy utilization for hospital accreditation, 1. Assess what systems are used to determine that autopsy
the number of hospital autopsies in the United States has training is adequate or inadequate.
steadily declined.1 At the 2014 meeting of the Association 2. Collect data on how autopsy training is accomplished
of Pathology Chairs (APC), some pathologists suggested that by the programs doing it most effectively.
it was time to remove autopsy from the training curriculum of 3. Identify more specifically what the limiting factors are
pathology residents. The rationale given for this recommen- where training is not effective.
dation was 2-fold. Firstly, fewer autopsies are performed by 4. Evaluate whether the requirement of 50 autopsies
resident and attending pathologists, which has challenged consumes a disproportionate amount of resident train-
some training programs to provide residents the number of ing time.
autopsies required by the American Board of Pathology 5. Review whether and how programs can continue to
(ABP) to sit for the basic qualifying examination in anatomic comply with the 50 autopsy requirement while autopsy
pathology (currently 50 autopsies).2 Secondly, removal of numbers are declining.
autopsy from the residency curriculum would provide several 6. Explore whether new paradigms of training could be
additional months for resident training in newer disciplines in deployed to support the weaker training programs.
pathology, such as molecular genetics and informatics. At the 7. Develop tools to assist programs in addressing relevant
same time, the ABP has received complaints from some estab- issues (live workshops, online resources, webinars).
lished pathologists that newly hired pathologists recently cer-
tified in anatomic pathology by the ABP are unable to perform The Autopsy Working Group addressed these tasks during
an autopsy when called upon to do so as a professional com- its existence from 2014 to 2016. In 2016, the Autopsy Working
ponent of their new position. Group submitted its final report to the APC. The members of
In response to the call to abolish autopsy from pathology the Autopsy Working Group have now revised their 2016
training on the one hand and the call for more rigorous autopsy report into a format suitable for publication as an article in
training on the other, the APC formed the Autopsy Working Academic Pathology, with the members of the Autopsy Work-
ing Group listed as authors of this article.
Group to examine the role of autopsy in pathology residency
training. The APC wanted this work to be a joint effort between
experts in pathology training and in autopsy practice, and so the Materials and Methods
group had 2 co chairs—Gayle L. Winters representing the
Pathology Program Directors (PRODS) and Gregory G. Davis The Autopsy Working Group met 14 times in person or by
representing the National Association of Medical Examiners conference call during its 2-year existence. Much of the infor-
(NAME). The members of the Autopsy Working Group repre- mation that the Autopsy Working Group had to work with was
sented various organizations concerned with resident training, anecdotal in nature, as formal studies of efficacious pathology
training are lacking. The members of the Autopsy Working
resident evaluation, and autopsy practice, including the
Group agreed on the following points.
Accreditation Council for Graduate Medical Education
(ACGME) and its Review Committee for Pathology, the ABP,
the APC with its Residency Program Directors Section Background
(PRODS), the College of American Pathologists (CAP), and
the NAME. By virtue of their appointments, various members 1. Performing an autopsy is the practice of medicine. As
were able to communicate the insights and concerns of the such, autopsy has been central to the development of the
Davis et al 3

science of medicine for centuries. Autopsy has been there no current information about how autopsy training is
similarly central to the development of pathology as a accomplished in US programs, but also there was no available
discipline and to pathology training for over a century. list of autopsy service directors who would be the parties with
2. An ABP survey indicates few practicing pathologists the most direct knowledge of how such autopsy training is
perform autopsies at a rate sufficient to maintain skills performed. To address these needs, the Autopsy Working
in autopsy practice. (Specifically, 46% perform no Group contacted 142 program directors of pathology residency
autopsies, and another 30% of all responders to the training programs in April 2016 using a SurveyMonkey poll
survey perform 1-5 autopsies per year.)3 from the APC to collect the names of persons serving the role
3. Autopsy is unsurpassed as a method of quality assur- of autopsy service director for their program. These 142
ance for assessing sensitivity and specificity of clinical requests led to 120 responses and 113 named autopsy service
diagnoses. directors. In June 2016, the named autopsy service directors
4. Autopsy is an essential component of competent med- were asked to complete a second SurveyMonkey poll about
icolegal death investigation. autopsy training in their programs. The 113 requests yielded
5. With a proper autopsy permit, an autopsy can allow 66 at least partial responses (58% of requests). Of the nonre-
training in aspects of pathology beyond autopsy alone, sponders, 4 opened the survey but did not respond, 42 did not
such as performing a bone marrow biopsy or a fine open the survey, and 1 request could not be delivered through
needle aspirate. e-mail. The autopsy service director survey was conducted with
6. After peaking in the mid-20th century at approximately the identity of the programs known, but not to be individually
50%, hospital autopsy rates have been decreasing since disclosed. Publically available statistics about the numbers of
then, in part related to the decision by The Joint Com- trainees in the programs were obtained from the web site of the
mission to remove the requirement of a minimum num- ACGME and were used to interpret the survey results. Of the
ber of autopsies for hospital accreditation.4 responding autopsy service directors, 28% also served in
7. Hospital autopsy has no direct method for reimburse- the capacity of residency program director.
ment from the Centers for Medicare & Medicaid The Autopsy Service Director Survey addressed several key
Services or third-party payers because it is “medically features of autopsy training, including annual case volumes and
unnecessary.” distribution of special autopsy types (fetal, pediatric, or foren-
sic—not intended to be mutually exclusive) on the main ser-
vice. Case volumes were combined with numbers of residents
Residency Training in the program to estimate the number of autopsies available
per resident on the main autopsy service (4  annual autopsy
1. Pathology is becoming more complex and more volume/total number of residents in the program on the
diverse. Subspecialties within pathology may have little ACGME roster). No correction was made for the number of
or no relationship to other subspecialties with respect to residents in a clinical pathology-only track who do not have an
daily practice. autopsy requirement, for residents in an anatomic pathology-
2. Assuming combined training in anatomic and clinical only track (who may progress through training in only 3 years),
pathology, residents have 48 months to train. More diver- for residents in an anatomic–neuropathology (who may prog-
sity means fiercer competition for training time during the ress through anatomic pathology in only 2 years), or for pro-
48 months. In one study, 16 programs (21% of all survey grams covering more than 1 autopsy service. The percentage of
respondents) no longer have a dedicated autopsy rotation, cases shared by 2 residents on the main service was solicited
but combine autopsy with other rotations.5 and was then used to extend the estimated number of autopsies
3. Currently, 50 autopsies are required to qualify for available per resident, multiplying the above estimate by (1 þ
ABP certification in anatomic pathology/clinical % of autopsies shared).
pathology (AP/CP), anatomic pathology/neuropathol- The survey collected specific data about who usually per-
ogy (AP/NP), or AP only. forms evisceration procedures and the most frequent dissec-
4. Fifty autopsies can require months of assigned rotations tion method used on the main service. For autopsy service
in some institutions, limiting months spent training in directors who were not also the residency program director,
newer, developing fields. the autopsy service director was asked whether a list of
5. Forensic pathology performs a vital function in an autopsy cases completed by each resident was transmitted to
ordered society. Resident training in hospital and for- the residency program director. Autopsy service directors
ensic autopsy pathology is vital to maintain an influx of were also asked about the fraction of autopsies on their ser-
interested and qualified trainees into the currently cri- vices where they serve as the attending pathologist of record.
tically short-staffed field of forensic pathology. Autopsy service directors were asked questions about the
roles of board-certified forensic pathologists and neuropathol-
In attempting to address its charge, the Autopsy Working ogists on the main autopsy service.
Group sensed the need to better understand the present state The sharing of teaching responsibilities on the autopsy ser-
of autopsy training in US residency programs. Not only was vice among various parties (autopsy service director, other
4 Academic Pathology

faculty, fellow, other residents, pathologist assistant, or other

staff) was solicited for a list of entrustable activities that com-
prise the performance of a complete autopsy. For each entrus-
table activity, the autopsy service director was asked to respond
which party most often teaches how to perform the entrustable
activity, with an additional option for each activity to respond
that the activity is not taught.
Autopsy service directors were asked about their opinions
about the current requirement that residents perform or share at
least 50 autopsies, whether the number 50 is too many, too few,
or about right, and then whether residents in their department
have trouble completing the required 50 autopsies. The results
of these opinion questions were further stratified by the number
of residents in the program, either fewer than 18 total residents Figure 1. What was the total autopsy case volume of the service you
or 18 or more residents. direct for calendar year 2015? Each bar on x-axis represents 1 of the
All members of the Autopsy Working Group agreed that 59 programs that responded to this question.
autopsy training is an essential component in transforming a
newly graduated physician into a competent anatomical pathol-
ogist. The various members of the Autopsy Working Group
differed in their opinions of how many autopsies were neces-
sary to achieve competency in autopsy, so much time was spent
discussing methods of documenting competency in performing
an autopsy. The current model requires 50 autopsies for all
residents, which includes a limited number of fetal and
single-organ autopsy examinations. This one number ignores
the truth that some residents are more naturally gifted at
autopsy techniques and assimilating data to enable a sound
diagnosis, but the number has the advantage of being a discrete
measure that is easily assessed. Some members of the Autopsy
Working Group strongly advocate developing a competency
model for assessing residents, so that a resident needs to per- Figure 2. Number of available autopsies in main service per resident.
form no more autopsies than are necessary for the resident to 4 total autopsies on main service/residents on Accreditation Council
for Graduate Medical Education (ACGME) roster. Each bar on x-axis
demonstrate competence in autopsy procedures and diagnosis.
represents 1 of the 59 programs that responded to this question.
No one could offer a satisfying model for a competency-based
training system, however.

Results
Soliciting responses about autopsy training from autopsy ser-
vice directors provided the Autopsy Working Group with
important perspectives from which to make its recommenda-
tions. The 66 responding programs ranged from over 35 resi-
dents to as few as 7 residents and showed good geographic
distribution (data not shown).
There was an enormous difference in the available num-
ber of autopsies for resident education, with a range from
900 to 14 cases per year (Figure 1). Normalizing the avail-
able number of autopsies per resident, without accounting Figure 3. Available autopsies on main service per resident, extended
for sharing, showed a range from 3 to 275 cases per resi- by sharing. Resident  (1 þ % shared). Each bar on x-axis represents 1
dent. Only 18 of the 59 programs responding to this ques- of the 59 programs that responded to this question.
tion attained 50 available autopsies on the main service
without sharing (Figure 2). autopsies per resident using the reported rate of sharing allowed
Sharing of autopsies at rates ranging from 1% to 100% was 24 of 59 programs to attain 50 autopsies per resident on the
reported by 42 of the 60 program respondents (data not shown). main service (Figure 3). It is clear that in order to reach the
Eighteen programs explicitly stated that there was no sharing of required 50 autopsies per resident, many programs depend
autopsies by residents. Extending the number of available upon residents sharing autopsies.
Davis et al 5

Figure 4. Distribution of fetal, pediatric, and forensic cases, percent

of total. The order of the programs is identical to that in Figure 1.
Because the categories are not mutually exclusive, some totals exceed
100%. Each bar on x-axis represents 1 of the 59 programs that Figure 5. Entrustable tasks most often taught by the autopsy service
responded to this question, including the space. director. Numbers shown in bold text under bold text heading.

Special autopsies (fetal, pediatric, and forensic) showed

striking variation in distribution (Figure 4). Although it is clear
that forensic autopsies predominate in most of the programs
with the highest case volumes, several programs with fewer
total autopsies have relatively high numbers of fetal cases and
few pediatric or forensic autopsies.
Questioning autopsy service directors about usual practices
on their services showed remarkable heterogeneity. Of 60
respondents, only 15 indicated that residents performed the
majority of eviscerations, with 45 indicating that the task is
usually performed by the assistant and 5 indicating that evis-
ceration is usually performed by the attending. The most com-
mon dissection technique was en bloc evisceration followed by
dissection for 42 of 60 respondents, organ-by-organ removal
for 17, and in situ examination of organs for 1 respondent. Of
the 46 responding autopsy service directors who were not also Figure 6. Entrustable tasks most often taught by the other faculty.
Numbers shown in bold text under bold text heading.
the residency program director, 28 (60.8%) stated that they do
not provide the residency program director a list of autopsies
completed by each resident.
The distribution of teaching activities on autopsy services
showed interesting trends. Autopsy service directors reported
that they were the main teachers for many of the entrustable
activities comprising a complete autopsy, with the exception of
sampling the brain, removal of the brain and spinal cord, and
restoration of the body for disposition (Figure 5). Other faculty
played important roles teaching neuropathology, histologic
diagnosis, and interpretation of laboratory results (Figure 6).
Other trainees in the program were most involved in teaching
how to review the medical record and to perform organ dissec-
tion and sampling (Figure 7). Support staff were instrumental
in teaching restoration of the body as well as opening and
evisceration techniques (Figure 8). The most common entrus-
table activities that were noted as not taught to residents were Figure 7. Entrustable tasks most often taught by other trainees.
those of interviewing caregivers, restoring the body, removing Numbers shown in bold text under bold text heading.
the brain and spinal cord, and reviewing the medical record
(Figure 9). indicated that forensic autopsies were a part of the experience
The role of forensic autopsies on the main teaching services on the main teaching service. Of the services performing for-
showed considerable variation. Only 25 of 60 respondents ensic autopsies, 12 indicated that a board-certified forensic
6 Academic Pathology

Figure 8. Entrustable tasks most often taught by support staff.

Figure 10. What is your view of the American Board of Pathology
Numbers shown in bold text under bold text heading.
(ABP) requirement that residents perform 50 autopsies? 58
responses; 8 skipped this question.

Figure 9. Entrustable tasks most often not taught. Numbers shown in

bold text under bold text heading.
Figure 11. Do your residents have trouble getting the required 50
cases? 60 responses; 6 skipped this question.
pathologist always supervised forensic autopsies, while 7
stated that a forensic pathologist never supervised the forensic had trouble reaching the required 50 autopsies (56 of 60
autopsies. The resident was responsible for completion of the respondents), 3 of the 4 responses indicating that residents had
forensic autopsy report in 16% of responding programs, had a trouble performing 50 autopsies came from larger programs
variable role for reporting in 37%, and had no role for reporting (Figure 11).
in 54%, with 6% indicating that the forensic experience was
purely observational.
The role of neuropathologists in teaching autopsy was more Discussion
uniform, with 92% of respondents either having a board- In surveying autopsy service directors, the Autopsy Working
certified neuropathologist on-site or visiting to support the Group has gathered the most complete information to date
autopsy service. about the current state of pathology resident autopsy education
The opinions of autopsy service directors about the current in the United States. Although the rate of response was favor-
rule requiring residents to complete 50 autopsies indicated that able, there are important limitations that must be considered in
most (41 of 58 responding) felt that 50 autopsies were about interpreting the data. First, the fact that the identity of respon-
right, with 12 indicating that 50 cases were too few and 7 that dents was used in order to include program size statistics in
50 autopsies were too many. The responses to these questions calculations and comparisons may create a bias in the
did not show significant differences between autopsy service responses, dissuading responses from struggling programs or
directors from small programs (fewer than 18 residents) or eliciting more positive responses in cases where the respondent
large programs (18 or more residents; Figure 10). Although may have doubts. The survey therefore reflects a reality that is
most autopsy service directors did not feel that their residents most likely not better than is reflected by the data presented.
Davis et al 7

Second, in surveying autopsy service directors, we understand Because, as noted above, it is likely that residents learn to
fully that the respondents in many cases have a vested career perform autopsies in more than one setting, they may be
interest in the role of the autopsy in graduate medical education exposed to a variety of procedures and philosophies.
in particular, as well as in the practice of medicine in general.
The Autopsy Working Group itself shares the same bias Resident Education in Autopsy is Not Conducted in a
because of its members. Any recommendations that are put
Standardized Fashion
forward should consider carefully the needs of other stake-
holders, including those of our trainees, the body of employers One predicate of a number-based criterion for assessing com-
of newly trained pathologists, clinical colleagues who from petency in autopsy is that each counted autopsy should have a
time to time request autopsy services with a reasonable expec- similar instructive value from resident to resident and from
tation that a satisfactory examination will be conducted, and program to program. In addition to the previously noted dif-
the public, who ultimately benefit from understanding ferences in available cases and in the mixture of specialized
advances in medicine through postmortem surveillance. Third, cases on different services, the very act of sharing cases
for the sake of simplicity, it was assumed in performing calcu- fundamentally changes the unit value of every case for an
lations and making comparisons based on the total numbers of individual resident. Sharing of cases has allowed for the
residents that all residents were in 4-year anatomic and clinical number-based criterion of 50 cases to last for a few
pathology programs. Even though the great majority of resi- more years on the basis that all residents are required to par-
dents do train in both anatomic and clinical pathology, the ticipate in 8 broadly defined component parts of the autopsy.6
residents in straight clinical pathology, who are not required The reality is that cases cannot be completely shared at the
to train in autopsy, have the effect of making more autopsies most basic level. Only one person in a team will remove the
available for their colleagues. Residents in straight anatomic brain, only one will run the bowel, and only one person will
pathology programs or in anatomic–neuropathology programs, draft the report for the first time. On the other hand, as noted
who are exposed to the incoming volume of autopsies for 3 or 2 above, sharing autopsies does permit for a positive social
years, respectively, rather than 4, have a slight opposite effect interaction among residents, who might otherwise be left
on the availability of cases for their colleagues. alone to complete the complex task.
With the limitations of the survey understood, the Autopsy Perhaps more importantly, our results show great variation
Working Group was able to draw several important conclusions: in the technical aspects of autopsy training among programs. It
is true that there is more than one way to perform an autopsy,
but familiarity with in situ examination only will ill serve a
There is Great Variation Among Programs in the resident hired by a pathology group that practices en bloc dis-
Availability of Autopsies for Training Residents sections. A number-based criterion that admits as equivalents
Our results indicate that not only do total numbers of autop- en bloc dissection as the standard protocol in one program,
sies on the main service under the direction of an autopsy organ-by-organ dissection as the standard protocol in a second
service director vary by orders of magnitude, there is also program, and in situ examination as the standard protocol in a
considerable variation in the mix of cases available for this third program seems to have missed its mark.
training. Both of these factors may affect the training of res-
idents on service. Even extending the numbers of autopsies There is Need for Accountability for Autopsy Training to
available per resident by the reported rates of sharing autop- the Residency Program Director
sies, most responding programs could not achieve the
required 50 autopsy quota on their main services alone. In A relative minority of autopsy service directors are also
many cases, residents are sent offsite to perform forensic residency program directors (28%). As a part of the present
autopsies, which do complement the hospital autopsy experi- application to sit for credentialing by the ABP, the resident
ence, but should not replace it. must represent, and the program director must attest to the
Board, that a stated number-based training requirement has
been met. It is concerning that the majority of autopsy ser-
Autopsy Training is a Team Sport vice directors who are not concurrently the residency pro-
In learning to perform autopsies, pathology residents should gram director do not provide a list of cases completed by
learn to master many component entrustable activities. Although each resident to the program director. More concerning still
much of the teaching comes from faculty (whether a service is that in trying to collect the list of autopsy service direc-
director and others), very significant teaching contributions are tors, it became apparent that no such person existed in some
made by other trainees, as well as by participating support staff. programs.
The sharing of responsibility for autopsy, instituted to
prolong the number-based criterion, has created a dialogue
between residents working together to complete their com-
Recommendations
plex task, which is a valuable team experience to be gained On the basis of the foregoing, the Autopsy Working Group
in residency. makes the following recommendations:
8 Academic Pathology

1. Autopsy Should Remain a Component of Anatomic  Demonstrate an ability to resolve disputes fairly with
Pathology Training diplomacy and tact.
 Believe in and advocate for value of the hospital autopsy
Although the numbers of autopsies have decreased in hospital to medical practice and public health.
settings, and many new-in-practice pathologists perform few
autopsies, if any, the ability to review a medical record, inter- A competent autopsy service director will provide hands-on
view clinical colleagues, perform a thorough examination, and teaching of residents in autopsy performance, from gathering
make a meaningful report add value to the education of resi- information prior to autopsy, to examination and evisceration
dents. This is their main contact with common entities in car- of the body, to the interpretation of findings, autopsy reporting,
diovascular pathology, neuropathology, and renal pathology including composition of the report, and communication of
that are less frequently seen in surgical pathology. The autopsy findings to treating physicians and at conferences. The director
provides excellent opportunity to review anatomy and gain will encourage research by the residents in training; manage
skills handling tissues of every type. quality assurance, staff, and supplies (ultimately); participate in
The Autopsy Working Group endorses the practice of maintaining laboratory and residency training accreditation;
autopsy as relevant to the practice of medicine and as an and keep abreast of the future direction of the autopsy, both
essential component of pathology training, now and in the scientifically and socially. A more complete description of the
future. Autopsy integrates medical knowledge with clinical ideal autopsy service director is presented in Supplemental
history, scientific observation, and pathological test results Appendix 1. Some of the qualities listed in Supplemental
more thoroughly than any other procedure in pathology. A Appendix 1 are aspirational—it is unlikely than any one person
solid foundation in autopsy practice catalyzes the transforma- would embody all these traits as an autopsy service director,
tion of a medical student into a practicing pathologist able to but these are the traits that a pathologist appointed as autopsy
assess data in a given case and synthesize this information so service director should work to embody.
that the appropriate analyses are performed to provide the To the extent that any number-based criterion exists in the
correct diagnosis. Autopsy remains essential for pathology future, the numbers of cases performed by each resident on
training because autopsy practice makes one a better pathol- each service where residents rotate should be independently
ogist in any aspect of anatomical pathology practice and verifiable by the program director.
informs clinical pathology practice too, for those pathologists The residency program director is accountable to the ABP
with combined training in AP and CP. For those individuals for resident training in autopsy pathology, including perfor-
focused on molecular genetics, remember that rapid autopsy mance of at least 50 autopsies currently and ensuring that the
allows pathologists to procure tumor samples for research in resident has appropriately participated in all aspects of the
molecular genetics. Without autopsy training and an active autopsy. Because the residency program director (unless also
autopsy service, this important component of molecular the autopsy director) may not have first-hand knowledge of
genetic research becomes impossible. residents’ autopsy experiences, enhanced communication
between an attending pathologist who has been involved with
the resident on the autopsy service and the program director is
2. A Training Program Must Have an Autopsy Service necessary. Ideally, this role should fall to the director of the
Director With Defined Responsibilities, Including autopsy service. It is in this specific way that the Autopsy
Accountability to the Program Director to Record Every Working Group recommends that an autopsy service director
Autopsy Performed by Every Resident be accountable to the program director.
The Autopsy Working Group recommends the form in Sup-
Proper autopsy training requires the participation of a team, and
plemental Appendix 2 as a means of enhancing communication
setting standards for the performance of autopsies and the edu-
between the residency program director and the autopsy direc-
cation of residents requires oversight to prevent progressive
tor or the director’s designee. This form was derived from
cutting of corners.
program requirements of the ABP and ACGME and is taken
An autopsy service director manages the autopsy service in a
from a program that has been using it successfully to succinctly
teaching hospital with a residency training program. The success-
document resident competence in autopsy performance.
ful autopsy service director is active in teaching, service work,
and research and is also the primary liaison for internal and exter-
nal questions and problems that must be resolved for the continu- 3. Specific Entrustable Activities Should be Defined That a
ing function of the autopsy service. The successful autopsy Resident Must Master in Order to be Deemed Competent
service director recognizes and acts upon appropriate opportuni- in Autopsy Practice, as Well as Criteria for Gaining the
ties for improvement in the autopsy service. At a minimum, an Trust to Perform the Activities Without Direct Supervision
autopsy service director should have the following qualifications:
Entrustable Professional Activities in pathology training have
 Be certified by the ABP in anatomic pathology recently been put forward as a more acceptable model for
 Possess experience in practicing anatomic pathology, defining and evaluating the progress of residents in pathology
preferably with recent experience in autopsy practice. residencies. Unlike many other clinical residencies, pathology
Davis et al 9

training comprises a number of mini-residencies with knowl- from one training program is equivalent to assertion of compe-
edge, skills, and attitudes that residents acquire in different tence from another training program.
orders based upon their program and rotation schedule. The
autopsy experience is only one of those areas where residents
gain the trust of their teachers. We have broadly defined several Future Developments
such skills in conducting and interpreting our survey (Figures
Any change as basic as altering the role of autopsy in pathology
5-9). Other sets of entrustable Professional Activities have been
training will require discussion among stakeholders, including
proposed (McCloskey et al and Supplemental Appendix 3.)7
some or all of the following:

4. Technical Standardization of Autopsy Performance Program directors

and Reporting Must be Improved Autopsy service directors
Department chairs
There is a great need to define the expectations of what it Private practice pathology groups
means to perform and report an autopsy as a resident. The Hospital administrators/Chief Quality Medical Officers
standards may be incorporated in definition of Entrustable Pro- Accreditation Council for Graduate Medical Education
fessional Activities, or in a procedure manual that can be American Board of Pathology
agreed upon by key stakeholders. Being able to compare the American Association of Neuropathologists
autopsies performed in one program with those performed in American Society for Clinical Pathology
another program is mandatory for any number-based criterion College of American Pathologists
that may be adopted in the future. National Association of Medical Examiners
A degree of standardization in the teaching and performance Society for Pediatric Pathology
of autopsies throughout the nation would help ensure adequate United States and Canadian Academy of Pathology
training of pathologists. The Autopsy Working Group recom- American College of Medical Quality
mends communicating the expectations for autopsy training in
America in a white paper concerning the role of autopsy in As stated above, the Autopsy Working Group endorses the
pathology training, including discussion of technical standards practice of autopsy as relevant to the practice of medicine and
of autopsy. as an essential component of pathology training, now and in the
Online learning modules can supplement resident training future. Competent autopsy practice requires integration of med-
and experience in autopsy practice, provide a means for self- ical knowledge with clinical history, scientific observation, and
assessment, and offer standard training education that would be pathological test results more thoroughly than any other consulta-
available to all pathology residents. Finally, incorporating tion in pathology. Autopsy has remained a bedrock for pathology
some questions that test knowledge of autopsy technique into training for this very reason—a thorough foundation in autopsy
the anatomic pathology examinations, such as the Resident In- practice makes one a better pathologist in any aspect of anatomi-
Service Examination or the examination by the ABP, will rein- cal pathology practice and informs clinical pathology practice
force the importance of learning autopsy techniques. too, for those pathologists with combined training in AP and CP.
Nevertheless, pathologists must reform autopsy to make it
5. The Current Minimum Number of 50 Autopsies relevant to current practice. An autopsy report that took a
month to complete may have been useful 50 years ago, but
Should Not be Reduced Until the Changes
medicine is practiced at a much faster pace today, and pathol-
Recommended Above have Been Implemented ogists must adjust to this demand by clinicians and patients
Whatever its shortcomings, the current minimum number of 50 alike. Reports that do not provide clinicopathological correla-
autopsies was endorsed as appropriate for residency training by tion are of little use to clinicians. Other users of autopsy data
the majority of autopsy service directors who responded to the exist—quality assurance officers, public health agencies, fam-
survey (70%). Any move to a competency-based model would ily members of the decedent, insurance companies, and attor-
require agreement among stakeholders on what constitutes neys—and autopsy reports should address all these users, not
competency in performing and reporting an autopsy. Agree- primarily pathologists and clinicians.
ment will require more standardization than currently exists Beyond this, pathology is positioning itself as the specialty
regarding the responsibilities of an autopsy service director, best suited to manage vast data through informatics and com-
the technical skills that constitute competency at performing putational algorithms. Autopsies generate a tremendous
an autopsy, a means of assessing these skills, and the respon- amount of data, but to be useful in the 21st century, these data
sibility and mechanism for reporting achievement of compe- must be reported electronically in a uniform format. Thus, it
tency to the ABP and ACGME. The Autopsy Working Group behooves pathologists to work together, and quickly, to agree
recommends retaining 50 autopsies as the minimum number to a standardized format for all autopsies so that all the autopsy
required for residency training until the stakeholders (see data from the nation can be converted to a structured data
below) agree to standards of autopsy performance and report- format that can be mined as a uniquely powerful database for
ing that will allow confidence that an assertion of competence improving health care.
10 Academic Pathology

All things change. Even the autopsy is undergoing a trans- References

formation. Medical examiner and coroner offices in the United 1. Hoyert DL. The changing profile of autopsied deaths in the
States are gaining access to computed tomography (CT) scan- United States, 1972-2007. NCHS Data Brief. 2011:1-8. http://
ners for use in their autopsy work. Research in the use of CT www.cdc.gov/nchs/data/databriefs/db67.pdf. Accessed July 7,
imaging as an adjunct to autopsy pathology to determine the 2018.
cause of death is still in its first decade, but it is already clear 2. American Board of Pathology. ABPath Booklet of Information.
that autopsy and CT are complementary examination modal- Tampa, FL: American Board of Pathology; 2018:27. http://www.
ities. Computed tomography imaging is superior to autopsy abpath.org/images/booklets/ABPath_BOI.pdf. Accessed July 7,
pathology in demonstrating some diseases and injuries, while 2018.
autopsy pathology is superior to CT imaging in showing other 3. Rebecca L, Johnson MD. 2015 Annual APC/PRODS Meeting,
types of disease and injury.8,9 Together these approaches to
PRODS Plenary Session I, American Board of Pathology Update.
examining a body provide the fullest account yet of the diseases
San Diego, CA; 2015.
and injuries that lead to death. These changes are coming to the
4. Levy B. Informatics and autopsy pathology. Surg Pathol Clincs.
hospital autopsy, too, as some of the offices acquiring CT
2015;8:159-174.
scanners are joint forensic and hospital autopsy services. This
5. Naritoku WY, Powell SZ, Black-Schaffer WS. Evolution of the
model of postmortem examination represents a possible future
pathology residency curriculum: preparing for a positive future.
in which pathologists and radiologists work more closely
Acad Pathol. 2016;3. doi:10.1177/2374289516667746.
together as diagnostic specialists.
6. Accreditation Council for Graduate Medical Education. ACGME
The autopsy still provides a wealth of information that can
Program Requirements for Graduate Medical Education in
benefit medical practice and improve patient care. It is inter-
Anatomic Pathology and Clinical Pathology. Focused revision
esting that a nonpathology specialty recognizes this truth.10
Autopsy remains relevant to medical practice, and therefore, February 4, 2018, effective July 1, 2018. Chicago, IL: Accredita-
it remains relevant to pathology training. The saying “Hic locus tion Council for Graduate Medical Education; 2018:18-19.
est ubi mors gaudet succurrere vitae.” remains true. https://www.acgme.org/Portals/0/PFAssets/ProgramRequire
ments/300PathologyCore2018.pdf?ver¼2018-02-19-085552-
Acknowledgments 553. Accessed July 7, 2018.
The authors thank the Association of Pathology Chairs and Dr Donald 7. McCloskey CB, Domen RE, Conran RM, et al. Entrustable
Karcher, who led the call for this work. professional activities for pathology: recommendations from
the College of American Pathologists Graduate Medical Edu-
Declaration of Conflicting Interests cation Committee. Acad Pathol. 2017;4. doi:10.1177/
The author(s) declare no potential conflicts of interest with respect to 2374289517714283.
the research, authorship, and/or publication of this article. 8. Rutty GN, Morgan B, Robinson C, et al. Diagnostic accuracy
of post-mortem CT with targeted coronary angiography ver-
Funding sus autopsy for coroner-requested post-mortem investiga-
The author(s) received no financial support for the research, author- tions: a prospective, masked, comparison study. Lancet.
ship, and/or publication of this article. 2017;390:145-154. doi:10.1016/S0140-6736(17)30333-1.
9. Decker LA, Hatch GM, Lathrop SL, Nolte KB. The role of
ORCID iD postmortem computed tomography in the evaluation of strangu-
Gregory G. Davis, MD, MSPH http://orcid.org/0000-0001-8932- lation deaths. J Forensic Sci. 2018. doi:10.1111/1556-4029.
296X 13760.
10. Aiello VD. The autopsy as the cornerstone for education and
Supplemental Material research in cardiology. Autopsy Case Rep. 2016;6:1-3. doi:10.
Supplemental material for this article is available online. 4322/acr.2016.055.