A Follow-up on the Autopsy

This follow up on the Autopsy has data and suggestions for returning the autopsy to a significant role in medical education (both initial and continuing education of physicians and all other health professionals).

Pathologists have reported autopsy findings since before the 18^th century. Autopsy findings contributed to major improvements in diagnosis in the North American Continent. In 1912, Richard Cabot, MD, from the Massachusetts General Hospital, reviewed the results of 3,000 autopsies and demonstrated that clinical diagnosis was frequently incorrect[i]. William Osler, MD, the author of one of the first comprehensive textbooks of internal medicine[ii] performed over 1,000 autopsies himself. It is said that a large portion of the knowledge he shared in his book was because of his experience at the autopsy table. Even after he stopped doing his own autopsies, Osler continued to attend autopsy organ review at Johns Hopkins because he still thought that the autopsy was important.

In the latter portion of the 20^th century, the autopsy seemed to have taken a lesser place in the education of the practitioner of medicine. There may be at least three reasons for this downplaying of the autopsy:

1. The autopsy takes time and specific skills to perform properly. Many physicians/pathologists believe that they have other, potentially more important (or remunerative) priorities
2. There is the expectation that the clinical diagnosis, with the aid of “advanced imaging”, is sufficiently accurate to explain a patient’s illness
3. In the era of “fee for service”, there is no third party payment for the performance of an autopsy. Thus, the patient’s family might be asked to pay for a post mortem examination

A corollary to these observations is that the autopsy is not always well done and the findings from an autopsy examination are not promptly relayed to the concerned physicians and family

There are data, which suggest that the rate of autopsies in hospital deaths was reasonably stable from the 1950’s through the early 1970’s[iii].

Reports since the beginning of the 21^st century continue to show that the clinical diagnosis before death is incorrect in a significant number of cases[iv]. In 2015, an abstract from Denver, CO suggested that the cause of death determined on clinical grounds wasn’t completely correct in 70% of autopsied cases and it was markedly different from autopsy findings in as many as 35% of instances[v]

The precursor organization to The Joint Commission (TJC) stopped requiring a specific autopsy rate as a quality indicator in approximately 1972[vi]. The reason for this change in policy is not entirely clear, but there may be several explanations why this change occurred. One might have been the introduction of advanced imaging, beginning with the CT scan (described by Hounsfield in 1972). These techniques provided pictures that initially seemed to obviate the need for anatomic confirmation^[vii]. However, there are multiple examples where the CT and MRI results are “wrong” in that they are not confirmed at surgery or autopsy. Some of these errors may be biases in the interpretation of a scan. A recent report showed an alarming level of missed diagnoses and “over diagnosis” in MRI evaluations[viii]. Results of studies like these should suggest to us that “advanced imaging” does not have all the answers and that some correlation/confirmation of diagnostic findings might be warranted.

Thus, as an NPR item in 2012 suggests, the autopsy would seem to provide information that should be of interest to both clinicians and public health officials[ix].

However, in the era of “Fee for Service”, insurers have taken a stance that they would/should not pay for the performance of an autopsy. One example of this policy was the implementation by Priority Health, a Michigan based Health Insurance provider. At its inception the company declared, “An autopsy is not medically necessary for the welfare of the patient; therefore, autopsies are not a covered benefit”. Interestingly there were NO references cited to support the policy[x].

The Joint Commission may not be likely to “require” a specific rate of autopsies, as they did in the 1970’s. One might postulate that there are too many “business considerations” relating to the requirement of a number/proportion of autopsies. Perhaps the most compelling reason is that there are now competing accrediting organizations that also do not require an autopsy, and if hospitals don’t want to have to pay for autopsies, they might simply switch to another less demanding accreditation organization[xi]. This would lead to a loss of revenue stream to TJC. Unfortunately this posture could potentially allow US hospitals to be subject to less strict oversight (and potentially reduced quality of care) than is currently the case. If the Centers for Medicare and Medicaid Services to require a certain percentage of autopsies be done on hospital deaths this argument may be overcome.

The cost of an autopsy is not inconsequential. If the cost were to be born by the family, without any other payment, cost might be prohibitive. However, the costs of having an autopsy done are, like most costs in medicine, not transparent. There are data that an autopsy would cost less than $5,000[xii]. A 2011 article published in the website Pro-publica suggests that in one survey of 8 states, “an autopsy costs about $1,275” [xiii]. An investment by hospital boards of only one half a million dollars to support quality through the autopsy would pay for between 200-400 autopsies a year. Certainly a hospital (or hospital board) that was deeply invested in care improvement could divert this amount to support the post mortem exam and demonstrate a commitment to quality in their institution.

The medical community could take a lead in quality improvement by asking our politicians to again ask certifying agencies to include some autopsy numbers as part of the “deeming”[xiv] process. If an autopsy standard were incorporated into the health care facility/provider inspection process, then it would again be part of the “checks and balances” along with other means of ensuring diagnostic accuracy. Several medical groups currently active including but not limited to: The American Medical Association, The American College of Physicians; The American College of Pathology and The Society to Improve Diagnosis in Medicine the could work on this initiative. Many other physicians, including George Lundberg, MD, Stephen Geller, MD, and Lee Goldman, MD have all lamented the fall of the autopsy. Now might be an opportune time to re-visit this issue and take political action.

Endnotes/References:

[i] Cabot, R.C.: Diagnostic Pitfalls Identified during a Study of Three Thousand Autopsies; JAMA, 1912, 59, 2295-2298. doi:10.1001/jama.1912.04270130001001: Accessed 6/8/2109

[ii] “The Principles and Practice of Medicine”, which is reputed to be the among the first textbooks of Medicine written and published in North America

[iii] Hoyert DL. The changing profile of autopsied deaths in the United States, 1972– 2007. NCHS data brief, no 67. Hyattsville, MD: National Center for Health Statistics. 2011.: https://www.cdc.gov/nchs/data/databriefs/db67.pdf Accessed 5/18/2019

[iv] Error rates of 15% to 20%in clinical diagnosis have been reported.

[v] Bol, K; Norton, D; Boyer, P; Low, R: Death Certification Inaccuracies and the Validity of Public Health Statistics MMWR; 2015:
https://www.cdc.gov/nchs/events/2015nchs/poster_abstracts.htm accessed 5/6/2019

^[vi] Prior to this, the JCAHO had required approximately 20% of deaths to undergo an autopsy as a quality measure: https://www.medicinenet.com/autopsy/article.htm#why_is_the_autopsy_rate_declining. Accessed 5/9/2019

^[vii] Many physicians and surgeons, however, cite a fairly high percentage of instances in which the CT scan suggested an anatomic process that was not there.

[viii] Herzog, R., Elgort, D.R., Flanders, A.E., Moley, P.J. “Variability in diagnostic error rates of 10 MRI centers performing lumbar spine MRI examinations on the same patient within a 3-week period,” The Spine Journal. Published online Nov. 17, 2016. http://dx.doi.org/10.1016/j.spinee.2016.11.009

[ix] https://www.npr.org/2012/02/05/146355717/fewer-autopsies-mean-crucial-info-goes-to-the-grave Accessed 5/9/2109

[x] The quote came form the newly organized company’s policy manual. The date of origin of the policy was recorded as June 30, 1988 – almost at the company’s inception. There are NO references supporting the assertion: https:www.priorityhealth.com/provider/manual/auths/~/media/documents/medical-policies/91054.pdf. Accessed 4/7/2019

[xi] Currently there are two accrediting organizations in addition to the Joint Commission (the original accrediting organization deemed acceptable by CMS): These are Det Norske Veritas (DNV) Healthcare (Norway) and Healthcare Facilities Accreditation Program (originally an Osteopathic organization)

[xii] Another PBS article on the autopsy cost between $3,000 and $5,000: https://www.pbs.org/wgbh/pages/frontline/post-mortem/things-to-know/autopsy-101.html.

There are also data suggesting in Cook County the autopsy done privately is between $1,000 and $5,000: http://autopsychicago.com/why-choose-acc/frequently-asked-questions/#cost

[xiii] https://www.propublica.org/article/without-autopsies-hospitals-bury-their-mistakes. Accessed 5/9/2019

[xiv] An agency is “deemed” to be an adequate surrogate for CMS inspections, themselves.