A Tale of Three Autopsies

The Autopsy used to be the basis of almost all knowledge in Medicine. Sir William Osler was said to have personally done well over 1,000 autopsies before he became Chief of Medicine at Johns Hopkins in 1889. His experience in the morgue and the knowledge of anatomy and pathology that he gained there were among the bases of his historic textbook[1]. In the 1960s Autopsy Rounds on Thursday afternoons at Cook County Hospital were mandatory. Today, the autopsy is almost a relic. Several groups, including the Society to Improve Diagnosis in Medicine believes that the autopsy should be revived to help with learning. Recently, I have become aware of several instances in which the autopsy changed the diagnosis of cause of death for patients and their survivors.

This will be a post in 2 Parts: The first will tell stories of three patients who died and how the autopsy helped improve knowledge, diagnostic accuracy, and improved reporting of data. The second part will be on how the Autopsy has declined in frequency and how we might restore its use. My opinion is that the Autopsy is a central component of the end of life event. Its use should help our professionals grow, keep our data more accurate, and help families understand their loved one’s illness. It could perhaps help them improve their own health behaviors.

Part: the first – Stories and their implications:

Episode One:

An elderly woman with spinal stenosis and known aortic & mitral valve disease had been seeing her internist for shortness of breath (SOB) of about a year’s duration. In addition to obesity, she had a mildly complicated past history[2].

Her SOB was largely on exertion, but there were occasional otherwise unexplained exacerbations even at rest.

Over her last 2 months, she had reduced most of her normal activity because of the SOB and back pain. She became even more sedentary.

Her physical exam was essentially unrevealing[3].

An echocardiogram was reported to show mild AS and MR with a normal LVEF[4] and no wall motion abnormalities. There was mild LVH (LV wall thickness was 12mm)

Neither her internist nor cardiologist thought that her heart disease warranted intervention.

One morning her husband found her dead in bed.

The medical examiner was ready to rule the cause of death – Cardiac arrest due to complications of Aortic valve disease.

At the urging of a family friend, an autopsy WAS performed. It showed recurrent recent and remote pulmonary emboli and one large clot in the right main pulmonary artery. There were thrombi in the left femoral vein.

Without the autopsy the main cause of death would have gone undiagnosed, the death certificate would have been incorrect, and the vital statistics of her county and state would have missed something that may have warranted a public education intervention.

Episode 2:

A middle-aged man, of the “baby boomer” generation, who had been diagnosed with Bipolar Disorder, became increasingly confused one day. His wife took him to the local hospital where he was found to be obtunded with no localizing physical findings. His past history was remarkable for multiple episodes of mania and severe depression. He was not consistent with his medications. Lab testing showed severe hyponatremia, which was treated aggressively at the local hospital. A lithium level 2 days after admission was twice the lab reference for upper limit of therapeutic. His exam was otherwise unremarkable. He never regained consciousness and was transferred to a referral center. In the ICU, his blood chemistries normalized, but he never recovered consciousness. Spinal fluid analysis initially showed no abnormalities. After two months, a second spinal fluid analysis showed potentially malignant cells. No primary was ever found. After two months, life saving measures were tapered and he died.

An autopsy showed diffuse hepatocellular carcinoma and cerebral edema.

The diagnosis of liver cancer would never have been made without an autopsy and the “official cause of death” would have been incomplete

Unfortunately, no HCV testing was done.

Episode 3:

A middle-aged man with known bicuspid aortic valve disease with mild stenosis was brought to the Emergency Room late one Friday afternoon because of chest pain. Other than the early systolic ejection click and systolic murmur, the exam was unremarkable. There was no evidence of myocardial infarction or myocardial ischemia on ECG, blood testing or a myocardial scan. A CT of the chest was reported as normal. The ER was especially busy that day and the short-staffed ER doctors didn’t look at the CT themselves. The patient was sent home to see a cardiologist for workup in 2 days.

His wife says he died that night next to her in bed.

An autopsy was performed and the pathologist told the family that they should sue because he had a descending aortic dissection.

The autopsy was helpful in teaching the staff who cared for this man, because it emphasized the need to look at diagnostic tests or at least review them with the radiologist or pathologist.

In each of these episodes, the autopsy clearly changed the diagnosis as to the cause of death, making statistics more likely to be representative of “truth”. In each instance the autopsy was of comfort to the family. In each instance, the medical staff could have used the information from the autopsy as a learning opportunity, even encouraging them to re-read the literature (online textbooks such as Merck Manual, Up To Date, or others as well as some “original literature”).

Some families may be hesitant to request an autopsy, because it “might” be disrespectful to the loved one. The autopsy doesn’t hurt the deceased[5]. There are no real strictures against the autopsy in the major religions practiced in the US[6]. In spite of this, many families still plead religious constraints

Part two of this post, describing the demise of the autopsy will follow soon.

Footnotes:
[1] His textbook, “The Principles and Practice of Medicine”, published in 1892 was the first of its kind in the US and helped establish Osler as the premier internist of his day.

[2] She also had back pain as a major complaint – she used a cane to help her ambulate.
Other than mild hypertension controlled on medications, she had no other major past medical/surgical history.
A review of symptoms is otherwise essentially unremarkable with one exception:
Her husband had noted that she had some foot pain on the left which is relieved when he massaged the foot.

[3] Exam showed a Temp of 98 F (oral), a HR of 90/min, a RR of 22, and BP of 134/86 (in both arms). There was 1+ SOA. The lung exam was unremarkable. The heart exam showed no JVD at 30 degree head up, nor HJR. The heart sounds were normal (S2 was widely physiologically split). There are systolic murmurs consistent with AS & MR. There were no diastolic murmurs. Abdominal exam was difficult because of obesity. Pulses were normal with no femoral or carotid bruits. There was bilateral 1-2+ ankle edema.

[4] Left Ventricular Ejection Fraction, which is a measure of how well the myocardium (heart muscle) works

[5] Many times family members will decline an autopsy, because they say, “He has suffered enough.”

[6] I have discussed this with clergy in Christian, Jewish, and Muslim faiths. None of them can find a biblical or other holy book such a constraint/restriction on performance of an autopsy on a diseased.