Death as Data: Autopsy and the Death Certificate

1
Death as Data Autopsy and the Death Certificate

• Two basic goals
• value of the autopsy
• proper use of the death certificate

2
Essential websites and URLS

• Bruce.case_at_mcgill.ca

National Association of Medical Examiners Death
Certificate Tutorials http//www.thename.org/Caus
eDeath/main.htm
A Canadian angle 1998 article from the
Canadian Medical Association Journal http//www.
cma.ca/cmaj/vol-158/issue-10/1317.htm
3
Outline

• Autopsy history in three periods
• The fall of the autopsy 1960 onward
• Evidence of continuing relevance
• Some attempts to explain the problem
• Effects of falling rates
• An example of the effect
• Death certificate what it is, how it should be
  approached

4
The autopsy in history

• Classical period test authority
• Pre-modern period (17-18C) emphasis on anatomy
• Modern period 19C on
• Rokitansky (gross autopsy)
• Virchow (added the microscope)
• Osler a modern example
• Ultimate recognition as prime goal a contribution
  to medical knowledge

5
Falling Autopsy Rates

• From 50 in the 1960s to
• Much lower than 10 today, despite
• (for example) of three U.S. studies, an incorrect
  diagnosis of malignant tumors was shown in
• 36.5 of cases (1923)
• 41 of cases (1972)
• 44 of cases (1998, Louisiana)

6
(No Transcript)
7
Falling Autopsy Rates RVH 1998
(figures from 1998) DEATHS AUTOPSIES RATE ()
ADULT INPATIENTS 721 92 12.8
EMERGENCY 43 3 7
INFANTS (OVER 500g) 19 12 63
8
JAMA 1998 Louisiana study

1. All autopsies 1986-95
2. Outcome measure discordance in clinical vs.
   autopsy for cancer
3. 1105 cases mean age 48 years (very atypical)
4. 443 neoplasms at autopsy 250 malignant
5. 111 wrong CLINICAL diagnoses of malignancy
   including 57 which caused death

9
So what is wrong? 1. Why do the rates keep
falling in the face of continuing evidence of
error?

1. Increasing reliance on imaging
2. Fear of lawsuits? May explain USA but not
   elsewhere

10
So what is wrong? 2. Changing patterns in
pathology

1. Changing patterns in pathology and pathologists
2. A new but worrying factor regard for autopsy
   practices as violating civil rights (lawsuit in
   UK over pediatric autopsies) reflects a constant
   fight over values over the years coupled with
   some abuses such as Burking

11
So what is wrong? 3. Poor communication

• Suggestions of poor communication between
  pathologists and clinicians
• Wherever a special effort is made to educate
  rates increase, although this may be transitory.
  Rates can reach 100 in some centres!

12
So what is wrong? 4.

• Suggestions of poor communication between
  pathologists and clinicians
• Poor pay, lack of curiosity, lack of professional
  attitude to reporting can lead to vicious
  circle of late reporting
• Clinical mortality rounds seem to result in
  higher rates when pathologists attend

13
Is anything right?

1. In academic centres cases with unknown cause
   still invoke requests for autopsy
2. This means that almost every increasingly rare
   autopsy has become more interesting both for
   the pathologist, for teaching, and for
   publication, BUT...
3. This applies only to academic centres

14
Effects of falling rates

1. Similar to effects of bad death certificate
   reporting
2. National health statistics wrong
3. Lack of Quality Control
4. Problems for analytical epidemiology (garbage in,
   garbage out)

15
Autopsy trends and their effect on disease
ascertainment an example.

1. 1. What is this lesion?
2. 2. How rare is it?
3. 3. Difficulties in Diagnosis
4. 1. Result in UNCERTAINTY (or
   guessing This could be X or possibly Y or)
5. 2. Result in outright error (mainly lung ca)

16
Special Procedures in Pathology Trends for 228
women with mesothelioma 1970-90

• Trends among 142 and 98 female cases diagnosed
  1970-1984 and 1985-1991, respectively.

17
Effect of Autopsy Rate on Reliability and
Accuracy in Two Diagnostic Eras
18
CMAJ ARTICLE

• 1. Improving the accuracy of death
  certification
• Eight case scenarios are presented
• Kathryn A. Myers, MD, EdM
• Donald R.E. Farquhar, MD, SM
• CMAJ 19981581317-23

19
Ontario Death Certificate
20
2. WRITING CAUSE-OF-DEATH STATEMENTS

• An On-Line Tutorial
• http//www.thename.org/CauseDeath/main.htm

21
Why learn this now?

• Often, a physician's first encounter with the
  death certificate occurs upon the physician's
  first patient death when he/she is handed the
  death certificate form and asked to complete it.
• This usually occurs during the first year of
  residency.
• Many, perhaps most, are not told how and
  never learn!

22
The cause-of- death statement contains two parts
Part I

• A)
• Due to, or as a result of
• B)
• Due to, or as a result of
• C)

PART I is designed so that a sequence of
conditions leading to death may be reported
23
The cause-of- death statement contains two parts
Part II

• Part II. OTHER SIGNIFICANT CONDITIONS Conditions
  contributing to death but not resulting in the
  underlying cause of death in Part I

EXAMPLES hypertension, diabetes, chronic
obstructive lung disease, renal diseasediseases
pre-existing or co-existing with the MAIN
UNDERLYING DISEASE but NOT related to it
24
PART I ONE CONDITION per line, starting with
the most recent condition on the top line and
going backward in time
PART I A. Most recent condition (e.g., Cardiac tamponade)
Due to, or as a consequence of B. Next oldest condition (e.g., Ruptured myocardial infarction)
Due to, or as a consequence of C. Oldest (original, initiating) condition (e.g., Atherosclerotic coronary artery)
25
An example
Part I A. Cerebral infarction Due to, or as a consequence of
B. Thrombo-embolism to right internal carotid artery Due to, or as a consequence of
C. Thrombo-embolism from bacterial endocarditis of mitral valve Due to, or as a consequence of
D. Floppy mitral valve syndrome (underlying cause of death-- the specific condition (disease or injury) that started the downhill course of events that led to death.)
26
Variants problems

• Single Line Part I Format (missing data) e.g. no
  autopsy, patient dies at home, known to have
  prostate carcinoma
• uncertainty or presumption use probable
• ALWAYS REPORT CANCER!
• Can cheat on part two to record risk factor
  (smoking, asbestos exposure)

27
Additional Information on the Death Certificate

1. Usually a space to record TIME since onset of
   event
2. Always indicate whether (a) an autopsy has been
   asked for and (b) whether the DC includes autopsy
   information
3. In some places, can record occupation retired
   is NOT an occupation!!!
4. Mandatory reporting violent death, certain
   infections varies with state

28
Multiple cause-of-death coding

1. All data to date are based on a SINGLE cause of
   death but
2. Modern national statistics programs record ALL
   information on the death certificate and can
   derive
3. multiple cause-of-death data

29
http//www.thename.org/CauseDeath/main.htm (This
is the web address for the tutorial on death
certificates)