Osler

1
Oslers Study of the Act of Dying (1900-1904)
An Analysis of the Original Data

• Paul S. Mueller, MD, MPH, FACP
• Associate Professor of Medicine
• Mayo Clinic Rochester

2
Disclosures

• I am a member of the Boston Scientific Patient
  Safety Advisory Board
• I am an associate editor for Journal Watch
• No off-label use of drugs or devices will be
  discussed

3
Objectives

• Explain why Osler did the study
• Explain why certain data were gathered
• Describe the results of the study
• Draw conclusions from the study
• Describe Oslers post-study views of death

4
Sir William Osler12 Jul 1849 - 29 Dec 1919

• Born in Canada
• Widely-adored physician who practiced in Canada,
  the US, and the UK
• Highly influential author, educator, researcher,
  reformer, and civil servant
• Family man

William Osler
5
Science and Immortality1904 Ingersoll Lecture at
Harvard University

• Annual lecture on topic of immortality (still
  given today)
• Osler invited to give lecture multiple times by
  then Harvard President Eliot
• Osler eventually, yet reluctantly, agreed and
  gave lecture 18 May 1904
• Later published as a book

6
Science and ImmortalityExcerpt
I have careful records of about five hundred
death-beds...Ninety suffered bodily pain or
distress of one sort or another, eleven showed
mental apprehension, two positive terror, one
expressed spiritual exaltation, one bitter
remorse. The great majority gave no sign one way
or the other like their birth, their death was a
sleep and a forgetting (p.19).
7
Why re-examine OslersStudy of the Act of
Dying?

• A detailed manuscript describing the study and
  its results was never published, yet the brief
  quote from Science and Immortality continues to
  be cited
• Analysis would shed light on the attitudes toward
  death and the dying experience and during Oslers
  era
• Few similar studies have been done

8
Methods of the current analysis

• Resources
• Data collection cards and a spreadsheet used by
  Osler and stored at McGill University
• JHH administrative records
• Oslers papers

• Methods
• Data abstracted from cards and compared to data
  recorded on Oslers spreadsheet

9
Features of Oslers study

• Done at the Johns Hopkins Hospital (JHH)
  1900-1904
• Observations of 486 dying patients recorded on
  data collection cards
• Data from cards transferred to a spreadsheet
  tallied by Osler

William Osler (1900)
10
Data collection card
Name blocked out
Purpose of the study
11
Name blocked out
Thomas McCrae
12
Purpose of the study

• Printed at bottom of data collection card
• The object of this investigation is to ascertain
  the relative proportion of cases in which (1) the
  death is sudden (2) accompanied by coma or
  unconsciousness (3) by pain, dread, or
  apprehension.

13
Oslers spreadsheet
Tallies written by Osler and cited by him in
Science and Immortality (Ninety suffered bodily
paineleven showed mental apprehension)
14
Results Who completed and signed the data cards?

• 429 (89) cards were signed
• 258 (60) by nurses
• 139 (32) by residents
• 32 (7) unidentified
• None of the cards were signed by Osler

• 149 unique individuals signed the cards
• 97 (65) nurses
• 34 (23) residents
• 18 (12) unidentified

15
Osler and nurse Bessie Boley (1904)
16
Life on the JHH wards1898 Nursing Annual Report,
Mary A. Nutting
The head nurses are responsible for the general
management of the wardsThey are held responsible
for every detail in connection with their wards,
of which they do not and cannot do the actual
work. They are an essential feature of a
properly organized and equipped ward.
M.A. Nutting
17
Results Acquisition of cases

• First case 3 Jan 1900
• Last case 29 Mar 1904 (just 6 weeks before Osler
  gave the Ingersoll Lecture)
• gt1-year gaps between card 400 (dated 18 Feb 1901)
  and card 401 (dated 26 Feb 1902) and between card
  463 (dated 20 Jan 1903) and card 464 (dated 5 Feb
  1904)

18
Results Acquisition of cases
Mortality 6.6 (1449 deaths/21926 patients)
19
Results Gender

• No place on the data card to record gender
  however, gender-specific descriptors (eg, he)
  and names (eg, Mary) used for most patients
• Breakdown
• Male 308 (63)
• Female 169 (35)
• Unknown 9 (2)

20
Results Nationality and religion

• Nationality
• American 360 (78)
• German 44 (9)
• Unknown 18 (4)
• Irish 13 (3)
• Polish 13 (3)
• Other 38 (7)

• Religion
• Protestant 171 (35)
• Catholic 69 (14)
• Jewish 19 (4)
• None 1
• Unknown 226 (47)

21
Results Age at death

• Recorded for 477 (98) patients (440 adults, 47
  children)
• Overall median (range) age at death 39.0
  (0.3-83) years
• Gender female 39.0 (0.8-75), male 40.0 (0.3-83)
• Race black 36.0 (0.3-82), nonblack 40.0 (0.8-83)

22
Results Length of illness before death

• Recorded for 428 (88) patients
• Overall median (range) length of illness before
  death 0.8 (0-81) months
• Gender female 0.8 (0-81), 0.9 (0-48)
• Race black 0.8 (0-26), nonblack 0.8 (0-81)
• Age child 0.3 (0-24), adult 1.0 (0-81)

23
Results Cause of death

• At the start of the study, the data card had no
  place to record diagnosis place added later
  (first 288 dated 6 Apr 1901)
• Overall, a diagnosis was recorded for 188 (39)
  patients

From card 288
24
Results Cause of death

• All (N188)
• Infection 60 (32)
• CV 33 (18)
• Surgical 31 (17)
• Cancer 21 (11)
• Nephritis 16 (9)
• Trauma/burns 13 (7)
• Other 14 (7)

• Children only (N17)
• Infection 9 (53)
• CV 1 (6)
• Surgical 2 (12)
• Cancer 1 (6)
• Burns 4 (24)

25
Results Cause of deathBy gender (adults only)

• Men (N110)
• Infection 37 (34)
• CV 27 (25)
• Surgical 9 (8)
• Cancer 11 (10)
• Nephritis 11 (10)
• Trauma/burns 5 (9)
• Other 10 (9)

• Women (N58)
• Infection 14 (24)
• CV 5 (9)
• Surgical 19 (33)
• Cancer 9 (16)
• Nephritis 5 (9)
• Trauma/burns 3 (5)
• Other 3 (5)

26
Results The act of dyingIf sudden

• Recorded for 472 (97) patients
• Sudden 340 (70)
• Observers specifically wrote gradual next to
  if sudden for 46 patients and quiet next to
  The act of dying for 42

27
Results The act of dyingDid respiration stop
before pulse?

• Recorded for 330 (68) patients
• Respirations before pulse 119 (24)
• Time available for 100 patients median (range) 1
  (0.1-10) minutes
• Pulse before respirations 187 (38)
• Time available for 145 patients median (range) 3
  (0.1-240) minutes
• Same time 24 (5)

28
Results The act of dyingComa or
unconsciousness

• Recorded for 458 (94) patients
• Coma affected 311 (64) patients
• Length of coma recorded for 296 patients
• Median (range) length of coma 3 (0-168) hours
• Alert 147 (30) patients

29
Results The act of dyingNumber of unique
patients experiencing discomforts
41 higher than recorded on Oslers
spreadsheet 79 higher than reported in Science
and Immortality
30
Results The act of dyingNumber of discomforts
experienced by patients
Excludes 24 patients for whom fear column
marked yes 34 higher than reported by
Osler 107 higher than reported by Osler plt0.05
v. spreadsheet
31
Results The act of dyingSpecific discomforts
experienced by patients
Total number of discomforts 215
32
Results Discrepant discomfort data

• Most discrepancies are not subtle
• 76 (16) cards had data (84 discomforts)
  discrepant with Oslers spreadsheet
• Physical discomforts 37 (pain 22, dyspnea 23)
• Mental discomforts 47 (fear 17, anxiety 15,
  irrational 12)

33
Results Discrepant discomfort data

• Physical (N37)
• Pain 22 (for 4 patients, pain was marked as
  yes on Oslers spreadsheet, but not on the
  cards)
• Dyspnea 12
• Convulsions 3

• Mental (N47)
• Fear 17
• Anxiety 15
• Irrational 12
• Mental yes 2
• Depression 1

34
Specific spreadsheet-data card discrepanciesCard
47
35
Specific spreadsheet-data card discrepanciesCard
194
36
Specific spreadsheet-data card discrepanciesCard
314
37
Specific spreadsheet-data card discrepanciesCard
408

For about 10 minutes before death the patient
was in a condition of acute mania. He fought
with the orderlies, bit one and was restrained
with difficulty.
Oslers spreadsheet also marked none for each
38
What conclusions can be drawn from the study
results?

• Nurses played a major role in the study Osler
  had little direct involvement
• Data collected suggest nearly constant bedside
  presence of observers who completed the data
  cards
• Dying patients were young
• Median length of illness before death was brief
  shorter for children

39
What conclusions can be drawn from the study
results?

• Most common cause of death infectious diseases
  causes differed by gender
• A majority experienced coma before death median
  duration of coma brief
• More patients (38) experienced discomforts
  (mostly physical) than reported by Osler (21) in
  Science and Immortality

40
What prompted Osler to do a study of
dying?Personal interest

• 1872 to investigate causes of death among the
  highest objects of the physician
• Entries in his pocket day books
• Plans for lectures and essays (eg, On the Hour
  of Death)
• Numerous volumes on death, dying and related
  topics in his personal library

41
What prompted Osler to do a study of
dying?Popular opinion and fears

• Death viewed as the King of Terror. Why?
• Decline of religious faith
• Reduced death rates, especially due to infectious
  diseases death primarily affects elders with
  chronic diseases
• Rise of science, and hospitals death medicalized
  and moved from home to hospital
• Death becomes foreign to most

42
What prompted Osler to do a study of dying?Osler
openly disputed the popular opinion

• Osler, like many physicians of his era, refuted
  the popular opinion that death was accompanied by
  suffering
• 1888 article death a sleep and forgetting
• 1893 lecture death accompanied with pleasure
  rather than with pain
• Nevertheless, empirical data about death were
  sparse Osler was a naturalist

43
Why were certain data gathered?

• Popular concerns about apparent death, the
  nature of the exact moment of death and about
  discomforts associated with the dying process
• Physician interest in modes of death
• Sudden versus gradual
• Cardiac versus respiratory
• Presence or absence of coma

44
Why were the results reported by Osler inaccurate?

• Possible reasons
• Osler did not carefully examine data cards (he
  did not sign any of them)
• Confusing layout of data cards
• Osler did not complete the spreadsheet
• Osler inaccurately tallied the spreadsheet
• Haste

45
Confusing layout of data cardsCard 41
46
Confusing layout of data cardsCard 164
47
Inaccurate tallying of spreadsheet by Osler
__
___ __ __ Actual figures 24
131 9
1
None of these 24 included in the mental column
48
Possible role of haste

• Short interval between last case (29 Mar 1904)
  and lecture (18 May 1904)
• Hectic time for Osler
• Very busy with consultative practice and
  scholarly activities
• Great Baltimore Fire of February 1904 nearly
  consumed Oslers home
• Observations of others and comments by Osler

49
The Great Baltimore FireFebruary 1904
View from Calvert and Baltimore Streets
50
Observations of others and comments by Osler

• Possible role of haste
• Cushing1 lecture written in a single morning
• Cushing biography notes2 lecture written in
  some haste...misquoted sources
• Osler letter to Miss Jewett2 a rash production

1Ann Med Hist (Summer, 1919) 2Cushing folders at
McGill University
Harvey Cushing
51
McCrae on Science and ImmortalityLetter to
Cushing (8 Jul 1920)
There is one point there in which I know you are
in error namely in saying that the address
Science and Immortality was written in a
morning. Up till that time he had never spent as
much time on an address and had written it
several times.
Thomas McCrae
52
Cushing on Science and ImmortalityThe Life of
William Osler (Oxford, 1925)
With all this pressure upon him the date of the
Ingersoll Lecture was approaching, and though he
had been making notes and giving thought to it
during the preceding months there had been scant
time for the sort of preparation the subjectHe,
indeed, had agonized over it perhaps more than
any of his previous drafts and it was rewritten
and redrafted many times. (p. 638)
53
Parts of Science and Immortality were written
years before the lecture

• Osler traveled to Minneapolis during 1898
• Document in Cushing folders at McGill
• Written chiefly in the train from Minneapolis to
  Boston copied Sept. 24th and 25th.
• Fragment written by W.O. (Script) on 1 West
  Franklin Street notepaper.
• Contains basic elements of Science and
  Immortality (but no mention of the dying study)

54
Oslers reporting of the study resultsProbable
sequence of events

• Osler wrote elements of Science and Immortality
  years before it was given
• Final touches were likely added in haste
• Hectic period for Osler
• Study statistics must have been added just prior
  to the lecture
• Spreadsheet and spreadsheet tally (the statistics
  cited in lecture) were inaccurate
• Observations of colleagues and Osler suggest haste

55
What were Oslers views of death after the study?

• Remained interested in the topic and did not
  waver in his opinion as expressed in Science and
  Immortality
• Added volumes on death and dying to his library
• 1911 letter to Spectator protests painful
  characterization of death in essay (Death)
  written by Nobel Prize (literature) recipient
  Maeterlinck
• 1919 wrote book review of Aspects of Death by F.
  Parkes Weber for Ann Med History

56
Oslers death at age 70 years29 Dec 1919

• Acute complication of chronic bronchitis
• Experienced painful bouts of coughing
• Shunt the whole pharmacopoeia, except opiumWhat
  a comfort it has been!
• Osler calmly accepted death
• The experience has been encouraging--discomfort
  of course, but no actual pain except for the
  worry about leaving dear ones, singularly free of
  mental distress

57
Conclusions

• A substantial number of dying patients in Oslers
  study experienced discomforts
• Osler was prompted to do the study because of
  popular fears and a lack of empirical data
  regarding death
• The historical context of the study is rich
• Lessons for today

58
Acknowledgements

• Osler Library of the History of Medicine at
  McGill University
• Alan M. Chesney Archives at the Johns Hopkins
  University
• The Chest Foundation
• Mayo Foundation

59
Questions?mueller.pauls_at_mayo.edu