Epidemiological Study of Sudden and
Unexpected Deaths due to
Arteriosclerotic Heart Disease
By LEWIS KULLERI, M.D., M.P.H.,DR. P.H., ABRAHAM LILIENFELD, M.D., M.P.H.,
AND
RUSSELL FISHER, M.D.
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
frequently to sudden deaths due to ASHD;
(2) studies of sudden deaths as one of the
manifestations of new disease in prospective
or cross-sectional studies of healthy populations10-17; (3) studies of sudden death in
prevalence studies of heart disease18 19; and
(4) reports of sudden death due to a variety
of other diseases.5 20-28
A STUDY of sudden and unexpected nontraumatic deaths in Baltimore residents
aged 20 to 64 years was conducted between
June 15, 1964, and June 14, 1965, in an attempt to determine the frequency and specific
causes of sudden death in a defined population. Differences between racial and socioeconomic groups could be measured by
studying a heterogeneous population. The
relationship of sudden death due to arteriosclerotic heart disease (ASHD), to prior history of heart disease and several other cardiovascular diseases and to prior medical
treatment was determined. Information was
also obtained on activity and place of onset
of events leading to sudden death, length of
hospitalization prior to death, and certain
clinical information at the time of admission
to the hospital.
Previous studies of sudden and unexpected
death have been of four types: (1) reports
by medical examiners in various cities of
deaths referred to their offices;1-9 these studies
were often restricted to autopsied deaths and
Methods
Definition of Sudden Death
There is no generally accepted definition of a
sudden death. In considering a definition, two
factors must be taken into account: (1) time
interval from the onset of the immediate event
until death, and (2) expectation of death prior
to the time of occurrence. The definition of sudden
death used in this study was death due to
natural causes of an individual who was not
restricted to his house, hospital, or other institution, or unable to function in the community
for more than 24 hours prior to death, and for
whom the time interval from onset of the fatal
event until death was less than 24 hours. The
sudden deaths were divided into several categories
according to the time interval from onset until
death, whether the death was witnessed or not,
and by prior history of the disease causing the
death.
From the Department of Chronic Diseases, Johns
Hopkins University School of Hygiene and Public
Health, and from the Office of the Chief Medical
Examiner for the State of Maryland and The Maryland Medical Legal Foundation, Baltimore, Mary-
Sampling
A stratified systematic sample of all natural
deaths in Baltimore residents in the 20 to 64
age group was obtained concurrently from the
Baltimore City Health Department. Eighteen
hundred and fifty-seven deaths representing
49.6% of the total deaths, were included in the
sample. This included all such deaths of persons
less than age 40, all such deaths outside of a
hospital, one quarter of the deaths of persons
aged 40 to 64 in a hospital which were certified
by the Medical Examiner, one third of the deaths
of persons aged 40 to 64 in a hospital which
were not certified by the Medical Examiner,
land.
This study was supported in part by the U. S.
Public Health Service Training Grants HE-5297 and
HE-5082 from the National Heart Institute, General
Research Support Grant GS 1501-FR 5445-04 from
the National Institutes of Health, and by a Research
Career Program Award K6-GM-13,901 from the
National Institute of General Medical Services.
Presented in part at the Conference on the Epidemiology of Cardiovascular Diseases, Chicago, Illinois, January 29 and 30, 1966.
1056
Circulation, Volume XXXIV, December 1966
DEATHS DUE TO ARTERIOSCLEROTIC HEART DISEASE
cc~~~~~~~~~~~l
~
CO
O£
bN
X0
-c M
-
'C
10
CZ
E
ns=g
't
CS
a~~~~~~~~0
6n oco_ Coo o
0~~~~
F-4
1- ,4
oo.
t~~~~~c m-
0
cct
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
rH b
X UQ
-e
rH~---0CO 00 00LO
CC
m
CO
63 O.
X E~~~~~~~~c
0
O
0
O
C) 0CC
m O -N
OO C
O.
C;0
H
a
Wt
XOCZ
0~
E-
(U
0~01m0
CQ0-0
00 ]
im
CD CD
-O CO cO
cO CDl--..
-
r
x-
I
ce
c)a)
.0)
C'
1-
co
C3
-
0)
o
cn* 0)_
X
4.4
c)
00 i
flt::)~ ~ 0 000
.
Q0)
C,O
-4
CO
C,)_
~4- C
0)
¢~~~~c
c
CZ
0
H-
Circulation, Volume XXXIV, December 1966
*
-I-
1057
and all deaths of Baltimore residents outside
of Baltimore City except deaths in state hospitals
(table 1).
After the sample of death certificates had been
obtained, the deaths included in the sample
were classified into groups with a high or low
probability of sudden death based on information
on the death certificate, including the estimated
length of hospitalization, cause of death, and
interval from onset of the disease until death.
In an earlier pilot study we found that this
separation was very sensitive and that practically
no sudden deaths were included in the low
probability group. The low probability deaths,
representing about 25% of the total deaths in
the sample, were then classified as not-sudden
except for a small sample processed with the
high probability group.
The high probability deaths were thoroughly
studied by reviewing all available medical information, including hospital records, medical
examiners' and physicians' reports, and autopsy
protocols, in order to determine (1) whether the
death was possibly sudden or not and (2) the accuracy of the diagnosis as reported on the death
certificate. In all deaths considered to be possibly sudden after reviewing the available information, the next of kin, other relative, or
friend of the deceased was interviewed in order
to find out whether the death was sudden and
to obtain further information about the death. Approximately 92% of the interviews were successfully completed. Of the 1,857 deaths in the
original sample, 589 were ascertained to have
been sudden and unexpected deaths.
Table 1 also contains the estimated number
of sudden deaths in the total population after
adjustments were made for the differences in
sampling fractions of the various subgroups of
total natural deaths.
In order to permit comparisons to be made
with regard to certain characteristics of the sudden deaths, two additional groups were obtained:
a probability sample of the white male Baltimore
population aged 40 to 64, and all ASHD deaths
that were not sudden.
Accuracy of the Diagnosis in Deaths
due to Arteriosclerotic Heart Disease
Since we were able to review hospital records,
autopsy protocols, and physicians' reports and
also to obtain information from nonmedical personnel who had witnessed a death, a more accurate diagnosis of cause of death was sometimes
possible than that available from the death certificate. After review of all available information,
ASHD was determined to be the most likely
cause of death in 489 deaths in Baltimore
City. In 453 (92.6%) ASHD had been originally
1058
CI;
KULLER ET AL.
ascertained to be the underlying cause, and in
36 (7.4%) as either the immediate or contributing
rnct; e,
*0
O~P-
X,::V
-o
cause on the death certificate. In table 2 the
criteria for the ASHD diagnosis are shown, according to the place of death, for the 489
deaths in Baltimore City due to ASHD, prior
to adjustment for sampling. The information re-
t":I"I
r-
C
.l
m0 Ct o C
Cf.
CD,
-d
0
lates only to the clinical information during the
events just prior to death. Thus, in deaths outside
of a hospital the deceased may previously have
had an ECG diagnostic of ASHD, but because
of the rapidity of death no diagnostic information
could be obtained immediately prior to death.
After adjusting for sampling fractions (by multiplying the hospital deaths of persons aged 40 to
64 certified by the Medical Examiner by 4 and
those of persons aged 40 to 64 not so certified
by 3, the inverse of the sampling fractions),
there were 683 sudden deaths due to ASHD;
194 (28.4%) were verified by autopsy, including
16 (72.7%) of 22 at less than 40 years of
age and 177 (26.8%) of the 661 at age 40 to 64.
Results
The results to be described are based on
the estimated total deaths after adjusting for
sampling fractions (as described) unless otherwise noted. The causes of death are those
ascertained in the study and not necessarily
the underlying causes on the death certificates.
0
0
-
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Co
co
0
¢
es
Co
4.
4Q,
.e4
Co
C3
9-
0F
0t
COD
-e
0
0
r.
v
En
o0
e-
,-- mO
C
4I CO
1, b ,
coCCo Co 1'
-
6
zi
c
C-.
;4
-°
C
co
ct
E4
c
z
.C
Number and Percentage of Sudden Deaths
'o co mq
C:
v,
C)~
0
_i
¢.!
C)
CZ!
C)
iFz
r-r
_
C's
_
z
C);
;-4
-Cd
o4
C
H-.1_ O E~~~~4.
ox*
-00
Of the 3,648 deaths, 1,178 or 32.37o were
sudden and unexpected (table 3). Although
the observed frequency of sudden death may
seem high, the definition of sudden death is
conservative, since deaths in a hospital after
24 hours were not classified as sudden. Deaths
in the operating room or cardiac arrests on a
ward were not considered sudden deaths in
this study. From table 3 we note that, in
spite of the differences in the frequency of
deaths by race and sex in different age
groups, the percentage of dea-ths that were
sudden is similar in the three age groups.
Also, higher percentages of deaths were sudden in males and in Negroes.
Arteriosclerotic heart disease accounted for
58.1% of the total sudden deaths. In the 40 to
64 age group, 61.4% and in the 20 to 39 age
group 21.8% of the sudden deaths were due
to ASHD. A much higher percentage of sudden deaths in white males were due to ASHD
Circulation, Volume XXXIV, Decemiber 1966
DEATHS DUE TO ARTERIOSCLEROTIC HEART DISEASE
"i cq cli ce)
-a
4C~
,4
C
co
co
ce
CO
0
0
I
10
C-tt
0
fEH-
1.0
cq in
c,
oo
--
t
eb
cs N
-
co X-
'0
E
Qo
Cr
00
t-
C
10t-: Co5
-C/
t--
c0
co c01
co
r-l0001
c
0
b)
'0
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
00i-000
0m
00
10
C
10)
o
e,
00
oo C
'IC
Co
Cou
c',
Co
4L)'
.2t4) 0
E
o
II
o cn;?
ba
00
cs
v
0
¢
)q
1-
01
00
1t> 0)0
E
00 00 1- Co
-0
0 01
'0$
cn t
X a
-0:
--I
;
1
06
S~~~~4)4)~ ~ ~ ~~C
tXU oo
C's
2)
C'
cl C-uto
C
~
Ciclain Vlm
XXV
~
than in the other three race-sex groups. Even
when the sudden deaths due to hypertensive
heart disease are added to those due to
ASHD, a higher percentage of sudden deaths
were still due to arteriosclerotic-hypertensive
heart disease in white males.
Although ASHD accounted for 61.4% of
the sudden deaths in the 40 to 64 age group,
only 437 (19.4%) of 2,249 not-sudden deaths
were due to ASHD. A higher percentage of
sudden deaths in the less-than-2-hour category were due to ASHD than in either the
unwitnessed or 2-to-24-hour category. This
was true in all four race-sex groups (table 4).
Of the 1,098 deaths due to ASHD at age 40
to 64, 661 (60.2%) were sudden and unexpected deaths, 203 (18.5%) were unwitnessed, 291
(26.5%) were in the witnessed and less-than-2hour category, and 167 (15.2%) witnessed occurred in 2 to 24 hours. A higher percentage
of ASHD deaths in white males were sudden
than in white females, while no sex differences
were noted among Negroes (table 5). Of 29
deaths due to ASHD at age less than 40, 22
(75.9%) were sudden and unexpected.
Death Rates due to Arteriosclerotic
Heart Disease
In order to compare the different race-sex
death rates were computed and are
presented in table 6. White males have the
highest age-specific death rates due to ASHD,
and white females have the lowest. Rates are
higher for white than Negro males, and for
Negro as compared to white females. The
sudden death rate accounts for a large part of
the absolute difference between white males
and females. As we go from the unwitnessed
and less-than-2-hour categories of sudden
death to the 2-to-24-hour witnessed and finally the not-sudden deaths, the differences
in the death rates between white males and
females become closer but never approach
unity. There is a somewhat similar trend in
Negroes but with smaller differences. It
should be noted that the sudden death rate
for witnessed deaths in less than 2 hours in
white males (202/ 100,000) is greater than
the total death rate due to ASHD in white
females (186/100,000). In white females,
groups,
c
4)
)
Dcmer16
k
ce
1059
1060
Clico
C6
Vb
-
1
t
CS
d-
6
Co
U]
Coo>100
_
0
tN
CO
- 1t01 oo 1- 10 co
4'
z16
,
CZ
0
010 O't
00 01
i
O
H
II
01
0
C aI
t COOCo) 0
I
Il0
00 10 I'l It 01i
Crn
IRD 0t
01
z
n:5
z1
-Ie
Co
OC) C
01
:
CO
I-e
0
6z
z4c
C
x.0
co Co
mmi
Ct
S
0
ds t-
t
CZ
C- 1-
000z
0
-
C)
1-
0
z
X
e
Co
'
60 q
r-4P-
C1
6
cS o6 in
r-
r-4
cc o101
i
017-CO
If
C) IV
uz
C)~~~~~~~0
0
co
6
z
COCSt
~17
0
0l:
IC:
O
C0
(0
11
0
0
0-
I
coo 0 - 01 ICof b
-
.0
>'ste1n
0
CCv
C°st
0 's lo
10 co ~ 0
o5
Hi
01
1t¢
cn
._
0
C3
10 r- c0'It
17CoI
r cC
Co
0<t
ct
(A
CO i
..~~
oII6
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Iz
z1
,: t: "4 1,
I _-I
0 0 co Ci
oo o c- 0I0
1l0
02
6
-
CCo) 1O
000 Co000i
0e
0
>
Y~0
Ic
C0
r01C
~
cq cq cli cq cq
-0)
0
i0
)
1C
CQ Co 00
Co
0
,J
1
IZVi
°
°°iN
'
0z
I-I
0
1.O
c01
0
Ci IlD,.
00
0100 1> 10 co
00 CO Co-
.0
00
O
C;tO
tls
C,cI0-0
ol1~
0z:
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
gC~ i
KULLER ET AL.
I)C X*x
-
cq
lo'1I
H;E0
,;
Circulation,. Volume XXXIV. December 1966
o
DEATHS DUE TO ARTERIOSCLEROTIC HEART DISEASE
.-
O
M1
C'sC
r-
.0
CC
0)
10
co
m2
m
0
m0
-4
_
-,
cl
C]
0
COD
6
t1c]
c] 00 1- u0
cli
.-
o 0 b
CCd
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
6
.
.
CO
9O.-
cs o
0 N
C
c
CO
0
-1-
c0
.0
cla
O
m6 "4
1.
C
co
cli
therefore, both the percentage of ASHD
deaths that were sudden and the sudden
death rate are lowest; while the highest
rate, the largest percentage of sudden deaths,
and the largest proportion of all ASHD deaths
that are sudden, especially within 2 hours,
are reported in white males.
To better appreciate the magnitude of the
unwitnessed and less-than-2-hour categories
of ASHD deaths in white males of age 40 to
64, it should be noted that the death rate for
these two categories combined is equal to the
total mortality due to neoplastic diseases for
the same age group.
Place and Activity at Onset
The onset of the series of events leading
to the fatality was defined as the time when
the individual was required to change his
activities. For 74.4% of the sudden deaths due
.CZ
En
cc
o
ux
c]
02
't
in
1CO
cliI-t~
I'l CO
CoC
00
000
:Z2
0
0
r-
c:
c
Cr'
,4
CCo
4)
0
-
-
6
-CC
-
0t
-
i
C]
Es
from work. Even if
were
C'
HQ)
X4- .=
4-
CO
'-4
c]
-
.,
C
C)
u0
V c 00 m
o oo cD
c]
1- r- Itv c]i
.0
o SN
o
b
'et
0
EH
6
z
Circulation,
Vulume
°
0 .0 02"cl
o-
XXXIV, December 1966
to ASHD in the age group 40 to 64, the onset
was at home. For 11.2% of white males the
onset occurred at work, and for 8.1% it occurred en route to work or returning home
we
adjust these
percen-
tages to include only the white males who
cc
~0
*0-
1061
Cc6
bO
working
on
the last regular working day
prior to death, there does not appear to be an
association between being at work and the
frequency of sudden deaths, since in only
15.3% of men employed at the time of their
deaths did the onset of the fatal event occur
while they were at work, as compared to an
estimated 25% to 30% of time normally spent at
work.
We also did not find any association between a specific activity and sudden ASHD
death, in either the 20 to 39 or the 40 to 64
age groups. Also, the distribution of the activities at onset were similar for those who
died suddenly from ASHD and from other
causes.
C's
Place of Death
Only 51.8% of the ASHD sudden and 17.2%
of the ASHD not-sudden deaths in Baltimore
City were certified by the Medical Examiner.
The place of death was defined as the location of death stated on the death certificate.
However, an individual may die outside of a
KULLER ET AL.
1062
Table 7
Numbers and Percentages of Arteriosclerotic Heart Disease Deaths in Baltimore City at Age 40-64 Distributed by Place of Death and Length of Hospitalization
Cumulative
Noncumulative
Place of death and length of hospitalization
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Died outside of hospital
Sudden
Not sudden
Total
Hospital deaths
Dead on arrival
Sudden
Not sudden
Total
Hospitalized less than 2 hours
Sudden
Not sudden
Total
Hospitalized 2-24 hours
Sudden
Not sudden
Total
Hospitalized more than 24 hours (not sudden)
Admitted with non-ASHD diagnosis
Admitted with diagnosis of ASHD
Grand total
hospital and be brought to the emergency
room only to be pronounced dead. In table 7
the distribution of all ASHD deaths in Baltimore City in the 40 to 64 age group by place
of death and length of hospitalization is reported. Of the total 1,030 persons dying of
ASHD in Baltimore City, 212 (20.6%) died
outside of a hospital, 476 (46.2%) were dead
on arrival, 57 (5.5%) survived less than 2
hours in the hospital, and 159 (15.4%) survived longer than 24 hours. Sixty-nine (43.4%)
of the 159 that survived longer than 24 hours
had been admitted to the hospital with another disease and had subsequently had a
new coronary event while in the hospital. An
estimated 45 (65%) of these 69 patients had
undergone a surgical procedure while in the
hospital. These deaths were often unexpected
and frequently have been considered sudden
in other studies.
Of the 57 who survived less than 2 hours in
the hospital, 33 (57.9%) had no detectable
systolic or diastolic blood pressure on admission to the hospital, as well as 18 (20.7%)
of the 87 who survived from 2 to 24 hours
No.
7%o
138
74
212
65.0
35.0
100.0
376
100
79.0
21.0
476
% of
total
No.
20.6
212
20.6
100.0
46.2
688
66.8
45
12
57
78.9
21.1
100.0
5.5
745
72.3
60
30
90
69
66.7
33.3
100.0
100.0
8.7
6.7
835
904
81.1
87.8
126
1030
100.0
12.2
1030
100.0
100.0
and only three (1.5%) of the 198 who survived longer than 24 hours.
Individuals contemplating the development
and utilization of procedures to reduce mortality should note that only 12.2% of deaths occur after the first 24 hours of hospitalization
for those admitted with the diagnosis of
ASHD and that 66.8% are either dead on arrival or die outside of a hospital.
Any therapeutic methods which are effective only after an individual reaches a hospital will have little effect on the total ASHD
mortality unless several other factors are
changed as well: namely, (1) better diagnostic methods so that patients with minimal
symptoms can be brought to a hospital earlier,
and (2) an effective system in the hospital
to handle the emergencies during the first 24
hours of hospitalization.
The Suddenness of Death
In the preceding discussion we determined
suddenness of death principally in terms of
the time interval from the onset of the fatality
Circulation, Volume XXXIV, Dc ember 1966
DEATHS DUE TO ARTERIOSCLEROTIC HEART DISEASE
1-
Co
Iq
I
Co
't
cl
r
t
Co
1063
I
0
O
0
CC
oo~
00
ooI
6
O;,
I O II
CO
co
co
C-I
0
c
0
Cc
6-
1- o
o0
Co
't
C\C0
00
C00
O
o °
'tCD
Co01Co
Co1
I
00
0
CA Co
a
C)
o, n
0.
6
ro
CO
OO
Cq Cq
CA
00 co
co Nt
01
Co
C)0
0
0
Q)
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
CZ
"0
CC
CC
01)
CC
o Ci
Co
4 C,5
cs
oo
cO
00
r-
t m
6 ci
t
co
"0CZ
cl
r--
o 4t
o -)
CC
cC
C)
04
CO
S
CoK
co 10
z
CD
't CID
"q
1
01
I~-0
CC
0C)
._
r.
>4)
Co
t
co
r-
10
c63o6
10 00
c6
4~6i
b
m
0co
o6 i6
Ct1010
0
CZ X
Co00
C)
0-0
rQ
"0
,4
Co
Co
_co
co
O1Co
CO
1O
co
0CC
0
t
C)
CO
*
ooc01-
t0O
C
0
CO _
co00
CD
P
co
0
4
Co
..
C )
0
bo
vCcl
,
00
0t
0o
u
u
0ou
"
CC
0
"0 0
X C)
~C)
cJ
0)1
En
Circulation, Volume
XXX1IV,
December 1966
-
Z
z
KULLER ET AL.
1064
until death. We shall now consider several
other ways of defining suddenness of death.
Past History of Heart Disease
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
A past history of heart disease was reported
for one half of all the cases of ASHD sudden
deaths between the ages of 40 and 64 (table
8), with little variation in the percentages
among the three categories of sudden death.
The history was more often positive in males
than in females, but no racial differences were
noted. A higher percentage of the deceased
in the not-sudden death group had a history
of heart disease than in the sudden-death
group. However, the differences are smaller
than anticipated, considering that those in the
not-sudden death group were supposedly incapacitated for at least 24 hours prior to
death. Also, 76 (11.4%) of the 666 individuals in the sudden-ASHD-death group and
95 (24.9%) in the 381 not-sudden-ASHDdeath group had a history of diabetes, and
185 (27.8%) in the sudden and 136 (35.7%)
in the not-sudden-death groups had a history of hypertension (table 8).
Finally, approximately two thirds of those
in the ASHD-sudden-death group had a history of at least one of the following four diseases: heart disease, diabetes, cerebrovascular disease, and hypertension (table 8). Thus
when we measure sudden death by two parameters, time from the onset of the fatal
and a past history of several diseases, a
somewhat different picture emerges than
when sudden deaths are determined only in
terms of the time interval.
event
Recent Medical Care
We shall now consider another parameter,
medical care. Approximately 24% of
the individuals in the ASHD-sudden-death
group had been seen by a physician within 7
days prior to the onset of the fatal event (table 9). There was little difference in the four
race-sex groups, although the percentage was
somewhat lower for Negro females. Thirtytwo per cent of the persons dying suddenly
with a history of heart disease and 17% of
those without such a history had seen a
physician within a week prior to death. In
contrast, only 7.5% of the probability sample
of white males had seen a physician within
a week prior to interview.
We previously noted that 60% of ASHD
deaths were sudden when only the time interval is considered. When the history of
heart disease and a recent visit to a physician
(that is, within 1 month prior to death) are
added to the criteria, only about 17% of
ASHD deaths are in the category of deaths
within 2 hours or unwitnessed in which the
deceased had not seen a physician within a
month and did not have a history of heart
disease. Finally, there were only 90 deaths
recent
Table 9
Distribution of Sudden Deaths due to Arteriosclerotic Heart Disease by Length of Time prior to Death
That Victim Had Been Last Seen by a Physician, and by Race and Sex (40-64 Years)
Race
&
sex
WM
WF
NM
NF
Less than
24 hr
No.
%
44
7
11.2
7.9
Length of time prior to death when last seen by a physician
1 Day to
7 Days to
7 days
6 mo
Over 6 mo
No.
%
No.
No.
%
%
54
18
7
7
86
Unknown*
No.
%
Total
13.8
20.2
6.2
9.9
12.9
156
39.8
55
14.1
83
21.1
392
21
23.5
28
31.4
15
16.8
89
22
19.3
46
40.3
12
10.5
27
23.7
114
17
23.9
7
9.9
40
56.3
71
Total
73
11.0
240
36.0
102
15.4
165
24.7
666
Probability sample
WM
7
3.3
9
4.2
87
40.8
83
39.0
27
12.7
213
*In a large number of the sudden deaths the informant did not know how long prior to the onset of events leading to death it had been since the victim had last seen a physician. In most cases
they stated that it was a "long time." Most of these deaths were also certified by the Medical Examiner, and
his office also failed to determine that there had been a recent visit to a physician. We have, therefore, assumed
that when the time was unknown, the deceased had not seen a physician within the month prior to death. We
may, therefore, have underestimated the frequency of recent visits to a physician.
Circulation, Volume XXXIV, December 1966
1065
DEATHS DUE TO ARTERIOSCLEROTIC HEART DISEASE
(8.2% of all the ASHD deaths) that had occurred within 2 hours or were unwitnessed,
when the deceased did not have a history of
cardiovascular or other vascular disease or
diabetes, had not seen a physician within 1
month prior to death, had not been hospitalized in the year prior to death, and was able
to work. These deaths may represent the sudden and unexpected cardiac deaths in apparently healthy individuals (table 10).
Estimation of the Case-Fatality Rate
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
No reliable figures are available on either
the incidence or the prevalence of coronary
artery disease among Baltimore residents.
Several of the prospective studies of coronary
artery disease have estimated that the incidence of coronary artery disease in middleaged white males is about 1% per year,'0' 29, 30
excluding angina pectoris and abnormal
exercise tests. Using this percentage, we estimated that there were 500 new cases of
coronary artery disease among the 51,000
white male Baltimore residents 50 to 64 years
of age during the period of the study. There
were 311 sudden deaths due to ASHD in
white males in this age group. Fifty-three per
cent of the 311 males had a past history of
heart disease as reported by the informants.
However, we had noted that only 80% of those
with histories of heart disease reported by
physicians were also reported by the informants. After adjusting for this underreporting, we estimated that 112 (36%) of the 311
sudden deaths occurred in males who did
not have a history of clinical coronary artery
disease. Therefore, 112 (22%) sudden deaths
occurred among the estimated 500 new cases
of coronary artery disease. After making a
similar adjustment, it was found that 43
(23.5%) of the 183 ASHD deaths that were
not sudden were not associated with a history
of heart disease. About 9% of the deaths in the
500 new cases are therefore not sudden. The
estimated total case-fatality rate is thus found
to be 31%, and 74% of these deaths are sudden.
In table 11 we have compared our estimates
with the results of several prospective studies
of ASHD. In spite of the differences in the
definitions of sudden death in several of the
studies, the percentage of new coronary
events, excluding angina pectoris and abnormal exercise tests, that are manifest only as
sudden deaths, is between 20% and 25% in
practically every study.
Several other studies have also estimated
that a similar high percentage of all ASHD
deaths occur within the first day or so after
the onset of a new coronary event.10' 14, 15, 19
Table 10
Distribution by Race and Sex of Arteriosclerotic Heart Disease Sudden Deaths at Age
40-64 in Which Death Was Either Unwitnessed or Had Occurred in Less Than 2 Hours
after Onset of Events and in Which the Deceased Did Not Have a History of Diabetes,
Heart Disease, Hypertension or Stroke, Had not Seen a Physician Within the Month
prior to Death, Had Not Been Hospitalized in the Year prior to Death, and Was Able
to Work
ASHD deaths
Sudden deaths in above category
Race,
sex
Total
WM
WF
NM
NF
610
Total
175
189
124
1098
Sudden
deaths
392
89
114
71
666
Unwitnessed
Witnessed
< 2 hrs
Total
% of
total
deaths
24
0
7
10
41
23
12
5
9
49
47
12
12
19
90
7.7
6.9
6.3
15.3
8.2
Table includes information obtained from
hospitals, as well as from the interview.
Circulation, Volume XXXIV, December 1966
% of
total
sudden
deaths
12.0
13.5
10.5
26.8
13.5
physicians' reports, Medical Examiner, and
KULLER ET AL.
1066
Table 11
Summary of Results of Studies on the Frequency
of Sudden Death as the Initial Mani ifestation of
Coronary Artery Disease in Previou sly Healthy
Individuals
Study*
New cases of
coronary artery
disease
deaths
Sudden
No.
%
1,331
135
336
26
25.2
19.0
74
229
36
332
252
18
45
24.0
20.0
7
19.0
DuPont Company15
North Dakota'7
Middlesex County,
Connecticut'2
Framingham3'
Albany, N. Y."1
H.I.P.16
Medical practitioners14
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Air Force pilots'3
Civil Service
116
100
employees29
Present study
500
3,105
Total
*All studies exclude
abnormal exercise test.
cases
of angina
25.6
85
76 30.0
(First week)
14.6
17
20
112
742
20.0
22.0
23.9
pectoris
and
Discussion
At present it is rather obvious thiat methods
of either primary prevention of a fir^st coronary
event, or secondary prevention oif a subsequent event, are much more likelyz to have a
demonstrable effect on coronary d isease mortality than therapeutic methods tlhat depend
on the individual surviving at leasst 24 hours
in the hospital.
Routine vital statistics reports inidicate that
approximately 80% of ASHD deatlhs occur in
a hospital, an estimate based on i the statements on the death certificates. TI iis suggests
that improved hospital treatmen t will significantly reduce total cardiovas cular mortality. However, the more intensiv e review of
the series of events surrounding ; death, as
was done in this study, clearly inclicates that
in a large proportion of such dealths the patients were dead on arrival (58.2
d
within 24 hours after admissio] n (18.0%).
Therefore, the potential benefits of hospital
treatment of cardiovascular diseasoe, in terms
of influencing total cardiovascular mortality,
may well be quite limited.
With respect to the possibility of the reduction of the ASHD mortality in the 40 to
64 age group, several important facts reported
in this study should be reemphasized: (1)
60% of deaths attributed to ASHD were sudden; (2) only 18.9% of the individuals who
died from ASHD in Baltimore City survived
longer than 24 hours in a hospital, and one
third of these had been previously admitted
with another diagnosis; (3) of those who
died suddenly from ASHD, 50% had a history
of heart disease, 11.4% of diabetes, 27.8% of
hypertension, and (4) approximately 24%
had seen a physician within a week prior to
death. However, only 90 (8.2%) of the 1,098
total ASHD deaths had either occurred within 2 hours or were unwitnessed among the
persons who did not have a history of diabetes, heart disease, hypertension or stroke,
and also had not seen a physician within a
month prior to death, or had not been hospitalized in the year prior to death and had
been also able to work (table 10).
Only primary prevention will be effective
in reducing the mortality in the 90 (13.5%)
of the 666 ASHD sudden deaths of persons
without either a history of ASHD or a recent
visit to a physician. For the remaining 235
sudden deaths without the deceased's having
a history of ASHD but after a fairly recent
visit to a physician and for the 339 with a
history of ASHD, either primary or secondary
prevention may be effective.
A reduction of the incidence of new coronary artery disease in white males aged 50
to 64 by 20% would have reduced the total
mortality also by 20%. A new treatment that
is effective after the first 24 hours of hospitalization would have to be 100% effective to
produce a similar reduction in mortality. We
have noted, however, that almost one quarter of the deceased had seen a physician
within a week prior to death and that rela-
tively few had not had medical care within a
period of about 6 months prior to death. The
introduction of improved diagnostic methods
so that more of these men would be brought
to the hospital prior to death, plus improved
therapeutic methods, may have a major effect
on mortality rate.
Circulation, Volume XXXIV, December 1966
DEATHS DUE TO ARTERIOSCLEROTIC HEART DISEASE
Summary
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
A study of sudden unexpected nontraumatic deaths was begun on June 1, 1964. A
sample of all nontraumatic deaths in Baltimore residents between the ages of 20 and
64 from June 15, 1964, to June 14, 1965, was
obtained. The deaths were then studied by
reviewing all available medical information
in order to determine: (1) whether the death
was possibly sudden or not and (2) the accuracy of the diagnosis reported on the
death certificate. The next of kin or other relative or friend of each deceased person who
died suddenly was then interviewed.
For comparison, information was obtained
on (1) a probability sample of the Baltimore
population, and (2) deaths due to arteriosclerotic heart disease (ASHD deaths) that
were found to be "not-sudden."
There were 1,857 deaths in the original
sample, of which 589 were sudden according
to the definition of sudden death. After adjustment for sampling, it was estimated that
1,178 (32%) of the total 3,648 deaths in Baltimore were sudden. Arteriosclerotic heart
disease (ASHD) accounted for 58% and the
cardiovascular group together for 69% of the
sudden deaths.
Sixty per cent of all ASHD deaths were
sudden. Of the 1,030 ASHD deaths in Baltimore City between the ages of 40 and 64,
20.6% occurred outside of a hospital and
46.2% represented deaths on arrival at a hospital. Only 18.9% of all ASHD deaths occurred
after the first 24 hours of hospitalization.
By use of data provided in several crosssectional and prospective studies, it was estimated that 22% of new coronary events were
sudden deaths and that the case-fatality rate
was 31%.
In approximately half of the ASHD sudden
deaths the deceased had a history of heart
disease prior to death and in 24% the deceased had seen a physician within the week
prior to death. Unfortunately we were not
able to determine the reasons for these visits.
In considering the implications of these findings with regard to the prevention of ASHD
deaths, it would appear that prevention of
Circulation, Volume XXXIV, December 1966
1067
only a comparatively small percentage (8.2%)
of ASHD deaths is completely dependent on
primary prevention. For the remaining ASHD
deaths a combination of both primary and
secondary prevention may be effective. Because of the rapidity of death and the high
frequency of these deaths either occurring
outside of a hospital or being called deaths
on arrival, hospital treatment may well have
little effect on reducing the ASHD mortality,
while, on the other hand, the combination of
better and earlier diagnosis and intensive
treatment in a hospital could conceivably reduce the mortality.
Acknowledgment
The study could not have been conducted without
the support of physicians, all of the hospitals in Baltimore City, and the Baltimore City Health Department. Part of the computations were done in
the computing center of the Johns Hopkins Medical
Institutions, which is supported by Research Grant
FR-00004 from the National Institutes of Health,
U. S. Public Health Service, and by educational contributions from the International Business Machines
Corporation.
References
1. ADELSON, L., AND HOFFMAN, W.: Sudden death
from coronary disease: Related to a lethal
mechanism arising independently of vascular
occlusion or myocardial damage. JAMA 176:
129, 1961.
2. HELPERN, M., AND RABSON, S. M.: Sudden and
unexpected natural death: General considerations and statistics. New York J Med 45:
1197, 1945.
3. JANES, L. R.: Common causes of unexpected
deaths. In Legal Medicine, vol. 1, edited by
R. B. H. Gradwohl. St. Louis, C. V. Mosby
Co., 1954, p. 157.
4. LEW, E.: Some implications of mortality statistics
relating to coronary artery diseases. J Chronic
Dis 6: 192, 1957.
5. SIMPSON, K.: The Investigation of Sudden
Death: Modem Trends in Forensic Pathology.
St. Louis, C. V. Mosby Co., 1953.
6. SPAIN, D., AND BRADESS, V.: Relationship of sex,
age and physical activity to sudden death
from coronary occlusion. In Work and the
Heart, edited by F. F. Rosenbaum and E. L.
Belknap. New York, Hoebner & Co., 1959, p.
283.
7. SPAIN, D., BRADESS, V., AND MOHR, C.: Coronary
atherosclerosis as a cause of unexpected and
unexplained death. JAMA 174: 384, 1960.
1068
8. SUGAI, M.: Pathological study of sudden and
unexpected death, especially on the cardiac
deaths autopsied by the medical examiners in
Tokyo. Acta Path Jap 9: 723, 1959.
9. WEINBERG, S. B., AND HELPERN, M.: Circum-
10.
11.
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
12.
13.
14.
15.
16.
17.
18.
stances related to sudden and unexpected
death in coronary heart disease. In Work and
the Heart: Section 3, Pathology, edited by
F. F. Rosenbaum and E. L. Belknap. New
York, Hoebner & Co., 1959, p. 288.
DAWBER, T. R., KANNEL, W. B., AND BARRY,
P.: Immediate mortality in coronary heart
disease: The Framingham Study. Memorias IV
Congreso Mundial de Cardiologia, IV B, Mexico City, 1963, p. 176.
DOYLE, J. T., HESLIN, A. S., HILLEBOE, H. E.,
AND FORMEL, P. F.: Early diagnosis of ischemic heart disease. New Eng J Med 261:
1096, 1959.
EISENBERG, H., FELTNER, W. R., PAYNE, G. H.,
AND HADDAD, C. A.: Epidemiology of coronary
heart disease in Middlesex County, Connecticut: Preliminary report on methodology and
the incidence of primary myocardial infarction. J Chronic Dis 14: 221, 1961.
MATHEWSON, F. A. L., BRECETON, C. C., KELTIE, W. A., AND PAUL, G. I.: University of
Manitoba follow-up study: Prospective investigation of cardiovascular disease. Canad Med
Ass J 92: 947, 1965.
MoRRIs, J. N., HEADY, M. A., AND BARLEY,
R. G.: Coronary heart disease in medical
practitioners. Brit Med J 1: 503, 1952.
PELL, S., AND D'ALONZO, C. A.: Iinmmediate
mortality and five year survival of employed
men with a first myocardial infarction. New
Eng J Med 270: 915, 1964.
SHAPIRO, S., WEINBLATT, E., FRANK, C. WV.,
AND SAGER, R. V.: The H.I.P. study of incidence and prognosis of coronary heart disease. J Chronic Dis 18: 527, 1965.
ZUKEL, W. L., ET AL.: Short-term community
study of the epidemiology of coronary heart
disease. Amer J Public Health 49: 1630,
1959.
B16RK, G.: Course and prognosis in some cardiac
diseases. J Chronic Dis 15: 9, 1962.
KULLER ET AL.
19. SPIEKERMAN, R. E., BRANDENBURG, J. T.,
ACHOR, R. W. P., AND EDWARDS, J. E.: Spectrum of coronary heart disease in a community
of 30,000: Clinicopatbologic study. Circulation
25: 57, 1962.
20. DURLACHER, S. H., MEIER, R. J., FISHER, R. S.,
AND LovITT, W. V.: Sudden death due to
pulmonary fat embolism in chronic alcoholics
with fatty livers. J Forensic Sci 4: 2, 1959.
21. HIRST, A. E., JOHNS, V. J., AND KIME, W. S.,
JR.: Dissecting aneurysm of the aorta: Review
of 505 cases. Medicine 37: 217, 1958.
22. MAXWELL, J.: Unexpected deaths in asthma.
Dis Chest 27: 208, 1955.
23. MOTE, C. D., AND CARR, J.: Dissecting aneurysm
of the aorta. Amer Heart J 24: 69, 1942.
24. NIELSEN, N. C.: Dissecting aneurysm of the
aorta. Acta Med Scand 170: 117, 1961.
25. PENTON, G. B., MILLER, H., AND LEVINE, S. A.:
Some clinical features of complete heart block.
Circulation 13: 801, 1956.
26. POLISAR, I., BURBANK, B., LEVITT, L., KATZ,
H., AND MARRIONE, T.: Bilateral midline fixation of cricoarytenoid joints as a serious medical emergency. JAMA 172: 901, 1960.
27. RATNOFF, 0. D., AND BRECKENRIDGE, R. T.:
Pulmonary embolism and unexpected death
in supposedly normal persons. New Eng J
Med 270: 298, 1964.
28. SECHER-HANSEN, E.: Subarachnoid hemorrhage
and sudden unexpected death. Acta Neurol
Scand 40: 115, 1964.
29. CHAPMAN, J. MI., AND MASSEY, F. J.: Interrelationship of serum cholesterol, hypertension,
body weight and risk of coronary disease:
Results of the first ten years' follow up in the
Los Angeles Heart Study. J Chronic Dis 17:
933, 1964.
30. PELL, S., AND D'ALONZO, A. C.: Acute myocardial infarction in a large industrial population. JAMA 185: 831, 1963.
31. KANNEL, W. B., KAGAN, A., DAWBER, T. R., AND
REVOTSKIE, N.: Epidemiology of coronary
heart disease: Implications for the practicing
physician. Geriatrics 17: 675, 1962.
Circulation, Volume XXXIV, December 1966
Epidemiological Study of Sudden and Unexpected Deaths due to Arteriosclerotic
Heart Disease
LEWIS KULLER, ABRAHAM LILIENFELD and RUSSELL FISHER
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2017
Circulation. 1966;34:1056-1068
doi: 10.1161/01.CIR.34.6.1056
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1966 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/content/34/6/1056
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles
originally published in Circulation can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column of
the Web page under Services. Further information about this process is available in the Permissions
and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/