Age and Ageing 2009; 38: 515520 C The Author 2009.

09. Published by Oxford University Press on behalf of the British Geriatrics Society.
All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org doi: 10.1093/ageing/afp119
Published electronically 15 July 2009
Age, invasive ventilatory support and outcomes
in elderly patients admitted to intensive care
units
JOSE MARCELO FARFEL
1
, SUELENE AIRES FRANCA
2
, MARIA DO CARMO SITTA
1
, WILSON JACOB FILHO
1
,
CARLOS ROBERTO RIBEIRO CARVALHO
2
1
Geriatrics Division, University of Sao Paulo Medical School, Sao Paulo, Brazil
2
Pulmonary Division, University of Sao Paulo Medical School, Sao Paulo, Brazil
Address correspondence to: J. M. Farfel. Av. Dr. Arnaldo 455, Room 1353, 01246-903, Sao Paulo, Brazil. Tel: (+55) 11 3061
8249; Fax: (+55) 11 3082 4973. Email: farfel@usp.br
Abstract
Background: although advancing age is associated with worse outcomes on mechanically ventilated elderly patients admitted
to intensive care units (ICU), this relation has not been extensively investigated on patients not requiring invasive ventilatory
support.
Objective: to determine the relationship between age and in-hospital mortality of elderly patients, admitted to ICU, requiring
and not requiring invasive ventilatory support.
Design: prospective observational cohort study conducted over a period of 11 months.
Setting: medical-surgical ICU at a Brazilian university hospital.
Subjects: a total of 840 patients aged 55 years and older were admitted to ICU.
Methods: in-hospital death rates for patients requiring and not requiring invasive ventilatory support were compared across
three successive age intervals (5564; 6574 and 75 or more years), adjusting for severity of illness using the Acute Physiologic
Score.
Results: age was strongly correlated with mortality among the invasively ventilated subgroup of patients and the multivariate
adjusted odds ratios increased progressively with every age increment (OR = 1.60, 95% CI = 1.012.54 for 6574 years old
and OR = 2.68, 95% CI = 1.584.56 for 75 years). For the patients not submitted to invasive ventilatory support, age was
not independently associated with in-hospital mortality (OR =2.28, 95% CI =0.995.25 for 6574 years old and OR =1.95,
95% CI = 0.824.62 for 75 years old).
Conclusions: the combination of age and invasive mechanical ventilation is strongly associated with in-hospital mortality.
Age should not be considered as a factor related to in-hospital mortality of elderly patients not requiring invasive ventilatory
support in ICU.
Keywords: age, intensive care unit, prognosis, invasive ventilatory support, elderly
Introduction
The ageing process results in a higher rate of hospitalisation
and admission to intensive care units (ICU) [13]. Intensive
care is a costly intervention and, although the number of
ICU beds increases in Brazil as it does in other countries, the
offer does not meet the rising demand. Additional research is
needed to optimally determine the prognostic factors related
to ICU outcomes and hence making the best selection of
patients. This is especially crucial in developing countries
where nancial resources allocated to healthcare are limited.
Although recent studies have investigated the factors
related to outcomes of elderly patients in the ICU, conicting
results were observed, especially regarding the role played by
age as a factor inuencing mortality [414]. It is commonly
acknowledged that elderly patients are more likely to die
in the ICU than younger ones. Nonetheless, it is yet to be
determined whether mortality of the elderly is independently
determined by age or by other factors. When admitted to an
ICU, elderly patients suffer from a relatively higher degree of
physiological impairment and from certain co-morbid con-
ditions that increase the risk of death [59].
515

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
1
,

2
0
1
0
a
g
e
i
n
g
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

J. M. Farfel et al.
Recent studies of patients requiring invasive mechanical
ventilationshowthat age is independently andpositively asso-
ciated with mortality of elderly patients admitted to ICUs but
have not provided precise quantitative descriptions on how
this association varies along with age [1521].
Earlier literature excluded non-ventilated patients from
the studies or included them together with mechanically ven-
tilated patients in mixed cohort populations. As patients not
submitted to invasive ventilatory support have not been
studied previously as an isolated subgroup, the relation-
ship between age and mortality for this population remains
unknown.
The aim of the current study is to examine prospec-
tively the relationship between age and in-hospital mortality
of elderly patients admitted to ICUs and to investigate the
degree to which this relationship may vary across the age
brackets within both subgroups of patients requiring and not
requiring invasive ventilatory support.
Methods
This prospective observational study consecutively recruited
patients over a period of 11 months (from 1 December 2000
to 31 October 2001) who were admitted to one of the ICUs
located at the Central Institute of the Hospital das Clinicas,
University of Sao Paulo Medical School, a tertiary university
hospital with 900 beds, 87 of which are ICUbeds divided into
12 different units (42% used exclusively for surgical patients,
36% for medical admission and the remaining 22% used for
both purposes). The sample of our study was dened as the
elderly population aged 65 years and over. Those aged 55 and
over were used as a reference population.
Patients who either died or were discharged within 24 h
after admission were excluded fromthe study to avoid poten-
tial bias resulting from the inclusion of terminally ill patients
and individuals submitted to elective surgery just recover-
ing from anaesthesia. If, within the same hospitalisation,
a patient was readmitted to the ICU, only data from the
rst ICU-admission were analysed. The patients with coro-
nary and orthopedic emergencies requiring intensive therapy
were admitted to other areas of the hospital. The local ethics
committee approved this study and, due to its observational
nature, informed consent was not required.
The following data were recorded using a predened form
within the rst 24 h of admission: age, gender, date of admis-
sion to hospital and ICU, primary cause for admission and
co-morbidity (according to the Simplied Acute Physiology
ScoreSAPS II criteria [22]), need for invasive mechanical
ventilation required for more than 24 h.
Most of the previous studies which analysed the relation-
ship between mortality and mechanical ventilation included
only invasively ventilated patients. Esteban et al. included
non-invasively ventilated subjects, but this subgroup repre-
sented only 4.9% of the sample [14]. In order to compare our
results with those obtained in the previous literature, patients
who were exclusively submitted to non-invasive ventilation
were included on the group of patients not requiring invasive
ventilatory support.
To assess the severity of illness on the rst day of admis-
sion to the ICU, the Acute Physiology Score (APS) was cal-
culated using only the 12 physiologic variables of the SAPS
II. The APS was used instead of the SAPS II because age,
type of admission and presence of co-morbidity were anal-
ysed as independent variables. The use of SAPS II could lead
to biased results since it is partially based on these param-
eters. The APS scores from 0 to 115 points. Higher scores
represent more severe illness. The organic dysfunction was
assessedusing the Logistic OrganDysfunctionScore (LODS)
system. The score ranges from 0 to 22 points [23].
Advanced directives such as do not resuscitate or intubate
were not analysed in the study. Usually, patients transferred
to ICUs in our institution are candidates for all advanced
measures needed, irrespective of their age. Withdrawal of care
or terminal weaning is an ICU team decision in agreement
with the family and was not inuenced by researchers.
Lengths of ICU and hospital stay were recorded. The
main measure of outcome used was survival status at the
time of hospital discharge.
The statistical analysis was performed using the Statistic
Package for Social Scientists (SPSS) version 13.0. The con-
tinuous variables were compared between groups using the
Student t-test for equality of means and non-parametric sim-
ilar tests when necessary, whereas categorical variables were
analysed using the chi-square test and odds ratio (OR) with a
95% condence interval (CI). The level of signicance used
was P <0.05. The variables that had a signicant association
with mortality to a value of P < 0.10 on univariate analysis
were entered into a logistic regression analysis performed for
the whole sample and for the isolated subgroups of ventilated
and non-mechanically ventilated patients in order to deter-
mine the relationship between age and in-hospital mortality
for these groups. Rather than using a continuous variable for
age, the relationship between age and in-hospital mortality
was examined using a dummy variable for three different
age intervals (5564, 6574 and 75 years). The rst cat-
egory of patients aged 55 to 64 years was considered as a
reference population for comparison. The three intervals of
age were considered the best and clearest solution to show
the evolution of mortality across the ageing process on an
elderly population.
Results
During the period of investigation, 2,118 patients were admit-
ted to the ICU, 975 being 55 years or older. From these, 135
(13.8%) stayed in the ICUfor <24 h or were discharged soon
after this period not allowing enough time for data collection.
A total of 840 subjects met the inclusion criteria and consti-
tuted the population of the present study. Table 1 illustrates
the main characteristics of the sample across the four age
intervals. The mean age standard deviation of the sam-
ple was 69.2 8.7 years (range 5599); there were 486 men
516

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
1
,

2
0
1
0
a
g
e
i
n
g
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

Age and outcomes in intensive care unit
Table 1. Demographic, clinical characteristics and outcomes of the sample (N= 840)
Age (years)
Characteristics 5564 6574 75 P-value
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Number of patients (%) 284 (34) 326 (39) 230 (27)
Age, mean (SD) 59.9 (2.8) 69.5 (2.7) 80.4 (4.7) < 0.001
Female, n (%) 115 (40) 123 (38) 116 (50) 0.009
Time between hospital admission and ICU (days), mean (SD) 6.5 (9.5) 6.8 (9.9) 6.0 (8.3) 0.50
Type of admission
Medical, n (%) 122 (43) 134 (41) 97 (42) 0.90
Scheduled surgery, n (%) 119 (42) 134 (41) 70 (30) 0.013
Unscheduled surgery, n (%) 43 (15) 58 (18) 63 (27) 0.001
Co-morbidity
Solid cancer, n (%) 80 (28) 96 (29) 63 (27) 0.86
Metastatic cancer, n (%) 21 (7) 24 (7) 18 (8) 0.97
Haematological cancer, n (%) 10 (3) 13 (4) 6 (3) 0.68
Invasive ventilation for 24 h, n (%) 147 (52) 195 (60) 140 (61) 0.06
APS, mean (SD) 15.6 (12.5) 17.2 (12.5) 18.8 (13.0) 0.02
LODS, mean (SD) 3.1 (2.6) 3.6 (2.8) 3.9 (2.8) 0.003
Neurological, mean (SD) 0.3 (0.7) 0.2 (0.6) 0.3 (0.7) 0.37
Cardiovascular, mean (SD) 0.2 (0.5) 0.4 (0.8) 0.4 (0.8) 0.03
Pulmonary, mean (SD) 0.6 (0.8) 0.6 (0.7) 0.6 (0.8) 0.88
Renal, mean (SD) 1.8 (1.9) 2.2 (1.8) 2.4 (1.9) < 0.001
Haematological, mean (SD) 0.1 (0.3) 0.1 (0.3) 0.1 (0.2) 0.80
Hepatic, mean (SD) 0.1 (0.3) 0.1 (0.3) 0.1 (0.3) 0.72
Length of ICU stay (days), mean (SD) 10.8 (16.0) 11.4 (16.0) 10.4 (12.9) 0.75
Length of hospital stay (days), mean (SD) 26.3 (24.2) 25.0 (24.6) 24.0 (19.6) 0.49
Died in ICU, n (%) 63 (24) 127 (39) 96 (42) < 0.001
Died in the hospital, n (%) 89 (31) 143 (44) 121 (53) < 0.001
SD = standard deviation; SAPS II = Simplied Acute Physiology Score II; APS = Acute Physiology Score; LODS = Logistic Organ Dysfunction Score.
(mean age 68.5 8.0) and 354 women (mean age 70.2 9.4).
Of the subjects examined, 42.0% were medical, 38.5% had
an elective surgery and 19.5%had an emergency surgery. The
mean time between hospitalisation and ICU admission was
6.5 9.4 days. Mechanically invasive ventilation for >24 h
was necessary for 57.4% of the patients. The mean score for
the SAPS II, APS and LODS, respectively, were 35.5 15.2,
17.1 12.7 and 3.5 2.8 points. The mean duration of
the ICU stay was 11.0 15.2 days, and the mean duration
of the hospital stay was 25.2 22.4 days. There were 353
deaths (42%), 292 (82.7%) in the ICUand 61 (17.3%) in other
departments.
Of the continuous variables, age (67.9 8.5 years for sur-
vivors vs 71.0 8.7 years for non-survivors, P <0.001), APS
(12.0 9.9 points vs 24.2 12.7, P < 0.001), LOD scores
(2.5 2.3 points vs 4.9 2.7, P < 0.001) and time between
hospital and ICU admission (5.9 8.2 days vs 7.2 10.8,
P =0.05) were signicantly associated with in-hospital mor-
tality on the univariate analysis. Of the categorical variables,
type of admission (57.5% mortality for medical admission,
47.6% for unscheduled surgery and 22.3% for scheduled
surgery, P<0.001) and invasive ventilation(63.3%of mortal-
ity, P < 0.001) were associated signicantly with in-hospital
mortality. The variables gender and presence of solid and
metastatic cancer were not associated with mortality.
The results of multivariate analysis on the overall sample
are shown in Table 2. Age, severity of the acute physiologi-
Table 2. Variables independently predictive of in-hospital
mortality by logistic regression
Variable Odds ratio (95% CI) P-value
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Age (6574 years) 1.68 (1.112.53) 0.013
Age (7584 years) 2.44 (1.563.80) < 0.001
Medical admission 3.30 (2.175.00) < 0.001
Unscheduled surgery 2.08 (1.273.39) 0.003
APS, points 1.05 (1.031.07) < 0.001
Invasive ventilatory support 6.12 (4.049.29) < 0.001
Time hospital/ICU (days) 1.03 (1.011.05) 0.011
CI = condence interval.
cal impairment measured through APS, invasive mechanical
ventilation, type of ICUadmission and time between hospital
and ICUadmission were determined as independent risk fac-
tors for in-hospital mortality. When LODS replaced the APS
in the analysis, it was determined as an independent factor
associated with mortality, a factor that remained unobserved
when it was introduced together with APS.
The relationship between age and in-hospital mortality
rates was examined using multivariate analysis of the odds
ratio of death for patients divided into three age brackets
(5564, 6574, 75 years). The rst category showed the
lowest mortality rate and was used as a reference group.
When adjusted for other variables, age remained strongly
correlated with in-hospital mortality in all categories and the
517

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
1
,

2
0
1
0
a
g
e
i
n
g
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

J. M. Farfel et al.
Table 3. Multivariate analysis for patients requiring and not requiring invasive mechanical ventilation
Requiring invasive mechanical ventilation Not requiring invasive mechanical ventilation
Mortality N (%) Odds ratio (95% CI) P-value Mortality N (%) Odds ratio (95% CI) P-value
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5564 years 75 (51) 1.00 14 (10) 1.00
6574 years 125 (64) 1.60 (1.012.54) 0.047 18 (14) 2.28 (0.995.25) 0.15
75 years 105 (75) 2.68 (1.584.56) <0.001 16 (17) 1.95 (0.824.62) 0.48
Scheduled surgery 80 (57) 1.00 11 (6) 1.00
Unscheduled surgery 74 (67) 1.94 (1.133.34) <0.001 4 (8) 1.16 (0.353.89) 0.81
Medical admission 170 (74) 2.63 (1.634.25) <0.001 33 (27) 3.74 (1.738.05) 0.001
APS, points
a
1.04 (1.021.06) <0.001 1.08 (1.041.12) <0.001
CI = condence interval.
a
APS = Acute Physiology Scoremortality columns for the APS are not fullled because APS is a numeric continuous variable. On multivariate odds ratio values
correspond to one point of increment on the score.
multivariate odds ratios increased progressively with every
age interval.
To examine the consistency of the relationship between
age and in-hospital mortality of elderly patients admitted to
the ICU and to identify the need for invasive mechanical
ventilation as a factor that could modify this relationship we
conducteda stratiedmultivariate analysis separating patients
requiring and not requiring invasive ventilatory support. The
results are shown in Table 3 and indicated that age was an
independent factor related to mortality for the invasively ven-
tilated patients and the multivariate odds ratios increased pro-
gressively with every age interval. However, for the patients
not requiring invasive ventilation, age was not correlated with
mortality in any interval.
Discussion
The results of our study suggest that the in-hospital mor-
tality of elderly patients admitted to an ICU is elevated and
furthermore affected by a variety of factors such as age,
invasive mechanical ventilation, organ dysfunction, severity
of impairment on admission, primary cause for admission
and time between hospital and ICU admission. Although
our results reinforce current evidence on predictive fac-
tors of mortality in older patients admitted to ICU [13, 14],
this study could clearly reveal differences between patients
receiving and not receiving invasive ventilatory support with
respect to the relationship between age and in-hospital
mortality.
This study evaluated the importance of the time inter-
val between hospitalisation and admission to the ICU for
the elderly population. The literature reports that patients
admitted from the emergency department have better prog-
nosis than those admitted from other hospital wards, proba-
bly because of the prompt transfer to the ICU [26, 27]. Our
results showed that a delayed transfer to the ICUis correlated
with higher rates of in-hospital mortality. The lack of vacant
beds on ICU was the major reason for delayed transfer.
Similar to other studies, age was strongly associated with
the severity of illness scores [8, 9, 24], but even when adjusted
for the degree of physiological impairment, age remained a
strong predictor of mortality. This nding is consistent with
previous studies, which evaluated the severity of illness upon
admissiontothe ICUthroughthe APS applicationand differs
frommany others that found no association between age and
mortality [5, 9, 10]. Most of these studies used the SAPS II
or the Acute Physiology and Chronic Health Evaluation III
(APACHEIII) [6, 8, 24, 25]. The use of these two scales could
lead to biased results since they are partially based on age. In
our study, the use of SAPS II in multivariate analysis instead
of APS removed age from the list of factors independently
related to mortality.
In the current study, the mortality rates veried for the
subgroup of invasively ventilated patients in all age groups
were slightly more elevated than those observed in other
similar studies. Although the severity of disease measured by
APS was comparable to that found in the literature, other fac-
tors not measured such as functionality or previous cognitive
status could have inuenced our results [6]. Methodological
aspects like the inclusion of patients requiring non-invasive
ventilatory support in this subgroup could explain the lower
mortality found in other studies [16].
Although age has frequently been examined as a prog-
nostic factor related to mortality of patients admitted to the
ICU, few prior studies have provided quantitative estimates
of increased risk associated with specic age intervals. Most
of these studies adopted different selection criteria when
dening an elderly population, mainly ranging from 60 to
85 years of age and not differentiating between various age
intervals [512]. Our study dened an elderly population fol-
lowing the most widely used criteria which considers elderly
every patient aged 65 and over and analysed the relationship
between age and in-hospital mortality in three intervals of age
(5564, 6574 and >75 years).
Esteban et al. created a different model to determine
a threshold of age that could best discriminate for ICU
survival [19]. A higher mortality for older patients, com-
pared with the younger according to Esteban cut points was
found in our sample for both subgroups of patients requir-
ing and not requiring invasive ventilatory support. How-
ever, our goal was not to compare an older and a younger
518

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
1
,

2
0
1
0
a
g
e
i
n
g
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

Age and outcomes in intensive care unit
subgroup of ICU patients, but to compare in-hospital
mortality occurring at different age stages in an elderly
population.
Different from the prior literature that contains a vari-
ety of studies on invasively ventilated patients [1521], the
present study evaluated the relationship between age and in-
hospital mortality for both patients receiving and not receiv-
ing invasive ventilatory support. We found that age is strongly
related to mortality of invasively ventilated patients and
that this relationship strengthens along with decades of age.
However, this effect was not observed among patients not
requiring invasive ventilatory support, animportant subgroup
of ICUadmittedpatients, whichrepresented42%of our sam-
ple. On this subgroup of patients, there was no age-related
increase in mortality both in the individuals between 6574
years old and in the individuals aged 75 years or older when
compared to the reference group, although a large sample
was included in the three age brackets.
Although there was a greater mortality on the group of
patients not requiring mechanical ventilation aged 65 years
and over (including those aged 75 years and over) when
compared to younger individuals, this comparison was not
the main objective of the current study, as stated before.
Furthermore, the age-related mortality found in the group
aged 75 years or more was even lower than that in the group
between 65 and 74 years old, showing that age should not
be considered as a factor related to in-hospital mortality in
elderly patients not requiring invasive ventilatory support in
ICUs.
The association of age and invasive mechanical ventila-
tion is strongly related to mortality. These results constitute
additional evidence for the liberal indication of non-invasive
mechanical ventilation and for the development of novel and
particular strategies of invasive mechanical ventilation for
elderly patients admitted to the ICU with respiratory acute
failure.
Some limitations of this study must be addressed. First,
although previous studies have shown relationship of in-
hospital mortality with other variables such as cognitive and
functional impairment and nutritional status prior to admis-
sion, these factors were not analysed in our study and could
affect the age-related impact in the outcome [6]. Second, we
did not examine the aggressiveness or appropriateness of
care, factors that could vary across different age groups (as
previously shown [2830]) and could be responsible for some
of the difference in age-related mortality. Nevertheless, there
were no statistical differences observed between the three
age groups regarding the need for invasive mechanical ven-
tilation, showing that the same procedures were indicated,
irrespective of the patients age. The demand for ICU beds
is higher than the offer in our institution and the selection of
candidates for admission results from general physicians or
general surgeons referral and acceptance by the intensivists.
The researchers did not interfere on this decision and a possi-
ble age-related bias of selection could have occurred both on
the referral and on the acceptance processes. Third, our study
was conducted at a single university institution with a spe-
cic sample excluding primary cardiologic and orthopedics
admissions.
Key points
Age worsens outcomes in elderly submitted to invasive
mechanical ventilation.
Age is not related to mortality in elderly not submitted to
invasive mechanical ventilation.
Severity of acute illness is a predictor of mortality in elderly
admitted to ICU.
Age is associated with severity of illness in elderly admitted
to ICU.
Acknowledgements
We are indebted to Carmen Saldiva Andre for statistic sup-
port and Natalie Oto for her assistance in the preparation
of the manuscript. The GRAPETI study group members
include: Andre Carlos K. Amaral, Andre Luiz D. Hovnanian,
Andre Luis P. Albuquerque, Anete A. Cionarele, Carla B.
Valeri, Carlos Toufen Junior, Daniel N. Forte, Eduardo L.
Queiroz, Eduardo R. Borges, Fabio M. Andrade, Guilherme
S. A. Azevedo, Gustavo F. J. Matos, Humberto B. Bogos-
sian, Idal Beer, Janaina Pilau, Jo ao Marcos Salge, Joz elio F.
Carvalho, Juliana C. Ferreira, Leonardo M. Tucci, Leonardo
Brauer, Liria Y. Yamaguchi, Marcel Y. Nishimura, Marcelo
Jorge J. Rocha, Marco Aurelio M. Pereira, Marcos Vini-
cius O. Costa, Marilia S. Almeida, Mateo S. Yaksic, Nelson
Ho, Teresa de Jesus J. Segovia, Pauliane V. Santana, Pedro
R. Genta, Pedro Medeiros Junior, Renato Testa, Ricardo
Henrique O. B. Teixeira, Ricardo R. Sanga, Sergio M. da Silva,
Tarso A. Accorsi, Telma Antunes, Valdelis N. Okamoto,
Vinicius Torsani and Vladimir R. P. Pizzo. The GRAPETI
study group participated on acquisition of data.
Conicts of interest
None.
Funding
This study was fully supported by the Geriatrics and Pul-
monary Divisions of University of Sao Paulo Medical
School, LIMHospital das ClinicasFMUSP and CAPES
(Coordenac aode Aperfeicoamentode Pessoal de Nvel Supe-
rior) in Brazil.
References
1. Jakob SM, Rothen HU. Intensive care 19801995: change in
patients characteristics, nursing workload and outcome. Inten-
sive Care Med 1997; 23: 116570.
519

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
1
,

2
0
1
0
a
g
e
i
n
g
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

J. M. Farfel et al.
2. Groeger JS, Guntupalli KK, Strosberg M et al. Descriptive
analysis of critical care unit in the United States: patients char-
acteristics and intensive care utilization. Crit Care Med 1993;
21: 27991.
3. Angus DC, Kelly MA, Schmitz RJ, White A, Popovich J. Cur-
rent and projected workforce requirements for care of the
critically ill and patients with pulmonary disease: can we meet
the requirements of an aging population? JAMA 2000; 284:
276270.
4. Fagon JY. Acute respiratory failure in the elderly. Crit Care
2006; 10: 151.
5. Chelluri L, Pinsky MR, Donohoe MP, Grenvik A. Long-term
outcome of critically ill elderly patients requiring intensive care.
JAMA 1993; 269: 311923.
6. Bo M, Massaia M, Raspo S et al. Predictive factors of in-hospital
mortality in older patients admitted to a medical intensive care
unit. J Am Geriatr Soc 2003; 51: 52933.
7. Van Den Noortgate N, Vogelaers D, Afschrift M, Colardyn
F. Intensive care for very elderly patients: outcome and risk
factors for in-hospital mortality. Age Ageing 1999; 28: 253
6.
8. Rockwood K, Noseworthy TW, Gibney RT et al. One-year
outcome of elderly and young patients admitted to intensive
care units. Crit Care Med 1993; 21: 68791.
9. Nicolas F, Le Gall JR, Alperovitch A, Loirat P, Villers D.
Inuence of patients age on survival, level of therapy and
length of stay in intensive care units. Intensive Care Med 1987;
13: 913.
10. Mayer-Oakes SA, Oye RK, Leake B. Predictors of mortality
in older patients following intensive care: the importance of
functional status. J Am Geriatr Soc 1991; 39: 8628.
11. Kass JE, Castriotta RJ, Malakoff F. Intensive care unit outcome
in the very elderly. Crit Care Med 1992; 20: 166671.
12. Boumendil A, Maury E, Reinhard I, Luquel L, Offenstadt G,
Guidet B. Prognosis of patients aged80 years andover admitted
in medical intensive care unit. Intensive Care Med 2004; 30:
64754.
13. Rosenthal GE, Kaboli PJ, Barnett MJ, Mitchell J, Sirio CA. Age
and the risk of in-hospital death: insights from multihospital
study of intensive care patients. J Am Geriatr Soc 2002; 50:
120512.
14. SUPPORT Principal Investigators. A controlled trial to
improve care for seriously ill hospitalized patients: the Study
to Understand Prognoses and Preferences for Outcomes and
Risks of Treatments (SUPPORT). The SUPPORT Principal
Investigators. JAMA 1995; 274: 15918.
15. Luhr OR, AntonsenK, KarlssonMet al. The ARFStudy Group.
Incidence and mortality after acute respiratory failure and acute
respiratory distress syndrome inSweden, DenmarkandIceland.
Am J Respir Crit Care Med 1999; 159: 184961.
16. Esteban A, Anzueto A, Frutos F et al. Mechanical Ventilation
International Study Group. Characteristics and outcomes in
adult patients receiving mechanical ventilation. A 28-day inter-
national study. JAMA 2002; 287: 34555.
17. Vincent JL, Akca S, De Mendonca A et al. SOFA Working
Group, Sequential Organ Failure Assessment: the epidemiol-
ogy of acute respiratory failure in critically ill patients. Chest
2002; 121: 16029.
18. Stauffer JL, Fayter NA, Graves B, Cromb M, Lynch JC, Goebel
P. Survival following mechanical ventilation for acute respira-
tory failure in adult men. Chest 1993; 104: 12229.
19. Esteban A, Anzueto A, Frutos-Vivar F et al. Mechanical Ven-
tilation International Study Group. Outcome of older patients
receiving mechanical ventilation. Intensive Care Med 2004; 30:
63946.
20. Cohen IL, Lambrinos J. Investigating the impact of age on out-
come of mechanically ventilation using a population of 41.848
patients from a statewide database. Chest 1995; 107: 1673
80.
21. Ely EW, Evans GW, Haponik EF. Mechanical ventilation in
a cohort of elderly patients admitted to an intensive care unit.
Ann Intern Med 1999; 131: 96104.
22. Le Gall JR, Lemeshow S, Saulnier F. A new simplied acute
physiology score (SAPS II) based on a European/North
American multicenter study. JAMA 1993; 270: 295763.
23. Le Gall JR, Klar J, Lameshow S et al. The logistic organ dys-
function systema newway to assess organ dysfunction in the
intensive care unit. JAMA 1996; 276: 80210.
24. Vosylius S, Sipylaite J, Ivaskevicius J. Determinants of outcome
in elderly patients admitted to the intensive care unit. Age and
Ageing 2005; 34: 15762.
25. Knaus WA, Wagner DP, Draper EA et al. The APACHE
III prognostic system. Risk prediction of hospital mortality
for critically ill hospitalized adults. Chest 1991; 100: 1619
36.
26. Doyle RL, Szaarsk N, Modin GW, Wiener-Kronish JP,
Matthey MA. Identication of patients with acute lung injury.
Predictors of mortality. Am J Respir Crit Care Med 1995; 152:
181824.
27. Goldhill DR, Sumner A. Outcome of intensive care patients in
a group of British intensive care units. Crit Care Med 1998; 26:
133745.
28. Hamel MB, Teno JM, Goldman L et al. SUPPORT Investiga-
tors. Patients age and decisions towithhold life-sustaining treat-
ments from seriously ill, hospitalized adults. Study to under-
stand prognoses and preferences for outcomes and risks for
treatment. Ann Intern Med 1999; 130: 11625.
29. Jayes RL, Zimmerman JE, Wagner DP, Draper EA, Knaus
WA. Do-not-resuscitate orders in intensive care units: current
practices and recent changes. JAMA 1993; 270: 22137.
30. Shepardson LB, Youngner SJ, Speroff T, O Brien RG, Smyth
KA, Rosenthal GE. Variation in the use of do-not-resuscitate
orders in patients with stroke. Arch Intern Med 1997; 157:
18417.
Received 7 May 2008; accepted in revised form 19 March 2009
520

b
y

g
u
e
s
t

o
n

O
c
t
o
b
e
r

3
1
,

2
0
1
0
a
g
e
i
n
g
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m