MAJOR ARTICLE
Severe Hypothermia Increases the Risk for
Intensive Care Unit–Acquired Infection
Kevin B. Laupland,1,2 Jean-Ralph Zahar,1,3 Christophe Adrie,4 Clémence Minet,5 Aurélien Vésin,1
Dany Goldgran-Toledano,6 Elie Azoulay,7 Maité Garrouste-Orgeas,1,8 Yves Cohen,9 Carole Schwebel,5 Samir Jamali,10
Michael Darmon,11 Anne-Sylvie Dumenil,12 Hatem Kallel,13 Bertrand Souweine,14 and Jean-Franc
xois Timsit1,5
1Team 11: Outcome of Respiratory Cancers and Mechanically Ventilated Patients, Integrated Research Center U823–Albert Bonniot Institute, Rond Point
de la Chantourne, La Tronche, France; 2Department of Critical Care Medicine, Peter Lougheed Centre, University of Calgary, Alberta, Canada; 3Infection
Control and Microbiology Unit, Necker University Hospital, 4Department of Cardiovascular Physiology, Cochin University Hospital, Paris, 5Medical
Polyvalent ICU, Grenoble University Hospital, 6Polyvalent ICU, Gonesse General Hospital, 7Medical ICU, University Hospital St Louis, 8Polyvalent ICU,
Groupe Hospitalier St Joseph, Paris, 9Medical-Surgical ICU, Avicenne University Hospital, Bobigny, 10Polyvalent ICU, General Hospital, Dourdan,
11Medical ICU, University Hospital, St Etienne, 12Surgical ICU, University Hospital Antoine Beclere, Clamart, 13Polyvalent ICU, General Hospital,
Cayenne, and 14Medical ICU, Gabriel Montpied University Hospital, Clermont Ferrand, France
Background. Although hypothermia is widely accepted as a risk factor for subsequent infection in surgical
patients, it has not been well defined in medical patients. We sought to assess the risk of acquiring intensive care unit
(ICU)–acquired infection after hypothermia among medical ICU patients.
Methods. Adults ($18 years) admitted to French ICUs for at least 2 days between April 2000 and November
2010 were included. Surgical patients were excluded. Patient were classified as having had mild hypothermia
(35.0°C–35.9°C), moderate hypothermia (32°C–34.9°C), or severe hypothermia (,32°C), and were followed for
the development of pneumonia or bloodstream infection until ICU discharge.
Results. A total of 6237 patients were included. Within the first day of admission, 648 (10%) patients had mild
hypothermia, 288 (5%) patients had moderate hypothermia, and 45 (1%) patients had severe hypothermia. Among
the 5256 patients who did not have any hypothermia at day 1, subsequent hypothermia developed in 868 (17%), of
which 673 (13%), 176 (3%), and 19 (,1%) patients had lowest temperatures of 35.0°C–35.9°C, 32.0°C–34.9°C, and
,32°C, respectively. During the course of ICU admission, 320 (5%) patients developed ICU-acquired bloodstream
infection and 724 (12%) patients developed ICU-acquired pneumonia. After controlling for confounding variables
in multivariable analyses, severe hypothermia was found to increase the risk for subsequent ICU-acquired infection,
particularly in patients who did not present with severe sepsis or septic shock.
Conclusions. The presence of severe hypothermia is a risk factor for development of ICU-acquired infection
in medical patients.
Hypothermia, commonly defined as a body temperature
,36°C, is common among patients who suffer critical
illness and is associated with adverse mortality outcome
[1–9]. Whereas the presence of hypothermia may be
a marker of severity of illness associated with mortality,
Received 28 July 2011; accepted 6 December 2011; electronically published 30
January 2012.
The manuscript was written on behalf of the OUTCOMEREA study group
(listed in the Appendix).
Correspondence: Jean-Francois Timsit, MD, PhD, Medical Polyvalent ICU,
University Hospital A Michallon, Grenoble, France 38043 ([email protected]).
Clinical Infectious Diseases 2012;54(8):1064–70
Ó The Author 2012. Published by Oxford University Press on behalf of the Infectious
Diseases Society of America. All rights reserved. For Permissions, please e-mail:
[email protected].
DOI: 10.1093/cid/cir1033
1064
d
CID 2012:54 (15 April)
d
Laupland et al
it may also result in complications that increase the risk
for subsequent adverse outcome. Infection is a widely
recognized complication of hypothermia and likely
arises, at least in part, due to decreased humoral and
cellular immunity associated with reduced body temperature [10–12]. In surgical patients, it is widely recognized that hypothermia increases the risk for wound
infection and that minimization of intraoperative hypothermia and long-term treatment at admission to the
intensive care unit (ICU) has been demonstrated to
reduce infectious complications and improve overall
outcome [13].
The epidemiology, clinical features, and outcomes
associated with temperature abnormalities are different among medical and surgical patients admitted to
ICUs [14]. Although a number of studies have identified that
the presence of hypothermia in patients who have sepsis or infection have worse outcomes [5, 7, 15, 16], it is not clear whether
hypothermia increases the risk for subsequent development of
nosocomial infection in critically ill medical patients. The objective of this study was to describe the occurrence of hypothermia
in adult medical patients admitted to ICU and assess the risk
for subsequent development of ICU-acquired pneumonia and
bloodstream infection.
METHODS
This study used an inception cohort design. All data were obtained using the OUTCOMEREA database [17]. First admissions among adults ($18 years) between April 2000 and
November 2010 with medical classification based on Simplified
Acute Physiology Score (SAPS II) criteria were included. Only
patients with ICU length of stay $2 days were included. Patients who were treated with therapeutic hypothermia were
excluded. Patients were monitored for the development of
hypothermia and infection until ICU discharge. According to
French law, this study did not require individual patient consent, because it involved research on a previously developed
and approved database.
OUTCOMEREA
OUTCOMEREA is an ongoing prospective observational collaborative study group that includes detailed clinical and outcome data on patients admitted to participating French ICUs
[17]. The lowest temperature is routinely recorded at presentation (the time between admission and 7:00 AM) and then
daily thereafter. Data included in the OUTCOMEREA database
have been collected by senior physicians or by trained monitors
in the participating ICUs. In some cases, participants in the
OUTCOMEREA group have enrolled consecutive patients admitted to ICU, and in others sampling has been performed in
which all consecutive admissions at a specific time during the
year or to certain ICU beds are included. For each patient, the
data were entered into an electronic case-report form using
VIGIREA and RHEA data-capture software (OUTCOMEREA,
Paris, France), and all case-report forms were then entered into
the OUTCOMEREA data warehouse. The data-capture software automatically conducts multiple checks for internal
consistency of most of the variables upon entry into the database. Queries generated by these checks were resolved with
the source ICU before incorporation of the new data into the
database. At each participating ICU, data quality was controlled
by having a senior physician from another participating ICU
check a 2% random sample of the study data. A 1-day coding
course is organized annually with the study investigators and
contrast research organization monitors.
Study Protocol
Once all study patients were identified, presentation and daily
low temperatures were recorded, and the admission age, gender,
and SAPS II and sepsis-related organ failure assessment scores
were extracted [18, 19]. Knaus scale definitions were used to
record preexisting chronic organ failures including respiratory,
cardiac, hepatic, renal, and immune system compromise [20].
The requirement for assisted ventilation, renal replacement
therapy, and use of corticosteroids was recorded. The presence
of severe sepsis and septic shock was established using standard
criteria [21].
Patients were classified according to a priori–defined criteria for the presence of hypothermia as defined as any low
temperature measured ,36°C. Once patients fulfilled criteria
for hypothermia, they were considered to be at subsequent
risk per that group unless criteria for a more severe category
were fulfilled. There was no requirement for a particular duration of hypothermia. Mild hypothermia was classified as a body
temperature of 35.0°C–35.9°C, and moderate and severe hypothermia were classified as low temperatures of 32.0°C–34.9°C
and ,32°C, respectively [4]. Admission temperatures were recorded until the end of day 1 admission to ICU. The presence
of infection was documented according to the standard definitions of the Centers for Disease Control and Prevention
[22]. Patients were deemed to have ICU-acquired infections
(pneumonia or bloodstream infection) if these first occurred
2 or more days after ICU admission.
Statistical Analysis
All data were analyzed using Stata 11.2 (Statacorp, College
Station, TX). Before analysis, the underlying distribution of all
continuous variables was assessed using histograms. Means with
standard deviations (SDs) were used to describe normally or
near normally distributed continuous variables, and skewed
variables were reported as medians with interquartile ranges
(IQR). Categorical variables were compared using Fisher’s exact
test for 2 3 2 tables, and the v2 test was used for multiple
categories.
Subdistribution proportional hazards [23] models were
developed to assess the independent effects of hypothermia
that occurred within the first 2 days of admission to ICU on
subsequent risk for ICU-acquired infection Discharge from
ICU or death were treated as competing events. The initial
models included temperature classification, ventilation status, antimicrobial therapy at admission, SAPS II, age, gender,
corticosteroid therapy, hemodialysis, sepsis syndromes, and
Knaus scale comorbidities. Stepwise variable elimination was
performed to develop the final models. The assumption of
proportional subhazards was assessed by testing for interactions with time. For all analyses, P , .05 was considered to
be statistically significant.
Hypothermia and ICU-Acquired Infections
d
CID 2012:54 (15 April)
d
1065
RESULTS
A total of 6237 patients were included. The median age was
63 years (IQR, 49–76), 3738 (60%) admissions were males, and
the median admission SAPS II score was 42 (IQR, 30–55). There
were 66 168 patient days of observation, 65 182 (99%) of which
were completed for temperature data.
Within the first day of admission, the mean (6SD) low
temperature was 36.6 6 1.1°C, and 648 (10%) patients had at
mild hypothermia (35.0°C–35.9°C), 288 (5%) patients had
moderate hypothermia (32.0°C–34.9°C), and 45 (1%) patients
had severe hypothermia (,32°C). Among patients with hypothermia at admission (within day 1), 630 of 648 (97%), 269 of
288 (93%), and 41 of 45 (91%) patients with mild, moderate,
and severe hypothermia had resolution to $36°C while alive
and admitted to ICU, with the median number of days to
resolution of 2 (IQR, 2–3), 2 (IQR, 2–3), and 3 (IQR, 2–4)
days, respectively.
Among the 5256 patients who did not have any hypothermia
at day 1, subsequent hypothermia developed in 868 (17%), of
which 673 (13%), 176 (3%), and 19 (,1%) patients had lowest
temperatures of 35.0°C–35.9°C, 32.0°C–34.9°C, and ,32°C,
respectively. The time to first temperatures ,36°C, ,35°C, and
,32°C in this cohort were 4 (3–8), 5 (3–11), and 6 (3–10) days,
respectively.
The cumulative incidence of at least 1 documented hypothermia episode ,36°C was 1849/6237 (30%). The cumulative incidence of mild hypothermia was 1242 (20%), moderate
hypothermia was 533 (9%), and 74 (1%) for severe hypothermia.
Among the total 66 168 ICU admission days, the incidence
density for mild, moderate, and severe hypothermia were 4.9,
1.4, and 0.1 per 100 days, respectively.
During the course of ICU admission, 320 (5%) patients developed ICU-acquired bloodstream infection and 724 (12%)
patients developed ICU-acquired pneumonia. The median time
to development of bloodstream infection or pneumonia was
8 (4–14) and 6 (3–12) days, respectively. The cumulative
incidence of pneumonia, bloodstream infection, or either was
related to the degree of hypothermia at admission as shown in
Table 1. Among the first 897 ICU-acquired infections, 683 were
pneumonias and 214 were bloodstream infections. No difference in etiology was observed in relation to the presence or
degree of hypothermia. Although there was no difference in
age (P 5 .684), higher median SAPS II scores were associated
with development of an ICU-acquired infection (48; IQR,
36–61 vs 41; IQR, 29–54; P , .0001). A number of categorical
factors were assessed for their association with development
of an ICU-acquired infection in univariate analyses and are
shown in Table 2.
The overall median length of ICU stay was 5 (IQR, 3–11)
days. The overall ICU and in-hospital case-fatality rates were
1120 of 6237 (18%) and 1434 (23%), respectively. Three
multivariable subdistribution proportional hazards models
(overall and 1 for patients presenting with severe sepsis/septic
shock and the other without) were developed with ICU discharge or death treated as a competing risk to assess the effect of
hypothermia occurring within the first 2 days of ICU admission
on subsequent risk for acquisition of an ICU-acquired infection
(pneumonia and/or bloodstream infection). After controlling
for confounding variables, severe hypothermia was associated
with a significant increased risk for later development of ICUacquired infection in the overall cohort and among patients
who did not present with severe sepsis or septic shock, as shown
in Tables 3 and 4, respectively. Hypothermia was not found to
be associated with an increased risk for ICU-acquired infection
among patients with severe sepsis or septic shock, as shown in
Table 5.
DISCUSSION
In this study, we demonstrate that hypothermia is a common
abnormal physical sign observed among critically ill medical
patients both at time of admission and subsequently during their
stay. We also show that hypothermia, especially when severe, is
Table 1. Risk for Development of Intensive Care Unit (ICU)–Acquired Bloodstream Infection and/or Pneumonia According to the
Presence of Hypothermia at Admission to ICU
No Hypothermia (n 5 5256)
Mild (n 5 648)
Moderate (n 5 288)
Severe (n 5 45)
P Value
Bloodstream infection
245 (4.7%)
49 (7.6%)
20 (6.9%)
6 (13.3%)
,.001a
Pneumonia
615 (11.7%)
71 (11.0%)
27 (9.4%)
11 (24.4%)
.030b
Either
741 (14.1%)
103 (15.9%)
39 (13.5%)
14 (31.1%)
.008c
a
Rates were not different among mild, moderate, and severe hypothermic patients (P 5 .322), but hypothermic patients were significantly different compared
with patients with no hypothermia (P 5 .0002).
b
Rates were not different among patients with no hypothermia, mild hypothermia, or moderate hypothermia (P 5 .434), but severe hypothermia was significantly
different compared with other conditions (P 5 .0157).
c
Rates were not different among patients with no hypothermia, mild hypothermia, or moderate hypothermia (P 5 .438), but severe hypothermia was significantly
different compared other conditions (P 5 .0041).
1066
d
CID 2012:54 (15 April)
d
Laupland et al
Table 2. Assessment of Factors Associated With Development of an Intensive Care Unit (ICU)–Acquired Infection (Pneumonia and/or
Bloodstream Infection) Among Medical ICU Patients (Univariate Analysis)
ICU-Acquired Infection
With Factor
Factor
ICU-Acquired Infection
Without Factor
Relative Risk
(95% Confidence Interval)
Hypothermia class
P Value
,.001
None
474/4388 (11%)
Admission mild
103/648 (16%)
Admission moderate
39/288 (14%)
Admission severe
14/45 (31%)
Male gender
Not ventilated
590/3738 (16%)
255/2917 (9%)
Noninvasive
307/2499 (12%)
1.28 (1.13–1.46)
,.0001
,.001
99/780 (13%)
Invasive
543/2540 (21%)
Day 1 renal replacement
39/321 (12%)
858/5058 (15%)
.84 (.62–1.13)
.288
Day 1 central vein catheter
434/2239 (19%)
463/3998 (12%)
1.78 (1.59–2.03)
,.0001
Day 1 antimicrobial therapy
561/3692 (15%)
336/2545 (13%)
1.15 (1.02–1.30)
.0277
Corticosteroids
185/1055 (18%)
712/5182 (14%)
1.28 (1.10–1.48)
.0017
Severe sepsis/shock
Diabetes
457/2326 (20%)
143/1003 (14%)
440/3911 (11%
754/5234 (14%)
1.75 (1.55–1.97)
.99 (.84–1.17)
,.0001
.922
Hepatic disease
Cardiovascular
Respiratory
Kidney
Immune compromise
58/373 (16%)
839/5864 (14%)
1.09 (.85–1.39)
.494
118/856 (14%)
779/5381 (14%)
.95 (.80–1.14)
.637
189/1104 (17%)
708/5133 (14%)
1.24 (1.07–1.44)
.0052
326/375 (13%)
848/5862 (14%)
.90 (.69–1.18))
.495
147/1059 (14%)
750/5178 (14%)
.96 (.81–1.13)
.631
Abbreviation: ICU, intensive care unit.
associated with the development of ICU-acquired infection.
Although there is ongoing debate on whether ICU-acquired
pneumonia significantly increases the attributable risk for death
among patients admitted to ICU [24, 25], this appears to be the
case for ICU-acquired bloodstream infection [26, 27]. Our
observation of an increased risk for infection may, at least in
part, explain the increased risk for mortality observed among
medical patients admitted to ICU with hypothermia.
It is biologically plausible that hypothermia increases the
risk for development of subsequent infection. Hypothermia
Table 3. Multivariable Modeling of Risk for Development of
Intensive Care Unit (ICU)–Acquired Infection Associated With
Hypothermia Among Patients Admitted to ICU (Overall Cohort)
Table 4. Multivariable Modeling of Risk for Development of
Intensive Care Unit (ICU)–Acquired Infection Associated With
Hypothermia Among Patients Admitted to ICU Without Severe
Sepsis or Septic Shock
Temperature
Classification
Temperature
Classification
No hypothermia
SHR
[95% CI]
1
Adj SHR
[95% CI]
Pa
,.0001 1
Pb
.004
No hypothermia
SHR
[95% CI]
1
Adj SHR
[95% CI]
Pa
,.0001 1
Mild hypothermia 1.28 [1.04–1.58]
.95 [.86–1.51]
Mild hypothermia 1.05 [.76–1.46]
.74 [.56–1.03]
Moderate
hypothermia
1.13 [.82–1.56]
.78 [.56–1.09]
Moderate
hypothermia
1.02 [.62–1.69]
.65 [.39–1.09]
Severe
hypothermia
3.54 [2.08–6.03]
2.42 [1.41–4.15]
Severe
hypothermia
4.52 [2.4–8.52]
2.76 [1.44–5.28]
Pb
.0009
Abbreviations: Adj, adjusted; CI, confidence interval; SHR, subdistribution hazard
ratio.
Abbreviations: Adj, adjusted; CI, confidence interval; SHR, subdistribution hazard
ratio.
a
a
b
Subdistribution hazard model.
Subdistribution hazard model adjusted for male sex and respiratory chronic
disease, at admission variables (symptoms, Simplified Acute Physiology Score
II), on 1st day of admission (and sepsis-related organ failure assessment score,
mechanical ventilation, central vein catheter, corticosteroids use, antibiotic use),
and stratified by center.
Subdistribution hazard model.
b
Subdistribution hazard model adjusted for male sex and respiratory chronic
disease, at admission variables (symptoms, Simplified Acute Physiology Score
II), on 1st day of admission (sepsis-related organ failure assessment score,
mechanical ventilation, central vein catheter, corticosteroids use, antibiotic use),
and stratified by center.
Hypothermia and ICU-Acquired Infections
d
CID 2012:54 (15 April)
d
1067
Table 5. Multivariable Modeling of Risk for Development of
Intensive Care Unit (ICU)–Acquired Infection Associated With
Hypothermia Among Patients Admitted to ICU With Severe Sepsis
or Septic Shock
Temperature
Classification
SHR
[95% CI]
Adj SHR
[95% CI]
Pa
No hypothermia
1
Mild hypothermia
1.40 [1.07–1.85]
.06
1.14 [.86–1.51]
1
Moderate
hypothermia
1.15 [.75–1.77]
.90 [.58–1.39]
Severe
hypothermia
1.90 [.71–5.13]
1.64 [.60–4.45]
Pb
.5
Abbreviations: Adj, adjusted; CI, confidence interval; SHR, subdistribution hazard
ratio.
a
Subdistribution hazard model.
b
Subdistribution hazard model adjusted for male sex and respiratory chronic
disease, at admission variables (symptoms, Simplified Acute Physiology Score
II), on 1st day of admission (sepsis-related organ failure assessment score,
mechanical ventilation, central vein catheter, corticosteroids use, antibiotic
use), and stratified by center.
has been demonstrated to suppress the release of proinflammatory cytokine levels, lower the expression of heat shock
protein 60, and reduce leukocyte count such that there is
a relative cellular and humoral immune suppression, which
in turn may lead to risk of infection [10–12]. Furthermore,
hypothermia may also lead to an increased risk of infection
related to a number of other mechanisms including, but not
limited to, cold-induced vasoconstriction and relative peripheral ischemia, decreases in level of consciousness, changes
in gastric motility, metabolic derangement, and increased
rates of bacterial colonization [13, 28, 29].
Although the body of literature is small, the results of previous studies have demonstrated an excess risk for development
of infection in neurology, trauma, and other surgical patients
associated with hypothermia [29, 30]. Perhaps the most compelling evidence is the clinical trial evidence, which demonstrated
that active intervention to reduce intraoperative hypothermia
reduces infection risk. In this study conducted by Kurz et al [13],
200 patients undergoing colorectal surgery were randomized
to routine intraoperative temperature management or to
additional warming. The mean (6SD) final intraoperative
temperature was 34.7 6 0.6°C in the routine management
and 36.6 6 0.5°C in the warming group. The subsequent risk
for wound infection was significantly higher in the routine
‘‘hypothermic’’ group compared with the additional warming group (19% vs 6%; P 5 .002).
It is notable that the effects of hypothermia and outcome
appear to be different in patients with and without severe
sepsis/septic shock. One possibility is that because these patients have severe sepsis or septic shock at admission, they may
be subsequently treated with antibiotics that reduce the later
1068
d
CID 2012:54 (15 April)
d
Laupland et al
risk for development of an ICU-acquired infection. Others
have found that the mortality outcome of severe sepsis/septic
shock is particularly influenced by the presence of hypothermia [5–7, 15, 16]. It is possible that similar factors
(ie, immune suppression) that contribute to adverse mortality outcome in hypothermic acutely septic patients may
also increase risk for subsequent infection in these patients.
However, this is a speculative hypothesis that requires further
investigation. It is also notable that, although not statistically
significant, moderate hypothermia was associated with a lower
risk for subsequent infection (Tables 3, 4, and 5). We do not
know why this may be the case and suggest that additional
studies are needed.
Whereas this study benefits from a relatively large sample
size obtained from a multicenter study, there are some
limitations that merit discussion. More importantly, because
this was a retrospective database study, we were limited by
the existing information available in the study database. We
were not able to ascertain the cause of the original hypothermia, and we are not able to report specifically what interventions may or may not have been performed to treat
patients with hypothermia. In addition, there are other
factors that increase the risk for ICU-acquired infection
(such as the use of invasive devices) that we did not incorporate into our analysis. In contrast, there are probably
other factors that may decrease the risk for infection such as
antimicrobial use and enhanced attention to procedural
technique. Finally, it is possible that hypothermia may have
been a result of incubating infection. However, we attempted
to minimize this discrepancy by excluding patients who were
admitted to the ICU for less than 2 days. In addition, in our
multivariable models, we only examined the effect of presence of hypothermia during the first 2 days on the development of later infection, which occurred, on average,
more than 1 week later.
In summary, this study identifies that hypothermia is a common complication of critical illness among medical patients
admitted to ICUs, and that severe hypothermia is a significant
risk for subsequent ICU-acquired infection. These data support
enhanced efforts to reduce the occurrence of hypothermia
among critically ill patients. Studies are needed to define the
optimal management of patients with temperature dysregulation
in the ICU.
Notes
Financial support. The study was funded by the OUTCOMEREA
organization.
Potential conflict of interest. All authors: No conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential
Conflicts of Interest. Conflicts that the editors consider relevant to the
content of the manuscript have been disclosed.
References
1. den Hartog AW, de Pont AC, Robillard LB, Binnekade JM, Schultz MJ,
Horn J. Spontaneous hypothermia on intensive care unit admission is
a predictor of unfavorable neurological outcome in patients after resuscitation: an observational cohort study. Crit Care 2010; 14:R121.
2. Waibel BH, Schlitzkus LL, Newell MA, Durham CA, Sagraves SG,
Rotondo MF. Impact of hypothermia (below 36 degrees C) in the rural
trauma patient. J Am Coll Surg 2009; 209:580–8.
3. Inaba K, Teixeira PG, Rhee P, et al. Mortality impact of hypothermia
after cavitary explorations in trauma. World J Surg 2009; 33:864–9.
4. Martin RS, Kilgo PD, Miller PR, Hoth JJ, Meredith JW, Chang MC.
Injury-associated hypothermia: an analysis of the 2004 National
Trauma Data Bank. Shock 2005; 24:114–18.
5. Tiruvoipati R, Ong K, Gangopadhyay H, Arora S, Carney I, Botha J.
Hypothermia predicts mortality in critically ill elderly patients with
sepsis. BMC Geriatr 2010; 10:70.
6. Peres Bota D, Lopes Ferreira F, Melot C, Vincent JL. Body temperature
alterations in the critically ill. Intensive Care Med 2004; 30:811–16.
7. Laupland KB, Davies HD, Church DL, et al. Bloodstream infectionassociated sepsis and septic shock in critically ill adults: a populationbased study. Infection 2004; 32:59–64.
8. Karalapillai D, Story DA, Calzavacca P, Licari E, Liu YL, Hart GK.
Inadvertent hypothermia and mortality in postoperative intensive care
patients: retrospective audit of 5050 patients. Anaesthesia 2009; 64:
968–72.
9. Capuzzo M, Moreno RP, Jordan B, Bauer P, Alvisi R, Metnitz PG.
Predictors of early recovery of health status after intensive care. Intensive Care Med 2006; 32:1832–8.
10. Ishikawa K, Tanaka H, Shiozaki T, et al. Characteristics of infection and
leukocyte count in severely head-injured patients treated with mild
hypothermia. J Trauma 2000; 49:912–22.
11. Biggar WD, Bohn D, Kent G. Neutrophil circulation and release from
bone marrow during hypothermia. Infect Immun 1983; 40:708–12.
12. Hashiguchi N, Shiozaki T, Ogura H, et al. Mild hypothermia reduces
expression of heat shock protein 60 in leukocytes from severely
head-injured patients. J Trauma 2003; 55:1054–60.
13. Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce
the incidence of surgical-wound infection and shorten hospitalization.
Study of Wound Infection and Temperature Group. N Engl J Med
1996; 334:1209–15.
14. Laupland KB, Shahpori R, Kirkpatrick AW, Ross T, Gregson DB,
Stelfox HT. Occurrence and outcome of fever in critically ill adults. Crit
Care Med 2008; 36:1531–5.
15. Pittet D, Thievent B, Wenzel RP, Li N, Auckenthaler R, Suter PM.
Bedside prediction of mortality from bacteremic sepsis. A dynamic
analysis of ICU patients. Am J Respir Crit Care Med 1996; 153:684–93.
16. Brivet F, Carras PM, Dormont J, et al. Hypothermia, a pertinent clinical
prognostic factor in severe systemic inflammatory response syndrome.
Crit Care Med 1994; 22:533–4.
17. OUTCOMEREA. Promotion et Development de la Recherche et de
l’Enseignement en Reanimation (French). Available at: www.outcomerea.
org. Accessed 31 January 2011.
18. Le Gall JR, Lemeshow S, Saulnier F. A new simplified acute physiology
score (SAPS II) based on a European/North American multicenter
study. JAMA 1993; 270:2957–63.
19. Vincent JL, de Mendonca A, Cantraine F, et al. Use of the SOFA score
to assess the incidence of organ dysfunction/failure in intensive care
units: results of a multicenter, prospective study. Working group on
‘‘sepsis-related problems’’ of the European Society of Intensive Care
Medicine. Crit Care Med 1998; 26:1793–800.
20. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II:
a severity of disease classification system. Crit Care Med 1985; 13:
818–29.
21. American College of Chest Physicians/Society of Critical Care Medicine
Consensus Conference: definitions for sepsis and organ failure and
22.
23.
24.
25.
26.
27.
28.
29.
30.
guidelines for the use of innovative therapies in sepsis. Crit Care Med
1992; 20:864–74.
Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;
16:128–40.
Fine J, Gray RJ. A proportional hazards model for the subdistribution
of a competing risk. J Am Stat Assoc 1999; 94:496–509.
Nguile-Makao M, Zahar JR, Francais A, et al. Attributable mortality of
ventilator-associated pneumonia: respective impact of main characteristics at ICU admission and VAP onset using conditional logistic regression and multi-state models. Intensive Care Med 2010; 36:781–9.
Muscedere JG, Day A, Heyland DK. Mortality, attributable mortality,
and clinical events as end points for clinical trials of ventilatorassociated pneumonia and hospital-acquired pneumonia. Clin Infect
Dis 2010; 51(Suppl 1):S120–5.
Laupland KB, Lee H, Gregson DB, Manns BJ. Cost of intensive care
unit-acquired bloodstream infections. J Hosp Infect 2006; 63:124–32.
Garrouste-Orgeas M, Timsit JF, Tafflet M, et al. Excess risk of death
from intensive care unit-acquired nosocomial bloodstream infections:
a reappraisal. Clin Infect Dis 2006; 42:1118–26.
Fries M, Stoppe C, Brucken D, Rossaint R, Kuhlen R. Influence of mild
therapeutic hypothermia on the inflammatory response after successful
resuscitation from cardiac arrest. J Crit Care 2009; 24:453–7.
Polderman KH. Mechanisms of action, physiological effects, and complications of hypothermia. Crit Care Med 2009; 37(7 Suppl):S186–202.
Peterson K, Carson S, Carney N. Hypothermia treatment for traumatic
brain injury: a systematic review and meta-analysis. J Neurotrauma
2008; 25:62–71.
Appendix
Members of the OUTCOMEREA Study Group Scientific
Committee. Jean-Franc
xois Timsit (Hôpital Albert Michallon
and INSERM U823, Grenoble, France), Elie Azoulay (Medical
ICU, Hôpital Saint Louis, Paris, France), Yves Cohen (ICU,
Hôpital Avicenne, Bobigny, France), Maı̈té Garrouste-Orgeas
(ICU Hôpital Saint-Joseph, Paris, France), Lilia Soufir
(ICU, Hôpital Saint-Joseph, Paris, France), Jean-Ralph
Zahar (Microbiology Department, Hôpital Necker, Paris,
France), Christophe Adrie (ICU, Hôpital Delafontaine,
Saint Denis, France), Michael Darmon (Medical ICU,
University Hospital, St Etienne, France), and Christophe
Clec’h (ICU, Hôpital Avicenne, Bobigny, and INSERM
U823, Grenoble, France).
Biostatistical and Informatics Expertise. Jean-Francois Timsit
(Hôpital Albert Michallon and Integrated Research Center
U823, Grenoble, France), Corinne Alberti (Medical Computer
Sciences and Biostatistics Department, Robert Debré, Paris,
France), Adrien Franc
xais (Integrated Research Center U823,
Grenoble, France), Aurélien Vesin (Integrated Research Center
U823, Grenoble, France), Stephane Ruckly (INSERM U823,
Grenoble, France), Christophe Clec’h (ICU, Hôpital Avicenne,
Bobigny, and INSERM U823, Grenoble, France), Frederik
Lecorre (Supelec, France), Didier Nakache (Conservatoire
National des Arts et Métiers, Paris, France), and Aurélien
Vannieuwenhuyze (Tourcoing, France).
Investigators of the OUTCOMEREA Database. Christophe Adrie
(ICU, Hôpital Delafontaine, Saint Denis, France and
Physiology, Hôpital Cochin, Paris, France), Bernard
Allaouchiche (ICU, Edouard Herriot Hospital, Lyon, France),
Caroline Bornstain (ICU, Hôpital de Montfermeil, France),
Hypothermia and ICU-Acquired Infections
d
CID 2012:54 (15 April)
d
1069
Alexandre Boyer (ICU, Hôpital Pellegrin, Bordeaux, France),
Christine Cheval (SICU, Hôpital Saint-Joseph, Paris, France),
Jean-Pierre Colin (ICU, Hôpital de Dourdan, Dourdan,
France), Michael Darmon (ICU, CHU Saint Etienne, France),
Anne-Sylvie Dumenil (Hôpital Antoine Béclère, Clamart,
France), Adrien Descorps-Declere (Hôpital Antoine Béclère,
Clamart, France), Jean-Philippe Fosse (ICU, Hôpital Avicenne,
Bobigny, France), Samir Jamali (ICU, Hôpital de Dourdan,
Dourdan, France), Hatem Khallel (ICU, Cayenne General
Hospital), Christian Laplace (ICU, Hôpital Kremlin-Bicêtre,
Bicêtre, France), Alexandre Lauttrette (ICU, CHU G
Montpied, Clermont-Ferrand, France), Thierry Lazard (ICU,
Hôpital de la Croix Saint-Simon, Paris, France), Eric Le Miere
(ICU, Hôpital Louis Mourier, Colombes, France), Laurent
Montesino (ICU, Hôpital Bichat, Paris, France), Bruno
1070
d
CID 2012:54 (15 April)
d
Laupland et al
Mourvillier (ICU, Hôpital Bichat, France), Benoı̂t Misset
(ICU, Hôpital Saint-Joseph, Paris, France), Delphine Moreau
(ICU, Hôpital Saint-Louis, Paris, France), Etienne Pigné
(ICU, Hôpital Louis Mourier, Colombes, France), Bertrand
Souweine (ICU, CHU G Montpied, Clermont-Ferrand,
France), Carole Schwebel (CHU A Michallon, Grenoble,
France), Gilles Troché (Hôpital Antoine, Béclère, Clamart,
France), Marie Thuong (ICU, Hôpital Delafontaine, Saint
Denis, France), Guillaume Thierry (ICU, Hôpital Saint-Louis,
Paris, France), Dany Toledano (CH Gonnesse, France), and
Eric Vantalon (SICU, Hôpital Saint-Joseph, Paris, France).
Study monitors. Caroline Tournegros, Loic Ferrand, Nadira
Kaddour, Boris Berthe, Samir Bekkhouche, Sylvain Anselme,
and Kaouttar Mellouk.