ArticlePDF Available

Lactate and base deficit are predictors of mortality in critically ill patients with cancer


Abstract and Figures

OBJECTIVE: Cancer patients frequently require admission to intensive care unit. However, there are a few dataregarding predictive factors for mortality in this group of patients. The aim of this study was to evaluate whetherarteriallactateorstandardbasedeficitonadmissionandafter24 hourscanpredictmortalityforpatientswithcancer.METHODS: We evaluated 1,129 patients with severe sepsis, septic shock, or postoperative after high-risk surgery.Lactate and standard base deficit collected at admission and after 24 hours were compared between survivors andnon-survivors. We evaluated whether these perfusion markers are independent predictors of mortality.RESULTS: There were 854 hospital survivors (76.5%). 24 h lactate .1.9 mmol/L and standard base deficit , -2.3 wereindependent predictors of intensive care unit mortality. 24 h lactate .1.9 mmol/L and 24 h standard base deficit, -2.3 mmol/Lwere independent predictors of hospital death.CONCLUSION: Our findings suggest that lactate and standard base deficit measurement should be included in theroutine assessment of patients with cancer admitted to the intensive care unit with sepsis, septic shock or after high-risk surgery. These markers may be useful in the adequate allocation of resources in this population.KEYWORDS: Lactate; Mortality; Cancer; Critical Care.
Content may be subject to copyright.
Basic Science
Adaptation analysis of di erent noninvasive ventilation interfaces
in critically ill patients
RMD Silva, KT Timenetski, RCM Neves, LH Shigemichi, SS Kanda,
CCRodrigues, RA Caserta, E Silva
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P1 (doi: 10.1186/cc10149)
Introduction Noninvasive ventilation is a safe and e ective method
to treat acute respiratory failure, minimizing the respiratory workload
and oxygenation. Few studies compare the e cacy of di erent types of
noninvasive ventilation interfaces and their adaptation.
Objective To identify the most frequently noninvasive ventilation
interfaces used and eventual problems related to their adaptation in
critically ill patients.
Methods We conducted an observational study, with patients older
than 18 years old admitted to the intensive care and step-down units of
the Albert Einstein Jewish Hospital that used noninvasive ventilation.
We collected data such as reason to use noninvasive ventilation,
interface used, scheme of noninvasive ventilation used (continuously,
periods or nocturnal use), adaptation, and reasons for nonadaptation.
Results We evaluated 245 patients with a median age of 82 years (range
of 20 to 107 years). Acute respiratory failure was the most frequent
cause of noninvasive ventilation used (71.3%), followed by pulmonary
expansion (10.24%), after mechanical ventilation weaning (6.14%) and
sleep obstructive apnea (8.6%). The most frequently used interface
was total face masks (74.7%), followed by facial masks in 24.5% of
the patients, and 0.8% used performax masks. The use of noninvasive
ventilation for periods (82.4%) was the most common scheme of use,
with 10.6% using it continuously and 6.9% during the nocturnal period
only. Interface adaptation occurred in 76% of the patients; the 24%
that did not adapt had their interface changed to improve adaptation
afterwards. The total face mask had 75.5% of interface adaptation,
the facial mask had 80% and no adaptation occurred in patients that
used the performax mask. The face format was the most frequent
cause of nonadaptation in 30.5% of the patients, followed by patient’s
related discomfort (28.8%), air leaking (27.7%), claustrophobia (18.6%),
noncollaborative patient (10.1%), patient agitation (6.7%), facial
trauma or lesion (1.7%), type of mask  xation (1.7%), and 1.7% patients
with other causes.
Conclusion Acute respiratory failure was the most frequent reason
for noninvasive ventilation use, with the total face mask being the
most frequent interface used. The most common causes of interface
nonadaptation were face format, patient-related discomfort and air
leaking, showing improvement of adaptation after changing the
interface used.
Exercise training reduces oxidative damage in skeletal muscle of
septic rats
CW Coelho
, PR Jannig
, AB Souza
, H Fronza Junior
, GA Westphal
, PM Silva
, F Dal-Pizzol
, E Silva
Programa de Pós-Graduação da FMUSP, São Paulo – SP, Brazil;
Joinville – SC, Brazil;
UDESC, Joinville – SC, Brazil;
CEDAP, Joinville – SC, Brazil;
Laboratorio de Fisiopatologia Experimental, UNESC, Criciuma – SC, Brazil;
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P2 (doi: 10.1186/cc10150)
Introduction Septic patients frequently develop critical illness
myopathies (CIMs) that may represent a crucial factor for prolonged
intensive care unit treatment and for ventilator weaning delay.
Experimental  ndings have identi ed that oxidative stress plays a role
in causing muscle depletion in chronic pathological states like sepsis.
It is well documented that regular moderate physical exercise can
decreased oxidative stress and enhance antioxidant functions.
Objective To investigate whether exercise training reduces oxidative
damage in septic rats induced by cecal ligation and perforation (CLP).
Methods Wistar rats were randomly assigned to three groups: Sham
(submitted to a fake surgery), CLP, and CLP that was previously trained
(CLPT). The exercise training protocol consisted of 8 weeks of running
on a treadmill, 5 days/week, for 60 minutes at 60% of the maximal
running speed obtained on the graded treadmill test. Rats were
subjected to CLP surgery; after 120 hours of surgical procedure they
were killed by decapitation. Oxidative damage of lipids (thiobarbituric
acid reactive species (TBARS)) and proteins (carbonyl groups) were
analyzed in Soleus (type I  ber) and plantaris (type II  ber) muscles.
Results See Table 1.
© 2010 BioMed Central Ltd
Sixth International Symposium on Intensive Care
and Emergency Medicine for Latin America
São Paulo, Brazil. 22–25 June 2011
Published: 22 June 2011
Table 1 (abstract P2). Levels of TBARS and carbonyl of soleus and plantaris muscles
Analysis Muscle Sham CLP CLPT
TBARS (nmol/mg protein x 10
) Soleus 43.0±5.4 (11) 60.2±5.9 (13)* 39.7±5.5 (6)**
Plantaris 31.3±2.6 (10) 55.3±7.3 (11)* 27.8±5.6 (5)**
Carbonyl (nmol/mg protein x 10
) Soleus 38.8±4.3 (11) 50.9±4.6 (12)
31.3±4.6 (6)**
Plantaris 28.8±3.9 (10) 49.7±5.1 (13)* 45.5±6.9 (6)
Values presented as mean±SEM. *P <0.05 vs. sham. **P <0.05 vs. CLP.
P = 0.06 vs. sham.
Critical Care 2011, Volume 15 Suppl 2
© 2011 BioMed Central Ltd
Conclusion TBARS and carbonyl analysis for CLPT are lower than for CLP
with statistical signi cance, except for carbonyl plantaris with P=0.06
(Table 1 and Figure 1). Our data supported that exercise training before
sepsis could decrease oxidative damage in both muscle  ber types.
Direct hepatic tissue PO
measurements in sepsis and tamponade
E Silva, P Rehder, AJ Pereira, F Colombari, LFP Figueiredo
Institute of Heart, University of São Paulo, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P3 (doi: 10.1186/cc10151)
Introduction Tissue hypoxia diagnosis at the bedside remains a huge
challenge for intensivists, and surrogate markers of tissue oxygen
utilization are used instead. The precise correlation between them is
not well de ned.
Objective To verify the correlation between portal blood  ow, O
gradients, hepatic lactate gradient with hepatic tissue PO
Methods This is an observational experimental study, in which 16
large, male, white pigs, about 35 kg, were allocated into two groups:
sepsis (n = 8), and tamponade (n = 8). All protocols were approved by
the institutional review board for animal experiments. Anesthesia: pre-
medication with intramuscular ketamine (10 mg/kg) plus midazolam
(0.25 mg/kg); induction with intravenous propofol 5 mg/kg (at maxi-
mum) followed by continuous iso urane (1.5%), fentanil 2.5 µg/kg/hour
and pancuronium 0.24 mg/kg/hour. Mechanical ventilation settings:
Vt 10 ml/kg, PEEP 5 cmH
O, respiratory rate set to normocapnia and
adjusted to arterial oxygen partial pressure 60 to 100 mmHg.
Continuous gas analysis was also performed. Electrocardiography,
invasive pressure in dissected femoral artery, right atria and ventricular
pressures after left internal jugular dissection; etCO
(by gas analyzer),
pulmonary artery catheter, portal vein  ow Doppler ultrasound, and
small bowel tonometry, after median laparotomy. Liver tissue pO
monitoring: pO
–  uorescence quenching optode – and LDF – laser
Doppler  uxometry – probes were directly inserted inside liver
parenchyma (Oxford-Optronix, UK). Other procedures: cistostomy (to
monitor diuresis), inferior vena cava (by femoral) and superior vena
cava (by right jugular) vein catheterizations. Portal vein catheter, after
liver hilus dissection (Seldinger) and  uoroscopy-guided right supra-
hepatic vein catheterization. After experiments, pigs were sacri ced
with sedative overdose and 20 ml KCl 19.1% injection. Sepsis was
induced by spread of 150 ml warm saline diluted 1g/kg feces in the
peritoneal cavity. Tamponade: mini-thoracotomy and a mono-lumen
intrapericardium catheter positioning to arouse cardiac tamponade,
targeting 20% of baseline decrease in cardiac output at each time
phase. Data were analyzed in Excel 2007.
Results In both groups, there was a progressive decrease in portal
blood  ow, an increase in jejune–portal CO
gap, and a decrease in
hepatic tissue PO
. Interestingly, there was a progressive hepatic lactate
consumption as hepatic tissue PO
decreases. Figure 1 (overleaf)
depicts the behavior of the above variables.
Conclusion Hepatic tissue PO
paralleled portal blood  ow and was
inversely related to the jejune tissue PCO
gap. Liver has increased
lactate consumption as hepatic tissue PO
Myocardial energy metabolism in sepsis and in anemic, stagnant
and hypoxic hypoxia
AJ Pereira, P Rehder, LFP Figueiredo, F Colombari, D Backer, E Silva
Instituto do Coração, Universidade de São Paulo, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P4 (doi: 10.1186/cc10152)
Introduction Tissue hypoxia and in ammation are the pillars of
multiple organ dysfunction. Current therapeutic interventions aimed
to improve systemic oxygen delivery are mediated by increases in
cardiac output, but myocardium energetic demand increases in
conditions of limited supply. Only scarce data are available on heart
oxygen utilization during hypoxic injuries.
Objective To understand the heart metabolism, challenged by di er-
ent tissue hypoxia models, by examining oxygen, lactate, and glucose
in vascular compartments, including coronary sinus.
Methods Thirty-seven pigs, fully monitored, were challenged with
di er ent injuries, including normovolemic anemia (n = 8), cardiac
tamponade (n = 8), hypoxic hypoxia (n = 8), peritonitis-induced sepsis
(n= 8) while  ve served as controls. In addition to global hemodynamics
and oxygen transport, we measured oxygen saturation, lactate and
glucose concentrations in arterial, pulmonary artery and coronary
sinus vascular compartments. Cardiac power output was calculated as
a surrogate marker of cardiac demand.
Figure 1 (abstract P2). Levels of TBARS of soleus (a) and plantaris (b); and levels of protein carbonyl of soleus (c) and plantaris (d). Values presented as
mean±SEM. *Signi cant di erence in relation to Sham group (P <0.05);
Signi cant di erence in relation to CLP group (P <0.05).
Critical Care 2011, Volume 15 Suppl 2
Results No signi cant alterations were found in the energetic pro le in
the stagnant group. There was both a decrease in lactate consumption
and an increase in glucose consumption in anemia (LAC changed
from –0.7 to +0.5 mmol/l, P = 0.018; GLU changed from –0.1 to
–0.4 mmol/l, P = 0.118) and in hypoxic hypoxia (LAC from –0.4 to
–0.2 mmol/l, P = 0.361; GLU from –0.25 to –0.5 mmol/l, P = 0.096)
groups. In sepsis, we observed a progressive increase in glucose (GLU
from –0.1 to –0.25 mmol/l, P = 0.618) and lactate (LAC from –0.26 to
–0.53 mmol/l, P = 0.105) consumption by the heart. The highest lactate
production was observed in late phases of anemia (+0.5 mmol/l)
and the highest glucose consumption (–0.5 mmol/l) in late phases of
hypoxic hypoxia. A similar and low CPO (between 3.31 and 4.4 W) was
achieved in di erent time points according to the hypoxia model, such
as a FiO
about 10%, a Htc about 7%, a 30% reduction of cardiac output
in tamponade, or 4 hours after fecal peritonitis induction, suggesting
that the heart better tolerates hypoxia and anemia than sepsis and
tamponade. See Figures 1 and 2 overleaf.
Conclusion Energetic substrate selection seems to be an important
adaptive mechanism in response to di erent types of tissue oxygen
delivery impairment, which may have implications on inotropic agent
Outcomes of 3,400 patients with cancer admitted to intensive care
unit: a Brazilian prospective study
LA Hajjar, F Galas, J Almeida, D Nagaoka, FA Duarte, RE Nakamura,
CSimoes, R Kalil-Filho, PM Ho , JOC Auler Jr
ICESP, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P5 (doi: 10.1186/cc10153)
Background Intensive care unit (ICU) admission of critically ill cancer
patients was controversial until recently. In the last years, advances in
the management of malignancies and organ failures have improved
outcomes of patients, resulting in higher rates of survival in the ICU. The
aim of this study is to prospectively evaluate the characteristics, short
and midterm outcomes of cancer patients requiring intensive care.
Methods During 2 years, we evaluated prospectively patients with
cancer admitted to the Instituto do Cancer do Estado de São Paulo.
A total of 3,400 patients were included in the study; and collected
data were baseline data, risk scores, clinical status, co-morbidities,
admission diagnosis, ICU interventions, ICU and hospital outcomes and
90-day outcomes.
Results From 3,400 patients, 52.8% had solid tumors and 47.2%
had hematologic malignancies. The most frequent reasons for ICU
admission were: sepsis (32%), postoperative care (27%) and respiratory
failure (21%). The mean APACHE II score value 24 hours after admission
was 23.1±7.8 (8 to 45). ICU mortality was 22%, hospital mortality was
31% and 3-month mortality was 44%. Logistic regression analysis
showed that need for mechanical ventilation (odds ratio = 7.76; 95%
CI= 4.56 to 12.85), presence of metastasis (odds ratio = 2.87; 95% CI =
2.06 to 5.28), occurrence of acute renal failure (odds ratio = 2.92; 95%
CI = 1.67 to 9.46) and higher SOFA scores 72 hours after admission
(odds ratio = 6.76; 95% CI = 5.56 to 13.85) were independently
associated with increased hospital mortality. The 3-month quality of life
of patients who survived was considered unchanged in 51% patients,
worse in 25% and better in 24%.
Conclusion This prospective analysis of 3,400 patients with cancer
needing intensive care shows high survival rates and good quality of
life after ICU admission. These data encourage intensive care treatment
in oncologic patients to prevent, detect and cure organ dysfunction.
Red blood cell transfusion after cardiac surgery does not result in
improvement of tissue perfusion in adult patients
F Galas, JL Vincent, J Fukushima, R Nakamura, R Kalil Filho, F Jatene,
JOCAuler Jr, L Hajjar
InCor, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P6 (doi: 10.1186/cc10154)
Background Most patients undergoing cardiac surgeries are exposed
to red blood cell (RBC) transfusions, in the operating room or in the
postoperative period. One of the main beliefs of this therapy is the
ability of the RBCs to improve tissue perfusion through oxygen supply.
However, recently, this concept is being questioned by some evidence
as RBC storage lesion and adverse outcomes in transfused patients.
Figure 1 (abstract P3). Portal blood  ow, hepatic CO
gap, hepatic lactate gradient, jejune CO
gap and hepatic tissue PO
over time in both groups, sepsis
and tamponade.
Critical Care 2011, Volume 15 Suppl 2
The aim of this study was to determine if RBC transfusion after cardiac
surgery results in improvement of tissue perfusion.
Methods From February 2009 to February 2010, a total of 502 patients
underwent cardiac surgery with cardiopulmonary bypass at InCor
– University of São Paulo. Arterial lactate, standard base de cit (SBD),
arterial bicarbonate and oxygen central venous saturation (ScVO
were collected immediately at the beginning and end of the procedure,
immediately postoperative (POI), after 24 hours (1PO), 48 hours (2PO),
Figure 1 (abstract P4). Systemic and coronary gradients of lactate and glucose – sham, cytopathic hypoxia (sepsis) and anemic hypoxia. DLACH = heart
lactate di erence = CS LAC – ART LAC; DLACS = systemic lactate di erence = PA LAC – ART LAC; DGLUH = heart glucose di erence = CS GLU – ART GLU;
DGLUS = systemic glucose di erence = PA GLU – ART GLU.
Critical Care 2011, Volume 15 Suppl 2
72 hours (3PO) and at ICU discharge. Mean values of these above-
mentioned parameters were compared in patients exposed to RBC
transfusions and patients not exposed through repeated-measures
variance analysis.
Results Hemoglobin values were di erent between groups since
before surgery until just before ICU discharge and in all periods, the
group not exposed to RBC transfusions presented higher values
compared with the exposed group (see Figure 1 overleaf).
Conclusion In this prospective study, red blood transfusion did not
result in improvement of tissue perfusion parameters. This  nding
brings to discussion the real role of blood transfusion in cardiac
Replacing fentanyl infusion by enteral methadone decreases
weaning time from mechanical ventilation
R Wanzuita, G Westphal, F Pfuetzenreiter, S Ayres, A Cavalcanti,
Medical School, University of Joinville, Joinville – SC, Brazil
Critical Care 2011, 15(Suppl 2):P7 (doi: 10.1186/cc10155)
Background Patients exposed to long-term infusion or high-dose
of opioids may develop physiological dependence and withdrawal
symptoms during discontinuation. In mechanically ventilated adult
patients, the occurrence of fentanyl withdrawal syndrome has been
associated with di culties in discontinuing ventilatory support and
with increased length of stay (LOS).
Objective We tested the hypothesis that replacement of fentanyl
infusion by enteral methadone decreases weaning time from
mechanical ventilation.
Methods A prospective, randomized and double-blind study involving
patients ful lling criteria to weaning from mechanical ventilation but
under high risk for fentanyl abstinence syndrome (de ned as continuous
fentanyl for more than 5 days or more than 5 g/kg/hour during
12 hours). Patients were randomized into two groups, methadone
(MET) group and control (CT) group, as follows: at  rst 24hours both
groups were given 80% of the original dose of fentanyl and received,
additionally, in the MET group enteral methadone (10mg each 6 hours)
or enteral placebo in the CT group. After the  rst 24 hours, the MET
group received enteral methadone and intravenous placebo while the
CT group received enteral placebo and intravenous fentanyl. In both
groups, the blinded intravenous solutions were reduced by 20% of the
original dose, every 24 hours. Any abstinence symptoms were treated
with a bolus of fentanyl. A Kaplan–Meyer curve was constructed and
the Student t test was used to compare groups in following criteria: (1)
weaning time from MV, (2) days under MV and (3) ICU LOS.
Results Of 75 randomized patients, seven were excluded and 68 were
analyzed: 37 at MET and 31 in CT. Between the beginning of weaning
and extubation, there was a greater probability of anticipation of
Figure 2 (abstract P4). Systemic and coronary gradients of lactate and glucose – stagnant hypoxia and hypoxic hypoxia. DLACH = heart lactate di erence=
CS LAC – ART LAC; DLACS = systemic lactate di erence = PA LAC – ART LAC; DGLUH = heart glucose di erence = CS GLU – ART GLU; DGLUS = systemic
glucose di erence = PA GLU – ART GLU.
Critical Care 2011, Volume 15 Suppl 2
extubation in the methadone group, but the di erence was not
signi cant (hazard ratio = 1.44; 95% CI = 0.81 to 2.56; P = 0.21). The
e ects of treatment on weaning time were time dependent, and we
observed that on the  fth day the probability of successful weaning
was 2.27 times greater in the MET (P vs. 13.28±12.85 days, P < 0.004).
There was no di erence between the two groups with respect to the
duration of mechanical ventilation and ICU LOS.
Conclusion These data show that replacement of fentanyl infusion
by enteral methadone reduces the weaning time from mechanical
Ultrasound-guided venous cannulation: a model of training
between medical students and emergency physicians
UAP Flato, HP Guimaraes, O Berwanguer
Instituto de Ensino e Pesquisa do Hospital do Coração – Associação do
Sanatório Sírio, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P8 (doi: 10.1186/cc10156)
Introduction Use of ultrasound introduced as part of intensive care
therapy makes viable bedside invasive procedures and diagnosis.
Due to portability, combined with team training, its use guarantees
less complications related to insertion, as well as patients’ safety.
It also reduces severe conditions related to the catheter, such as
pneumothorax among others. The aim of this study was to evaluate the
accuracy related to ultrasound-guided venous catheter insertion in a
low-cost famtoma among medical students of third-year graduation
compared with experienced doctors and medical residents. We
evaluated the success rate of insertion, the number of puncture
attempts and the time related to the insertion of the needle from
Figure 1 (abstract P6). Comparison between groups exposed or not to red blood cell transfusions considering hemoglobin (Hb) values and perfusion tissue
parameters (lactate, oxygen venous central saturation, standard base excess and bicarbonate). *P <0.005.
Critical Care 2011, Volume 15 Suppl 2
contact with the surface of the phantom and its correct placement in
the vein.
Methods Study participants were 25 undergraduate students of
medicine (third year) participating in the curriculum of emergency
medicine and intensive care, nine medical residents (internal medicine)
and nine critical care physicians. All participants had no previous
experience with ultrasound-guided procedures, and medical students
had no previous experience with central venous access puncture.
There was a lecture prior to the study of 2 hours in ultrasound-guided
venous cannulation. Evaluation of the average time between groups
was performed by ANOVA using data processing in rank due to lack of
homogeneity and the Tukey test for multiple comparisons. A possible
relationship between the time needed until the puncture is performed
and length of experience was assessed by Spearman correlation, due to
lack of normality in the data.
Results We found a success rate of 100% in the insertion of a catheter in
phantom among all participants, a longer time in the group of graduate
students (Table 1), as well as the number of punctures (mean of 2).
Table 1 (abstract P8). Time (seconds) to cannulation in each group
Group n Mean Standard deviation P value
Inexperienced 25 19.60 13.778
Residents 9 12.44 5.525 0.003
Experts 8 9.88 1.553
Total 42 16.21 11.638
Conclusion The use of ultrasound-guided cannulation is a reliable
method of training associated with a high of success among graduate
students and experienced professionals.
Adverse events associated with long-term ketamine use in pediatric
septic shock
CMF Mangia, AFCF Martins, AP Loretti, RM Sousa, MC Andrade
Universidade Federal de São Paulo, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P9 (doi: 10.1186/cc10157)
Objective Ketamine hydrochloride is a noncompetitive antagonist
of the NMDA receptors and produces a dissociative state described
as a ‘functional and neuro-physiological dissociation between the
neocortical and limbic systems’ [1,2].
Methods We describe long-term use of ketamine in the pediatric
intensive care unit (PICU) inducing pyramidal liberation in a septic
shock patient.
Case A 15-month-old boy with congenital cardiopathy and develop-
mental delay without previous chronic encephalopathy history. He
was admitted with septic shock and during the PICU stay received
association of multiple analgesic-sedative agents and high doses of
ketamine intravenous infusion (Figure 1). The patient presented after
10 days of PICU stay symptoms associated with pyramidal liberation:
deep hyperre exia with sinre exia, Babinski sign on both sides,
opisthotonus, trismus. The clinical signs were not associated with new
metabolic or structural intracranial lesion. The patient was discharged
from hospital after 36 days receiving pericyazine that was interrupted
1week after hospital discharge.
Conclusion The ketamine side e ects after short-term use include
[1,2]: hypertension, apnoea, laryngospasm, emergence phenomena,
vomiting, nystagmus, ataxia, myoclonus, random limb movements,
opistho tonus, transient facial rash or  ushing, intracranial hypertension.
The long-term-use side e ects are unknown. This is the  rst report of
pyramidal liberation-associated intravenous ketamine for a prolonged
1. Emerg Med J 2004, 21:275-280.
2. Anesthesiology 1982, 56:119-136.
Critically ill patients with cancer and sepsis: clinical course and
prognostic factors
LSCF Rabello
, M Rosalem
, T Lisboa
, P Caruso
, R Costa
, J Leal
, J Salluh
M Soares
Instituto Nacional de Câncer, Rio de Janeiro – RJ, Brazil;
Hospital de Clínicas,
Universidade Federal do Rio Grande do Sul, Porto Alegre – RS, Brazil;
A. C. Camargo, São Paulo – SP, Brazil;
D’Or Institute for Research and
Education, Rio de Janeiro – RJ, Brazil
Critical Care 2011, 15(Suppl 2):P10 (doi: 10.1186/cc10158)
Introduction Sepsis is a frequent complication in patients with cancer
associated with adverse outcomes. The aim of this study was to
evaluate the clinical course and to identify independent predictors of
mortality in these patients.
Methods We performed a secondary analysis of a prospective cohort
study conducted at an oncological medical–surgical ICU. Logistic
regression was used to identify predictors of hospital mortality.
Results A total of 563 patients (77% solid tumor; 23% hematological
malignancies) were included over a 55-month period. The most
frequent sites of infection were the lung, abdomen and urinary tract;
91% patients had severe sepsis/septic shock. Gram-negative bacteria
were responsible for more than half of the episodes of infection; 207
(38%) patients had polymicrobial (>1 infectious agent) infections.
ICU, hospital and 6-month mortality rates were 51%, 65% and 72%. In
multivariate analyses, sepsis in the context of medical complications,
active disease, compromised performance status, presence of
three or four SIRS criteria, and the presence of respiratory, renal and
cardiovascular failures were associated with increased mortality.
Adjusting for other covariates, patients with urinary tract infection had
better outcomes. Patients could be strati ed into categories of risk for
death according to the number of clinical predictors.
Conclusion Our results can be of help to assist intensivists in clinical
decisions and counseling of patients and families, and to contribute
with future research to improve characterization and risk-strati cation
in these patients.
Evaluation of knowledge of nurses in intensive care, semi-intensive
care and ready for a private hospital of St Paul on sepsis
CDC Gambin, DFM Junior, SCL Shiramizo, ACMA Gonçalves, E Silva
Hospital Israelita Albert Einstein, Morumbi, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P11 (doi: 10.1186/cc10159)
Background Sepsis is an in ammatory response secondary to an
infectious process with presumed or known [1] focus that can lead to
involvement of multiple organs and death. The incidence of severe
sepsis and septic shock among patients admitted to intensive care
units (ICUs) in Brazil was 36 and 30 per 1,000 patient-days, respectively
[2]. ICUs in other countries reported an incidence of severe sepsis of 21
Figure 1 (abstract P9). Daily total doses of analgesic-sedative agents.
Critical Care 2011, Volume 15 Suppl 2
cases per 100 admissions in Paris [3] and 16 cases per 100 admissions
in the United States [4].
Method Field research, descriptive and exploratory, transversal,
prospective, level I, with a quantitative approach. We approached
nurses working in intensive care, emergency care and semi-intensive
work during the day or night. The study was conducted in a large
private hospital in São Paulo. A semi-structured questionnaire was
developed with multiple-choice questions containing personal data
and information on knowledge about sepsis.
Results We found 82 respondents with 33 nurses from the ICU, 30 from
semi-intensive units (USI) and 19 of the health care unit (APU); there is a
predominance of females and training time, being an average of 80%.
Over 60% of respondents were postgraduates. The APU was found to
have the greatest number of correct classi cations of sepsis, more than
50% of respondents; the ICU was in second place, with an average of
40% hits; and the USI averaged 30% correct.
Conclusion Of the nurses responding to the questionnaire, 66 (80%)
are female and 74 (90%) have worked for more than 1 year and are
trained well, and 22 (26%) hold a postgraduate program graduation.
According to the results, we can observe that the better performance
was seen in the emergency care units and intensive care. This does not
exclude such units from a proposal for continuing education, since the
primary concern relates to the retention of clinical symptoms.
1. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, et al.: De nitions
for sepsis and organ failure and guidelines for the use of innovative
therapies in sepsis. The ACCP/SCCM Consensus Conference Committee.
American College of Chest Physicians/Society of Critical Care Medicine.
Chest 1992, 101:1644-1655.
2. Silva E, Pedro MA, Sogayar AC, Mohovic T, Silva CL, Janiszewski M, et al.:
Brazilian sepsis epidemiological study (BASES study). Crit Care 2004,
3. Guidet B, Aegerter P, Gauzit R, Meshaka P, Dreyfuss D: Incidence and impact
of organ dysfunctions associated with sepsis. Chest 2005, 127:942-951.
4. Longo M: Evaluation of the knowledge of professionals in emergency
medicine about the criteria for SIRS, sepsis, severe sepsis and septic shock.
PhD thesis, Santa Catarina; 2004.
First results of a sepsis protocol at Diadema State Hospital
FM Gazoni, ILS Braga, E Giordanni, LS Vendrame
Hospital Estadual de Diadema, Diadema – SP, Brazil
Critical Care 2011, 15(Suppl 2):P12 (doi: 10.1186/cc10160)
Introduction Sepsis is an important cause of death at Diadema State
Hospital, therefore a sepsis protocol was designed.
Objective To reduce sepsis prevalence, morbidity, the mortality rate
and its high cost.
Methods An audit was conducted in the period of April to September
2010 with data collected through hospital records.
Results Sixty-three patients were enrolled. Analyzed was each item of
the package of 6 hours according to the designed protocol, including
total adherence to the package of 6 hours, mortality of eligible patients
and mortality of patients who adhered to the package of 6 hours. Of
63 patients, 28 patients were discharged and 35 evolved to death, only
one case not correlated with death from septic shock. Mortality due to
sepsis at our service was 56%, which is consistent with the mortality
rate in Brazil (57.3%, according to ILAS) and in public hospitals (63.9%).
Adherence to the package of 6 hours recommended by the SSC was
only 21 of the 63 cases. Of these 21 cases, 11 patients survived and
10 died. Thirty cases of all had some compliance with the protocol
of 6 hours, and of these 17 were discharged and 13 died. Disrupting
the total mortality (35 cases, 56%), it was found that mortality among
patients who adhered to the package of 6 hours was lower (48%) when
compared with those who did not join (60%).
Conclusion The results show a lower mortality rate in cases where
there was total adherence to the package of interventions in the  rst
6 hours, but we still have low level of adherence to this package (33%).
The average length of stay decreased dramatically from 2008 to 2010
(73% vs. 62%) when we compared the patients who died with those
who survived, which is still high but has fallen over time, surpassing
the survival rates measured in other public hospitals in Brazil (data
from ILAS). After these  rst results, improvements were made to be
implemented in 2011 such as review and redrafting of the protocol
ow; training di erent categories of professionals (technicians,
nurses, physiotherapists, doctors, pharmacists); realignment with
ILAS, including manager selection protocol with capacity-building and
training for use of the international database for comparative analysis;
review the recommendation of antimicrobials for the second focus
of infection with sepsis; and regular monitoring of results, including
average length of stay and mortality. The challenge now is to decrease
deaths, aiming to achieve levels comparable with the best institutions
in the world. In partnership with ILAS, the project SPDM against Sepsis,
our team has strived to achieve this goal.
Georeferencing sepsis in São Paulo city
E Silva, AS Cypriano, LF Lisboa, M Cendoroglo, F Colombari, BFC Santos
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P13 (doi: 10.1186/cc10161)
Introduction Sepsis is a worldwide disease with heterogeneous
outcome. The main factors related to prognosis are age, associated
comorbidities, invader virulence, and time to therapeutic initiation.
Data related to social–economical attributes have been scarcely
Objective To evaluate the distribution of sepsis-associated deaths in
São Paulo city using a geographic information system (GIS); to verify
whether there is any correlation between socioeconomic status and
number of deaths.
Methods GIS is a system for input, storage, manipulation, and output
of geographic information. GIS allows one to know the socioeconomic
conditions of the region studied, including provision of health services,
spatial data (rivers, parks, and so on), population data (age and sex),
and estimated demand for health services. Thus, GIS could support
health managers for planning, monitoring, priority setting and
decision-making. Sepsis was identi ed through death certi cates using
several International Disease Codes including, but not restricted to,
sepsis, septicemia, pneumonia, urinary tract infection, wound surgical
infection, bloodstream infection, meningitis, and multiple organ failure
among others.
Results Figure 1 (overleaf) depicts every death according to the
location of residence.
Conclusion Death secondary to sepsis is widely distributed throughout
the regions of São Paulo, and further analysis needs to be done in
di erent subgroups for better characterization and contrast of this
syndrome in distinct regions and socioeconomic strata of the city.
Impact of the Surviving Sepsis Campaign implementation on severe
sepsis outcome
E Silva
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P14 (doi: 10.1186/cc10162)
Introduction Sepsis is associated with high morbi-mortality rates and
evidence-based strategy implementation could improve outcome.
Objective To evaluate the impact of the 6-hour Surviving Sepsis
Campaign (SSC) bundle on mortality in a tertiary hospital.
Methods A multifaceted intervention to facilitate compliance with
selected guideline recommendations in the intensive care unit,
emergency department, and wards in our hospital was implemented.
Data were collected in two periods, before and after implementation
of the protocol, from July 2005 to December 2008. The  rst period was
called the Control Group from July 2005 to March 2006 and the Protocol
Group from April 2006 to December 2008. SSC was implemented in
April 2006. Compliance to the 6-hour SSC bundle was measured in both
periods, as well as outcome.
Results A total of 414 patients were enrolled, 92 in the Control Group
and 322 in the Protocol Group. Mean age was 66 ± 19 years, mean
APACHE II score was 24.1 ± 7.5, and 42% were female. Hospital LOS
Critical Care 2011, Volume 15 Suppl 2
in the Control Group was 37±44 days and in the Protocol Group was
47±90 (P = 0.36). ICU LOS were similar, 14±17 days and 14±35 days,
respectively. The 6-hour bundle adherence has signi cantly increased
from 10% in the Control Group to 28% in the Protocol Group (P = 0.001).
Di erences between 6-hour bundle variables are shown in Table 1. The
mortality rate decreased after protocol implementation from 57% to
38% (P = 0.001).
Conclusion Implementation of the SSC 6-hour bundle was associated
with lower mortality.
Importance of glycated hemoglobin in hyperglycemia diagnosis of
patients with sepsis
ACM Simioni
EPM-UNIFESP, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P15 (doi: 10.1186/cc10163)
Introduction Hyperglycemia is a frequent event in patients hospitalized
in intensive care units (ICUs) and was attributed to endocrine metabolic
stress related to acute disease. However, the interference of diabetes
mellitus (DM) or undiagnosed glucose intolerance in hyperglycemia
pathogenesis in critically ill patients is not well established.
Objective To correlate the presence of DM with hyperglycemia or
glucose intolerance, not previously diagnosed in patients with severe
sepsis/septic shock in the ICU, using the new standards of the American
Diabetes Association (ADA) for classi cation of glycated hemoglobin
(HbA1c) [1].
Figure 1 (abstract P13).
Table 1 (abstract P14). Compliance to each SSC 6-hour bundle variable
Variable (%) Overall (n = 414) Control (n = 92) Protocol (n = 322) P value
Blood culture 69 55 73 0.001
Lactate measurement 83 70 87 <0.0005
Central venous oxygen saturation 62 55 65 0.068
Central venous pressure 65 48 70 <0.0005
Early large spectrum antibiotics 89 86 89 0.46
All variables 24 10 28 0.001
Critical Care 2011, Volume 15 Suppl 2
Methods A study prospectively evaluating patients admitted to the
ICU between the January 2007 and August 2009. We included patients
with severe sepsis or septic shock, with less than 48 hours from organ
dysfunction onset. Severe sepsis and septic shock were de ned based
on International Sepsis De nitions Conference criteria [2]. Exclusion
criteria were: previous diagnosis of DM, insulin infusion at the time of
evaluation, sepsis within <30 days and refusal to participate. According
to new ADA classi cation, patients were considered normal with
HbA1c ≤5.6%, glucose intolerant with HbA1c between 5.7% and 6.4%
and diabetic those with HbA1c≥ 6.5% [1]. Statistical analysis used the
t test, chi-square and correlation coe cient and was made using SPSS
15.0 software.
Results Our sample included 59 patients, mean age 60 ± 18 years,
62.7% were male. By classifying patients according to HbA1c, although
denying a history of DM, only 37.3% had normal HbA1c. About 28.8%
had undiagnosed diabetes and 33.9% had glucose intolerance.
Analyzing the HbA1c as a continuous variable, we found only a
statistically signi cant correlation with blood glucose levels at inclusion
(P = 0.04), serum insulin at inclusion (P = 0.02) and insulin resistance
at inclusion (P = 0.02). Studying the population characteristics, an
association between HbA1c change and presence of comorbidities was
observed (P = 0.004). Furthermore, patients with HbA1c changes were
older (P = 0.02), had higher blood glucose at inclusion (P = 0.03) and
higher lactate after 24 hours of inclusion (P = 0.03). See Figure 1.
Conclusion In this sample of patients with sepsis without previous
history of DM a high incidence of patients with diabetes and glucose
intolerance undiagnosed was found. Therefore, HbA1c measurement
in the ICU may be useful in the investigation of patients with
1. American Diabetes Association: Diagnosis and classi cation of diabetes
mellitus. Diabetes Care 2010, 33(Suppl 1):S62-S69.
2. Levy MM, Fink MP, Marsahll JC, Abraham E, Angus D, Cook D, Cohen J, Opal
SM, Vincent JL, Ramsay G; SCCM/ESICM/ACCP/ATS/SIS: International Sepsis
De nitions Conference. Crit Care Med 2003, 31:1250-1256.
In uence of vasopressor agent in pediatric septic shock mortality
CMF Mangia, LPS Jose, FL Monteiro, AT Fernandes, P Biasi, F Menezes,
ELLima, C Oliveira, F Bueno, P Paiva, MC Andrade
Universidade Federal de São Paulo, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P16 (doi: 10.1186/cc10164)
Objective To clarify the impact of the choice of vasopressor support on
mortality in pediatric septic shock (SS).
Methods A retrospective study based on the institutional database
analyzing 1,050 patients admitted from October 1999 to January 2005.
We studied children with SS after the neonatal period admitted to
the pediatric intensive care (PICU) and we assessed the vasopressor
support in the  rst 24 hours, PICU and hospital (HSP) length of stay
(LOS), number of vasoactive drugs used, association between drugs
and HSP mortality.
Results There were 101 consecutive patients with SS, mean age
41 months (95% CI = 30 to 52 months); mean of PICU LOS 16.73
days (95% CI = 11.18 to 22.28) and hospital LOS 55.46 days (95% CI =
43.16 to 67.75). PICU mortality was 32% and HSP mortality after PICU
discharge was 10.8%. Of these, 33% patients received dobutamine and
26% patients dopamine as the only vasoactive drug. Dopamine plus
dobutamine was used in 17.8%; dobutamine plus norepinephrine in
18% and dopamine plus norephinephrine in 3.9%. The HSP mortality
associated with dobutamine was 29.4%; dopamine 53.8%; dopamine
plus dobutamine 50%; dopamine plus norepinephrine 25%. The
dopamine and dopamine plus dobutamine groups had higher hospital
mortality (66% vs. 34%). Dopamine was associated with hypertensive
state (odds ratio, 0.433; 95% CI = 0.192 to 0.976; P = 0.047), hypoxemia
(odds ratio, 0.190; 95% CI = 0.040 to 0.909) and mechanical ventilation
utilization (odds ratio, 2.625; 95% CI = 1.085 to 6.327; P = 0.035).
Conclusion Adrenergic support for pediatric patients with SS remains
controversial. A prospective randomized controlled trial will be
important to determine which subgroups of SS patients will bene t
most with each drug.
Incidence and risk factors for sepsis in surgical patients: a cohort
AAFS Georgeto, ACGP Elias, MT Tanita, CMC Grion, LTQ Cardoso, P Verri,
CFF Veiga, ÁRG Barbosa, AZ Dotti, T Matsuo
Hospital Universitário de Londrina, Universidade Estadual de Londrina,
Londrina – PR, Brazil
Critical Care 2011, 15(Suppl 2):P17 (doi: 10.1186/cc10165)
Introduction Surgical patients are vulnerable to infectious compli-
cations during hospitalization due to several factors. Sepsis seems to
be a common complication in the postoperative period, and prompt
recognition combined with early interventions is an e ective way of
reducing mortality in this condition.
Objective To evaluate risk factors for sepsis in surgical patients
admitted to the intensive care unit (ICU).
Methods Prospective data collected from a cohort of surgical patients
from January 2005 to December 2007. We analyzed the incidence of
sepsis and certain variables from the preoperative, intraoperative and
postoperative period as risk factors for sepsis.
Results We studied 648 surgical patients. The mortality rate was 19.3%
and mean age was 53.2±18.8 years. The incidences of severe sepsis and
septic shock were 6.6% and 12.7%, respectively. Multivariate analysis
showed that the following variables were associated with sepsis:
urgent surgery (OR = 6.92, 95% CI = 4.34 to 11.03), emergency surgery
(OR = 5.36, 95% CI = 2.86 to 10.05), POSSUM physiologic variables (OR =
1.03, 95% CI = 1.01 to 1.06), POSSUM surgical variables (OR = 1.09, 95%
CI = 1.05 to 1.13), mechanical ventilation (OR = 7.20, 95% CI = 3.78 to
13.71) and Sequential Organ Failure Assessment at ICU admission (OR=
1.13, 95% CI = 1.05 to 1.22).
Conclusion The present study detected a high incidence of infectious
complications in surgical patients that resulted in high mortality rates.
Risk factors associated with sepsis during the perioperative period were
easily detectable and knowledge of these can be useful for prevention
strategies and early identi cation of complications.
Lactate and base de cit are predictors of mortality in critically ill
patients with cancer
LA Hajjar, JL Vincent, FRBG Galas, JP Almeida, FB Jatene, PC Bueno,
JTFukushima, RE Nakamura, CM Silva, R Kalil Filho, JOC Auler Jr
ICESP – FMUSP, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P18 (doi: 10.1186/cc10166)
Objective Cancer patients frequently require admission to the intensive
care unit (ICU); however, there are few data regarding predictive factors
for mortality. The aim of this study was to evaluate whether arterial
lactate or standard base de cit (SBD) on admission and after 24 hours
can predict ICU and hospital mortality for patients with cancer.
Figure 1 (abstract P15). Presence of diabetes or glucose intolerance
Critical Care 2011, Volume 15 Suppl 2
Methods We evaluated 1,129 patients with severe sepsis, septic shock,
or postoperative after high-risk surgery. Lactate and SBD collected
at admission and after 24 hours were compared between survivors
and nonsurvivors. We evaluated whether arterial lactate and SBD are
independent predictors of ICU and hospital mortality.
Results There were 854 hospital survivors (76.5%). Twenty-four-hour
lactate >1.9 mmol/l (OR = 4.02, CI = 2.7 to 5.97) and SBD <–2.3 (OR =
2.4, CI = 1.64 to 3.52) were independent predictors of ICU mortality.
Twenty-four-hour lactate >1.9 mmol/l (HR = 2.63, CI = 1.99 to 3.47)
and 24-hour SBD <–2.3 mmol/l (HR = 1.74, CI = 1.33 to 2.27) were
independent predictors of hospital death.
Conclusion Our  ndings suggest that lactate and SBD measurement
should be included in the routine assessment of patients with cancer
admitted to the ICU. These markers may be useful in the adequate
allocation of resources in this population.
Plasma levels of IL-6 and IL-10 in septic patients at admission and
during follow-up and association with clinical outcomes
RT Costa
, MKC Brunialti
, F Machado
, E Silva
, O Rigato
, R Salomão
Division of Infectious Diseases, Escola Paulista de Medicina, Federal University
of São Paulo, São Paulo – SP, Brazil;
Hospital Sírio-Libanês, São Paulo – SP,
Discipline of Anesthesiology, Escola Paulista de Medicina, Federal
University of São Paulo, São Paulo – SP, Brazil;
Israelita Albert Einstein
Hospital, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P19 (doi: 10.1186/cc10167)
Introduction Sepsis is a systemic in ammatory syndrome triggered by
infection. It has been recognized that a dynamic interaction between
proin ammatory and anti-in ammatory response is present in this
syndrome, which is balanced by as yet unknown mechanisms. We and
others showed that in ammatory cytokines are upregulated in the
early phase and downregulated in the late phases of sepsis, while anti-
in ammatory cytokines are preserved. However, there are few data
about the dynamics of these cytokines during follow-up of patients
and their relation with clinical outcome. The aim of this study was to
evaluate the plasma levels of a proin ammatory, IL-6, and an anti-
in ammatory, IL-10, cytokine in septic patients.
Methods This prospective study included 53 septic patients (SP) and 29
healthy volunteers (HV) as a control group. Patients were admitted to
the intensive care units of São Paulo, Sirio-Libanes and Israelita Albert
Einstein hospitals. Samples were collected during the  rst 48 hours of
organ dysfunction or sepsis (D0). A second sample was collected after
7 days from 35 SP (D7). The plasma levels of cytokines were measured
using the cytometric bead array method (limit detection 2.0 pg/ml) by
ow cytometry.
Results IL-6 and IL-10 plasma levels were higher in SP (median 170.8pg/
ml, range 3.53 to 16,028.52 pg/ml; and median 6.6 pg/ml, range 0.0 to
1,698.92 pg/ml, respectively) than HV (median 2.3 pg/ml, range 0.0 to
19.92 pg/ml for IL-6; and median 2.4 pg/ml, range 0.0 to 12.7 pg/ml for
IL-10) (P = 0.0001 and P = 0.007, respectively). Plasma levels of IL-6 and
IL-10 at D7 were not signi cant di erent from those at D0 (P = 0.85 and
P= 0.59, respectively). IL-6 and IL-10 admission plasma levels were higher
in nonsurvivors (median 284.76 pg/ml, range 9.16 to 16,028.52 pg/ml;
and median 17.6 pg/ml, range 0.0 to 1,698.92 pg/ml, respectively) than
in survivors (median 103.57, range 3.53 to 9,745.43pg/ml; and median
9.91 pg/ml, range 0.0 to 313 pg/ml; P=0.02 and P= 0.003, respectively).
Conclusion Our results show that both proin ammatory and anti-
in ammatory cytokines are detected during sepsis and a higher level
of both cytokines at admission is associated with worst outcomes.
Relevance of eosinopenia as an early sepsis marker
EB Moura, MO Maia, JA Araújo Neto, FF Amorim
Hospital Santa Luzia, Brasília – DF, Brazil
Critical Care 2011, 15(Suppl 2):P20 (doi: 10.1186/cc10168)
Introduction Early diagnosis of sepsis based on biomarker values has
been evaluated. However, there is no ideal marker for this purpose yet.
Objective To evaluate eosinopenia as an early sepsis marker.
Methods A retrospective study, on a 40-bed surgical–medical intensive
care unit (ICU). Data from 300 charts of patients consecutively admitted
(between January and March 2009) were collected. The patients were
classi ed as negative (no systemic in ammatory response syndrome
(SIRS)), SIRS, sepsis, severe sepsis or septic shock, according to the
criteria of the American College of Chest Physicians/Society of Critical
Care Medicine. Patients who died or were discharged within 24 hours
after admission, with previous hematological disease and those whose
data were incomplete were excluded from the study. We compared
the eosinophil cell count (hematology analyzer ABX Pentra DF 120;
Horiba Medical, Montpellier, France) on the day of admission to the
ICU between the non-infected group (negative and SIRS) and the
infected group (sepsis, severe sepsis and septic shock). The normality
of the distribution was tested by the Kolmogorov–Smirnov test and the
comparisons were made utilizing the Mann–Whitney test. Statistical
analyses were done utilizing SPSS 19 version.
Results Three hundred patients were admitted to the ICU in the period,
mean age 58.6 ± 20 years. The mean length of stay was 9.2 ± 15.7
days, the mean APACHE II score was 9.4±6.5. Eighteen patients were
excluded (one because of discharge within 24 hours; 11 patients
because of previous hematological disease; six because of incomplete
data). The remaining 282 patients were enrolled into the study,
classi ed as follows: negative (158 patients – 56%), SIRS (25 – 8.8%),
sepsis (44 – 15.6%), severe sepsis (23 – 8.2%) and septic shock (32 –
11.4%). At the time of admission, 99 (35.1%) patients had an infection.
The mean±SD eosinophil count was 167.6±131.5, 153.6±129 and
153.7±135.6 cells/mm
in the total, non-infected and infected groups,
respectively (P = 0.46; Figure 1). At a cut-o value of 100 cells/mm
, the
eosinophil count yielded a sensitivity of 35%, a speci city of 71%, a PPV
of 40% and a NPV of 66%.
Conclusion Eosinopenia was not a good early diagnostic marker for
sepsis in this population.
Th17 lymphocytes and alternatively activated monocytes are
upregulated in clinical sepsis
R Salomão, M Brunialti, M Santos, O Rigato, F Machado, E Silva
Escola Paulista de Medicina, UNIFESP, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P21 (doi: 10.1186/cc10169)
Introduction Sepsis is a systemic in ammatory response triggered
by infection. In ammatory response is modulated during sepsis and
Figure 1 (abstract P20). Eosinophil count comparison between non-
infected and infected patients.
Critical Care 2011, Volume 15 Suppl 2
upregulation and downregulation of cellular activity is observed,
depending on the cells and functions evaluated. Nevertheless, the
interaction of innate and adaptative immune responses has been little
studied in clinical sepsis.
Objective The aims of this study were to evaluate the presence of TCD4
lymphocytes Th1, Th17, regulatory (Treg) and alternatively activated
monocytes in septic patients and their association with prognosis.
Methods Septic patients were enrolled at admission (D0, n = 67) and
after 7 days of therapy (D7, n = 33). Thirty-two healthy volunteers
matched for age and gender were included as controls. PBMC were
obtained by the Ficoll gradient method. Th1 and Th17 lymphocytes
were identi ed by the intracellular detection of IFNγ and IL-17,
respectively, and Treg cells were identi ed by Foxp3
expression. Monocytes were evaluated for CD206 and
CD163 expression.
Results Spontaneous production of IFNγ and IL-17A was increased in
TCD4 cells of septic patients when compared with healthy volunteers.
After PMA/Io stimulation, the percentage of TCD4 lymphocytes
producing IFNγ was lower and IL-17 was higher in septic patients than
in healthy volunteers. The results based on absolute TCD4
counting showed a lower proportion of Th1 cells and double the
proportion of Th17 cells in septic patients compared with healthy
volunteers while the proportion of Treg remained unchanged. In
follow-up samples, a higher percentage of IFNγ and a lower percentage
of IL-17 producing cells were observed compared with D0 samples. A
higher percentage of spontaneously producing IFNγ was found in D7
compared with D0 samples from patients who died and a decreased
percentage of PMA/Io-induced IL-17 producing cells between patients’
samples of follow-up (D7) compared with admission samples was
found in survivors. Septic patients showed a markedly increased
proportion of alternatively activated monocytes, which was sustained
in both patients’ samples.
Conclusion We found a decreased proportion of Th1 and increased
proportion of Th17 in septic patients, and an impressive increase in
the percentage of monocytes expressing CD206 and CD163, indicating
di erentiation towards wound healing and regulatory or inhibitory
monocytes, which may underscore the previous studies showing a
reprogramming of monocytes’ function in sepsis.
Acknowledgments This work was supported by Fundação de Amparo
a Pesquisa de Estado de São Paulo (FAPESP – grants 2006/58744-1 and
Epidemiology of hospitalized pediatric bacterial sepsis in Brazil:
atrend analysis from 1992 to 2006
CMF Mangia, N Kissoon, OA Branchini, MC Andrade, BI Kopelman,
Universidade Federal de São Paulo, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P22 (doi: 10.1186/cc10170)
Objective To determine the epidemiology, costs and outcome of
hospitalized pediatric sepsis in Brazil (1992 to 2006) and to compare
mortality caused by sepsis with that caused by other major childhood
Methods We performed a population-based cohort study using a
government database of all hospitals a liated with the Brazilian health
system. We studied all hospitalizations in children from 28 days through
19 years with diagnosis of bacterial sepsis de ned by the criteria of the
International Classi cation of Diseases.
Results From 1992 through 2006, the pediatric hospital mortality rate
was 1.23%. There were 556,073 pediatric admissions with bacterial
sepsis, with a mean mortality rate of 19.9%. The incidence of sepsis
decreased 64% from 1992 to 2006 (P <0.001); however, the mortality
rate remained unchanged (from 1992 to 1996, 20.5%; and from 2002 to
2006, 19.7%). The sepsis hospital mortality rate was substantially higher
than pneumonia (0.5%), HIV (3.3%), diarrhea (0.3%), undernutrition
(2.3%), malaria (0.2%) and measles (0.7%). The Human Development
Index and mortality rates by region were: North region 0.76 and 21.7%;
Northeast region 0.72 and 27.1%; Central–West region 0.81 and 23.5%;
South region 0.83 and 12.2%; and Southeast region 0.82 and 14.8%,
Conclusion Sepsis remains an important health problem in children
in Brazil. The institution of universal primary care programs has been
associated with substantially reduced sepsis incidence and therefore
deaths; however, hospital mortality rates in children with sepsis remain
unchanged. Implementation of additional health initiatives to reduce
sepsis mortality in hospitalized patients could have great impact on
childhood mortality rates in Brazil.
Role of nurses in the early recognition of sepsis
P Padilha, B Almeida, BC Derico, FCM Elmiro, MFF Jesus, VL Sousa
Centro Universitário Ítalo-Brasileiro, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P23 (doi: 10.1186/cc10171)
Introduction Sepsis is considered one of the most challenging
diseases of all time [1]. During many years the concept of sepsis was not
the same inside the medical court, which resulted in a heterogeneous
population [2]. Its incidence has been growing dramatically over the
past decades, having advanced age of patients, increase of invasive
procedures, frequent use of immunosuppressive drugs and the increase
of infections caused by multiresistant bacteria as the main contributors
[3]. Nurses have an important role in early recognition of sepsis.
Objective We investigated whether nurses are able to early recognize
signals and symptoms of sepsis.
Methods The methodological strategy was quantitative, exploratory
and multicentric, with four small private hospitals involved. Thirty
nurses working for medical–surgical clinic, semi-intensive, intensive
and emergency units participated in the survey.
Results Only 23.3% of nurses considered variation in leukocytes, cardiac
and respiratory frequency and axillary temperature as classifying sepsis
clinical signals. In one case with sepsis signals, only 10% of answers were
correct. When trying to establish a di erentiation pattern among sepsis
stages, a new case was developed highlighting severe sepsis, showing
36.6% of right answers, making it clear that there is a confusion facing
this syndrome, where 10% chose sepsis, 26.6% septic shock and 26.6%
infection caused by a surgical wound. Only 30% of the nurses pointed
out that treatment is e ective within hours of its recognition. In a  nal
question 70% a rmed that it is important to recognize sepsis early.
Conclusion The study showed that there are di culties on the part
of nurses in recognition of sepsis. With the present results, it can be
concluded that the development of nursing care protocols with the early
recognition of sepsis signals by the nurse can help the patient’s recovery.
Training of nurses working in an ICU, and their team as a whole, can help
to reduce deaths in hospitals, improving the assistance and making
patients’ permanence in ICUs shorter, which can not only bene t patients
but can also lead to a reduction in costs for the institution.
1. Henkin SC, et al.: Sepsis: and actual vision. Sci Med 2009, 19:135-145.
2. Assunção MSC: Evaluation of medical knowledge on the concepts of
infection, syndrome of systemic in ammatory responses and sepsis in
their di erent ways of clinical presentation. Thesis, Universidade Federal de
São Paulo – Escola Paulista de Medicina, São Paulo; 2008.
3. Mazza BF, et al.: Impact of organic dysfunction period in the prognostic of
patients with severe sepsis and septic shock. Clinics 2008, 63:483-488.
Respiratory muscle weakness in acute heart failure patients
LHR Gonçalves, P Veríssimo, K Timenetsky, T Figueiredo, A Yang, T Andre,
M Nagano, C Alexandre, A Goedert, R Caserta, E Silva
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P24 (doi: 10.1186/cc10172)
Background Respiratory muscle weakness has been arbitrarily de ned
as a maximum inspiratory pressure lower than 70% of the predictive
value. Patients with chronic heart failure have 30 to 50% prevalence
of respiratory muscle weakness, and so far there is no evidence of this
prevalence in patients hospitalized with acute heart failure.
Critical Care 2011, Volume 15 Suppl 2
Objective To evaluate maximum inspiratory pressure and the preva-
lence of respiratory muscle weakness in hospitalized patients with
acute heart failure.
Methods A cohort study, performed at Hospital Israelita Albert
Einstein in acute heart failure patients admitted to our hospital. We
excluded patients with chronic pulmonary disease, neurological and
neuromuscular disorders, postoperative period and those that needed
an orotracheal tube. Patients after respiratory and hemodynamic
stability were submitted to a maximum inspiratory pressure (MIP)
measurement by a manuvacuometer. Measurement was performed
using a facial mask and unidirectional valve with the patient positioned
at 45°. We also collected demographic data, brain natriuretic peptide
hormone (BNP), ejection fraction estimated by echocardiogram and
use of non-invasive ventilation. MIP was measured at two moments,
the  rst measurement as soon as patients were clinically stable and the
second measurement before hospital discharge.
Results We evaluated 50 patients, with a mean age of 75 years (95%
CI= 72 to 78.8), mostly male patients (78%, 39 patients), mean ejection
fraction of 0.33 (95% CI = 0.31 to 0.35), and 93.5% had ejection fraction
lower than 0.45. At hospital admission, 24 patients used NIV (55.8%),
and the BNP median value was 726.5 pg/ml (range of 217 to 2,283 pg/
ml). The rst MIP measurement showed a median of –52 cmH
O (range
of –20 to –120 cmH
O), with 35 patients (70%) presenting MIP lower
than 70% of the predictive value. Time to the  rst measurement had
a median of 3.5 days (range of 1 to 22 days). At hospital discharge,
the median MIP was –53 cmH
O (range of –20 to –150 cmH
O), and
maintained 70% of patients with MIP lower than 70% of the predictive
value. There was no signi cant di erence between initial and hospital
discharge MIP (P = 0.806). Median hospital length of stay was 11 days
(range of 4 to 36 days).
Conclusion Hospitalized patients with acute heart failure have a high
prevalence of respiratory muscle weakness, and maintain weakness
even after clinical stabilization.
Three-dimensional and two-dimensional echocardiography
and biochemical analysis in patients with ST-segment elevation
myocardial infarction percutaneously treated: relationship between
LV function, remodeling and serum cardiac markers
MLC Vieira, WA Oliveira, AF Cury, A Cordovil, ACT Rodrigues, GNaccarato,
CG Mônaco, LPRV Costa, RB Romano, JR Calatróia, TR Afonso,
REUAzevedo, GMP Tavares, L Guimarães, EB Lira Filho, MA Perin,
CHFischer, SS Morhy
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P25 (doi: 10.1186/cc10173)
Introduction The prognosis of patients with acute myocardial infarc-
tion (MI) concerns multiple aspects that demonstrate myocardial
aggression (such as serum markers of cardiac damage), and also adap-
tative mechanisms relative to the acute event (ventricular remodeling).
Objective The aim of the study was to assess the relationship of serum
markers of cardiac damage and tridimensional echocardiographic (3D
Echo) parameters as well as echocardiographic bidimensional (2D
Echo) left ventricular ejection fraction (LVEF) in patients with acute ST-
elevation MI.
Methods A prospective study of 23 patients (17 males, mean age of
57 ± 13 years), with acute ST-elevation MI, primarily percutaneously
treated (stent). Serum cardiac markers (CK-MB, troponin I, myoglobin)
and serum BNP were compared with echocardiographic parameters
(volumes, LVEF, 3D dissynchrony index, 3D sphericity index (3D SPI)).
The 3D SPI is de ned as: LVEDV / (4/3π(D/2))
, where D is left ventricular
diastolic diameter on 4CH apical view. 3D SPI was compared with
a group of 20 normal volunteers (normal values: 0.29 ± 0.08). The
statistical analysis was performed using Pearsons correlation coe cient
(r), 95% CI, P <0.05, linear regression equation and Bland–Altman test.
Results 3D SPI ranged from 0.29 to 0.45 (0.35±0.0 8); 3D LVEF ranged
from 0.36 to 0.70 (0.50±0.06); 3D EDV ranged from 72 to 159 (100±27)
ml; 2D LVEF ranged from 0.40 to 0.71 (0.54±0.08); 2D EDV ranged from
57 to 165 (104±32) ml. Troponin I ranged from 2.3 to 33 (12.9±9) ng/
ml; CKMB ranged from 5.7 to 258 (94.4±78) ng/ml; BNP ranged from 25
to 1,058 (264±128) pg/ml. Pearsons correlation coe cient (r), relative
to 3D LVEF: 1 – BNP: r = –0.7427, P = 0.4800.
Conclusion In this series, stronger correlation was observed relative to
serum CK-MB, BNP and 3D Echo LVEF, when compared with 2D Echo
LVEF. We did not observe association concerning LV remodeling and
cardiac damage assessed by serum cardiac markers.
PK–PD correlation of anti-infective agents for dose adjustment in
one severe burn child with sepsis
EV Campos
, DS Gomez
, RP Azevedo
, A Despinoy
, MC Ferreira
, C Vieira Jr
, CS Giraud
, SRCJ Santos
Hospital das Clinicas da Faculdade de Medicina da Universidade de São
Paulo, São Paulo – SP, Brazil;
School of Pharmaceutical Sciences, University of
São Paulo, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P26 (doi: 10.1186/cc10174)
Introduction Altered pharmacokinetics in patients with major burns
may result in anti-infective plasma concentrations below those
required to be e ective against the common pathogens encountered
in burn patients. Altered  uid volumes and increased renal blood  ow
in these patients are the main factors responsible for pharmacokinetic
Table 1 (abstract P26). Pharmacokinetics and PK–PD correlation for six anti-infective agents for a burn child [AQ2]
Drug e cacy (%)
/MIC (mg/l)
(hours) CL (ml/minute/kg) Vd (l/kg)
Drug Follow-up periods Obtained Reference
Obtained Reference
Obtained Reference
Fluconazole 1 100% 100% 100% 22.10 27 to 37 0.27 0.20 to 0.34 0.53 0.50 to 0.70
8 mg/ml 16 mg/ml 32 mg/ml
Imipenem 3 100% 100% 100% 1.90 0.8 to 1.0 4.62 2.6 to 3.1 0.76 0.18 to 0.28
0.5 mg/ml 1 mg/ml 4 mg/ml
Linezolid 2 100% 50% 0% 3.45 4.5 to 5.4 3.22 1.14 to 2.08 0.95 0.57 to 0.86
1 mg/ml 2 mg/ml 4 mg/ml
Meropenem 8 100% 100% 62% 2.00 1.0 3.81 2.7 to 4.3 0.60 0.17 to 0.28
0.5 mg/ml 2 mg/ml 8 mg/ml
Sulphamethoxazole 2 50% 50% ND 19.65 7.5 to 12.7 0.74 0.24 to 0.38 1.34 0.22 to 0.30
32 mg/ml 64 mg/ml ND
Vancomycin 5 80% 60% 60% 2.10 5.0 to 11.0 1.46 1.3 to 1.5 0.30 0.33 to 0.45
1 mg/ml 2 mg/ml 4 mg/ml
Parameters PK–PD for in vivoin vitro correlation: AUC
0 to 24
/ MIC or %T >MIC.
Eucast, 2011.
Goodman and Gilman, 2006; Micromedex, 2010.
Critical Care 2011, Volume 15 Suppl 2
changes that require higher doses, reduction on time dose intervals or
both of them.
Objective Anti-infective plasma measurements in one burn patient
with sepsis to determine whether drug e cacy was achieved, thereby
improving the likelihood of infection control.
Methods A male burn child, 8 years old, 40 kg with severe thermal plus
inhalation injuries (petrol), 45% total burn surface area, was investi-
gated. He has received six anti-infective agents during the 88-day
period in the ICU. Drug plasma monitoring, pharmacokinetics and the
PK–PD correlation were done by blood sample collection, and drug
plasma measurements were performed by high-performance liquid
Results Since in burns pharmacokinetics is unpredictable for all agents
investigated, drug e cacy was based on PK–PD correlation (Table 1).
Dose adjustment was performed for vancomycin (from 0.5 g 6-hourly
to 1 g 8-hourly), meropenem (from 0.75 to 1 g 8-hourly) and linezolid
(from 0.3 to 0.6 g 12-hourly).
Conclusion PK–PD correlation was applied to investigate changes
on dose regimen to reach the e cacy for all anti-infective agents.
Dose adjustments were required only for vancomycin, linezolid, and
meropenem to guarantee drug e cacy.
Acknowledgements The authors are grateful to the Brazilian
Foundation CAPES, CNPq and FAPESP for  nancial support.
Preventing ventilator-associated pneumonia: a new methodology
for bed head control 24 x 7
G Vaz, RV Gonçalves
Hospital Quinta D’Or, Rio de Janeiro – RJ, Brazil
Critical Care 2011, 15(Suppl 2):P27 (doi: 10.1186/cc10175)
Background Among the measures for preventing ventilator-associated
pneumonia (VAP) in patients at risk, strict control of the bed head
above 30° stands as the single one with better cost bene t [1]. While
the semi-recumbent position is intended to be an inexpensive and
easily performed action by the intensive care unit team, the smart beds
currently available are not the reality for the vast majority of hospitals
around the world because of the high cost. Therefore, the simple
theoretical principles for its execution are contradicted by its di cult
practical application.
Objective We propose a new methodology for continuous control of
the bed head, thus making possible the appropriate compliance to the
semi-recumbent position, seeking a reduction in the VAP rates.
Methods A retrospective observational study with 41 mechanically
ventilated patients over a 7-month period starting in May 2010, in a
neurointensive critical care unit of a private tertiary hospital. There
was a historical control as reference during 3 months before the
intervention made in August, and measurements for the same time
after it as a means to con rm its appropriate implementation, based
on the National Nosocomial Infections Surveillance System (NNISS) as
a parameter. Applied was a technique for an hourly basis positioning of
the head of bed angle in such a manner that it never remained below
30º for over 1 hour in the 24 hours daily. It was turned into a mandatory
item in the prescription and its execution was performed by the nursing
sta , through reading of a speci c angulation marking adhesive in the
side head rail, and annotation in the usual sheet for recording the vital
signs, followed by the prompt adjustment to the right position. Other
items of the institutional bundle of VAP were not modi ed.
Results There was a trend towards reduction in the ventilator-
associated respiratory infection rate (Figure 1) after the implementation
of the methodology, bringing it to zero despite the elevation in device
utilization (Figure 2).
Conclusion This unsophisticated and low-cost method for controlling
heads of beds in an intensive care unit allowed its adequate employment,
thus seeming to cause an impact in the incidence of VAP when comparing
respiratory infection rate and device utilization, despite limitations about
the small case series and the short following period.
1. Drakulovic MB, Torres A, Bauer TT, Nicolas JM, Nogué S, Ferrer M: Supine
body position as a risk factor for nosocomial pneumonia in mechanically
ventilated patients: a randomised trial. Lancet 1999; 354:1851-1858.
Ventilator-associated pneumonia: microbiological pro le and
G Castro, JAT Lisboa, JIV Carvalho, JN Bacelar, MM Ferreira, CLVM Miranda,
ACP Carvalho, KRS Cruz, RM Salazar
Centro Universitário do Maranhão (UNICEUMA), São Luis – MA, Brazil
Critical Care 2011, 15(Suppl 2):P28 (doi: 10.1186/cc10176)
Background Ventilator-associated pneumonia (VAP) is the second
most frequent infection in American intensive care units (ICUs) and
the most frequent in European ICUs, and its incidence and mortality
rates are still high, despite the continuous advances in diagnosis and
treatment techniques. Although its multiple etiologies and complex
diagnosis breed divergence about its management approach.
Objective To evaluate the microbiological pro le of patients with VAP
admitted to the ICUs of two hospitals in São Luís – MA.
Methods A descriptive, analytic, retrospective study, with 1,072
patients admitted to ICUs of the hospitals Dr Carlos Macieira and
Centro Médico Maranhense between January 2008 and December
2009. The patients were strati ed by age, sex, infection type, identi ed
pathogens and ICU stay outcome. Data were analyzed by the software
Epi Info® (version 3.5.1; 2008) and so was calculated the chi-square (χ
nonparametric test, with 5% signi cance level adopted.
Results It was veri ed that 31.6% of the patients had a polymicrobial
infection and 68.4% acquired infection by monobacteria. Gram-
negative bacilli showed up as the most common pathogens overall. The
multidrug-resistant bacteria incidence was 51.3% and its correlation
with VAP mortality and the means of days under mechanical ventilation
of infected patients did not present statistical signi cance respectively.
Conclusion VAP has been pointed out as a manifold etiology disease,
with high morbi-mortality indexes that do not change according to the
etiologic agents.
1. Torres A, Ewig S, Lode H, Carlet J: De ning, treating and preventing hospital
acquired pneumonia: European perspective. Intensive Care Med 2009,
2. Tejerina E, Esteban A, Fernández-Segoviano P, Frutos-Vivar F, Aramburu J,
Ballesteros D, Rodríguez-Barbero JM: Accuracy of clinical de nitions of
ventilator-associated pneumonia: comparison with autopsy  ndings. J Crit
Care 2010, 25:62-68.
3. Combes A, Figlioni C, Troillet J-L, Kassis N, Wolf M, Gilbert C, Chastre J:
Incidence and outcome of polymicrobial ventilator-associated
pneumonia. Chest 2002, 121:1618-1623.
Figure 1 (abstract P27). Ventilator-associated respiratory infection rate
per 1,000 days of mechanical ventilation.
Figure 2 (abstract P27). Device utilization for mechanical ventilation.
Critical Care 2011, Volume 15 Suppl 2
Analysis of the outcome of mechanical ventilation in patients with
acute renal failure
JAA Neto, RB Fernandes, FB Lima, DA Castro, EB Moura, MO Maia
Hospital Santa Luzia, Brasilia – DF, Brazil
Critical Care 2011, 15(Suppl 2):P29 (doi: 10.1186/cc10177)
Background Mechanical ventilation (MV) is a factor that may induce
or worsen lung injury and also contribute to the failure of other organs.
An early manifestation of multiple organ failure in the ICU is acute
renal failure (ARF), with a prevalence ranging from 4 to 16%, which is
associated with increased rates of mortality.
Objective The aim of this study was to analyze the outcome of
mechanical ventilation in patients with ARF in the ICU.
Methods This is a retrospective and analytical study that included
patients aged >18 years, hospitalized in the ICU of HSL under MV for
more than 24 hours, from June 2009 to June 2010. Patients with chronic
renal failure were excluded. The AKIN criteria were used to stratify
patients into three groups: non-ARF, ARF and dialysis ARF. The variables
analyzed were age, gender, APACHE II, length in ICU, length of MV, MV
outcome and mortality. Statistical analysis used chi-square and ANOVA,
with a signi cance level of 5%.
Results The sample consisted of 131 patients, 51.1% women, mean age
65.6±20.0 years. According to the criterion AKIN, 69.5% of patients had
ARF, dialysis was 31.9%. APACHE II was higher in ARF (17.6±7.7) and IRA
dialysis (18.6± 11.0), compared with the group non-ARF (13.2± 7.7),
P= 0.01. The ICU stay was similar between groups (non-ARF 21.8±32.5
days; ARF 20.8±19.9 days; dialysis ARF 27.1±23.4 days, P = 0.53). The
duration of MV was higher in the dialysis ARF (non-ARF 5.5±4.7 days;
ARF 6.9±7.6 days; dialysis ARF 14.2±15.1, P <0.01). See Table 1 and
Figures 1 and 2.
Conclusions In the sample studied, we observed a high prevalence of
MV and ARF, and the presence of renal failure is associated with a lower
success rate of weaning and higher mortality.
1. Slutsky AS, Tremblay LN: Multiple system organ failure. Is mechanical
ventilation a contributing factor? Am J Respir Crit Care Med 1998,
2. Ranieri M, Giunta F, Suter PM, et al.: Mechanical ventilation as a mediator of
multisystem organ failure in acute respiratory distress syndrome. JAMA
2000, 284:43-44.
Fluid removal in critically ill patients during hemodialysis: is there a
role for functional hemodynamic monitoring?
RH Passos, PB Batista
Hospital São Rafael, Salvador – BA, Brazil
Critical Care 2011, 15(Suppl 2):P30 (doi: 10.1186/cc10178)
Introduction Renal replacement therapy is frequently required
in critically ill patients with acute kidney injury. With intermittent
hemodialysis, large volumes of  uid need to be removed over a
relatively short period of time, jeopardizing hemodynamic stability
in already hemodynamically compromised patients. Established
methods of dry weight estimation are not practical in critical care
and the estimation of excess body  uid removable by hemodialysis
constitutes a particular change in these patients. Dynamic parameters
of  uid responsiveness are increasingly being used to guide  uid
therapy in critical care, but their suitability to monitor  uid removal
with hemodialysis is not known.
Objective The aim of our study was to analyze changes in a dynamic
parameter of  uid responsiveness (pulse pressure variation) in critically
ill patients submitted to intermittent hemodialysis.
Methods Changes in pulse pressure variation, central venous pressure,
median arterial pressure, and cardiac index were analyzed every hour
over intermittent hemodynamics using a minimally hemodynamic
monitoring device (LIDCO plus) in 28 mechanically ventilated patients.
Additional measurements of lactate and central venous saturation
were measured at the same time.
Results Median dialysis duration was 4.5 hours, and a median of
2,900 ml  uid was removed. There were 102 hypotensive episodes.
The median arterial blood pressure was 72 mmHg. Median CVP was
16±6 and pulse pressure variation was 9±6 just before hemodialysis.
There was a signi cant increase in the pulse pressure variation over
the dialysis treatment (15±4) and a decrease in the CVP value (13±6).
Comparing the group of patients already  uid responsive (Pp >13%)
just before the start of hemodialysis with the group non uid reponsive
(Pp <13%), the median values of lactate (2.1 x 1.9, P = 0.78) and central
venous saturation (0.74 x 0.72, P = 0.94) were not signi cantly di erent,
but at the end of the procedure a signi cant di erence in lactate was
observed (4.2 x 2.5, P <0.2).
Conclusion In our study the rate of ultra ltration during hemodialysis
was re ected by the changes in the pulse pressure variation. In
patients already  uid responsive (Pp >13%) just before hemodialysis,
the impact of  uid removal at the end of the procedure in perfusion
parameters was signi cantly higher. Dynamic parameters of volemia
could be useful to guide  uid removal and avoid hypoperfusion in acute
renal failure patients mechanically ventilated during hemodialysis
Table 1 (abstract P29). Characteristics of subjects with AKIN criteria
No ARF ARF Dialysis ARF
(n = 40) (n = 62) (n = 29) P value
Age (years) 57.7±20.1 70.3±18.9 66.3±19.1 <0.01
APACHE II 13.2±7.7 17.6±7.7 18.6±11.0 0.01
SAPS II 40.0±14.6 45.6±12.6 44.3±16.4 0.14
Length in ICU (days) 21.8±32.5 20.8±19.9 27.1±23.4 0.53
Length of stay (days) 28.3±36.2 25.2±22.8 29.6±24.0 0.74
Duration of MV (hours) 131.8±112.1 166.7±182.0 341.6±363.4 <0.001
Figure 1 (abstract P29). Distribution of patients who progressed or not
to wean from MV. *P <0.01.
Figure 2 (abstract P29). Outcomes of weaning. *P <0.01.
Critical Care 2011, Volume 15 Suppl 2
Impact of positive  uid balance on survival in critically ill cancer
JP Almeida, H Palomba, L Hajjar, V Torres, F Galas, FA Duarte, D Nagaoka,
RE Nakamura, J Fukushima, L Yu
ICESP, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P31 (doi: 10.1186/cc10179)
Introduction Fluid overload has recently been linked to adverse
outcomes in critically ill patients, but its impact on the outcomes of
cancer patients admitted to intensive care units (ICUs) has not been
previously described.
Methods A total of 234 cancer patients admitted to the medical
ICU in a 6-month period were prospectively evaluated for survival.
Univariate and multivariate analyses were used to study ICU admission
parameters associated with ICU mortality. Exclusion criteria were ICU
stay <24 hours and chronic renal failure on dialysis.
Results Overall mortality was 21%. The mean age of all patients was
62.7 ± 11.6 years and 55% were male. Postoperative care (45%) and
sepsis (35%) were the major reasons for admission to the ICU. The
mean APACHE II score value at 24-hour ICU was 21.2±6.4 and the mean
Karnofsky score before ICU admission was 75.2±17.2. At multivariate
analysis, the following variables at ICU admission were signi cantly
associated with ICU mortality in cancer patients: Lung Injury Score >2
(OR = 3.3; 95% CI = 1.32 to 8.23) and positive  uid balance (for each
100ml/24 hours) (OR = 1.03; 95% CI = 1.01 to 1.06).
Conclusions Fluid overload is independently associated with increased
mortality in critically ill cancer patients. Further studies are necessary
to determine the impact of positive  uid balance on acute organ
dysfunction and overall prognosis of cancer patients.
Previous renal support is a predictor for chronic renal replacement
therapy after orthotopic liver transplantation
MCC Andreoli, MPV Coelho, ACC Matos, ÉB Rangel, NKG Souza, MÂ Góes,
AL Ammirati, TN Matsui, IJ Iizuca, FD Carneiro, ACMS Ramos, MA Souza,
RC Afonso, B Ferraz-Neto, MS Durão, MC Batista, JCM Monte, VG Pereira,
OFP Santos, BC Santos
Einstein Dialysis Center, Albert Einstein Jewish Hospital, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P32 (doi: 10.1186/cc10180)
Background In the Model for End-Stage Liver Disease (MELD) era of
organ allocation, renal replacement therapy (RRT) has been done in
many liver transplant patients. In this setting the time and probability
of kidney function recovery is essential for patient and transplant
program management.
Methods In this study we evaluated a sample of stable post-intensive
care dialysis patients from a group of 297 adults who were submitted
to orthotopic liver transplantation (OLT) in an urban tertiary medical
center from 1 June 2005 to 31 December 2009. We evaluated the
average time of renal function recovery (out of need for RRT) in OLT
patients on post-intensive care hemodialysis (HD) and determined risk
factors for chronic dialysis support during a 1-year follow-up period.
Patients were censored at recovery of kidney function, death on HD or
end of the follow-up period. The Cox proportional hazards model was
used to compare the relative risk (RR) of remaining or not in HD after
1year and predictor variables.
Results We evaluated the clinical records of 83 patients (50±14 years,
64% male, 22% pre-OLT diabetes mellitus (DM), 31% HCV-related
disease, MELD 27.5± 11.8, 17% acute re-OLT, 37% pre-OLT RRT, pre-
OLT serum creatinine 1.5±1.4 mg/dl, 28% pre-OLT proteinuria). During
the study period, 70 (84%) patients were removed from dialysis;
of these, six (7%) remained on HD for more than 90 days until renal
function recovery, 184 days being the longest period required. Nine
(11%) patients died on HD and only four (5%) patients were on HD after
1year. The median of recovery time was 28 days (from 6 to 184 days).
Classic risk factors for renal disease, like age and DM, acute re-OLT
requirement and pre-OLT RRT, were signi cant predictors of chronic
RRT. In the multivariate analysis, the most important prognostic factor
for chronic RRT was the presence of pre-OLT RRT (HR = 1.89, 95% CI =
1.145 to 3.129, P = 0.013).
Conclusion Given the shortage of available organs, kidney trans plan-
tation after or concomitant to OLT must be considered cautiously,
especially in OLT patients who were not submitted to pre-OLT RRT.
Prognostic factors for acute kidney injury development in critically
ill cancer patients
LA Hajjar, H Palomba, J Almeida, J Fukushima, RE Nakamura, F Galas,
VTorres, R Kalil Filho, L Yu
ICESP, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P33 (doi: 10.1186/cc10181)
Introduction Acute kidney injury (AKI) in cancer patients is a
complication that causes substantial morbidity and mortality.
Methods A total of 1,500 cancer patients admitted to the medical
intensive care unit (ICU) between November 2008 and February–March
2011 were evaluated for AKI, de ned as an increase in serum creatinine
(SCr) >0.3 mg/dl over the baseline value, according to the AKIN stage
I de nition. Univariate analysis was used to study ICU admission
parameters associated with AKI occurrence during the ICU stay.
Results AKI incidence was 31%, with a mortality rate of 42%, compared
with 20% for non-AKI patients. The mean age of all patients was
63.1±11.3 years and 55% were male. Sepsis (44.8%) and respiratory
failure (24.8%) were the major reasons for admission to the ICU. At
univariate analysis, the following variables at ICU admission were
signi cantly associated with AKI in cancer patients during the ICU stay:
need for vasopressors (74.3% vs. 25.7%; P = 0.004), serum potassium
(4.2, 3.6 to 4.6 mEq/l vs. 3.8, 3.5 to 4.2 mEq/l; P = 0.006), serum pH (7.35,
7.3 to 7.39 vs. 7.39, 7.34 to 7.42; P = 0.006), base excess (–5.5, –9.2 to
–1.8 vs. –2, –5 to 0.1; P = 0.003), serum phosphorus (3.9, 3.4 to 4.6
mg/dl vs. 2.9, 2.4 to 3.9 mg/dl; P = 0.0001), baseline serum creatinine
(1.2, 0.7 to 1.8 mg/dl vs. 0.6, 0.4 to 0.8 mg/dl; P = 0.01). At multivariate
analysis, the following variables at ICU admission were associated with
AKI: serum creatinine >1.0 mg/dl (OR = 9.2; 95% CI = 2.3 to 35.8), pH
<7.38 (OR = 5.1; 95% CI = 1.6 to 15.6) and need for vasopressors in
the  rst 24 hours (OR = 3.4; 95% CI = 1.2 to 9.6). Variables previously
thought to be indicative of a poor prognosis (advanced age, metastatic
or progressive disease, recent chemotherapy and performance status)
were not predictive of AKI.
Conclusions AKI is frequent in critically ill cancer patients and has a
great impact on mortality. AKI incidence can be better estimated by an
evaluation of the acute organ dysfunction at ICU admission than by the
characteristics of the underlying malignancy.
Serum soluble-Fas, in ammation and anemia in acute renal failure
and critical illness
MA Góes
, MA Dalboni
, BMR Quinto
, IJ Iizuka
, JC Monte
, OFPavão
dos Santos
, VG Pereira
, M de Souza Durão Jr
, MC Batista
Federal University of São Paulo – UNIFESP, São Paulo – SP, Brazil;
Israelita Albert Einstein, São Paulo – SP, Brazil;
New England Medical Center,
Tufts School of Medicine, Boston, MA, USA
Critical Care 2011, 15(Suppl 2):P34 (doi: 10.1186/cc10182)
Introduction Soluble Fas (sFas) levels are associated with anemia and
erythropoietin (Epo) hyporesponsiveness in chronic kidney disease.
Anemia is also a common feature in patients with acute renal failure
(ARF) and in critically ill patients. Therefore, it is possible that sFas levels
are also associated with anemia and increased need for serum Epo
levels in ARF and critical illness in order to maintain hemoglobin (Hgb)
Objective To investigate the relationship between serum levels of sFas,
Epo, in ammatory cytokines and Hgb levels in patients with ARF and
critically ill patients.
Methods We studied 72 critically ill patients with ARF on continuous
hemodia ltration (CVVHDF group; n = 53) or without ARF (non-ARF
group; n = 19), 29 chronic hemodialysis patients (ESRD group) and
29 healthy volunteers (Healthy group). The CVVHDF dose was 30 ml/
kg per hour or higher. We investigated among the four groups the
Critical Care 2011, Volume 15 Suppl 2
relationships between Hgb and serum levels of sFas, Epo, TNFα, IL-6,
IL-10 and iron status.
Results The CVVHDF and non-ARF groups had higher serum levels of
Epo, IL-6, IL-10 and ferritin than the other groups. Hgb levels were lower
in the CVVHDF group than in the other groups. Serum sFas levels were
higher in uremic patients (CVVHDF and ESRD groups; P<0.001). When
all critically ill patients were pooled together, Hgb levels correlated
negatively with serum levels of IL-6 (r=–0.55, P=0.001), sFas (r=–0.40,
P=0.001), TNFα (r=–0.37, P<0.001), iron (r = –0.28, P = 0.02), ferritin
(r = –0.35, P = 0.004) and transferrin saturation (r = –0.30, P = 0.01).
In multivariate analysis, after adjusting for markers of iron store
and in ammation, levels of IL-6 (P<0.001), sFas (P<0.001) and TNFα
(P=0.01) correlated negatively with Hgb in critically ill patients.
Conclusion Our  ndings demonstrate that sFas is associated with
anemia in ARF and critically ill patients. Serum sFas and Epo levels
were higher and Hgb levels were lower in critically ill patients with ARF,
suggesting that sFas may be associated with Epo hyporesponsiveness
in ARF and critical illness.
Applying a new weaning index in ICU older patients
LM Azeredo, SN Nemer, JB Caldeira, B Guimarães, R Noé, CP Caldas,
Hospital de Clínicas de Niterói, Niterói – RJ, Brazil
Critical Care 2011, 15(Suppl 2):P35 (doi: 10.1186/cc10183)
Introduction With the increase in life expectation, more admissions to
hospital, use of mechanical ventilation (MV) and weaning trials in older
patients have been observed.
Objective To evaluate the variables associated with successful weaning
from mechanical ventilation in older patients.
Methods We evaluated a cohort from September 2004 to January 2008
with 479 patients. We excluded one patient aged under 18 years, 35
tracheostomized and 112 with neurologic diseases, resulting in 331
patients. Besides the conventional weaning indexes, we evaluated the
performance of a new integrative weaning index (IWI). The study was
approved by the Ethics Committee of Pedro Ernesto University Hospital
(2206-CEP). The chances of successful weaning were investigated using
relative risk and logistic regression. The Hosmer–Lemeshow goodness-
of- t test was used to calibrate and the C statistic was calculated to
evaluate the association between predicted probabilities and observed
proportions in the logistic regression model.
Results Prevalence of successful weaning in the sample was 83.7%.
There was no di erence in mortality of older and nonolder patients
(P = 0.16), in the days of mechanical ventilation (P = 0.22) and days
of weaning (P = 0.55). In older patients, the IWI was the only variable
associated with respiratory weaning in this population (P <0.0001). See
Tables 1 to 5.
Table 1 (abstract P35). Etiology and population
Etiology Population (%)
COPD, n (%) 98 (29.6)
Pneumonia, n (%) 68 (20.54)
Postoperative, n (%) 63 (19.03)
Sepsis, n (%) 39 (11.78)
ARDS/ALI, n (%) 25 (7.55)
Trauma without brain injury, n (%) 11 (3.32)
Acute pulmonary edema, n (%) 10 (3.02)
Miscellaneous, n (%) 17 (5.13)
Total 331
ALI, acute lung injury; ARDS, acute respiratory distress syndrome; COPD, chronic
obstructive pulmonary disease.
Table 2 (abstract P35). Analysis of the outcome variables by sample and by age
Variable Category Total (%) Age ≥70 (%) Age <70 (%) P value
Result Success 277 (83.7) 125 (80.7) 152 (86.4) 0.16
Failure 54 (16.3) 30 (19.4) 24 (13.6)
Death 17 (5.1) 15 (9.7) 2 (1.1) 0.002
Evolution Discharge 277 (83.7) 125 (80.7) 152 (86.4)
Return 37 (11.2) 15 (9.7) 22 (12.5)
Days of MV Mean±DP/median 9.1±7.6/ 7 9.2±8.6/6 9.1±6.7/7 0.22
Days of weaning Mean±DP/median 2.7±2.3/2 2.8±2.6/2 2.6±2.0/2 0.55
APACHE II Mean±DP/median 16.0±5.6/15 16.9±5.8/16 15.3±5.2/14 0.009
Signi cant P <0.05.
Table 3 (abstract P35). Analysis of the respiratory variables according to the results by age
Age ≥70 Age <70
Variable Success (%) Failure (%) RR 95% CI Success (%) Failure (%) RR 95% CI
P/F ≥255 59.2 46.7 1.11 0.94 to 1.30 63.2 16.7 1.30 1.13 to 1.50 ≥30 80.8 23.3 1.83 1.38 to 2.43 84.2 29.2 1.62 1.25 to 2.10
IWI ≥25.5 97.6 3.3 10.6 3.60 to 31.1 96.1 8.3 4.60 2.26 to 9.36
P 0.1 ≤3.1 78.4 36.7 1.53 1.19 to 1.97 77.0 20.8 1.48 1.21 to 1.81
f ≤29 72.8 33.3 1.43 1.16 to 1.77 68.4 20.8 1.33 1.14 to 1.56
Vt ≥320 76.8 23.3 1.67 1.31 to 2.14 73.7 29.2 1.34 1.13 to 1.60
f/Vt*P 0.1 ≤270 80.0 33.3 1.64 1.25 to 2.14 77.0 16.7 1.52 1.24 to 1.86
f/Vt ≤100 81.6 23.3 1.87 1.40 to 2.51 78.3 25.0 1.47 1.20 to 1.81
RR, relative risk. Signi cant P <0.0001, except for P/F on age ≥70.
Critical Care 2011, Volume 15 Suppl 2
Table 4 (abstract P35). Logistic regression to the success of weaning by age
Signi cant
Age variable Coe cient SE P value RR 95% CI
≥70 Intercept –2.2687 0.606 0.0002
IWI ≥25.5 7.0727 1.173 <0.0001 1,179.3 118 to 11,752
<70 Intercept –3.1147 1.035 0.003
IWI ≥25.5 6.0547 1.187 <0.0001 426.1 41.6 to 4,364
APACHE ≤17 3.4249 1.159 0.003 30.7 3.2 to 298
CI, con dence interval; RR, relative risk; SE, standard error of coe cient.
Table 5 (abstract P35). Estimated probability of success according to the
logistic model by age group
Age IWI ≥25.5 APACHE ≤17 probability (%) 95% CI
≥70 No 9.4 3.06 to 25.4
Yes 99.2 94.5 to 99.9
<70 No No 4.3 0.58 to 25.2
No Yes 57.7 27.5 to 83.1
Yes Yes 95.0 81.8 to 98.8
Yes No 99.8 98.0 to 100.0
CI, con dence interval: lower limit (%) to upper limit (%).
Conclusion The IWI was the main independent variable in weaning
of the older patient population, and it can contribute to this critical
CPAP with variable  ow is comparable with Bubble CPAP in preterm
CM Rebello, ACZ Yagui, LA Vale, LB Haddad, C Prado, FS Rossi, AD Deutsch
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P36 (doi: 10.1186/cc10184)
Background nCPAP has many bene ts to treat respiratory distress in the
newborn. It has been shown that in devices with variable  ow, nCPAP
reduces work of breathing and increases lung recruitment, compared
with continuous  ow; nevertheless, there are few randomized
controlled trials comparing these di erent CPAP apparatuses regarding
respiratory outcomes.
Objective To evaluate the e cacy of nasal CPAP using a device with
variable  ow or Bubble CPAP, regarding CPAP failure, occurrence of air
leaks, total CPAP time and main complications of prematurity.
Methods A randomized clinical trial. Newborns admitted to the
Hospital Israelita Albert Einstein’s NICU (São Paulo, Brazil) with birth
weight ≥1,000 g, without previous mechanical ventilation and with
respiratory distress requiring nCPAP were randomized into two
study groups: Variable Flow (Servo-I; Siemens Elema Inc., Sweden) or
Bubble CPAP (Fisher and Paykel Healthcare, Auckland, New Zealand).
Both groups used the same interface (BC 161; Fisher and Paykel
Healthcare), with a target SatO
of 88 to 94%. Gestational age, birth
weight, Apgar 5 minutes, diagnosis of respiratory distress, CPAP failure,
the main complications of prematurity and total CPAP and oxygen
time were recorded. Continuous variables were analyzed by Student
t test, categorical variables were analyzed by Fisher’s exact test. The
signi cance level was set at P = 0.05.
Results A total of 40 infants were randomized. One baby was excluded
from the Variable Flow Group because we were obligated to change
the nasal prong interface due to the development of nasal injury and
damage to the septal mucosa. There were no di erences between
groups regarding birth weight (Variable Flow: n = 19, 2,602 ± 585g;
Bubble CPAP: n = 20, 2,518 ± 598 g; P = 0.663); gestational age
(35.8 ± 0.5 weeks and 35.7 ± 0.4 weeks; P = 0.863); gender (male:
68.4% and 70.0%; P = 0.915); Apgar5 (9.4±0.6 and 9.6±1.0; P = 0.246);
prenatal steroids (31.6% and 10.0%; P = 0.127); time for CPAP installation
(120/90/203 minutes and 135/50/225 minutes; P = 0.978); CPAP failure
(21.1% and 20.0%; P = 1.000); air leak syndrome (10.5% and 5.0%; P =
0.605); total CPAP time (22.0/8.00/31.00 hours and 22.0/6.00/32.00
hours; P = 0.822); and total oxygen time (24.00/7.00/85.00 hours and
21.00/9.50/66.75 hours; P = 0.779). Values are mean±SD or percentage
or median/interquartile ranges (for time for CPAP installation, total
CPAP and total oxygen time).
Conclusion In this small randomized clinical trial the use of a device
with variable  ow was comparable with Bubble CPAP regarding the
occurrence of the main variables analyzed.
Estimated work of breathing in PAV-plus ventilation in ICU patients
LP Couto
, A Thompson
, F Gago
, R Sera m
, F Saddy
, CSV Barbas
Hospital das Clinicas da FMUSP, São Paulo – SP, Brazil;
Hospital Copa D’Or
Rio de Janeiro, Rio de Janeiro – RJ, Brazil
Critical Care 2011, 15(Suppl 2):P37 (doi: 10.1186/cc10185)
Background The purpose of the new PAV-plus ventilation is to guaran-
tee a better patient ventilator synchrony allowing the measurement of
respiratory system mechanics and the estimation of the patient work
of breathing.
Objective To verify whether ICU patients recovering from acute
respiratory failure can be maintained well in PAV-plus ventilation and if
the PAV-plus ventilatory mode can estimate respiratory mechanics and
work of breathing in ICU clinical practice.
Methods We studied 20 stable ICU patients that were recovering
from acute respiratory failure and could be ventilated comfortably in
pressure support of 15 cmH
O. After 20 minutes in PSV of 15 cmH
O we
measured the tidal volume, respiratory rate, minute ventilation, PaCO
and asked the patients to give a note from 0 to 10 on a visual comfort
scale. Then, we changed the patients to PAV-plus ventilation with
65% support and after 20 minutes we measured the same mentioned
parameters plus the respiratory system compliance, resistance and
the patients work of breathing. The same procedure was made after
changing the patients to PAV-plus ventilation of 50% support. We
established the association between the estimated work of breathing
by the ventilator and the measured respiratory parameters (P <0.05).
Results Twenty ICU patients recovering from acute respiratory failure
were studied, mean age 71.7 ± 9 years, 12 females. Mean minute
ventilation at 15 cmH
O of pressure support ventilation was 8.4±2.0
l/minute and mean PaCO
was 36.8 ± 5.96 mmHg. Mean minute
ventilation was maintained at 8.5±2.0 and 9.48±3.0 l/minute in PAV-
plus of 65% and 50%, respectively (P = NS). Mean PaCO
was 39±6.6
mmHg in PAV-plus of 65% and 40.65±6.8 mmHg in PAV-plus of 50%
(P = NS). During PAV-plus 65% the mean estimated patient work of
breathing was 0.3±0.1 J/l, and in PAV-plus 50% was 0.4±0.1 J/l (P =
Figure 1 (abstract P37). Correlation between resistance and WOB in
Critical Care 2011, Volume 15 Suppl 2
NS). Mean compliance during PAV-plus 65% was 53.8±20.4 and PAV-
plus 50% was 55.3 ± 18.8 (P = NS). Mean respiratory resistance was
9.6±4.2 in PAV-plus 65% and 8.7±3.3 in PAV-plus 50% (P = NS). Mean
comfort scale was 8.45±1.8 in PSV of 15 cmH
O and 8.1±1.4 in PAV-
plus 65% and 8.1±1.2 in PAV-plus 50% (P = NS). Patient’s estimated
work of breathing signi cantly associated with respiratory resistance
(P <0.0001; Figure 1) and inversely with respiratory compliance (P =
0.03; Figure 2) and was not associated with the comfort scale (P = 0.8),
minute ventilation (P = 0.5), PaCO
levels (P = 0.5), tidal volume (P = 0.3)
or respiratory rate (P = 0.8).
Conclusion ICU patients recovering from acute respiratory failure could
be maintained comfortably in PAV-plus ventilation of 65% and 50%
compared with PSV of 15 cmH
O and their estimated work of breathing
correlated negatively with patients compliance and positively with
patient’s resistance.
In uence of the equipment used for manual ventilation over the
variability of respiratory mechanics in rabbits
CM Rebello, R Lorenzetti, LB Haddad, LA Vale, RS Mascaretti, R Quinzani,
AD Deutsch
Hospital Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P38 (doi: 10.1186/cc10186)
Introduction A self-in ating bag is used for newborns’ manual venti-
lation, using an oxygen concentration close to 100%, unknown peak
inspiratory pressure (PIP), no end positive expiratory pressure (PEEP)
and high tidal-volume (Vt). Manual ventilation using a T-piece device
allows better control of PIP, use of PEEP and probably less variation in
pulmonary ventilation.
Objective To compare using an experimental model with the adult
rabbit, the variability of PIP, PEEP, Vt, minute ventilation (VMin),
respiratory rate (RR), inspiratory time (Tins), expiratory time (Tex) and
ratio Tins/Total time during manual ventilation using a self-in ating
bag or T-piece device.
Methods Adult New Zealand White rabbits were manually ventilated
by 21 individuals using a self-in ating bag (LIFESAVER® Neonate
Manual Resuscitator; Tele ex Medical, Research Triangle Park, NC, USA)
or a T-piece device (Babypu ®; Fanem Ltd, São Paulo, Brazil). Before
ventilation each animal was sedated with intramuscular ketamine–
acepromazine solution (10 mg/kg and 0.1 mg/kg, respectively) and
anesthetized (1% lidocaine, s.c.) at the site of incision for tracheostomy
and carotid cannulation. After curarization (pancuronium 1 mg/kg, i.v.)
the ventilation was started and ventilatory data (PIP, PEEP, Vt, minute
volume, inspiratory and expiratory time) were continuously recorded
until sacri ce with sodium pentobarbital (100 mg/kg, i.v.), after
10-minute ventilation. For each variable analyzed a variability index
was calculated, de ned as the standard deviation of the mean values
of each variable during the 10-minute ventilation. Statistical analysis
was by t test or Mann–Whitney test, signi cance was set at P = 0.05.
Table 1 (abstract P38)
T-piece (n = 21) Self-in ating-bag (n = 21) P value
Vt (ml/kg) 0.7±0.7 2.5±1.6 <0.001
VMin (ml/kg) 28.7±14.6 122.3±86.2 <0.001
PIP (cmH
O) 0.34±0.21 3.3±2.1 <0.001
O) 0.2±0.2 0.0±0.0 <0.001
RR (bpm) 2.6±1.3 9.9±25.6 0.087
Tins (seg) 0.12±0.05 0.13±0.23 <0.001
Tex (seg) 0.13±0.06 0.20±0.22 0.902
Tins/Tt 0.04±0.01 0.06±0.07 0.034
Data are mean±SD.
Results The variability indices for all variables analyzed during the
10-minute ventilation are shown in Table 1.
Conclusion The authors conclude that the use of a T-piece device
allows lower variability during manual ventilation, with the exception
of respiratory rate and expiratory time. We speculate that this lower
variability could result in lower lung injury during manual ventilation.
Maximum recruitment strategy revealed e ciency and a larger
recruitable lung in a prospective series of early ARDS patients
GFJ Matos, F Stanzani, RH Passos, MF Fontana, R Albaladejo, RE Caserta,
DCB Santos, JB Borges, MBP Amato, CSV Barbas
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P39 (doi: 10.1186/cc10187)
Introduction A recent meta-analysis demonstrated that higher levels
of PEEP were associated with improved survival among the subgroup
of patients with ARDS. The maximum recruitment strategy (MRS)
guided by thoracic CT scan is capable of reversing alveolar collapse
almost completely, allowing PEEP titration to sustain lungs almost fully
open, homogenizing tidal ventilation and possibly reducing ventilator-
induced lung injury.
Objective To test the e ciency, feasibility and side e ects of MRS; to
compare the amount of non-aerated tissue during MRS and calculate
lung recruitability.
Methods A case series report in a general medical/surgical private
and academic ICU with 42 beds at Albert Einstein Hospital, São Paulo,
Brazil. Fifty-one severe ARDS patients were included. Early and severe
ARDS patients were submitted to MRS guided by thoracic CT. The
protocol consisted of two parts: recruitment phase to calculate the
opening pressure (PEEP 10 to 45 cmH
O and constant driving pressure
O); PEEP titration phase (PEEP 25 to 10 cmH
O) to maintain the
lungs open. Patients were followed until hospital discharge or death.
Results Fifty-one severe ARDS patients were included and followed,
of whom 84% had primary ARDS. The median maximum recruitment
PEEP level was 45 (IQR: 43 to 45) cmH
O and the median maximum
recruitment plateau pressure was 60 (IQR: 58 to 60) cmH
O, and the
median titrated PEEP after MRS was 25 (IQR: 25 to 25) cmH
O. Median
global nonaerated parenchyma decreased signi cantly from 53.6%
(IQR: 42.5 to 62.4) to 12.7% (IQR: 4.9 to 24.2) (p
ratio increased
from 125 (IQR: 86 to 164) to 307 (IQR: 236 to 373)) (P<0.01).
Conclusion The MRS was an e cient, feasible and safe ventilatory
strategy to reverse nonaerated lung and hypoxemia in early and severe
ARDS patients with multiple organ failure, revealing a larger recruitable
lung. No major complications except for transitory changes in blood
pressure were noted.
Figure 2 (abstract P37). Correlation between compliance and WOB in
Critical Care 2011, Volume 15 Suppl 2
Mechanical ventilation pro le in an adult ICU in Brazil
CSV Barbas, C Saghabi, C Taniguchi, K Timenetsky, S Calegaro,
CSAAzevedo, TP Stuchi, RA Caserta, E Silva
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P40 (doi: 10.1186/cc10188)
Background Adult critically ill patients need invasive mechanical venti-
lation support due to distinct causes that vary from an elective high-
risk surgery to post cardiorespiratory arrest.
Objective To know the mechanical ventilation pro le in an adult
medical–surgical ICU in Brazil. To study adults patients that needed
more than 24 hours invasive mechanical ventilation support in an adult
ICU in Brazil.
Methods We analyzed all patients that needed more than 24 hours
invasive mechanical ventilatory support admitted to Albert Einstein
Adult medical–surgical 36-bed ICU from December 2008 to April 2010.
We studied patients age, sex, APACHE II score, cause of intubation/
mechanical ventilation, duration of ventilatory support, maximum
inspiratory pressure (MIP; mmHg), maximum expiratory pressure
(MEP; mmHg) and respiratory shallow breathing index (RSBI; l), rate of
extubation success and ratio of reintubation.
Results A total of 252 patients were studied, mean age 63± 19 years,
35% females, mean APACHE II score 22 ± 6. The main cause of
intubation/mechanical ventilation was acute hypoxemic respiratory
failure (35%) followed by depressed level of consciousness (34%),
post high-risk surgery (19%), airway obstruction (3%), hemodynamic
instability (5%) and respiratory fatigue (01%). The mean duration of
invasive mechanical ventilation was 114±4 hours (27 to 566 hours).
Before a spontaneous breathing trial to check readiness for extubation,
mean MIP was 48±12 (20 to 120) mmHg, mean MEP was 45±15 (12 to
120) mmHg and mean RR/TV (l) was 53±20 (5 to 190). The extubation
success rate was 87.3%. We used non-invasive ventilation immediately
after extubation in 66% of our patients. In total, 12.7% patients needed
reintubation. The hospital mortality rate was 8.75% (22 patients). There
were no di erences in regard to age, gender, mechanical ventilation
time, MIP, MEP, RSBI, use of non-invasive mechanical ventilation and
reintubation rate between patients that survived and those that died.
Conclusion In our ICU the main causes for invasive mechanical
ventilatory support were hypoxemic respiratory failure and post high-
risk surgery. The mean duration of invasive support was 4.7 days and
the reintubation rate was 12.7%.
Positive end-expiratory pressure can increase brain tissue oxygen
pressure in hypoxemic severe traumatic brain injury patients
SN Nemer, R Santos, J Caldeira, P Reis, B Guimarães, T Loureiro, R Ramos,
EFarias, D Prado, R Turon
Hospital de Clínicas de Niterói, Niterói – RJ, Brazil
Critical Care 2011, 15(Suppl 2):P41 (doi: 10.1186/cc10189)
Introduction Brain tissue oxygen pressure (PtiO
) re ects brain oxy-
gena tion and is a useful tool in traumatic brain injury (TBI) patients.
Increases in inspired oxygen fraction (FiO
) are related to improvement
on PbrO
, but other approaches that aim to improve oxygenation, like
increasing positive-end expiratory pressure (PEEP), were not deeply
evaluated in humans.
Objective The aim of this study was to evaluate the e ects of three
di erent PEEP levels on PbrO
of hypoxemic severe TBI patients.
Methods From February 2007 to February 2011, 36 severe TBI patients
admitted to our intensive neurological unit were monitored with PtiO
through the Licox device (Integra Neuroscience). Seventeen patients
remained in the study according to the following inclusion criteria:
ratio of arterial oxygen tension to fraction of inspired oxygen (PaO
ratio) <300; cerebral perfusion pressure (CPP) >60 mmHg; intracranial
pressure (ICP) <20 mmHg; PtiO
>20 mmHg; absence of any signal
of brain deterioration. These patients were submitted to PEEP levels
of 5, 10 and 15 cmH
O, each one for at least 20 minutes. During the
three PEEP levels, PtiO
, pulse oxygen saturation (SpO
), ICP and CPP
were monitored and statistically analyzed by ANOVA and Bonferroni
methods. P <0.05 was considered statistically signi cant.
Table 1 (abstract P41). Baseline characteristics of the evaluated patients
Baseline characteristic Mean SD
Age 28.6 8.4
APACHE II 19.2 3.2
Glasgow 6.1 0.9
55.9 11.8
ratio 154 46.6
27.7 6.5
ICP 8.3 4.4
CPP 94.8 8.2
95.5 2.1
APACHE II, Acute Physiology and Chronic Health Evaluation.
Results The increase of PEEP level from 5 to 15 cmH
O increased SpO
from 95.5 ± 2.1 to 98.6 ± 1.2 (P = 0.0001) and PtiO
from 27.8 ± 6.5
mmHg to 33.9 ± 6.7 mmHg, respectively (P = 0.0001). On the other
hand, ICP and CPP did not present statistical signi cance according to
the increase of PEEP levels (8.29±4.44 mmHg to 8.65±4.42 mmHg;
P= 0.14 and 94.8±8.2 to 94.6±8.0 mmHg; P = 0.78, respectively). The
main characteristics of the evaluated patients are described in Table 1.
Changes in PtiO
and CPP according to the PEEP levels are represented
in Figures 1 and 2.
Conclusion In hypoxemic severe TBI patients, increasing PEEP levels
from 5 to 10 and 15 cmH
O increased PtiO
, without increasing ICP
and/or decreasing CPP. Increasing PEEP levels can be an alternative
ventilatory approach to improve brain oxygenation besides FiO
Figure 1 (abstract P41). Changes in PtiO
according to the PEEP levels.
Figure 2 (abstract P41). Changes in CPP according to the PEEP levels.
Critical Care 2011, Volume 15 Suppl 2
Pressure transmitting device: a simple and safe method of
continuous aspiration of subglottic secretions during orotracheal
CES Almeida, C Saghabi, AR Marra
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P42 (doi: 10.1186/cc10190)
Introduction Many interventions are known to decrease the incidence
of ventilator-associated pneumonia, which has great impact on
mortality, length of stay and costs in intensive care units. One of them
is the aspiration of the secretions that pool above the cu of the
endotracheal tube [1]. It is a simple device but its use is not free from
complications [2], being, most of them, bleedings and obstructions
due to lesions of tracheal mucosa. The maintenance of a constant
suction, without wide pressure variation, is an important point to
minimize these complications. The common manometers do not have
enough precision to set an adequate aspiration pressure, because of
its broad scale, and are not able to avoid or to limit pressure variations
in case of partial occlusions, by secretion, for example, facilitating
lesions occurrence. Pressure transmitting devices (Figure 1), usually
used for continuous aspiration of pleural drainage, have those helpful
characteristics. It can be set in an adequate aspiration pressure
(20mmHg ~ 27 cmH
O) by setting the water column height. It avoids
suction pressure variations since the air bubbles up on the water,
balancing pressure inside the system.
Methods Pressure transmitting devices were tested in 12 patients with
subglottic aspiration on their orotracheal tubes. They were watched for
complications and the  ndings are reported. The aspiration pressure
used was set at 20 cmH
Results The proposed system was used for periods that lasted from 3 to
14 days in each patient. It was able to remove the subglottic secretions
in all tested cases. There were two episodes of system obstruction due
to thick secretions, one of them was a blood clot (the patient had an
abundant oral bleeding), easily treated with gentle suction using a 5-ml
syringe. There was one case of obstruction resolved with air injection
through the subglottic suction lumen. There was no bleeding related
to subglottic suction. There was no ventilator-associated pneumonia.
Conclusion In those reported cases, the subglottic suction system
using a pressure transmitting device seemed to be e ective, without
serious complications. This study of cases is not able to a rm these
conclusions. It is just an initial test of a new method. For better evidence,
this system has to be compared with other devices, like manometers,
that are usually used for aspiration pressure control.
1. Dezfulian C, Shojania K, Collard HR, Kim HM, Matthay MA, Saint S: Subglottic
secretion drainage for preventing ventilator associated pneumonia: a
meta-analysis. Am J Med 2005, 118:11-18.
2. Harvey RC, Miller P, Lee JA, Bowton DL, MacGregor DA: Potential mucosal
injury related to continuous aspiration of subglottic secretion device.
Anesthesiology 2007, 107:666-669.
Pumpless extracorporeal lung assist in a pregnant woman with
severe ARDS
HFRD Cunha
, AP Coscia
, A Longo
, L Campioni
, EG Martins
, ED Meis
, R Costa
Hospital Quinta D’Or, Rio de Janeiro – RJ, Brazil;
Instituto Nacional do
Cancer – INCA, Rio de Janeiro – RJ, Brazil;
Hospital Copa D’Or, Rio de
Janeiro– RJ, Brazil
Critical Care 2011, 15(Suppl 2):P43 (doi: 10.1186/cc10191)
Introduction Acute respiratory distress syndrome is characterized
by acute-onset, refractory hypoxemia, bilateral in ltrates on chest
radiographs and PAOP <18 mmHg or absence of clinical signs of left
atrial hypertension. The protective ventilatory strategy limiting plateau
pressure to lower than 28 cmH
O, driving pressure below 15 cmH
and tidal volume between 4 and 6 ml/kg using a PEEP level to sustain
the open lung approach usually results in hypercapnia. However, it is
the mainstream supportive therapy that can modulate survival in this
Methods We describe a case report where a 31-year-old woman who
was admitted to the intensive care unit with fatigue, shortness of
breath and hypoxemia. She was 24 weeks pregnant and acute myeloid
leukemia, subtype M3 was diagnosed 5 days before admission. Non-
invasive ventilatory support, chemotherapy (doxirubicin and all-trans
retinoic acid) and blood components (red blood cells, fresh frozen
plasma, cryoprecipitate and platelets) were implemented. After 4 days
the clinical scenario was out of control and she was intubated. Renal
function deteriorated and hemodialysis was required.
Results Controlled mechanical ventilation using neuromuscular
blocking (NMB) agents was set to limit plateau pressure, driving
pressure, tidal volume and high level of PEEP (15 cmH
O). However,
oxygenation progressively deteriorated despite the instituted therapy
and on the eighth day on mechanical ventilation the intraabdominal
pressure (IAP) was 20 mmHg, the driving pressure was 20 cmH
O and
Vt was 5 ml/kg, which resulted in PaO
of 90, pH 7.15, PaCO
Figure 1 (abstract P42).
Critical Care 2011, Volume 15 Suppl 2
115 mmHg. Interventional lung assist (iLA; Novalung, GmbH, Talheim,
Germany), a pumpless arterio-venous extracorporeal membrane for
removal, was connected without systemic anticoagulation. After
20 minutes using iLA with 9 l/minute O
, a PEEP level of 20 cmH
O, Vt
of 4 ml/kg, driving pressure of 20 cmH
O, I:E of 1:1 resulted in a PaO
of 175, PaCO
of 57 mmHg and pH 7.35. Hemodynamics were
stable and vasopressor agents were not needed. The blood  ow in
the circuit was 1.4 l/minute. After 14 hours on iLA the NMB agent was
interrupted and assisted ventilatory support with Bivent + PSV (Servo
i Maquet, Solna, Sweden) was started, sustaining a driving pressure of
15 cmH
O. After 48 hours on iLA the baby was born naturally and the
IAP decreased to 7 mmHg. Respiratory system mechanics and the PaO
ratio improved: 56% and 64%, respectively. CPAP + PSV was started
on day 8 after iLA implementation and it was surgically removed on the
day after when the PaCO
was sustained below 40 mmHg.
Conclusion We present the  rst case so far where iLA was safely used
during 9 days in a pregnant woman with severe ARDS and multiple
organ dysfunction syndrome under continuous hemodialytic support
that allowed us to set a protective ventilatory strategy using an assisted
ventilation mode.
Strategies for reducing the time of mechanical ventilation and
ventilator-associated pneumonia
BLDS Guimaraes, SN Nemer, LM Azeredo, JB Caldeira, GM Souza,
FRodriguez, E Guimarães, LPPCPSM Damasceno
Hospital de Clínicas de Niterói, Niterói – RJ, Brazil
Critical Care 2011, 15(Suppl 2):P44 (doi: 10.1186/cc10192)
Introduction Ventilator-associated pneumonia (VAP) is one of the
most frequent causes of nosocomial infection and complication in the
intensive care unit (ICU). VAP is associated with increased mortality and
morbidity, as well as increased costs of intensive therapy.
Objective To compare the prevalence of VAP and the duration of
mechanical ventilation in a general ICU, before and after implantation
of a bundle of four and  ve measures.
Methods A prospective study made in the general ICU, from December
2007 to November 2009, with a total of 432 patients. The measures
adopted in the bundle of VAP were: daily sedative interruption,
Table 1 (abstract P44)
Control group Group 1 Group 2
% VAP 27.3 8.7 1.5
% VAP/1,000 days on MV 25.7 10.6 2.2
% death 63.3 42.9 41.5
Figure 1 (abstract P44).
Table 2 (abstract P44)
Control group Group 1 Group 2
Gender (men/women) 77/76 82/67 64/66
Age (years) 63.4±19.3 70.7±14.5 66±15.1
APACHE II 17.3±7 15.9±13.3 25±8.8
Diagnosis on admission to intensive care, n (%)
Stroke 15 (9.8%) 4 (2.6%) 3 (2.3%)
SDRA 2 (1.3%) 0 (0%) 0 (0%)
Cardiorespiratory arrest 5 (3.2%) 9 (6%) 7 (5.3%)
Sepsis 25 (16.3%) 15 (10%) 12 (9.2%)
Pneumonia 24 (15.6%) 28 (18.7%) 22 (16.9%)
COPD 12 (7.8%) 16 (10.7%) 16 (12.3%)
Postoperative 31 (20.2%) 30 (20.1%) 30 (23%)
abdominal surgery
Oncologic 13 (8.4%) 18 (12%) 14 (10.7%)
Miscellaneous 26 (16.9%) 29 (19.4%) 26 (20%)
Total 153 149 130
Critical Care 2011, Volume 15 Suppl 2
elevation of the head of the bed to 45°, deep venous thrombosis
prophylaxis, peptic ulcer disease prophylaxis. The  fth measure used
was the daily interruption of sedatives with spontaneous breathing
trials (SBTs). The control group was the group without the VAP bundle.
Group 1 was with the VAP bundle. Group 2 was the group of VAP bundle
with daily interruption of sedatives and SBTs.
Results Control group: 153 patients were ventilated from December
2006 to November 2007, with a mean ventilation time of 10.8±2.2 days,
as 41 patients were with VAP, 27.3% of VAP with 53.3% mortality. Group
1: 149 patients were ventilated from December 2007 to November
2008, with a mean ventilation time of 8.3 ± 2.3 days, as 13 patients
were with VAP, 8.7% of VAP with 42% mortality. Group 2: 130 patients
were ventilated from December 2008 to November 2009, with a mean
ventilation time of 7±2 days, as two patients were with VAP, 1.5% of
VAP with 41.5% mortality. All VAP cases on 15 patients happened after
the fourth day of MV; that is, all of them were cases of late VAP. See
Tables 1 and 2 and Figure 1.
Conclusion Implementation of a daily bundle with SBTs is associated
with reduction of mechanical ventilation time, and it is the determinant
factor to have lower indexes of VAP.
Impact of daily evaluation and spontaneous breathing test on
the duration of pediatric mechanical ventilation: a randomized
controlled trial
F Foronda, EJ Troster, JA Farias, CSV Barbas, AA Ferraro, LS Faria, A Bousso,
FF Panico, AF Delgado
Hospital das Clinicas da FMUSP, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P45 (doi: 10.1186/cc10193)
Objective To assess whether the combination of a daily evaluation and
application of a spontaneous breathing test (SBT) could shorten the
duration of mechanical ventilation (MV), as compared with weaning
based on our standard of care. Secondary outcome measures included
extubation failure rate and need for non-invasive ventilation (NIV).
Methods A prospective randomized controlled trial in two pediatric
intensive care units at university hospitals in Brazil. The trial involved
children between 28 days and 15 years of age who were receiving
MV for at least 24 hours. Patients were randomly assigned to one of
two weaning protocols. In the test group, children underwent a daily
evaluation to check readiness for weaning and a SBT with pressure
support of 10 cmH
O and PEEP of 5 cmH
O for 2 hours, with the SBT
repeated on the next day in children failing it. In the control group,
weaning was performed according to the services routine.
Results A total of 294 children were randomized, 155 to the test group
and 139 to the control group. The time to extubation was shorter in
the test group, in which the median duration of MV was 3.5 (95% CI =
3.0 to 4.0) days, in comparison with 4.7 (95% CI = 4.1 to 5.3) days in the
control group (P = 0.0127). This signi cant reduction in the duration of
MV in the intervention group was not associated with increased rates of
extubation failure or NIV, and represents a reduction of 30% in the risk
of remaining under MV (hazard ratio of 0.70).
Conclusion In children under MV for more than 24 hours, a daily
evaluation to check readiness for weaning combined with a SBT
reduced the duration of MV, without increasing the extubation failure
rate or the need for NIV.
Use of lung ultrasonography in the detection of pneumothorax
among medical students and emergency physicians
UAP Flato, HP Guimarães, G Petisco, F Bezerra, AB Cavalcante,
Instituto de Ensino e Pesquisa do Hospital do Coração – Associação do
Sanatório Sírio, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P46 (doi: 10.1186/cc10194)
Introduction The use of lung ultrasound in the detection of pneumo-
thorax is becoming routine in emergency departments and intensive
care units in the United States and Europe [1]. The interposition of the
visceral and parietal pleura (pleural–lung interface) produces pulmonary
artifacts easily visualized by ultrasound and described initially by
Lichtenstein and Meziere [2]. In evaluating the lung for pneumothorax,
the most important  nding is the presence or absence of lung sliding.
The presence of pleural sliding essentially rules out a pneumothorax
in the analyzed region and the absence of lung sliding indicates a high
suspicion of disease. Organizations such as the American College of
Emergency Physicians (ACEP) have demonstrated the short learning
curve and prompt application to clinical practice of this use of lung
ultrasound. There is already evidence, both in Brazil and beyond, that
knowledge retention based on an educational model using computer
simulation would be particularly useful in training Brazilian physicians in
lung ultrasound if it was proven to be e ective.
Objective To evaluate the sensitivity and speci city of diagnosis of
medical students compared with emergency physicians (experts) in
identifying pneumothorax by lung ultrasound.
Methods Students of 3 years of medical graduation participating in
the module Radiology Emergency Medicine (n = 40) and emergency
physicians (n = 11) with training in emergency medicine and intensive
care, called experts, were invited to participate. The study subjects were
assessed for the correct diagnosis of 20 cases of pneumothorax after
training through classroom teaching of lung ultrasound lasting 2 hours
addressing the recognition of artifacts in the lung and identi cation of
pneumothorax Lung Sliding Lines B. Prior to training, medical students
and emergency physicians had no prior knowledge or practice in
emergency ultrasonography. We used video-clips of 10 positive and
10 negative real cases of pneumothorax obtained by an experienced
examiner in lung ultrasound. The comparison between the two groups
was described by the mean and standard deviation of hits in each group
and tested by the nonparametric Mann–Whitney test. The agreement
between raters overall and in each group was estimated by the
kappa correlation coe cient. The di erence between the agreement
observers in each group was tested by Z test for proportions.
Results Students and experts did not have statistically di erent test
scores as shown in Table 1. There was a high degree of agreement
between raters both overall and in each isolated group.
Table 1 (abstract P46)
Sensibility Speci city PPV VPN Accuracy
Appraiser (%) (%) (%) (%) (%)
All 87.8 92.0 91.6 88.3 89.9
Undergraduates 87.3 91.0 90.7 87.7 89.1
Experts 90.0 95.5 95.2 90.5 92.7
Conclusion Medical students and medical experts are able to
accurately identify pneumothorax, despite an abbreviated training
time with no previous knowledge of ultrasound lung. Therefore the
use of a simulation model based on lung ultrasound videos can be
implemented in a systematic way to help health professionals and
medical students in their training.
1. Soldati G, Sher S: Bedside lung ultrasound in critical care practice. Bedside
lung ultrasound in critical care practice. Minerva Anestesiol 2009,
2. Lichtenstein D, Meziere G: Ultrasound probably has a bright future in the
diagnosis of pneumothorax. J Trauma 2002, 52:607.
Integral patient care: mental health in a critical patient service
LR Guastelli, ALM Silva, A Mafuf Neto, CMP Araújo, RP Blaya, ALL Camargo,
E Silva
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P47 (doi: 10.1186/cc10195)
Objective To determine the impact of the introduction of multi disci-
plinary meetings on mental health in the identi cation of psychiatric
and psychological risks in a unit of critical patients.
Critical Care 2011, Volume 15 Suppl 2
Introduction Mental healthcare in hospital wards for critical patients
is necessary both for individuals with psychological or psychiatric
disorders that require intensive medical care and for those individuals
who develop these disorders during hospitalization, often in the same
function, illness or treatment. These disorders may cause negative
impact on adherence to clinical care, well-being, psychosocial rehabili-
tation and patient safety during hospitalization. In our department
there is a psychologist working in conjunction as part of the healthcare
team, aiming to identify psychological risk factors that may impact
on treatment and help the team in handling di cult situations
psychologically. To identify patients with psychiatric risk, we developed
a protocol for Psychiatric Risk Assessment, whereby the presence of
11 items identi ed by the nurse initiates the discussion of a case with
a psychiatrist at the Center for Psychosomatic Medicine of Hospital
Israelita Albert Einstein, which directs care and/or suggests mental
health interventions. Driving this protocol is the need to ask the
nurse to discuss with the mental health professional based on the
identi cation and recovery of behavioral changes that may be missed
and/or be identi ed only when there is already an exacerbation of
psychiatric conditions or occurrences related to them. Aiming to assist
the nursing sta on early identi cation of these risks and organize
actions during the stay in the ward and at discharge, a multidisciplinary
meeting weekly was implemented to discuss cases and situations
related to them.
Methods Implementation of a multidisciplinary meeting consisting of
nurses, psychologists, psychiatrists, medical and nursing coordinators
in November 2010. Conducting a weekly meeting with the purpose
of discussing situations related to behavioral changes in patients
hospitalized in the unit, planning, multidisciplinary care and manage-
ment of cases. After the meeting, the nurse forwarded to the treatment
team a summary that included a description of what quali es as a
psychological or psychiatric risk factor for each case, the guidelines
for the team for management of the situation and suggestions for the
doctor when involving medical management.
Results There were 68 psychiatric risks in the semi-intensive unit in
the second half of 2010. Of these, 31 cases were reported in December,
the month following the beginning of the multidisciplinary meeting.
Whereas 12 cases were reported in October and 12 cases in November,
there was an increase of 158% in the number of cases reported in
December. Regarding reports of psychological risk, we observed that
the multidisciplinary meeting to discuss the risks promotes to the nurse
the understanding of all aspects involved, allowing the discrimination
of the psychological aspects and relevance to specialist interventions
as well as instrumentalizing the team to handle the patient and family.
Discussion The discussion of disciplinary cases seems to have enabled
an understanding, appreciation and discrimination of which behaviors
observed by the nurse should be accompanied by the psychology
team as the protocol of psychiatric risk. The discussion of mental health
with professionals may have a orded the team a better idea of how
these professionals can help provide routine care, promoting the early
identi cation of psychological and psychiatric risks. Other studies
should be performed to con rm the e ectiveness of this intervention.
Conclusion A multidisciplinary meeting was e ective to assist the
team in early detection and recovery of his observations of psychiatric
disorders in hospitalized patients in a semi-intensive unit.
Tissue plasminogen activator-treated patients with acute ischemic
stroke in the pioneer public service of Rio de Janeiro: a comparative
pro le with the NINDS study
H Missaka
, JE Almeida
, PC Figueiredo
, CL Nogueira
, VR Julião
JLL Alencar
, RS Lannes
, SL Fernandes
, G Caetani
, J Abrantes
, V Antonucci
, PHCF Pinto
, S Divan-Filho
CTI2 – Hospital Municipal Souza Aguiar, Rio de Janeiro – RJ, Brazil;
Filho University School of Medicine, Rio de Janeiro – RJ, Brazil;
Souza Marques
School of Medicine, Rio de Janeiro – RJ, Brazil;
Rio de Janeiro State University
School of Medicine, Rio de Janeiro – RJ, Brazil
Critical Care 2011, 15(Suppl 2):P48 (doi: 10.1186/cc10196)
Introduction Strokes are the leading cause of death in Brazil, with
an incidence of 108/100,000 inhabitants [1], 31% lethality, and
beyond they are causes of disability and high social costs. The National
Institute of Neurological Disorders and Stroke study (NINDS) [2] and
the 3rd European Cooperative Acute Stroke Study [3] demonstrated
that intravenous tissue plasminogen activator (t-PA) improved clinical
outcome at 3 months. This study recognized the patient’s pro le
attending a pioneer public Stroke Team – trained at Albert Einstein and
Mãe de Deus Hospitals – comparing and analyzing its results with NINDS.
Objective To evaluate the pro le – age, door-to-needle time (Dt), NIHSS
and mortality – in patients with acute ischemic stroke (AIS) treated with
t-PA in Hospital Municipal Souza Aguiar (HMSA). To compare the results
with NINDS’ reference data.
Methods An observational series and analysis of cases treated with
t-PA on HMSA. A review of recent literature.
Results From May 2006 through November 2010, 71 patients received
t-PA therapy and underwent this study (Table 1). Comparing with NINDS
(on average) we obtained: age: HMSA = 61.8 years (18 to 88), NINDS =
67 years. Dt HMSA = 2.65 hours (1 to 5.5). Patients obtained treatment
within 1.5 hours (Dt <1.5 hours): HMSA = 11 (16%); NINDS = 71 (49%).
Table 1 (abstract P48)
HMSA (n = 71) NINDS (n = 177)
Age (years) 61.8 (18 to 88) 67
Dt <1.5 hours (patients) 11 (15.49%) 71 (49%)
NIHSS at admission 13,4 (5 to 24) 14 (1 to 37)
Mortality 9 (13%) 24 (17%)
Conclusions The study reported an early presentation of AIS, which may
be associated with di cult access to primary care in this city. The entry
NIHSS was similar in both studies. In the NINDS, 50% of the patients
received t-PA within 1.5 hours, and only 16% in the HMSA at this time.
Pre-hospital quick reference and rapid diagnosis in the emergency
room could diminish the Dt. Symptomatic hemorrhage (13% HMSA)
was similar if we take into account only deaths from the use of t-PA
therapy. Finally, we demonstrated bene ts with t-PA treatment in AIS
in Rio de Janeiro and recognized limitations that, when overcome, will
allow improving the treatment of such severe disorder.
1. Brazilian consensus for the thrombolysis in acute ischemic stroke. Arq
Neuropsiquiatr 2002, 60:675-680.
2. The NINDS: rt-PA Stroke Study Group. Tissue plasminogen activator for
acute ischemic stroke. N Engl J Med 1995, 333:1581-1587.
3. Hacke W, et al.: Thrombolysis with alteplase 3 to 4.5 hours after acute
ischemic stroke. N Engl J Med 2008, 359:1317-1329.
Incidence of delirium in three critical care units of a teaching
hospital in Brazil
APN Okada, RP Azevedo, FR Machado
Universidade Federal de São Paulo, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P49 (doi: 10.1186/cc10197)
Introduction Delirium is a disturbance of consciousness in which
there is a sharply global de cit of attention associated with change
in cognition that cannot be attributed to a pre-existing dementia.
Its relevance is not only due to the high incidence, but above all
its consequences, such as in uence on mortality, morbidity, and
prolonging the period of hospitalization.
Objective The aim of this study was to assess the incidence of delirium
in patients admitted to three intensive care units (ICUs) of a teaching
hospital through the diagnostic tool CAM-ICU.
Methods The patients were evaluated through the daily application
of the CAM-ICU by the same physician. We evaluated the correlation
of clinical suspicion of attending physicians, medical residents and
nurses to perform the diagnosis of delirium compared with the CAM-
ICU, the median time to development of delirium, and risk factors for
developing delirium, and compared the outcome between patients
who progressed to delirium and those who had no delirium during
the study period. Standard descriptive statistics were used. Continuous
Critical Care 2011, Volume 15 Suppl 2
variables were reported as the mean and standard deviation.
Interobserver agreement was assessed using Cohens kappa statistic
(κ). All statistics and their 95% con dence intervals were computed
using SPSS software and Medcalc software.
Results In the period of 39 days, 106 patients were screened, and 42
patients ful lled inclusion criteria and were enrolled in the study. The
incidence of delirium was 21.4% (nine patients). The average time
to development of delirium was 62.67 hours (± 33.76), and 88.9%
of patients developed delirium in the  rst 5 days in the ICU. The
agreement of clinical diagnoses in relation to the CAM-ICU method
was moderate, with the best agreement assigned to nurses. A trend for
increased length of ICU and hospital stay was found between patients
who developed delirium. The average time in the ICU for patients with
delirium was 12.11 days (± 15.44) and patients without delirium was
5.75 days (± 7.13), P = 0.0821. The average time of hospitalization for
patients with delirium was 29 days (± 28.99) and without delirium was
21.69 days (± 22.83), P = 0.428. See Table 1.
Table 1 (abstract P49). Correlation of clinical suspicion of attending
physicians, medical residents and nurses to perform the diagnosis of
delirium compared with the CAM-ICU
κ Delirium Hypoactive or mixed
Attending physicians 0.610 NA 0.009
Medical residents 0.656 0.025 0.035
Nurses 0.690 0.038 0.057
Kappa values for delirium (three subtypes), and only the hypoactive or
hypoactive and mixed. NA, not available: no agreement between CAM-ICU and
the evaluation by attending physicians.
Conclusion Delirium is a common disorder in ICUs. Speci c tests
should be used regularly in order to optimize the correct diagnosis and
treatment of this disturbance.
Validity and reliability of the Brazilian–Portuguese version of three
tools to diagnose delirium: CAM-ICU, CAM-ICU Flowsheet and
D Gusmao-Flores
, JI Salluh
, F dal-Pizzol
, LR Santana
, RM Lins
, GV Serpa
, J Oliveira
, RAChalhub
, MA Lima
, MT Pitrowsky
LC Quarantini
Hospital Universitário Prof. Edgar Santos, Universidade Federal da Bahia,
Salvador – BA, Brazil;
D’Or Institute of Research and Education, Rio de Janeiro
– RJ, Brazil;
Instituto Nacional do Câncer, Rio de Janeiro – RJ, Brazil;
Nacional de Ciência e Tecnologia Translacional em Medicina, Universidade
do Extremo Sul Catarinense, Criciúma – SC, Brazil;
Instituto de Ciências da
Saúde, Universidade Federal da Bahia, Salvador – BA, Brazil
Critical Care 2011, 15(Suppl 2):P50 (doi: 10.1186/cc10198)
Introduction Delirium is a frequent form of acute brain dysfunction in
critically ill patients. Several detection methods have been developed
for use in these patients. This study has the objective to validate the
Brazilian–Portuguese CAM-ICU and to compare the sensitivity and
speci city of three diagnostic tools (ICDSC, CAM-ICU and CAM-ICU
Flowsheet) for delirium in a mixed population of critically ill patients.
Methods The study was conducted between July and November 2010
in four intensive care units (ICUs) in Brazil. Patients were screened for
delirium by a psychiatrist or neurologist as the reference rater using the
Diagnostic and Statistical Manual of Mental Diseases, Fourth Edition
(DSM-IV), and subsequently by an intensivist rater using a Portuguese
translation of the CAM-ICU, CAM-ICU Flowsheet and ICDSC (Intensive
Care Delirium Screening Checklist).
Results One hundred and nineteen patients were evaluated: 38.6%
were diagnosed with delirium by the reference rater. The CAM-ICU had
sensitivities of 72.5% (95% CI = 55.9 to 84.9%) and speci city 96.2%
(95% CI = 88.5% to 99.0%), the CAM-ICU Flowsheet had sensitivities
of 72.5% (95% CI = 55.9 to 84.9%) and speci city 96.2% (95% CI =
88.5% to 99.0%), and the ICDSC had sensitivities of 96.0% (95% CI =
81.5 to 99.8%) and speci city 72.4% (95% CI = 58.6 to 83.0%). High
agree ment occurred between CAM-ICU and CAM-ICU Flowsheet
(kappa coe cient= 0.96).
Conclusion The CAM-ICU Brazilian–Portuguese version is a valid and
reliable instrument for the assessment of delirium among critically
ill patients. The three instruments CAM-ICU, CAM-ICU Flowsheet and
ICDSC are good diagnostic tools in critically ill ICU patients and the
CAM-ICU was the most speci c. In addition, the CAM-ICU Flowsheet
presented an excellent correlation with the CAM-ICU and may be
employed in general ICU patients.
Epidemiology/Quality of Life/Administration
Adverse e ects of physiotherapy using the passive bicycle in the ICU
MM Martins, RY Sasai, MS Fole, B Rocha, EE Aquim, M Maturana
Instituto Inspirar, Curitiba – PR, Brazil
Critical Care 2011, 15(Suppl 2):P51 (doi: 10.1186/cc10199)
Introduction The present study aimed to analyze the adverse e ects
of the therapy using the passive bicycle in the intensive care unit (ICU).
Methods This was a longitudinal, experimental, non-randomized
controlled trial study. Performed with patients hospitalized in the ICU
from Vita Curitiba and Batel Hospitals, and the Institute of Neurology
from Curitiba, between 10 March and 30 June 2010. The total sample
was 41 patients, with a total of 215 events, of both genders, being 23
men and 18 women, with an average age of 64 years, Glasgow average
11±3 and APACHE II average score was 19± 6. Of the total sample,
only two patients were evaluated according to the Ramsay scale, with
an average of 4±0.7. The passive bicycle activity was performed while
the patient was in a bed or chair. The hemodynamic variables (heart
rate, respiratory rate, mean arterial pressure and oxygen saturation)
were collected at the beginning (before start of activity), 3 minutes
after the start, and at the very end of the activity, and there was no pre-
established activity time. The adverse e ects accidental extubation;
monitoring loss, like electrode, pulse oximetry and non-invasive blood
pressure measures; change of balance, as lack of trunk control; fall;
probe removal (nasogastric, nasoenteral and/or bladder); peripheral
venous/arterial access were observed during the whole therapy time.
The passive bicycle activity was performed 113 times in a chair (53%),
and 102 times in bed (47%), having an average of 7.8±2.29 minutes.
Results For the 215 events, were observed seven monitoring loss
(3.27%) and one for skin lesion (0.467%), and there was no statistic
signi cant from the proportion test. The adverse e ects fall, probe
removal, change of balance and extubation did not occur during the
activity application. For the hemodynamic variables, using the Student
t test (P <0.05), mean arterial pressure, heart rate and respiratory rate,
did not have signi cant change, without any hemodynamic instability
during the activity (see Figure 1).
Conclusion The results show that using the passive bicycle in the ICU
as a physiotherapy feature is secure and has a low risk of adverse e ects
related to ICU conduct.
Figure 1 (abstract P51). Adverse e ects during the passive bicycle
Critical Care 2011, Volume 15 Suppl 2
C-reactive protein/albumin ratio at ICU discharge as a predictor of
post-ICU death: a new useful tool
LCP Azevedo, OT Ranzani, LF Prada, FG Zampieri, JV Pina , LC Battaini,
YCSetogute, DN Forte, LC Azevedo, M Park
Hospital das Clínicas, Faculdade de Medicina da Universidade de São Paulo,
São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P52 (doi: 10.1186/cc10200)
Introduction There are classical predictors of death after ICU
discharge, such as age, severity of disease and level of nursing care.
CRP concentrations at discharge have also been reported as a predictor
of in-hospital outcome, but with controversial results. Considering that
albumin is a negative acute-phase protein and its decrease may be an
indicator of disease severity, we hypothesized that the CRP/albumin
ratio could be a marker of unfavorable outcomes in the post-ICU period.
Objective This study aimed to investigate whether the CRP/albumin
ratio at ICU discharge may be a predictor of post-ICU death. We also
evaluated which is the best cut-o value of the CRP/albumin ratio to
predict mortality.
Methods Patients discharged from the ICU after at least 72 hours of stay
were retrieved from our prospective collected database. A multivariate
analysis was performed using a backward-LR binary logistic model
taking in-hospital death as a dependent variable and age, APACHE II
at admission, comorbidities, ICU length of stay (LOS), support during
ICU, SOFA at ICU discharge, admission characteristics and CRP/albumin
ratio as independent variables. ROC curves and the Youden index were
used to calculate the best cut-o value of the CRP/albumin ratio.
Results We retrieved 548 patients. Mean age was 49±19 years, median
APACHE II score at admission was 16 (10 to 21) and median SOFA score
at discharge was 2 (1 to 3). The main causes of admission were septic
syndromes and respiratory failure. The in-hospital mortality after ICU
discharge was 18.6%. The ICU length of stay was 7 (4 to 11) days. At
the moment of ICU discharge the median CRP was 47 (22 to 109) mg/l,
albumin 27 (23 to 31) g/l and the mean of CRP/albumin ratio was 3. The
multivariate analysis resulted in the following independent in-hospital
death predictors: age (OR = 1.028, 95% CI = 1.014 to 1.043, P <0.001).
Conclusion We demonstrated that the CRP/albumin ratio, a possible
marker of residual in ammation, in addition to classical variables, could
be a useful and objective tool to support the clinical judgment on the
ICU discharge decision process. The best value of the CRP/albumin
ratio to predict death after ICU discharge is 2. Further prospective
investigations are necessary to con rm these  ndings.
1. Fernandez R, et al.: Crit Care 2006, 10:R179.
Daily multidisciplinary rounds reduce ICU length of stay
ES Pacheco, IP Campos, JF Seixas Junior, S Conejo, HP Vieira, SRG Mazutti,
CFP Garcia, DT Noritomi
Hospital Paulistano, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P53 (doi: 10.1186/cc10201)
Introduction Daily multidisciplinary rounds (DMR) can be helpful to
improve communication, share common goals and result in better
patient outcome [1].
Objective To evaluate the impact of the institution of DMR in clinical
outcomes in a mixed ICU of a private hospital.
Methods DMR were instituted in our mixed tertiary 16-bed ICU in
October 2010. Using our patient data bank (Epimed
) we retrieved
admission clinical and demographic data and outcome information
in two di erent admission periods: 1 year before and 1 year after
institutions of DMR. Four independent multivariate analysis were
performed with the ICU length of stay (LOS), hospital LOS, ICU mortality
and hospital mortality as dependent variables. The independent
variables were: period (previous to DMR and post DMR), age, SAPS
III score, Charlson score and type of admission (clinical vs. scheduled
surgery vs. unscheduled surgery).
Results From October 2008 to October 2010, 1,600 patients were
admitted to our ICU: 656 in period 1 (before DMR) and 944 in period
2 (after DMR). There was no gender or age di erence between the
two periods. However, there were signi cant di erences in the type of
admission (more urgent surgery in period 1, P <0.01), greater SAPS III
(53.4 vs. 46. 4; P <0.01) and Charlson score (2.9 vs. 1.7; P <0.01) in period
1 in comparison with period 2. In the multivariate linear analysis, the
ICU LOS was independently associated with the SAPS III (standardized
beta = 0.17; P <0.01) and period 2 – after DMR (standardized beta =
–0.07; P = 0.01). Only the SAPS in uenced hospital LOS (standardized
beta = 0.27; P <0.01). ICU mortality was only independently associated
with SAPS III (standardized beta = 1.11; P <0.01). Hospital mortality
was independently associated with SAPS III (standardized beta = 1.09;
P<0.01) and Charlson score (standardized beta = 1.07; P = 0.02).
Conclusion The institution of multidisciplinary rounds was indepen-
dently associated with a reduction in the ICU length of stay, without
any signi cant e ect in hospital outcomes.
1. Kim MM, Barnato AE, Angus DC, Fleisher LF, Kahn JM: The e ect of
multidisciplinary care teams on intensive care unit mortality. Arch Intern
Med 2010, 170:369-376.
E cacy of a palliative care program in critically ill patients
SCF Ribeiro, ASRG Mazutti, SC Furlan, IP Campos, JF Seixas Junior,
APSTavares, DT Noritomi
Hospital Paulistano, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P54 (doi: 10.1186/cc10202)
Introduction For some critically ill patients, ICU treatment is more
burdensome than bene cial and is inconsistent with these people’s
values, goals and preferences. Integration of palliative care in the ICU
can help address this issue. We developed in our hospital a proactive
palliative care program focusing on critically ill patients with chronic
limitations aiming at: timely implementation of care plans that are
realistic, appropriate and consistent with the patients’ preferences and
reductions in use of nonbene cial treatments, thus also reducing the
lengths of stay in ICU.
Objective To describe onset of a new palliative care program in a
private hospital ICU.
Methods All critically ill patients with Karnofsky score <40% were
evaluated by the multidisciplinary team when ICU transfer was
considered or, when this was not possible, soon after arriving in the ICU.
In the  rst familiar conference, the main family surrogate was identi ed
and the patient’s or family surrogate’s preferences concerning advanced
life support and end of life were discussed according to the clinical
situation. Revaluation of these decisions was performed whenever
necessary, according to the multidisciplinary team or family surrogate.
Speci c forms were  lled to ensure adequate communication with the
remaining hospital sta and for data acquisition.
Results Between November 2010 and January 2011, 61 patients were
included in our palliative care program. The patients’ median age was
78 (range: 38 to 101) years, with a slight predominance of women (54%).
The main reason for palliative care was severe dementia. All patients
had severe cognitive impairment, so all decisions were discussed solely
with family surrogates. The program was started at the ICU in 52 (85%)
and at the ER in nine (15%) cases. In the nine cases started outside the
ICU, an ICU admission was avoided in eight (89%) of them. In 79% of the
cases, decisions were made to withhold or withdraw some kind of life
support. There were three family-assisted withdrawals of mechanical
ventilation in the ICU. Several family surrogates reported that their
decision was based on previously expressed patient’s wishes.
Conclusion A proactive palliative care program focusing on critically ill
patients with chronic and irreversible limitations is feasible and results
in a better alignment of a patient’s or their familys wishes and medical-
related decisions and attitudes.
Critical Care 2011, Volume 15 Suppl 2
Impact of mechanical ventilation on the functional status in
patients admitted to the intensive care unit
JA Araújo Neto, RF Bom m, FB Lima, DA Castro, EB Moura, MO Maia
Hospital Santa Luzia, Brasília – DF, Brazil
Critical Care 2011, 15(Suppl 2):P55 (doi: 10.1186/cc10203)
Introduction Many ICU survivors report limitations in physical function
that, despite showing slow improvement over time, may be long-
lasting. As a complication of critical illness, weakness frequently slows
and even dominates the course of recovery from critical illness. Patients
requiring mechanical ventilation (MV) often have substantial weakness
of the respiratory and limb muscles that further impairs their functional
status and health-related quality of life.
Objective The aim of this study was to evaluate the impact of the use
of MV on the functional status.
Methods This is an observational, retrospective and analytical study
that included patients aged >18 years who were discharged from ICUs
from July 2010 to December 2010. We excluded patients transferred to
another hospital and who had not been evaluated by the physiotherapy
team at the time of discharge. Functionality was assessed at discharge
from the ICU and at discharge from the hospital through the Functional
Independence Measure (FIM) scale. The following variables were
considered: age, gender, APACHE II, length of ICU, length of stay, length
of MV and FIM. We used the normality tests, Mann–Whitney test and
Wilcoxon test.
Results The sample consisted of 158 patients, 51.9% female, mean
age 62.5 ± 19.8 years. Of these patients, 30.6% used mechanical
ventilation in the ICU. The length of ICU and hospital stay was higher
among patients who received MV (length of ICU: 28.3±24.2 days vs.
9.58±16.5 days, P = 0.001; length of stay: 37.6±27.4 days vs. 18.9±28.6
days, P = 0.001). APACHE II was also higher in this group (13.9 ± 8.3
vs. 10.8±6.57, P = 0.02) (Table 1). The functional status was lower in
the group undergoing MV at discharge from the ICU (65.3± 37.5 vs.
89.2±37.6, P = 0.001) and at discharge from hospital (74.6±41.9 vs.
94.3±37.7, P = 0.008) (Figure 1).
Conclusions In this population we observed that patients submitted
to MV have a lower functional status, and higher APACHE II, length of
ICU and length of stay.
1. Chiang L, et al.: E ects of physical training on functional status in patients
with prolonged mechanical ventilation. Phys Ther 2006, 86:1271-1281.
2. Gri ths RD, et al.: Intensive care unit-acquired weakness. Crit Care Med 2010,
3. Desai SV, Law TJ, Needham DM: Long-term complications of critical care.
Crit Care Med 2011, 39:371-379.
Improved outcome of critically ill patients treated by the Rapid
Response Team outside the intensive care unit
AAFS Georgeto, MT Tanita, PS Taguti, PS Pariz, D Kamiji, MF Sacon,
KPAraújo, LTQ Cardoso, CMC Grion
Hospital Universitário de Londrina, Universidade Estadual de Londrina,
Londrina – PR, Brazil
Critical Care 2011, 15(Suppl 2):P56 (doi: 10.1186/cc10204)
Introduction Due to the limited number of intensive care unit (ICU)
beds in Brazilian public hospitals, many critically ill patients are treated
in hospital wards while waiting to be transferred to the ICU. Care for
these patients is provided by ward sta , while waiting for ICU bed
availability. These healthcare providers are not trained in critical care
and are not as experienced in caring for ICU patients. In the Londrina
University Hospital, the Rapid Response Team (RRT) sta is composed
of intensivist healthcare providers who help to deliver specialized care
to critically ill patients in general hospital wards.
Objective To compare clinical outcomes of critically ill patients treated
in general hospital wards in two periods of time, before and after the
implementation of a RRT.
Methods A prospective longitudinal study developed in two periods:
from January to December 2005 before RRT implementation and
from January to December 2010 after the RRT is already performing
outreach care for critically ill patients. Patients entered the study on
the  rst day an ICU bed was requested and were followed until ICU
admission, death or the request for ICU was cancelled due to clinical
improvement. The chi-square test was used for statistical analyses.
Results We analyzed 699 patients in the  rst period of 2005 and 889 in
the second period of 2010. There was no di erence in mortality of these
patients comparing the two study periods. We observed an increase in
the proportion of patients who presented clinical improvement and
had their ICU bed request cancelled in the year 2010 compared with
the year 2005 (28.57% vs. 19.03%, P <0.001). There was a decrease in
the proportion of patients admitted to the ICU after waiting for bed
availability in the second period (45.67 vs. 59.80%, P <0.001) compared
with the  rst period. We also observed the inclusion of end-of-life
discussions during routine rounds in these patients outside the ICU and
decisions to withhold or withdraw treatment were the reason to cancel
an ICU bed request in 34 (3.82%) patients in the year 2010. See Table 1.
Table 1 (abstract P55). Characteristics of the subjects
MV (n = 37) Without MV (n = 121) P value
Age (years) 62.6±17.8 62.4±20.4 0.84
APACHE II 13.9±8.3 10.8±6.57 0.02
SAPS II 38.03±14.4 32.5±11.9 0.02
Length of ICU (days) 28.3±24.2 9.58±16.5 0.001
Length of stay (days) 37.6±27.4 18.9±28.6 0.001
FIM at discharge from ICU 65.3±37.5 89.2±37.6 0.001
FIM at discharge from hospital 74.6±41.9 94.3±37.7 0.008
Figure 1 (abstract P55). Values for FIM in patients on MV. *P < 0.01.
Critical Care 2011, Volume 15 Suppl 2
Table 1 (abstract P56). Number of patients according to clinical outcome
2005 2010
n % n %
Patient transfer to another institution* 24 3.43 5 0.56
Death** 124 17.74 190 21.37
Clinical improvement* 133 19.03 254 28.57
ICU admission* 418 59.80 406 45.67
Withhold/withdraw treatment* 0 0 34 3.82
Total 699 100.00 889 100.00
Number of patients according to clinical outcome in Londrina University
Hospital, Londrina, Paraná State, Brazil, January to December 2005 and 2010.
*P= 0.08. **P <0.001.
Conclusion We observed improvement in clinical outcome of critically
ill patients after the implementation of outreach intensive care support
delivered by a RRT in a teaching hospital. This e ect apparently
decreased the need for ICU beds, since more patients improved before
an ICU bed was available. We also observed the inclusion of end-of-life
discussions in the routine care of these patients.
Nursing in the ICU: comparison of the NAS and time on bedside
DF Moura Jr, NM Lucinio, A Pardini, S Shiramizo, MR Ribas, RA Morbek,
ELPCA Rosa, MAA Yamashita, MR Guerra, E Silva
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P57 (doi: 10.1186/cc10205)
Introduction The increasing costs of treatment in intensive care units
(ICUs) and the need to use resources e ciently require adequacy
between nursing sta and nursing workload, as a high cost is attributed
to the nurse sta of ICUs. The intensity of the nursing work e ort should
be considered because sta ng needs vary according to the amount of
patients being cared for, as well as the type of care provided for each
of those patients. As the intensity of the nursing work e ort increases,
the amount of nursing sta required to properly care for patients also
Objective To analyze the adequacy of nursing sta according to NAS,
and compare the time of care according to NAS and time of care
according to Nurse Call.
Methods An exploratory, descriptive prospective study was performed
in an adult 32-bed ICU of a private general hospital in São Paulo, Brazil.
In our study we included 18 beds for which the Nurse Call System by
Austco was available. The Nurse Call System by Austco enables nurses
to provide prompt and e ective responses to patients’ calls at all times.
For the analysis of the adequacy of the nursing sta , the mean NAS
expressed as percentage time was initially converted into hours
considering a 6-hour shift (6 hours equivalent to an NAS of 100%).
Results Follow-up of 1,710 patients who were admitted to the ICU
between July and December 2009 resulted in 4,592 NAS assessments.
Analysis of the nursing workload showed a mean NAS of 90.1±4.4%
(ranging between 82.9 and 93.7%). The number of patients ranged from
26.5 to 34.7 in the ICU. The ICU occupation rate  uctuated between 82.8
and 113.9%, during the study, suggesting that managing of the unit
was suboptimal. The hours available for nursing care in the 6-hour shift
remained constant throughout the studied period and represented
a total of 156 hours per shift-day. This number was the same for the
entire study period, as the number professionals was  xed. According
to the NAS, during half of the studied period (July to September) there
was a need for an increased number of nursing professionals, as there
was an average de cit of 30 hours (range 4.4 and 48.9 hours). In the
second half of the study (October to December) the number of nurses
available exceeded that considered necessary by NAS. This surplus
was of 14.2hours on average (range 9.0 and 22.5). The time required
for nurse care per patient per day was very similar between the two
assessment tools (NAS and Nurse Call). While for NAS the mean time
required by patient was 5.4 hours per day (ranging between 5.0 and
5.6), for the Nurse Call this time was 5.3 hours per day (ranging between
4.9 and 5.5).
Conclusion The Nurse Call System can help the ICU nurse manager
on the sta required, showing us a new strategy for managing the
nurse sta . Regarding it being more easy to use, it can be adequately
evaluated in the ICU.
Patients readmitted to intensive care: who they are and what
happens to them?
AP Nassar Junior, LD Salles, L Brauer
Hospital e Maternidade São Camilo, Unidade Pompeia, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P58 (doi: 10.1186/cc10206)
Introduction There is growing interest in quality-of-care indicators in
the ICU. Readmission is one of the proposed indicators to be measured.
Objective To investigate the incidence of, outcomes and possible
risk factors for readmission in a large cohort of patients in a medical–
surgical ICU and to evaluate the accuracy of Simpli ed Acute Physiology
Score III (SAPS III) and Acute Physiologic and Chronic Health Evaluation
IV (APACHE IV) to predict readmissions.
Methods We conducted an analysis of prospectively collected data
from all patients admitted between January 2009 and December 2010
who survived their  rst ICU stay. Patients aged <18 years, patients
transferred to another hospital and those who were not yet discharged
until 1 February 2011 were excluded from the analysis. The following
variables were evaluated as possible risk factors for readmission:
sex, age, type of admission (medical vs. surgical), SAPS III, APACHE III
score, APACHE IV mortality predicted risk, ICU length of stay (LOS), ICU
discharge at night and on weekends. Accuracies of SAPS III and APACHE
IV mortality predicted risk were assessed by calculating the area under
the receiver operating characteristic curve. Categorical variables are
presented as absolute numbers and percentages. Continuous variables
are presented as medians and interquartile ranges.
Results A total of 3,993 patients were admitted during the study period
and 3,637 ful lled study inclusion criteria. Two hundred and eighty-
three (7.8%) had at least one readmission. Patients’ characteristics are
displayed in Table 1. In the multivariate analysis, SAPS III (OR = 1.020; P=
0.008), APACHE III score (OR = 1.015; P <0.001).
Conclusion Readmitted patients were older, had longer ICU LOS and
higher severity scores at admission. Readmission was an independent
factor associated with in-hospital mortality. SAPS III and APACHE IV at
rst admission had only moderate ability to predict readmissions.
Table 1 (abstract P58). Characteristics and outcomes of readmitted and
nonreadmitted patients to ICU
Readmitted Nonreadmitted
patients patients P
(n = 283) (n = 3,354) value
Male sex 142 (50.2) 1,722 (51.3) 0.707
Age 73 (56 to 82) 63 (47 to 77) <0.001
Type of admission 0.379
Medical 243 (85.9) 2,813 (83.9)
Surgical 40 (14.1) 541 (16.1)
SAPS III 48 (38 to 57) 41 (33 to 49) <0.001
APACHE III score 36.5 (23 to 51) 26 (18 to 36) <0.001
APACHE IV risk (%) 8.36 (3.07 to 18.51) 2.94 (1.18 to 7.10) <0.001
ICU LOS 3.92 (2.20 to 8.39) 2.45 (1.58 to 3.85) <0.001
Discharge at night 62 (21.9) 589 (17.6) 0.068
Discharge on weekend 67 (23.7) 939 (28.0) 0.119
In-hospital mortality 81 (28.6) 43 (1.3) <0.001
Critical Care 2011, Volume 15 Suppl 2
Systemic in ammatory response syndrome and organ dysfunctions
are early predictors for ICU readmission
AM Japiassú, H Falcão, AP Coscia, DM Costa, RBN Silva
Hospital Quinta D’Or, Rio De Janeiro – RJ, Brazil
Critical Care 2011, 15(Suppl 2):P59 (doi: 10.1186/cc10207)
Introduction Previous studies have indicated risk factors for ICU re-
admission; sepsis, respiratory insu ciency, medical admission, organ
dysfunctions and age are associated with this outcome. Speci c
physiological and laboratory data were explored in some studies, but
no association was shown with readmission. Our hypothesis is that
in ammation and organ dysfunctions are more important for this
outcome than demographic data or type of admission.
Methods We selected all consecutive patients admitted to the  ve ICUs
of a tertiary hospital. All patients discharged from the ICU at least once
were included. Demographic, physiological and laboratory data were
collected on the  rst day after the  rst admission and organ support
resources (mechanical ventilation, use of vasopressors and renal
dialysis) were researched throughout the ICU stay. Organ dysfunctions
were de ned as they are in the SOFA score. A logistic regression was
made with all of the parameters with P <0.2 in the univariate analysis.
Results There were 1,073 patients admitted to all  ve ICUs during the
study period. Seventy patients died during the  rst admission and were
excluded, resulting in the analysis of 1,003 patients. There were 160 ICU
readmissions from 130 patients. The readmission rate was 13%. Sepsis
and respiratory or cardiovascular decompensation were the most
common causes of readmission. ICU readmitted patients were more likely
to be older (median 75 x 69 years, P = 0.004), medical rather than surgical
type (70 x 61%, P = 0.04), originated from the ward or intermediate care
unit (15 x 6%, P <0.001), with any infection on ICU admission (38 x 29%,
P= 0.03), and higher Charlson index (1 x 0 point, P <0.001).
Conclusion Systemic in ammatory response syndrome with organ
dysfunctions are predictors for ICU readmissions, despite the patient’s
origin, type of admission and the presence of infection at admission.
Hyponatremia severe and symptomatic in a critically ill infant
CMF Mangia, RM Sousa, AP Loretti, AFCF Martins, NF Oliveira, EL Lima,
Universidade Federal de São Paulo, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P60 (doi: 10.1186/cc10208)
Objective To report a case of severe symptomatic hyponatremia
secondary to previously undiagnosed congenital adrenal hyperplasia.
Case A 37-day-old infant, born at 38 weeks gestation, presented with
hypoactivity, weight loss, poor feeding and vomiting in the hospital.
The main clinical features were irritability, dehydration, hyponatremia,
hyperkalemia and ambiguous genitalia. The biochemical data are
presented in Table 1. The patient received isotonic  uids ( rst day) and
treatment for severe chronic hyponatremia (developing over more than
48 hours) calculated to 125 mEq/l under slow correction in 96 hours.
The sodium levels did not exceed 0.5 mEq/l/hour or 12 mEq/l/day. On
the  rst day was initiated hydrocortisone (100 mg/m
) and afterwards
50 mg/m
. There were no complications of treatment and the child was
discharged 2 weeks later without sequels. The karyotype was 46,XX.
Conclusion Hyponatremia is a frequent electrolyte disorder. It is
considered severe (<115 mEq/l) and chronic when the duration is
>48 hours or the installation time is unknown. Irreparable harm can
happen when abnormal serum sodium levels are corrected too
quickly or too slowly. The correct diagnosis and understanding of
the pathophysiology and mechanisms associated with hyponatremia
allows establishing safe treatment criteria and consequently avoiding
the sequels.
1. Bornstein SR: Predisposing factors for adrenal insu ciency. N Engl J Med
2009, 360:2328-2339.
Time to meet energy requirements in enteral nutrition and its
impact on patient tolerance and clinical outcomes in the ICU
FVCD Marco, JM Souza, C Harmbacher, ALG Guimares
Hospital Vivalle, São Jose Dos Campos – SP, Brazil
Critical Care 2011, 15(Suppl 2):P61 (doi: 10.1186/cc10209)
Introduction Delivering early nutrition support therapy, primarily
using the enteral route, is seen as a strategy that may reduce disease
severity, diminish complications, decrease length of stay in the ICU,
and favorably impact patient outcome. SCCM and ASPEN guidelines
support that after the initiation of enteral feeding we have 10 days
to meet 100% of predicted energy requirements before we consider
supplementation with parenteral nutrition (PN). There are scarce data
about the clinical e ects of using a more accelerated approach to reach
full caloric adequacy with enteral nutrition (EN).
Objective The aim of this observational study is to evaluate whether a
diminished time to target caloric goal is associated with more patient
intolerance and clinical bene ts in ICU patients receiving EN.
Methods From January 2010 to June 2010 we prospectively followed
all consecutive ICU patients receiving EN. We collected epidemiological
data, APACHE II score, LOS (ICU and hospital), need for mechanical
ventilation, incidence of nosocomial infection and hospital mortality.
We also collected data on nutrition therapy as the time to target
caloric goal (120 hours), total time on nutrition therapy, incidence of
diarrhea and other signs of EN intolerance (vomits, abdominal pain and
distension). For statistical analysis we used the Kolmogorov–Smirnov
test, Student’s
t test and Pearsons correlation coe cient.
Results We enrolled 32 patients (17 male/15 female) in the study. The
mean age was 66±18 years, mean APACHE II score 21±9, mean ICU
and hospital LOS were 21.3 and 35 days respectively, incidence of
nosocomial infection was 21.8%, mean total time in nutrition therapy
was 18.3±14 days and hospital mortality was 28%. There was need for
mechanical ventilation in 56%. There was need for PN supplementation
in 9.4% (n= 3) of patients. Comparing the di erent groups (120hours,
n = 16) we were unable to detect any di erence with statistical
signi cance regarding incidence of diarrhea, EN intolerance, need for
MV, total time on nutrition therapy, incidence of nosocomial infection,
ICU and hospital LOS and hospital mortality.
Conclusion These preliminary data have shown no correlation of
a diminished time to meet energy requirements in EN with patient
tolerance to nutrition therapy and clinical bene ts.
Table 1 (abstract P60). Serum electrolyte concentration during the  rst week
mEq/l Admission 1st
day 2nd day 3rd day 4th day 5th day 6th day 7th day
101 108 120 122 133 129 127 130
7.9 6.1 4.7 3.3 4.5 5.8 5.3 5.6
78 84 94 96 103 95 95 94
Critical Care 2011, Volume 15 Suppl 2
Large venous–arterial PCO
is associated with poor outcomes in
surgical patients
JM Silva Junior, AMRR Oliveira, VPl Maia, AMP Ferreira, DO Toledo,
ERezende, LMS Malbouisson
HSPE, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P62 (doi: 10.1186/cc10210)
Introduction This study evaluated whether a large venous–arterial CO
gap (PCO
gap) during the preoperative period is associated with poor
surgical outcome.
Methods A prospective observational study that included high-risk
surgical patients who were 18 years of age or older. Palliative surgery,
Child B and Child C cirrhosis, and class IV heart condition patients or
ejection fraction <30% were excluded. The patients were divided
into two groups: wide [P(v-a)CO
] versus narrow [P(v-a)CO
]. In order
to determine the best value to discriminate hospital mortality, the
receiver operating characteristic curve was used for the [P(v-a)CO
values collected during the preoperative period, and the most accurate
value was chosen as a cut-o to de ne the groups.
Results The study included 66 patients. The preoperative [P(v-a)CO
value that best discriminated hospital mortality was 5.0 mmHg, area=
0.73. Preoperative patients with [P(v-a)CO
] of more than 5.0 mmHg
presented a higher hospital mortality (36.4% vs. 4.5%, P = 0.004), higher
prevalence of circulatory shock (56.8% vs. 22.7%, P = 0.01) and acute
renal failure in the postoperative period (27.3% vs. 4.5%, P = 0.02), and
longer length of hospital stays (20.0 (14.0 to 30.0) vs. 13.5 (9.0 to 25.0)
days, P = 0.01). The groups did not present any di erences regarding
demographic and physiological data.
Conclusion The PCO
gap values of more than 5.0 mmHg in the pre-
operative period were associated with worse postoperative outcome.
Analysis of head trauma management in a secondary hospital
without neurosurgical service
M Steinman, L Lenci, C Kirschner, P Rogeri, S Possa, N Akamine
Hospital Municipal Dr. Moyses Deutsch, Instituto de Ensino e Pesquisa,
Hospital Israelita Albert Einstein, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P63 (doi: 10.1186/cc10211)
Introduction Head injuries are one of the most common causes of trauma
patient admission. A key part of the management of these patients
is airway control, rapid transport to appropriate trauma care facilities
and prompt resuscitation. Many trauma patients who have su ered a
head injury are initially taken to non-neurosurgical (NS) centers. In most
instances, patients with severe head injury have to be transferred to a
NS unit. Theoretically, the reason to transfer is the potential need for
immediate surgical intervention. The purpose of the study was to evaluate
head trauma patients who were transferred to NS units to determine the
incidence of this occurrence, patients’ pro le and criteria adopted.
Methods A 6-month retrospective study was conducted from January
through July 2010 at Hospital Municipal Dr. Moses Deutsch, located in
Jardim Angela, south of São Paulo, 30 kilometers away from downtown.
It is the only hospital within a radius of 7 miles and serves a population
of approximately 600,000 inhabitants. It is a secondary hospital that
provides medical sta in the emergency room for 24 hours as well
as on-site computed tomographic (CT) scanning capability and the
intensive care unit. All head trauma patients who were transported to
NS were included. Data collected were demographics, mechanism of
injury, Glasgow Coma Scale (GCS), clinical  nding, CT  ndings, transfer
times and returns from the NS.
Results There were 17,880 ED patient admissions and 2,255 were
trauma related. A total 296 were head-injured patients requiring
hospitalization. Eighteen seven patients demanded interhospital
transfer, because of CT  ndings and clinical picture. The main mecha-
nism of injury was falls (59.4%). The median transport delay to the
neurosurgical service site was 10±1.2 hours. Mean GCS were 12 and
56% of the CT had abnormal  ndings. Seventy- ve percent returned
after NS evaluation.
Conclusion Most of the cases are referred for assessment because
of lack of local expertise leading to unnecessary transfers. This often
resulted in the inappropriate transfer of ill patients and the unnecessary
occupation of neurosurgical beds in a tertiary center. Furthermore,
after assessment, many of these patients are sent back to the original
hospital. Emergency neurosurgical teleconsultation may have an
important role in the remote care of patients with head injuries and
other neurosurgical emergencies.
Outcome of surgical patients who present acidosis postoperatively
JM Silva Jr, AM Oliveira, YN Marti, TB Gonzaga, AMP Ferreira, VPl Maia,
HSPE, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P64 (doi: 10.1186/cc10212)
Introduction Acidosis is a very frequent disorder in surgical patients.
In this patient set there remains uncertainty of the clinic implications
from acidosis and characteristics postoperatively. Therefore, it is very
important to evaluate the role of acidosis in outcome for high-risk
surgical patients.
Methods A prospective observational study was performed in  ve
specialized intensive care units (ICUs) in surgical patients from three
di erent hospitals. The patients who needed postoperative intensive
care were involved in the study consecutively. Patients with low
life expectancy (cancer without treatment), hepatic failure (Child
B or C), renal failure (clearance of creatinine <50 ml/min or previous
hemodialysis), and diabetic diagnosis were excluded. The patients
were strati ed by admission from the ICU related to kind of acidosis
in the immediately postoperative period. The strati cation evaluated
metabolic acidosis by base excess <2 mmol/l and anion gap and
lactate, both >12 and 2 mmol/l, respectively.
Results The study involved 188 patients during 3 months. The
incidence of acidosis was bigger, but 52 (27.6%) presented a high anion
gap without hyperlactatemia, 50 (26.6%) showed a high anion gap with
hyperlactatemia, 48 (25.5%) a normal anion gap and in 38 (20.2%) there
was no metabolic acidosis. Overall, gastric surgery presented higher
percentages from metabolic acidosis (46.2% vs. 11.1% nonacidosis,
P<0.05). However, patients did not present di erence in severity (SAPS
III, SOFA and ASA), age and length of surgery. Patients with high anion
gap and hyperlactatemia immediately postoperative showed greater
complications, mainly shock, in comparison with only high anion
gap patients, normal anion gap patients and nonacidosis patients,
respectively 66%, 48.1%, 47.9% and 39.5% (P <0.05). The same was
veri ed in related to mortality rate, respectively 14.5%, 10.2%, 6.1%
and 2.0% (P = 0.04).
Conclusion Metabolic acidosis in surgical patients is a very important
complication postoperatively, mainly in gastric surgery. Patients
who developed metabolic acidosis with a high anion gap and
hyperlactatemia presented worst outcomes compared with patients
with other kinds of acidosis or patients with nonacidosis.
Red blood cell transfusion is an independent risk factor for
cardiovascular complications in adult patients undergoing cardiac
surgery: a propensity score-matched analysis
JP Almeida, F Galas, JL Vincent, JT Fukushima, RE Nakamura, R Kalil Filho,
FB Jatene, JOC Auler Jr, LA Hajjar
InCor, São Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P65 (doi: 10.1186/cc10213)
Introduction Red blood cell (RBC) transfusion is associated with a
higher occurrence of clinical complications after cardiac surgery.
However, the cause–e ect relationship is confounded by other risk
factors for worse outcomes as advanced age, valve or combined
procedure, high EuroSCORE, redo surgery, longer bypass time and
previous anemia. The objective of this study was to evaluate the e ect
of RBC transfusion in a propensity score-matched case–control analysis.
Critical Care 2011, Volume 15 Suppl 2
Methods A total of 502 patients who underwent cardiac surgery with
cardiopulmonary bypass from February 2009 to February 2010 were
evaluated. We performed a propensity score-matching analysis in
264 patients, considering the following risk factors for cardiovascular
complications: sex, age, type of procedure, EuroSCORE, redo surgery,
bypass time and previous hemoglobin.
Results Cardiovascular complications occurred in 39 patients (30%)
exposed to red blood cell transfusion, and in 22 patients (17%) not
exposed. The propensity score-matched analysis showed an odds ratio
of 2.1 (95% CI = 1.2 to 3.8) for cardiovascular complications in patients
exposed to RBC transfusion (Table 1).
Conclusion RBC transfusion after cardiac surgery increases the risk of
cardiovascular complications in a group of patients paired for other
risk factors. These  ndings bring into perspective the importance
of an adoption of a restrictive strategy of RBC transfusion to avoid
cardiovascular complications.
Risk factors for intra-abdominal hypertension and abdominal
compartment syndrome in patients admitted to the ICU
M Assunção, FS Oliveira, BF Mazza, F Freitas, M Jackiu, FR Machado
Escola Paulista de Medicina, Federal University of São Paulo, UNIFESP, São
Paulo – SP, Brazil
Critical Care 2011, 15(Suppl 2):P66 (doi: 10.1186/cc10214)
Introduction Intra-abdominal hypertension (IAH) and abdominal
com part ment syndrome (ACS) as well as their risk factors were de ned
recently by consensus. These diseases have a high incidence and
morbi-mortality in patients admitted to the ICU and represent a huge
problem among critically ill patients.
Objective To determine the incidence of IAH or ACS in patients
admitted to a university hospital ICU with two or more risk factors.
Methods All patients admitted to the ICU were evaluated daily.
Those with at least two risk factors were submitted to intraabdominal
pressure (IAP) monitoring by intravesical pressure method once
daily, during 7days or until death or ICU discharge. In each measure,
the abdominal perfusion pressure (APP) (that is, IAP – mean arterial
pressure) was recorded. Demographic data, APACHE II, ICU and hospital
length of stay and mortality were determined. Results are presented as
the percentage or mean±standard deviation (Table 1).
Results Patients were assessed from February 2010 to October 2010
and 164 were enrolled. Thirty-two patients ful lled criteria for IAP
monitoring (mean age 62±17 years, 37% (12) female, mean APACHE II
score 18±4). Among these 32 patients, 62% (20) had at least one IAH
episode and 12.5% (four) developed ACS. Only patients with ACS had
APP <60 mmHg. Hospital LOS was 60±55 days, ICU LOS was 14±15
days. The 28-day, ICU and hospital mortalities were 31% (10), 38% (20)
and 62% (20), respectively.
Conclusion Risk factors have a high incidence in our ICU. IAH/ACS
patients present a high mortality and a long LOS.
Table 1 (abstract P66). Characteristics of the study participants and
Age 66 (± 17)
Female gender, % (n) 34 (11)
APACHE II score, mean (±SD) 18 (± 4)
Emergency surgery, % (n) 0 (13)
Elective surgery, % (n) 22 (7)
Medical, % (n) 38 (12)
% (n) 62 (20)
ACS, % (n) 12.5 (4)
Length of stay (days)
Hospital 60 (± 55)
ICU 14 (± 15)
Mortality (%)
28-day 31
ICU 38
Hospital 62
Table 1 (abstract P65). Comparison between propensity-matched patients
groups with and without red blood cell transfusion after cardiac surgery
RBC transfusion
Variable No (n = 132) Yes (n = 132) P value
Sex (female) 34 (26%) 48 (36%) 0.063
Age (years), mean (95% CI) 59 (57 to 61) 60 (58 to 63) 0.295
CABG 89 (67%) 87 (66%) 0.873
Valve 36 (27%) 36 (27%)
CABG + valve 7 (5%) 9 (7%)
EuroSCORE, median (IQR) 4 (3 to 6) 4 (3 to 6) 0.683
Redo surgery 13 (10%) 14 (11%) 0.839
Bypass time (minutes), median (IQR) 89 (75 to 110) 91 (75 to 112) 0.418
Hemoglobin (g/dl), mean (95% CI) 13.3 (13.1 to 13.6) 13.1 (12.9 to 13.3) 0.192
Cardiovascular complications 22 (17%) 39 (30%) 0.013
OR = 2.1 (95% CI = 1.2 to 3.8)
CABG, coronary artery bypass surgery; CI, con dence interval; IQR, interquartile
range; OR, odds ratio. Statistical tests: Mann–Whitney and chi-square.
Cite abstracts in this supplement using the relevant abstract number, e.g.:
Assunção M, et al.: Risk factors for intra-abdominal hypertension and
abdominal compartment syndrome in patients admitted to the ICU
[abstract]. Critical Care 2011, 15(Suppl 2):P66.
Critical Care 2011, Volume 15 Suppl 2
... Limited studies evaluated the value of LA in predicting outcomes in patients with cancer having sepsis, and the results have been inconsistent. 18,19 Therefore, this study was conducted to examine the ability of LA to predict mortality in patients with cancer having septic shock. The main objective of this study was to evaluate the predictive validity of LA single measurements over the first 24 hours, as well as LA clearance, in predicting hospital mortality in patients with cancer having septic shock. ...
... Two other studies evaluated the predictive value of LA in patients with cancer having sepsis but reported inconsistent findings. 18,19 Although Hajjar et al 18 reported LA at 24 hours as an independent predictor of hospital mortality and suggested including it in the routine assessment of patients with cancer admitted with sepsis. Keçe et al 19 concluded that LA cannot be recommended alone to predict the poor clinical outcomes of adult patients with cancer having sepsis. ...
... Two other studies evaluated the predictive value of LA in patients with cancer having sepsis but reported inconsistent findings. 18,19 Although Hajjar et al 18 reported LA at 24 hours as an independent predictor of hospital mortality and suggested including it in the routine assessment of patients with cancer admitted with sepsis. Keçe et al 19 concluded that LA cannot be recommended alone to predict the poor clinical outcomes of adult patients with cancer having sepsis. ...
Full-text available
Purpose: Limited studies evaluated the predictive value of serum lactate (LA) in critically ill patients with cancer. The main objective of this study was to evaluate the predictive validity of LA single measurements as well as LA clearance in predicting mortality in patients with cancer having septic shock. The study also aimed to determine the LA measurement over the first 24 hours with the highest predictability for hospital mortality. Materials and methods: A retrospective cohort study of adult patients with cancer having septic shock and LA measurements during the first 24 hours. Three receiver-operating characteristic (ROC) curves were constructed to evaluate the predictive validity for hospital mortality of LA at baseline, at 6 hours and at 24 hours after identifying septic shock. The ROC with the largest area under the curve was analyzed to determine LA level with the highest predictability for hospital mortality. In addition, the ability of LA normalization (LA <2 mmol/L at 6 hours and at 24 hours) and the degree of LA elimination (>10% and >20% at 24 hours) to predict hospital mortality were evaluated by determining the predictive values for each clearance end point. Results: The study included 401 patients. LA >2.5 mmol/L at 24 hours showed the largest area under the ROC curve to predict hospital mortality (ROC area: 0.648; 95% confidence interval: 0.585-0.711) with a sensitivity of 58.4% and specificity of 62.8%. The LA normalization, LA clearance >10%, and LA clearance >20% were also predictors of hospital mortality, with the highest sensitivity for LA normalization at 6 hours (74%) and LA normalization at 24 hours (73.4%). Conclusion: In patients with cancer having septic shock, LA >2.5 mmol/L at 24 hours of septic shock had the highest predictability for hospital mortality. The LA normalization and clearance were also predictors of hospital mortality. However, all LA end points were not strong predictors.
... 17 In addition, lactate levels reported in septic cancer patients were not different from the mean levels of septic patients without malignancy in the literature. 18,19 In this case, we believe that our findings based on measured lactate levels in cancer patients in this study are valuable. ...
... 9 Among 1129 septic cancer patients in an ICU, Hajjar et al reported that the mortality was 28.7%, mean lactate levels were different between surviving and dead patients (2.4 mmol/L vs 3.7 mmol/L), and found that the lactate levels could predict mortality in multiple regression analysis. 18 The lactate levels of sepsis patients in our study were lower than that. This result was expected since our sepsis positive patient group was composed of milder forms since we did not group according to the severity of sepsis. ...
Full-text available
Objectives: Differentiating sepsis from other noninfectious causes of systemic inflammatory response syndrome (SIRS) in cancer patients is often challenging. Although lactate and procalcitonin have been studied extensively regarding sepsis management, little is known about their utility in cancer patients. This study aimed to compare the diagnostic and prognostic utility of lactate and procalcitonin for sepsis in cancer patients. Material and methods: This prospective case-control study was conducted with adult cancer patients presenting to emergency department (ED) with at least two SIRS criteria. The infection status of each patient was determined retrospectively. Main diagnostic variables were calculated for diagnostic and prognostic utilities of lactate and procalcitonin. Results: Among 86 patients, mean age was 61. Twenty-two (25.6%) were determined in the sepsis group. In the ROC analysis, a lactate value of 1 mmol/L predicted sepsis with 86.36% (95%CI: 65.1%-97.1%) sensitivity and 28.12% (95%CI: 17.6%-40.76%) specificity. A procalcitonin value of 0.8 ng/mL yielded a sensitivity of 63.64% (95%CI: 40.7%-82.8%) and 76.56% (95%CI: 63.4%-86.2%) specificity for differential diagnosis of sepsis in cancer patients. Lactate and procalcitonin showed similar abilities in differentiating sepsis from non-infective SIRS in cancer patients [AUROCs of 0.638 (95%CI:0.527-0.739) vs 0.637 (95%CI:0.527-0.738), respectively. p = 0.994]. They were also similar in predicting poor clinical outcome with AUROCs of 0.629 (95%CI:0.518-0.731) and 0.584 (95%CI: 0.473-0.69), respectively (p = 0.577). Conclusions: The results of this study indicated that, none of the lactate and procalcitonin can be recommended alone to differentiate sepsis from non-infectious SIRS and to predict the poor clinical outcomes in adult cancer patients with SIRS in the ED.
... Sabemos bien que la base (déficit de base en este caso) medida al ingreso a la UCI e