See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/45278019

Survey on physicians' knowledge of sepsis: Do

they recognize it promptly?

Article in Journal of critical care · December 2010

DOI: 10.1016/j.jcrc.2010.03.012 · Source: PubMed

CITATIONS READS

19 359

9 authors, including:

Murillo Santucci Cesar Assuncao Nelson Akamine

Hospital Israelita Albert Einstein Universidade Federal de São Paulo
97 PUBLICATIONS 776 CITATIONS 66 PUBLICATIONS 637 CITATIONS

SEE PROFILE SEE PROFILE

Patricia V C Mello Alvaro Rea-Neto

Universidade Estadual do Piauí (UESPI) Universidade Federal do Paraná
15 PUBLICATIONS 211 CITATIONS 60 PUBLICATIONS 2,095 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Brazilian surgical identification study- Brasis study View project

Sepsis protocol View project

All content following this page was uploaded by Murillo Santucci Cesar Assuncao on 13 April 2017.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
Journal of Critical Care (2010) 25, 545–552

Sepsis

Survey on physicians' knowledge of sepsis:

Do they recognize it promptly?☆,☆☆
Murillo Assunção MD a,b,⁎, Nelson Akamine MD b , Guttemberg S. Cardoso MD a ,
Patricia V.C. Mello MD c , José Mário M. Teles MD b,d , André Luis B. Nunes MD e ,
Marcelo Oliveira Maia MD f , Álvaro Rea-Neto MD b,g , Flavia Ribeiro Machado MD a,b
for the SEPSES study group
a
Disciplina de Anestesiologia, Dor e Terapia Intensiva. Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
b
Latin American Sepsis Institute, Brazil
c
Hospital de Terapia Intensiva, Universidade Estadual do Piaui, Teresina, Piauí, Brazil
d
Hospital Português, Salvador, Bahia, Brazil
e
Hospital São Camilo, São Paulo, São Paulo, Brazil
f
Hospital Santa Luzia, Brasília, Distrito Federal, Brazil
g
Hospital de Clínicas, Universidade Federal do Paraná, Curitiba, Paraná, Brazil

Keywords: Abstract
Sepsis;
Purpose: In Brazil, sepsis has a high mortality; and early recognition is essential in outcome. The aim of
Severe sepsis;
the study was to evaluate physicians' knowledge about systemic inflammatory response syndrome
Septic shock;
(SIRS), sepsis, severe sepsis, and septic shock concepts.
Professional practice;
Methods: This was a prospective, observational study performed in 21 hospitals in Brazil, which
Education and diagnoses
enrolled physicians working in the participant institutions. A previously validated questionnaire was
applied to physicians including 5 clinical cases.
Results: Twenty-one Brazilian institutions enrolled 917 physicians. The percentage of physicians
correctly recognizing SIRS, infection, sepsis, severe sepsis, and septic shock was 78.2%, 92.6%, 27.3%,
56.7%, and 81.0%, respectively. Intensivists performed better in all diagnoses. There was a significantly
higher rate of correct answers for SIRS (P b .001), sepsis (P = .001), and severe sepsis (P = .032) among
physicians from university hospitals as compared with those from public hospitals. A mean global score
of 3.36 ± 1.08 was found, with better performance for residents (P = .012) and intensivists (P b .001);
but no difference was found for emergency physicians (P = .875).

☆
The author(s) declare that they have no competing interests.
☆☆
Authors' contribution: MA, FRM, and NA conceived and participated in the design and coordination of the study. All authors contributed to data
collection. MA performed the statistical analyses. MA and FRM drafted the manuscript. GSC, PVCM, JMMT, ALBN, MOM, and ARN revised the article.
All authors read and approved the final manuscript.
⁎ Corresponding author. Disciplina de Anestesiologia, Dor e Terapia Intensiva, Universidade Federal de São Paulo, Rua Napoleão de Barros, 715-6° andar,
Vila Clementino, 04024-002 São Paulo, Brazil. Tel.: +55 1155764069; fax: +55 1155757768.
E-mail addresses: murilloassuncao@gmail.com (M. Assunção), akamine@einstein.br (N. Akamine), gscar1@ig.com.br (G.S. Cardoso),
patmello03@yahoo.com (P.V.C. Mello), zemario@atarde.com.br (J.M.M. Teles), murilloassuncao@gmail.com (A.L.B. Nunes), murilloassuncao@gmail.com
(M.O. Maia), reaneto@uol.com.br (Á. Rea-Neto), fmachado.dcir@epm.br (F.R. Machado).

0883-9441/$ – see front matter. Crown Copyright © 2010 Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.jcrc.2010.03.012
546
Conclusion: The prompt recognition of sepsis and its severity is not satisfactory. This difference is
probably due to the difficulty in the recognition of organ dysfunction, which hampers early
identification of septic patients.
Crown Copyright © 2010 Published by Elsevier Inc. All rights reserved.
1. Introduction
Sepsis incidence and mortality are high all over the world
[1-4]. The Brazilian Sepsis Epidemiological Study involved
patients in intensive care units (ICUs) in Brazilian private
and public hospitals, with mortality rates of 33.9%, 46.9%,
and 52.2% for sepsis, severe sepsis, and septic shock,
respectively [5]. In 2008, Sogayar et al [6], while assessing
the costs of sepsis treatment in Brazilian hospitals, reinforced
the data found in the Brazilian Sepsis Epidemiological Study
on the high general mortality rate (43.8%). A further series
involved 75 Brazilian hospitals; and sepsis, severe sepsis,
and septic shock related mortality was 16.7%, 34.4%, and
65.3%, respectively [7]. In a study conducted using data on
patients with severe sepsis and septic shock from several
countries, Brazil presented a 55.0% mortality rate, being the
country with the highest mortality rate [8].
Early treatment through protocols with well-established
therapeutic targets can reduce sepsis-related morbidity and
mortality [9-14]. However, early intervention calls for prompt
recognition by the team managing the patient. The few studies
that have assessed health professionals' capacity to recognize
patients with sepsis suggested knowledge of the concepts of
sepsis and its clinical forms of presentation to be limited
among the health care professionals, with very low rates of
correct answers [15-19]. Most of them involved a limited
series [15-19], and only one [15] assessed the theoretical
knowledge of the 1992 American College of Chest Physi-
cians/Society of Critical Care Medicine consensus definitions
[20]. In that study, a multicenter and international trial with
1058 physicians from several specialties; as few as 22.0% of
the intensivists and 5.0% of the other specialties physicians
were able to define sepsis correctly (P b .0001) [15].
Sepsis can affect any patient, regardless of the presence or
type of comorbidity. Thus, physicians in all specialties
should be aware of the early signs of sepsis. This would
make an early diagnosis and appropriate treatment possible,
leading to a better prognosis in these patients. This study
aimed to assess the capacity of the practicing physicians in
Brazilian hospitals to recognize clinical cases of infection,
systemic inflammatory response syndrome (SIRS), sepsis,
severe sepsis, and septic shock according to the definitions in
the 1992 consensus conference, as well as to correlate the
level of accuracy in such recognition with the demographic
and professional characteristics of the physician.
2. Material and methods
This is a multicenter study involving physicians
practicing in several Brazilian states. Representative hospi-
M. Assunção et al.
tals from all regions were invited for a nationwide
characterization of the physician's sepsis knowledge profile.
Public, private, and university hospitals participated in the
study. Public hospitals were considered to be those in which
the main paying source was the Brazilian Sistema Único de
Saúde (Unified Healthcare System), even if these hospitals
also handle private patients or if these hospitals are with
partnerships. Hospitals whose paying source was the patient
himself/herself or his/her health plan were considered
private. University hospitals were considered to be only
those connected to public or private schools of medicine and
not the ones with medical residency program.
All physicians, including residents, were allowed to
participate in the study by completing a single question-
naire. There were no exclusion criteria. Physicians practic-
ing in the ICU could participate as long as they would not
exceed, in absolute numbers, half the sample provided.
The following data were collected to characterize the
responding physicians' year of graduation, medical residen-
cy background, specialty, and type of hospital in which they
work the most, using the same definitions used in the study
center classification. Intensivist physicians were considered
to be those that specified intensive care as their primary
specialty, whether or not they were board certified in critical
care medicine by the Associação Medicina Intensiva
Brasileira (AMIB; Brazilian Intensive Care Association)
and whether or not they had completed a critical care
medicine fellowship. Work in the emergency department
(ED) was defined as performance of a systematic on-call
duty in the ED, although this was not their main professional
activity. Intensivists and current residents were not included
in this group.
Confidentiality was strictly maintained, and the subjects
were informed that completion of the questionnaire would be
equivalent to obtaining an informed consent. The study was
approved by the coordinating center' Research Ethics
Committee at Federal University of São Paulo.
The questionnaire was composed of demographic data; the
medical activity characteristics; and, in separate sheets, 5
fictitious medical cases involving patients that presented with
infection, SIRS, sepsis, severe sepsis, and septic shock (see
additional data file 1 for the original questionnaire). Before
the study began, 5 AMIB board-certified intensivists, all of
them considered experts in sepsis, validated the question-
naire. Those physicians achieved 100% correct diagnosis.
The principal investigator in each center was instructed
on the application of the questionnaires through a manual.
Every center received 50 questionnaires, with no autonomy
to reproduce them; additional copies when needed were
requested from the coordinating center. The questionnaire's
Survey on physicians' knowledge of sepsis 547

sheets were handed to the participant by the center Four were public; 9, private; and 8, university hospitals. One
coordinator or someone he or she assigned after training. thousand and two hundred questionnaires were sent to these
The sheets were distributed in a strict order: identification hospitals. Nine hundred seventeen (76.4%) questionnaires
card, medical cases of SIRS, septic shock, sepsis, severe were completed between January and April of 2005.
sepsis, and infection. Upon completion of each sheet, the The characteristics of the population studied are demon-
participant would put it in an envelope. The participant was strated in the Table 1. Physicians specialties and number of
not allowed to review the answers. As soon as the last sheet participants per specialty were as follows: intensive care, 228
was completed, the envelope was sealed with no identifi- (25%); surgery, 175 (19.2%); internal medicine, 154
cation of the participant. This method was adopted to (16.9%); cardiology, 107 (11.7%); anesthesiology, 34
ensure confidentiality. (3.7%); nephrology, 22 (2.4%); gynecology, 21 (2.3%);
gastroenterology, 16 (1.8%); pulmonary, 15 (1.6%); hema-
2.1. Statistical analysis tology, 7 (0.8%); rheumatology, 7 (0.8%); endocrinology, 6
(0.7%); and other specialties, 119 (13%). Among the
The study sample was based on a predicted incidence of nonintensivist physicians, 414 (60.1%) had on-call duties
correct answer considering an infinite population and not in in the ED, 140 (33.9%) of which were residents. Therefore,
comparison between different groups. The hypothesis was 274 made up the group considered as working in the ED.
that 70% of this infinite population would provide an The percentage of correct answers according to the
incidence of 70% correct answers, with a 3% sampling diagnosis is demonstrated in Table 2. Sepsis was most
error and 95% confidence interval. The number of samples frequently misdiagnosed as infection (66.5%), that is,
was estimated at 896. misdiagnosed as the presence of infection without clinical
In regard to the questions, only one was considered and laboratory signs of inflammatory response. In cases of
correct. The erasure answers were considered incorrect. The severe sepsis, misdiagnosis varied from sepsis (17.5%),
results were demonstrated using percentages and number of septic shock (12.1%), and infection (11.12%).
correct answers, varying from 1 to 5. In addition, in an Univariate analysis showed a significantly better perfor-
assessment as a continuous variable, the global score was mance in recognizing all diagnosis by intensivists than
considered to be the sum of correct answers to the 5
questions, each one of which was assigned a score of 1 or Table 1 Demographic and professional characteristics of the
0, resulting in a highest global score of 5. Those values responding physicians
were expressed as mean ± standard deviation and as median Variables
(25%-75% interval). Age (y) 35.0 ± 8.19
The global score variable was subject to a variance Sex
homogeneity test (Bartlett test) and to the D'Agostino and Male 68.2% (623)
Pearson normality test, identifying nonnormal distribution. Female 31.8% (290)
Thus, in the analyses of subgroups, the Mann-Whitney or Graduation
Kruskal-Wallis test, followed by the Dunnett test in case of Before or in 1992 32.7% (298)
more than 2 groups, was used. Pearson correlation was used After 1992 67.3% (613)
to correlate the global score with the respondent's age. In the Previous medical residency
categorized analysis (number of correct answers), as well as Yes 93.4% (854)
in the percentage of correct answers to each of the specific No 6.6% (60)
Resident physicians
questions, either Pearson χ2 test or Fisher exact test was
Yes 18.6% (174)
used, followed by orthogonal partitioning of χ2 . The
No 72.8% (668)
variables correlated with correct answers to the question Medical specialty
about sepsis and severe sepsis in the univariate analysis (P b Intensivist 24.9% (228)
.05) underwent multivariate analysis by logistic regression AMIB boarded 53.5% (121)
with the forward method. The odds ratio was calculated for Non–AMIB boarded 46.5% (105)
each of the professional categories involved. Nonintensivist 75.1% (789)
Results with values of P b .05 were considered Activity in the ED a
statistically significant. GraphPad (La Jolla, CA) Prism 5 Yes 56.8% (274)
for Windows version 5.0-2007 and SPSS (Chicago, IL) 15.0 No 42.5% (205)
package for Windows were used for statistical analysis. Prevailing dedication
Public institution 24.6% (224)
University institution 39.9% (364)
3. Results Private institution 35.5% (324)
a
Intensivist physicians and residents excluded. All results are
A total of 21 hospitals participated in this study, 17 shown in percentage (number) except for age, which is shown in
mean ± standard deviation.
located in the state capitals and 4 in the interior of the states.
548 M. Assunção et al.

Table 2 Rate of correct answers for each diagnosis according to professional characteristics
Diagnosis Infection SIRS Sepsis Severe sepsis Septic shock
Global 92.6% (849) 78.2% (717) 27.3% (250) 56.7% (520) 81.0% (743)
Medical specialty
Intensivist 96.9% (221) 91.7% (209) 48.2% (110) 71.9% (163) 86.0% (196)
Nonintensivist 91.1% (628) 73.7% (508) 20.3% (140) 51.7% (356) 79.4% (547)
P .004 ⁎ b.001 ⁎ b.001 ⁎ b.001 ⁎ .028 ⁎
Intensivist
Boarded 96.7% (117) 93.4% (113) 56.2% (68) 76.9% (93) 82.6% (100)
Nonboarded 97.1% (102) 90.5% (95) 39.0% (41) 66.7% (70) 90.5% (95)
P N.999 † .420 ⁎ .010 ⁎ .088 ⁎ .088 ⁎
Activity in ED
Yes 94.5% (259) 70.4% (193) 16.1% (44) 51.8% (142) 77.7% (213)
No 88.3% (181) 74.1% (152) 21.0% (43) 49.8% (102) 76.1% (156)
P .014 .371 .167 .654 .673
Resident physicians
Yes 90.8% (158) 88.5% (154) 31.6% (55) 59.2% (103) 85.6% (149)
No 93.1% (622) 77.5% (518) 26.9% (180) 56.7% (379) 79.5% (531)
P .299 ⁎ .001 ⁎ .222 ⁎ .559 ⁎ .067 ⁎
Type of institution
Public 92.0% (206) 70.1% (157) 19.6% (44) 50.0% (112) 77.2% (173)
Private 70.1% (338) 76.9% (280) 26.6% (97) 58.5% (213) 80.5% (293)
University 92.6% (300) 85.5% (277) 33.0% (107) 59.3% (192) 81.0% (273)
P .922 ⁎ b.001 ⁎ .002 ⁎ .066 ⁎ .113 ⁎
Data shown in percentage (number) of correct answers.
⁎ Pearson χ2.
†
Fisher exact test.

nonintensivists. Boarded intensivists performed better than
nonboarded in sepsis diagnosis. There was a significant
difference between physicians with or without activity in ED
only in promptly recognizing infection. Resident physicians
performed better than nonresidents when diagnoses were
SIRS and septic shock.
Regarding institutions in which physicians work, partition
χ2 analysis showed a significantly higher rate of correct
answers for the SIRS diagnosis among physicians from
university hospitals as compared with public (P b .001) and
private ones (P = .004). For sepsis and severe sepsis, the rate
of correct answers was higher among physicians from
university hospitals as compared with public ones (P = .001
and P = .032, respectively), with a difference also in favor of
private hospitals in relation to the latter (P = .053 and P =
.044, respectively).
In the multivariate logistic regression, 2 variables
remained related to sepsis recognition: the specialty
(AMIB board-certified intensivists: OR = 5.031 [3.440-
7.358] and not AMIB board-certified: OR = 2.512 [1.645-
3.838], in relation to nonintensivists) and the type of
institution (university: OR = 2.017 [1.353-3.006] and
private: OR = 1.486 [0.994-2.221], in relation to public).
Similarly, in regard to severe sepsis, both the institution and
the specialty were considered significant (institutions—
private: OR = 1.411 [1.010-1.971] and university: OR =
1.455 [1.033-2.049]; specialty—boarded: OR = 3.107
[2.017-4.786] and nonboarded: OR = 1.871 [1.220-2.869]).
Categorized distribution of scores among all physicians
was as follows: score 0, 0.2% (2); score 1, 3.7% (34); score 2,
18.3% (168); score 3, 32.1% (294); score 4, 29.0% (266);
and score 5, 16.7% (153). Between resident and nonresident
physicians, there was a difference in the distribution of
scores (P = .006), specifically for scores 2 (P = .001) and 4
(P = .001). A significant difference was also found among
board-certified intensivists, non–board-certified intensivists,
and other specialties physicians (P b .001, Fig. 1), but not
among physicians with and without ED activity (P = .861).
The distribution of scores in relation to the type of hospital
was significantly different (P b .001, Fig. 2).
Considering the global score, a mean value of 3.36 ± 1.08
was obtained. There was no correlation between mean score
and age (r = −0.235, P b .0001) or sex (median, 3.0 [3.0-4.0]
and 3.0 [3.0-4.0], respectively; P =.870). Resident physicians
performed better in relation to nonresidents (medians, 4.00
[3.00-4.00] and 3.00 [3.00-4.00], respectively; P = .012), as
did intensivists compared with nonintensivists (medians,
4.00 [3.00-5.00] and 3.00 [2.00-4.00], respectively; P b
.001) (Fig. 3). There was no difference in the global score in
relation to physicians with and without ED activity (median,
3.00 [2.00-4.00] and 3.00 [2.00-4.00], respectively; P =
.875). When comparing the scores among physicians from
different types of hospital, a significant difference was noted
favoring university hospitals (public, 3.00 [2.00-4.00];
private, 3.00 [3.00-4.00]; university, 4.00 [3.00-4.00]; P =
.005) (Fig. 4).
Survey on physicians' knowledge of sepsis
Fig. 1 Score distribution according to medical specialty.
Nonintensivists vs boarded intensivists (score 2: *P = .001,
score 3: **P = .010, and score 5: ***P b .001) and vs
nonboarded intensivists (score 2: #P b .001, score 3: ##P = .001,
and score 5: ###P b .001). Orthogonal partition of χ2.
4. Discussion
Our results demonstrated that physicians' knowledge of
sepsis and severe sepsis concepts is unsatisfactory, even
among intensivists and those working in the ED. This study
showed lower rates of correct answers in regard to sepsis
(27.3%) and to severe sepsis (56.7%). Despite the differences
in the study design and in the sample number, those rates are
better than the ones reported by Poeze et al [15], in which as
few as 22.0% of the intensivists and 5.0% of the physicians
from other specialties were able to define sepsis correctly
[15,20]. Fernandez et al [17] showed that 31.4% of the
sample was able to recognize SIRS. Pizzolatti et al [16]
showed, in a small sample of Brazilian responding
physicians (n = 25), that 62.0% and 74.0% were able to
correctly define SIRS and sepsis, respectively [16]. Those
data, different from the ones herein, can be explained once
they involved mainly boarded intensivists in university
Fig. 2 Score distribution according to type of hospital. University
vs public hospitals (score 2: *P b .001, score 4: **P b .001, score 5:
***P = .008), university vs private (score 2: #P = .081 and score 4:
##
P b.001), and public vs private (score 5: ∫P = .020). Orthogonal
partition of χ2.
549
Fig. 3 Global score comparison according to medical specialty.
Results shown in median, 25% to 75%, and minimum to maximum.
Kruskal-Wallis: P b .001. *Dunnett: P b .001 for nonintensivists
(3.00 [2.00-4.00]) vs nonboarded intensivists (4.00 [3.00-5.00]);
P b .001 for nonintensivists vs boarded intensivists (4.00 [3.00-
5.00]); and P = .314 for nonboarded intensivists vs boarded.
centers. Besides, the small size of the sample may have
compromised a more accurate analysis.
The low rate of correct answers for severe sepsis (56.7%)
must be related to the difficulty in recognizing organ
dysfunction as an indicator of the severity of illness because
the validation process and the presence of 2 well-
characterized dysfunctions preclude problems related to
the clinical case proposed in the questionnaire. Interesting is
the finding that what many believe to be sepsis is actually
the disease with the presence of organic disorder, that is,
severe sepsis, as most of the mistakes in the severe sepsis
question occurred through the answer “sepsis.” It is true that
many clinicians use the words sepsis and severe sepsis
synonymously, although this mistake does not necessarily
lead to equal treatment in both situations. Although we did
not specifically ask respondents what they thought the
patient's risk of death was or what should be the clinical
conduct in each case, this finding may be a sign that
physicians have difficulty in recognizing the presence of
organ dysfunction as an alert for emergency care. As this
Fig. 4 Global score comparison according to work place.
Kruskal-Wallis: P = .005. *Dunnett: P b .001 for public (3.00
[2.00-4.00]) vs university (4.00 [3.00-4.00]); P = .039 for public vs
private (3.00 [3.00-4.00]); and P = .007 private vs university.
550
possibility exists, in practical terms, it can represent a
challenge in terms of educational strategies aimed at sepsis
mortality reduction.
In regard to sepsis, the issue probably is the differentiation
from infection with no signs of inflammatory response,
which was the selection made by the majority of those who
gave the incorrect diagnosis. However, it is not clearly
shown that the presence of SIRS signs per se would imply
increased morbidity and death risk, although a study
demonstrated that the association of 3 or 4 SIRS criteria
would imply increased mortality [21]. However, other
studies already showed that the presence of SIRS did not
change mortality rates of infection [22-24]. Could that
actually be considered a risk factor, this limited knowledge
of the sepsis concept may also have more severe implications
in prognostic terms, assuming relevance also in regard to
educational policies.
The intensivist physicians presented a better performance
as compared with nonintensivists. This had been expected
because sepsis is a disease of high prevalence and incidence
in the Brazilian ICU. However, the rates of correct answers
were quite low for sepsis (48.2%), severe sepsis (71.9), and
even septic shock (86.0%). These results are better than the
ones reported by Poeze et al (2004), where as few as 22.0%
of intensivist physicians were able to define sepsis as per the
consensus. The differences in the study design, fictitious
medical cases attempting to depict the reality instead of
telephone interviews focusing on the concepts, may have
facilitated the development of the diagnostic hypothesis; that
is, perhaps, mastering a concept is more difficult than
assessing a medical case. Differently from the intensivists, it
was not possible to detect any difference between physicians
with and without ED activity in regard to recognizing sepsis
and severe sepsis concepts. Such situation can be worrisome,
as it might reflect a difficulty in recognizing the clinical
scenario of organ dysfunction. This category of physicians,
like the intensivists, is in contact with a large portion of the
population with sepsis.
Ziglam et al [18] showed that 48.0% and 67.0% of the
resident physicians were able to identify severe sepsis and
septic shock; and these results are less satisfactory than the
ones found in our study (59.2% and 85.6%, respectively).
Furthermore, in regard to the global score, resident
physicians performed better as compared with nonresidents,
probably for being in the educational phase, for a large
portion of them, in university centers. This suggests that the
current medical education process may become responsible
for improved knowledge and, as a result, for recognition of
the different clinical forms of presentation of this disease; it
may even play a role in mortality reduction in the future.
The physicians from public hospitals showed a lower
global score than the other ones. In Brazil, several studies
have demonstrated high sepsis-related mortality [5,7,25,26],
particularly in public hospitals [5,25]. There are many
possible reasons for this higher mortality in public hospitals.
However, disease severity and treatment resources did not
M. Assunção et al.
seem to be different, as stated in those studies [5,25],
although another Brazilian study of a small sample of
patients clearly showed that delay in sepsis recognition in a
public hospital was associated with higher mortality [27].
This may be one of the possible reasons for this finding, and
we can hypothetically suppose that absence of knowledge
may be one of the reasons for this delayed recognition of
organ dysfunction.
This study has some strengths. First, the questionnaire
validation by 5 board-certified intensivists was, in our
opinion, adequate. This validation process occurred in just
one of the studies cited, but with no description of the
process itself [17]. Likewise, the application process was
strictly established. Aiming to maintain data homogeneity,
each center's coordinator was instructed on how to apply the
questionnaire. Second, this is one of the largest reported
series on physicians' knowledge of sepsis concepts. Previous
studies have assessed from 25 up to 144 individuals [15-19].
This one achieved almost 1000 physicians assessed,
accounting for 76.4% of the total of questionnaires
forwarded, with minimum losses, which makes the data
quite robust when compared with other studies [28,29].
Third, it evaluated the assimilation of concepts and not only
their definitions. Most previous studies have assessed mainly
the theoretical concepts of knowledge, not evaluating its
correlation with the clinical practice through the use of
clinical cases [15-18]. Fourth, this study approached, in a
single sample, residents, intensivists, and nonintensivist
physicians and those with ED activity, representing a large
portion of the medical team directly responsible for medical
assistance. The subjects included in the previous studies
were variable. Some included nonresidents, intensivists, and
nonintensivist physicians [15]; one study included only
resident physicians [18]; one included ED and intensive care
physicians [16]; one included resident and nonintensivist
physicians [17]; and one study studied the knowledge of
ward nurses [19].
On the other hand, the study has some limitations. First,
assessing knowledge through clinical cases, particularly
because of the small number of questions, may be
considered inadequate from the educational point of view.
Although clinical cases are more realistic than a test of
theoretical knowledge, we cannot firmly infer that the
performance of the physicians that completed the question-
naire would be the same within a real setting. Moreover,
our study did not measure physicians' knowledge about
sepsis or its management. It did measure the ability of
physicians to articulate the specific diagnostic criteria of the
1991 consensus conference, which is not necessarily a sign
of a lack of knowledge or a lack of recognition of the
clinical entity. However, it has to be pointed out that
knowing the concepts may be an important step to
recognizing the disease in the clinical scenario. Taking
this into account, it is also a limitation regarding the stated
superior performance of intensivists, as it might not be a
reflection of greater diagnostic acumen but simply a greater
Survey on physicians' knowledge of sepsis
familiarity with the specific criteria of the American
College of Chest Physicians/Society of Critical Care
Medicine consensus conference. However, the low rates
of correct answers suggest an actual knowledge deficiency.
Second, our series may not adequately represent the
national reality considering that the centers were located
in the capitals and hospitals were previously selected to
participate in a multicenter study. In a way, this may mean
a better quality of the medical staff in those hospitals in
relation to the national reality.
5. Conclusions
Our main conclusion is that, although the medical
knowledge of the concepts of SIRS, infection, and septic
shock is satisfactory, the same cannot be said in regard to the
concepts of sepsis and severe sepsis. This difference is
probably due to the difficulty in the recognition of organ
dysfunction per se and to the lack of knowledge that the
presence of organ dysfunction defines severe sepsis. The
intensivist physicians and residents in training had a better
performance when compared with ED physicians. The
physicians practicing mainly on Brazilian public hospitals
had the worst performance, which may play a role in the
higher sepsis-related mortality found in this type of hospitals.
Acknowledgments
SEPSES study group: Hospital São Paulo, São Paulo-SP,
Murillo Assunção. Hospital Israelita Albert Einstein, São
Paulo-SP, Fabio Machado Santana. Hospital São Lucas,
PUC, Porto Alegre-RS, Fernando Dias. Hospital Português,
Salvador-BA, José Mário Meira Teles. Hospital Universi-
tário da Faculdade de Medicina de Catanduva, Catanduva-
SP, Jorge Valiatti. Hospital Sírio Libanês, São Paulo-SP,
Guilherme Schettino. Hospital de Clínicas-UFRP, Curitiba-
PR, Álvaro Réa Neto. Hospital das Nações, Curitiba-PR,
Nazah Youssef. Hospital do Trauma, Curitiba-PR, Mirella
Cristine de Oliveira. Hospital Bandeirantes, São Paulo-SP,
Mário Lúcio Baptista Filho. Hospital de Terapia Intensiva-
UESPI, Teresina-PI, Patrícia Veiga de C. Mello. Hospital de
Base São José do Rio Preto, São José do Rio Preto-SP,
Suzana Lobo. Irmandade Santa Casa de Misericórdia de
Cruzeiro, Cruzeiro-SP, Bruno Mazza. Hospital Santa Luzia,
Brasília-DF, Marcelo Maia de Oliveira. Irmandade Santa
Casa de Misericórdia de Passos, Passos-MG, Marcel
Resende. Hospital das Clínicas-USP, São Paulo-SP, Roberto
de Cleva. Hospital Beneficência Portuguesa, São Paulo-SP,
Haggeas Silveira Fernandes. Irmandade Santa Casa de
Misericórdia de Porto Alegre, Porto Alegre-RS,Gilberto
Friedman. Hospital Universitário-UFPB, João Pessoa-PB,
Ciro Leite Mendes. Hospital São Camilo, São Paulo-SP,
André Nunes. Hospital Pró-Cardíaco, Rio de Janeiro-RJ,
Rubens Costa Filho.
551
Appendix A. Supplementary data
Supplementary data associated with this article can
be found, in the online version, at doi:10.1016/j.jcrc.2010.
03.012.
References
[1] From the Centers for Disease Control. Increase in National Hospital
Discharge Survey rates for septicemia—United States, 1979-1987.
JAMA 1990;263:937-8.
[2] Annane D, Aegerter P, Jars-Guincestre MC, et al. Current epidemi-
ology of septic shock: the CUB-Rea Network. Am J Respir Crit Care
Med 2003;168:165-72.
[3] Alberti C, Brun-Buisson C, Burchardi H, et al. Epidemiology of sepsis
and infection in ICU patients from an international multicentre cohort
study. Intensive Care Med 2002;28:108-21.
[4] Padkin A, Goldfrad C, Brady AR, et al. Epidemiology of severe
sepsis occurring in the first 24 hrs in intensive care units in
England, Wales, and Northern Ireland. Crit Care Med 2003;31:
2332-8.
[5] Silva E, Pedro Mde A, Sogayar AC, et al. Brazilian Sepsis
Epidemiological Study (BASES study). Crit Care 2004;8:
R251-60.
[6] Sogayar AM, Machado FR, Rea-Neto A, et al. A multicentre,
prospective study to evaluate costs of septic patients in Brazilian
intensive care units. Pharmacoeconomics 2008;26:425-34.
[7] Sales J, David C, Souza P, et al. An epidemiological study of sepsis in
ICUs: Sepsis Brazil Study. Critical Care 2006;10:P111.
[8] Beale R, Reinhart K, Brunkhorst FM, et al. Promoting Global
Research Excellence in Severe Sepsis (PROGRESS): lessons from an
international sepsis registry. Infection 2009.
[9] Ventilation with lower tidal volumes as compared with traditional tidal
volumes for acute lung injury and the acute respiratory distress
syndrome. The Acute Respiratory Distress Syndrome Network. N
Engl J Med 2000;342:1301-8.
[10] Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of
recombinant human activated protein C for severe sepsis. N Engl J
Med 2001;344:699-709.
[11] van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin
therapy in the critically ill patients. N Engl J Med 2001;345:
1359-67.
[12] Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign
guidelines for management of severe sepsis and septic shock. Crit Care
Med 2004;32:858-73.
[13] Dellinger RP, Levy MM, Carlet JM, et al. Surviving Sepsis Campaign:
international guidelines for management of severe sepsis and septic
shock: 2008. Intensive Care Med 2008;34:17-60.
[14] Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in
the treatment of severe sepsis and septic shock. N Engl J Med
2001;345:1368-77.
[15] Poeze M, Ramsay G, Gerlach H, et al. An international sepsis survey: a
study of doctors' knowledge and perception about sepsis. Crit Care
2004;8:R409-13.
[16] Pizzolatti ML, Moritz RD, de Andrade J. Avaliação do conhecimento
dos profissionais da área de medicina de urgência sobre os critérios de
definição de SIRS, sepse, sepse grave e choque séptico. RBTI
2004;16:210-4.
[17] Fernandez R, Boque M, Rodriguez G, et al. Sepsis: a study of
physician's knowledge about the Surviving Sepsis Campaign in Puerto
Rico. Critical Care 2006;10:P126.
[18] Ziglam HM, Morales D, Webb K, et al. Knowledge about sepsis
among training-grade doctors. J Antimicrob Chemother 2006;57:
963-5.
552
[19] Robson W, Beavis S, Spittle N. An audit of ward nurses' knowledge of
sepsis. Nurs Crit Care 2007;12:86-92.
[20] Bone RC, Balk RA, Cerra FB, et al. Definitions 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;101:1644-55.
[21] Rangel-Frausto MS, Pittet D, Costigan M, et al. The natural history of
the systemic inflammatory response syndrome (SIRS). A prospective
study. JAMA 1995;273:117-23.
[22] Alberti C, Brun-Buisson C, Goodman SV, et al. Influence of systemic
inflammatory response syndrome and sepsis on outcome of critically
ill infected patients. Am J Respir Crit Care Med 2003;168:77-84.
[23] Alberti C, Brun-Buisson C, Chevret S, et al. Systemic inflammatory
response and progression to severe sepsis in critically ill infected
patients. Am J Respir Crit Care Med 2005;171:461-8.
[24] Dremsizov T, Clermont G, Kellum JA, et al. Severe sepsis in
community-acquired pneumonia: when does it happen, and do systemic
View publication stats
M. Assunção et al.
inflammatory response syndrome criteria help predict course? Chest
2006;129:968-78.
[25] Sogayar AMC, Machado FR, Rea-Neto A, et al. COSTS—
multicenter, prospective study to evaluate costs of septic patients in
Brazilian intensive care units. Pharmacoeconomics, in press.
[26] Beale R, Reinhart K, Silva E, et al. Comparison of PROGRESS
severe sepsis registry patients to INDEPTH integrated severe sepsis
clinical trial database placebo patients. Chest 2004;126:864S.
[27] Freitas FG, Salomao R, Tereran N, et al. The impact of duration of
organ dysfunction on the outcome of patients with severe sepsis and
septic shock. Clinics (Sao Paulo) 2008;63:483-8.
[28] Sivayoham N. Management of severe sepsis and septic shock in the
emergency department: a survey of current practice in emergency
departments in England. Emerg Med J 2007;24:422.
[29] Srinivasan A, Song X, Richards A, et al. A survey of knowledge,
attitudes, and beliefs of house staff physicians from various specialties
concerning antimicrobial use and resistance. Arch Intern Med
2004;164:1451-6.