Alqarni et al.

Critical Care (2018) 22:175

https://doi.org/10.1186/s13054-018-2099-5

RESEARCH Open Access

Clinical characteristics and prognosis of

bacteraemia during postoperative intra-
abdominal infections
Adel Alqarni1,2, Elie Kantor1,3, Nathalie Grall4,5, Sebastien Tanaka1,6, Nathalie Zappella1, Mathieu Godement1,
Lara Ribeiro-Parenti3,7, Alexy Tran-Dinh1,3,8 and Philippe Montravers1,3,9*

Abstract
Background: Bloodstream infections of abdominal origin are usually associated with poor prognosis. We assessed
the clinical and microbiological characteristics of critically ill patients admitted to the intensive care unit (ICU) for
postoperative intra-abdominal infection (PIAI) and analysed the influence of bacteraemia on their outcome.
Methods: All consecutive PIAI patients admitted to the ICU between 1999 and 2014 were prospectively analysed.
Bacteraemic patients (at least one positive blood culture in the 24 h preceding/following surgery) were compared
with non-bacteraemic patients. Demographic characteristics, underlying disease, severity scores at the time of
reoperation, microbiological results, therapeutic management, outcome, and survival were recorded. Results are
expressed as median (interquartile range (IQR)) or proportions.
Results: Overall, 343 patients (54% male, 62 (49–73) years old) with PIAI were analysed, including 64 (19%)
bacteraemic patients. Immunosuppression and cancer were more frequent in bacteraemic patients (p < 0.001 in both
cases). No difference between groups was observed for the characteristics of initial surgery. Time to reoperation, site,
and cause of PIAI were similar in both groups. At the time of reoperation, Sequential Organ Failure Assessment (SOFA)
score was higher in bacteraemic patients (8 (6–10) versus 7 (4–10); p < 0.05). A predominance of Gram-positive (34%)
and Gram-negative (47%) bacteria were recovered from blood cultures (polymicrobial bacteraemia in 9 (14%) patients
and bacteraemia involving multidrug-resistant organisms in 14 (22%) patients). In multivariate analysis, risk factors for
bacteraemia were immunosuppression or cancer, high SOFA score, and E. coli in peritoneal samples. Bacteraemia did
not impact the management (with similar results for the adequacy of antibiotic therapy, anti-infective agents used,
de-escalation or duration of therapy in both groups). Neither hospital mortality nor morbidity criteria differed between
groups. Risk factors for mortality in multivariate analysis were urgent initial surgery, high Simplified Acute Physiology
Score (SAPS) II score and documented antifungal therapy, but not perioperative bacteraemia.
Conclusions: In this ICU population, bacteraemia did not change the overall management of patients with PIAI. Our
data suggest that bacteraemic patients do not require a specific management.
Keywords: Intra-abdominal infection, Postoperative infection, Bacteraemia, Outcome

* Correspondence: [email protected]
1
Département d’Anesthésie-Réanimation, CHU Bichat-Claude Bernard, APHP,
Paris, France
3
Université Paris Diderot, Press Sorbonne Cité, Paris, France
Full list of author information is available at the end of the article

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Alqarni et al. Critical Care (2018) 22:175 Page 2 of 10

Background surgery for PIAI. In case of positive blood culture for

Bacteraemia is a major cause of morbidity and mortality coagulase-negative staphylococci (CNS), two positive
in critically ill patients [1–3], defined as primary when no blood cultures with the same microorganism and the
source of infection is identified or secondary to dissemin- same antibiotic susceptibility were required to establish
ation of pathogens from other sites of infection [4]. One the diagnosis of bacteraemia or one positive blood culture
of the most common causes of secondary bacteraemia is matching the same microorganism with the same anti-
intra-abdominal infection, accounting for 11–15% of all biotic susceptibility in the peritoneal samples. Medical
secondary cases in critically ill patients [4, 5]. charts were retrospectively reviewed. Demographic char-
Bacteraemia secondary to intra-abdominal surgical site acteristics, underlying medical conditions, and their sever-
infection is not uncommon, with a reported incidence of ity according to the McCabe score [13] were recorded.
10–26% at the time of reoperation [6, 7] and increased The type of initial surgery and presence of antimicrobial
rates in the case of delayed diagnosis [8], septic shock therapy prior to PIAI surgery were also recorded. Ana-
[6], or multidrug-resistance [9]. Patients who develop tomical location above or below the transverse mesocolon
postoperative intra-abdominal infection (PIAI) are at and source of PIAI were noted.
higher risk of adverse outcome and mortality. Interest- Severity scores (Simplified Acute Physiology Score
ingly, no specific data are available in the literature con- (SAPS) II and Sequential Organ Failure Assessment
cerning bacteraemia during PIAI. In a retrospective (SOFA) scores) were obtained at the time of reoperation
single-centre cohort study of 96 critically ill patients, De for PIAI [14, 15]. Temperature, white blood cell counts,
Waele et al. found that the mortality of patients with serum creatinine, and SOFA score were assessed on the
bacteraemia of abdominal origin was high (62.5%) and day of reoperation (day 0) and at days 3 and 7 after sur-
higher than the Acute Physiology and Chronic Health gery for patients still in the ICU. Differences (delta) be-
Evaluation (APACHE) II-based expected mortality rate tween the SOFA scores observed on day 3 and day 0
[10]. These investigators found that older age and renal were calculated. Patients meeting the following three cri-
failure were independent risk factors of death. However, teria on day 3 were arbitrarily defined as improving:
only 39% of their population had secondary peritonitis SOFA score ≤ 3 points, white blood cell counts ≥ 5000
and the proportion of PIAI cases was not specified. and ≤ 15,000 cells/mm3, and temperature ≥ 36 °C
Another controversial issue concerns the duration of and ≤ 38 °C. Similar analyses were performed at day 7 to
antibiotic therapy in bacteraemic patients. Prolonged ther- compare changes in these criteria between day 7 and day 0.
apy has been reported in patients treated for bacteraemia
[11], which can be explained by the fear of increased treat- Microbiological examinations
ment failure rates in the case of short-course antibiotic Blood cultures were systematically collected during
therapy [12]. However, insufficient evidence is available in the management of PIAI. All blood culture bottles
patients with peritonitis. were handled using the laboratory’s blood culture sys-
The objective of this study was to analyse the impact of tem (BD BACTEC™; Becton-Dickinson). Positive blood
bacteraemia on morbidity and mortality in critically ill PIAI cultures drawn within 24 h before and after reopera-
patients. We hypothesised that outcome is determined by tion for PIAI were noted. Diagnostic procedures on
both clinical and microbiological factors in this population. all positives bottles included Gram stain and bacterial
isolation and identification using standard bacteriology
Methods techniques. The antibiotic susceptibility of each strain
Study population was determined using the disk diffusion method
Between January 1999 and December 2014, all consecu- (http://www.sfm-microbiologie.org).
tive intensive care unit (ICU) patients admitted to our Peritoneal fluid samples were systematically collected
surgical ICU for management of a first episode of PIAI during surgery and were immediately sent to the bacteri-
were prospectively included in a database. This observa- ology laboratory. Samples were processed according to the
tional study was conducted according to the terms and laboratory’s standard methods. Plates were incubated for
regulations of the local institutional review boards 48 h at 35 °C. All morphologically distinct colonies were
(CEER CHU Bichat, Paris Diderot University, Paris, identified and tested for antibacterial susceptibility by the
agreement 10–008). According to French Law, no in- disk diffusion method (http://www.sfm-microbiologie.org)
formed consent was required since this observational [7]. Multidrug-resistant (MDR) bacteria were defined as
study did not modify the physician’s treatment decisions. those resistant to three or more antimicrobial classes [16].

Data collection Anti-infective therapy

Bacteraemic patients were defined as having at least one Empirical antibiotic therapy (EAT) was started at the
positive blood culture in the 24 h preceding or following time of reoperation for PIAI. The choice of therapy took
Alqarni et al. Critical Care (2018) 22:175 Page 3 of 10

into account the severity of the case, previous antibiotic slightly more severe at the time of diagnosis as illus-
therapies, and local epidemiology. According to French trated by the SOFA score (Table 1). The median SAPS II
guidelines, EAT included a combination of broad- score slowly increased from 46.0 in 1999 to 54.0 in 2014.
spectrum beta-lactams such as piperacillin/tazobactam or
imipenem/cilastatin associated with aminoglycosides and Microbiological results
vancomycin [17]. EAT, changes at the time of documented Overall, 102 positive blood cultures were collected in
therapy including escalation and de-escalation, and total bacteraemic patients, yielding 64 microorganisms with a
duration of therapy were recorded [7]. Empirical antifun- median of 1 (range 1–8) positive blood cultures per
gal agents were administered when a high risk of fungal patient (Table 2): 21 (33%) patients with aerobic
infection was suspected and adapted to the identification Gram-positive bacteraemia, 30 (47%) patients with aer-
results of peritoneal samples [17, 18]. obic Gram-negative bacteraemia, 10 (16%) patients with
anaerobic bacteraemia, and 6 (9%) patients with fungae-
Outcome mia. Polymicrobial bacteraemia was observed in 9 (14%)
The primary outcome was ICU mortality from any cause. cases (including 2 aerobic Gram-negative organisms
Medical and surgical complications, suspicion of persist- (n = 2), 2 aerobic Gram-positive organisms (n = 1), 2 an-
ent sepsis [19, 20], additional reoperations for persistence aerobes (n = 2), and 2 fungi (n = 1)). Among these organ-
of the initial infection or superinfections (including MDR isms, 14 bacteraemic patients had MDR bacteria: 6
bacteria), and death between the day following surgery Gram-positive bacteria (3 methicillin-resistant CNS, 1
and discharge from hospital were assessed. Duration of methicillin-resistant S. aureus, and 2 E. faecium) and 8
ICU stay and mechanical ventilation were also assessed. Gram-negative bacteria (6 ESBL or AmpC-producing
Enterobacteriaceae, and 2 Pseudomonas spp). Among the
Statistical analysis microorganisms isolated from blood culture, 29 organisms
Results are expressed as median and interquartile range in 27 patients were not isolated from surgical operative
(IQR) or number and proportions. Statistical analysis was samples: E. coli n = 4, Enterobacter spp. n = 1, Pseudo-
performed with SAS© 9.4 (SAS Institute, Cary, NC, USA). monas spp. n = 3, enterococci n = 1, streptococci n = 4,
Comparisons between bacteraemic and non-bacteraemic CNS n = 1, S. aureus n = 3, and anaerobes n = 10.
patients used Chi-square test and Fisher’s exact test for A total of 1022 microorganisms were isolated from
discrete variables and unpaired Wilcoxon tests for quanti- peritoneal surgical samples (Table 2). Polymicrobial peri-
tative variables. Risk factors for bacteraemia and death toneal cultures were reported in 58/64 (91%) bacter-
were assessed by univariate analysis and unadjusted odds aemic and 228/279 (82%) non-bacteraemic patients,
ratio (OR) and 95% confidence intervals (CIs) were calcu- respectively (p = 0.09). The frequency of bacteraemia
lated. Variables with p values less than 0.15 in univariate was increased in patients with Enterobacteriaceae, E.
analysis were introduced as dependent factors into two coli, and E. faecalis isolated from peritoneal samples
multivariate logistic regression analyses using a stepwise (OR 2.25, 95% CI 1.12–4.52, p = 0.022; OR 2.12, 95%
selection method, and the collinearity between predictors CI 1.22–3.70], p = 0.007; and OR 1.98, 95% CI 1.12–3.52,
was analysed. Bacteraemia constituted the outcome of the p = 0.019; respectively). The proportions of MDR bacteria
first model, and ICU mortality constituted the outcome of cultured from peritoneal samples were similar between
the second model. Bacteraemia was forced in the model of bacteraemic and non-bacteraemic patients (26 (41%) ver-
ICU mortality as our primary factor of interest until the sus 113 (41%) patients, respectively; p = 0.50).
end of the selection process. In addition, the year of inclu-
sion was forced in both analyses to take the time effect Risk factors for bacteraemia
into account. Logistic models were evaluated for discrim- The risk factors for bacteraemia identified in univariate
ination with the C-statistic and for calibration with the analysis are presented in Table 3. In multivariate ana-
Hosmer-Lemeshow test. lysis, after adjustment for the year of inclusion, we iden-
tified three risk factors independently associated with an
Results increased risk of bacteraemia: immunosuppression or
Patient characteristics cancer, SOFA score at the time of reoperation, and E.
Overall, 343 PIAI patients including 64 (19%) bacter- coli isolated from surgical peritoneal samples (Table 3).
aemic patients were analysed. The demographic charac-
teristics and underlying diseases at the time of diagnosis Anti-infective therapy
were similar in both groups, except for increased pro- Empirical therapy did not differ between groups (Table 4).
portions of immunosuppression and cancer in bacter- Similar proportions of adequacy of EAT were achieved in
aemic patients (Table 1). Surgical characteristics were both groups. At the time of documented therapy, propor-
similar in both groups, while bacteraemic patients were tions of escalation, de-escalation, and combination therapy
Alqarni et al. Critical Care (2018) 22:175 Page 4 of 10

Table 1 Characteristics and clinical findings of patients with positive blood cultures and patients without bacteraemia
Variables All patients Bacteraemic patients Patients without Odds ratio 95% CI p value
(n = 343) (n = 64) bacteraemia (n = 279)
Demographic characteristics and underlying diseases
Age (years), median (IQR) 62 (49–73) 63 (52–74) 61 (48–73) 0.92 (0.53–1.58) 0.56
Male gender, n (%) 184 (54) 32 (50) 152 (54) 0.83 (0.49–1.44) 0.52
Fatal underlying disease, n (%) 110 (32) 29 (45) 81 (29) 2.02 (1.16–3.53) 0.012
Immunosuppression, n (%) 121 (35) 37 (59) 84 (30) 3.25 (1.85–5.71) < 0.001
Diabetes mellitus, n (%) 59 (17) 12 (19) 47 (17) 1.14 (0.56–2.29) 0.72
Cancer, n (%) 126 (37) 35 (55) 91 (33) 2.49 (1.43–4.33) < 0.001
Initial surgery
Upper gastrointestinal tract surgery, n (%) 150 (44) 26 (41) 124 (44) 0.85 (0.49–1.48) 0.57
Septic or contaminated surgery, n (%) 123 (36) 25 (42) 98 (37) 1.20 (0.68–2.12) 0.53
Emergency procedure, n (%) 135 (39) 22 (34) 113 (41) 0.77 (0.43–1.35) 0.37
Antibiotics before reoperation for PIAI, n (%) 215 (63) 43 (67) 172 (62) 1.25 (0.70–2.22) 0.45
Interval between initial surgery and 7 (4–12) 8 (4–15) 7 (4–11) 1.09 (0.63–1.89) 0.25
reoperation (days), median (IQR)
Severity criteria at the time of reoperation
SAPS II score, median (IQR) 47 (35–59) 50 (38–63) 47 (34–58) 0.89 (0.52–1.54) 0.076
SOFA score, median (IQR) 8 (4–10) 8 (6–10) 7 (4–10) 1.10 (0.59–1.75) 0.034
a
Haemodynamic failure , n (%) 224 (65) 46 (73) 178 (64) 1.53 (0.83–2.81) 0.16
Respiratory failurea, n (%) 154 (45) 30 (48) 124 (44) 1.13 (0.65–1.96) 0.65
a
Renal failure , n (%) 83 (24) 16 (25) 67 (24) 1.05 (0.56–1.98) 0.86
Source of postoperative peritonitis and surgical observations
Generalised infection, n (%) 79 (23) 20 (31) 59 (21) 1.69 (0.92–3.09) 0.08
Gastroduodenal source of infection, n (%) 74 (22) 11 (17) 63 (23) 0.71 (0.35–1.44) 0.34
Small bowel source of infection 88 (26) 13 (20) 75 (27) 0.69 (0.35–1.34) 0.27
Colonic or rectal source of infection, n (%) 97 (28) 20 (31) 77 (28) 1.19 (0.66–2.15) 0.56
Anastomotic leak, n (%) 118 (34) 25 (40) 93 (33) 1.31 (0.74–2.29) 0.34
Bowel perforation, n (%) 119 (35) 16 (25) 103 (37) 0.57 (0.30–1.05) 0.07
Abscesses, n (%) 61 (18) 15 (23) 46 (16) 1.55 (0.80–2.99) 0.19
No demonstrated cause, n (%) 49 (14) 11 (17) 38 (14) 1.31 (0.63–2.74) 0.46
a
According to levels 3 or 4 of the SOFA score
CI confidence interval, IQR interquartile range, SAPS Simplified Acute Physiology Score, SOFA Sequential Organ Failure Assessment

for the targeted therapy were not different between not reveal any significant difference on day 3 and day 7
groups. In addition, the median duration of therapy was (Table 5). A trend toward increased ICU mortality rate
not different between groups (Table 4). was observed in bacteraemic patients (p = 0.08).
In the group of bacteraemic patients, empirical therapy The mortality rate was similar in bacteraemic patients
did not adequately target all bloodstream microorganisms with MDR bacteraemia and those without MDR bacter-
in 11 (17%) cases. The organisms most frequently not aemia (5/15 (33%) versus 22/49 (45%); p = 0.55). More-
adequately treated were methicillin-resistant CNS (n = 5), over, similar proportions of medical and surgical
penicillinase-producing Enterobacteriaceae (n = 2), Can- complications were observed in both groups. Mortality
dida spp. (n = 2), and E. faecium (n = 1). Interestingly, was observed in 3/6 (50%) cases of fungaemia versus
8 of these patients also had inadequate therapy for 23/58 (40%) of the other bacteraemic patients (p = 0.68).
peritoneal pathogens. No significant difference in terms of outcome was ob-
served between the 11 bacteraemic patients who did not
Patient outcome and risk factors of death receive adequate EAT against bloodstream isolates versus
Comparison of the clinical characteristics of bacteraemic the adequately treated cases: persistent intra-abdominal
patients and those without positive blood cultures did sepsis in 7 (64%) versus 31 (58%), p = 1; one or more
Alqarni et al. Critical Care (2018) 22:175 Page 5 of 10

Table 2 Microbiological results of peritoneal samples and antimicrobial therapy in bacteraemic and non-bacteraemic patients
Microbiological results Bacteraemic patients (n = 64) Patients without bacteraemia (n = 279)
Blood cultures Surgical samples Surgical samples
Total number of microorganisms 75 215 807
Gram-negative aerobic bacteria, n (%) 32 (43) 97 (45) 312 (39)
Enterobacteriaceae, n (%) 27 (36) 86 (40) 254 (31)
Escherichia coli, n (%) 13 (17) 42 (20) 124 (15)
Klebsiella spp., n (%) 3 (4) 10 (5) 34 (4)
Enterobacter spp., n (%) 8 (11) 17 (8) 44 (5)
Other Enterobacteriaceae, n (%) 3 (4) 17 (8) 52 (6)
Non-fermenting Gram-negative bacteria, n (%) 4 () 10 (5) 44 (5)
Pseudomonas aeruginosa, n (%) 3 (4) 9 (4) 39 (5)
Miscellaneous Gram-negative bacilli, n (%) 1 (1) 1 (0.5) 14 (2)
Gram-positive aerobic bacteria, n (%) 22 (29) 73 (34) 308 (38)
Staphylococcus aureus, n (%) 5 (7) 4 (2) 26 (3)
Coagulase-negative staphylococci, n (%) 5 (7) 8 (4) 50 (6)
Streptococcus spp., n (%) 8 (11) 21 (10) 75 (9)
Enterococcus spp., n (%) 3 (4) 38 (18) 151 (19)
Miscellaneous Gram-positive cocci, n (%) 1 (1) 2 (1) 6 (1)
Anaerobic bacteria, n (%) 14 (19) 22 (10) 76 (9)
Bacteroides spp., n (%) 12 (16) 18 (8) 46 (6)
Other anaerobes, n (%) 2 (3) 4 (2) 30 (4)
Multidrug-resistant bacteria, n (%) 14 (19) 26 (12) 113 (14)
Fungi, n (%) 7 (9) 23 (11) 110 (14)
Candida albicans, n (%) 3 (4) 13 (6) 71 (9)
Non-albicans Candida spp., n (%) 3 (4) 9 (4) 32 (4)
Proportions calculated as number of isolates

Table 3 Risk factors for bacteraemia in univariate and multivariate analysis

Univariate analysis Multivariable analysis
Variables Odds ratio 95% CI p value Adjusted odds ratio 95% CI p value
Demographic characteristics and underlying diseases
Fatal underlying disease 2.02 (1.16–3.53) 0.012
Cancer or immunosuppression 2.64 (1.51–4.62) < 0.001 2.59 (1.46–4.59) 0.001
Severity criteria at the time of reoperation
SOFA scorea 1.10 (0.59–1.75) 0.034 1.19a (1.02–1.39) 0.002
Source of postoperative peritonitis and surgical observations
Generalised peritonitis 1.69 (0.92–3.09) 0.08
Interval between initial surgery and reoperationa 1.02 0.98–1.05 0.247
Microbiological results of peritoneal surgical samples
Enterobacteriaceaeb 2.52 (1.23–5.20) 0.010
b
Escherichia coli 2.13 (1.22–3.71) 0.008 2.10 (1.19–3.73) 0.011
The model was adjusted on year of admission
Multivariate logistic regression with stepwise selection: the C-index of the model was 0.69 95% CI 0.61–0.77 and the Hosmer-Lemeshow test p value was 0.377
CI confidence interval, SOFA Sequential Organ Failure Assessment
a
Per 1 day
b
Per 2 points
Alqarni et al. Critical Care (2018) 22:175 Page 6 of 10

Table 4 Empirical and documented therapies in bacteraemic immunosuppression, ongoing cancer, and high SOFA
and non-bacteraemic patients score. The microorganisms isolated from the surgical site
Anti-infective regimens Bacteraemic Patients without p value did not differ between bacteraemic and non-bacteraemic
patients bacteraemia patients except for an increased proportion of E. coli in
(n = 64) (n = 279)
bacteraemic patients. Adequate empirical antimicrobial
Empirical therapy
therapy rates and duration of antibiotic therapy did not
Combination of EAT, n (%) 49 (77) 226 (81) 0.42 differ between bacteraemic and non-bacteraemic patients.
Imipenem/cilastatin, n (%) 22 (34) 69 (25) 0.12 High severity scores, initial emergency surgery, and anti-
Piperacillin tazobactam, n (%) 36 (56) 173 (62) 0.39 fungal therapy were independent risk factors for mortality
Vancomycin, n (%) 30 (47) 122 (44) 0.65 in multivariate analysis.
Few data are available assessing the clinical consequences
Aminoglycosides, n (%) 31 (48) 140 (50) 0.80
of secondary bloodstream infections in peritonitis. Most of
Antifungals, n (%) 21 (33) 101 (36) 0.61
the published reports do not distinguish community-ac-
Fluconazole, n (%) 17 (27) 89 (32) 0.45 quired infections from healthcare-associated infections.
Caspofungin, n (%) 3 (5) 8 (3) 0.43 During bacteraemia in surgical ICU patients, the factors
Adequacy of EAT, n (%) 43 (67) 190 (68) 0.89 associated with poor prognosis are the severity of acute
Documented therapy illness at the onset of bacteraemia, the presence of organ
dysfunction, and Gram-negative bacteria or Candida spp.
Antibiotic treatment 37 (58) 163 (58) 0.93
de-escalation, n (%) infections [21]. One-third of cases of undiagnosed abdom-
Antibiotic treatment 15 (23) 72 (26) 0.69
inal septic focus presented breakthrough infection or re-
escalation, n (%) appearance of positive blood cultures during antibiotic
Combination of EAT, n (%) 38 (59) 190 (68) 0.19 therapy [22]. Polymicrobial bacteraemia is usually suggest-
ive of intra-abdominal sepsis, but other sites are also
Imipenem/cilastatin, n (%) 23 (36) 56 (20) 0.007
associated with several polymicrobial samples [10, 11].
Piperacillin tazobactam, n (%) 20 (31) 98 (35) 0.55
Blood cultures yielding Gram-negative bacilli, enterococci,
Vancomycin, n (%) 16 (25) 90 (32) 0.25 and anaerobes are indicative of gastrointestinal flora, but
Aminoglycosides, n (%) 12 (19) 36 (13) 0.22 staphylococcal bacteraemia is also reported [6–9].
Antifungals, n (%) 20 (31) 117 (42) 0.11 In our cohort, the identified criteria for recognising
Fluconazole, n (%) 16 (25) 104 (37) 0.06 bacteraemic patients during PIAI were poorly discrim-
inative and did not allow any specific management. As
Caspofungin, n (%) 2 (3) 7 (3) 0.78
expected, the most severe cases were those at risk of
Duration of antibiotic therapy 10 (8–14) 10 (8–14) 0.89
in daysa, median (IQR)
bacteraemia. This point has been previously reported, as
bacteraemia associated with intra-abdominal infections
EAT Empirical antibiotic therapy, IQR interquartile range
a
Variable measured in surviving patients appears to be more frequent in patients with septic
shock [6]. The size of the inoculum could also be of
reoperations in 4 (36%) versus 26 (49%), p = 0.52; and importance, as suggested by the increased risk of
death in the ICU in 6 (55%) versus 21 (40%), p = 0.50. bacteraemia in the case of generalised peritonitis and
The risk factors for mortality identified in univariate peritoneal samples yielding E. coli cultures. Severe
analysis are presented in Table 6. In multivariate ana- underlying disease and immunosuppression have been
lysis, after adjustment for the year of inclusion and the previously described as possible risk factors for bacter-
presence of bacteraemia, the following factors were aemia, but these criteria were not assessed in surgical
independently associated with a higher mortality rate: postoperative infections.
urgent initial surgery (p = 0.0003), SAPS II > 47 at the In our PIAI patients, bacteraemia was associated with
time of reoperation (p < 0.0001), and empirical antifungal limited clinical consequences. Clinical recovery, analysed
therapy (p = 0.002), while de-escalation of documented on day 3 and day 7 after surgery, did not differ between
therapy (p = 0.003) had a protective role. groups. Our data confirm the observations of Havey et
al., who reported early improvement of clinical parame-
Discussion ters in patients treated for bacteraemia [11]. Our obser-
In this single-centre cohort study of patients with PIAI, vations suggest an incidental nature of bacteraemia
the incidence of perioperative bacteraemia was 19%. A rather than a major threat. This assumption is also con-
trend toward increased mortality rate in bacteraemic firmed by the absence of any specific morbidity criteria
patients was observed compared with non-bacteraemic in the postoperative course of bacteraemic patients com-
patients, with no major changes in morbidity criteria. The pared with non-bacteraemic patients. Additional obser-
factors associated with bacteraemia were the presence of vations, such as the high mortality rates reported in the
Alqarni et al. Critical Care (2018) 22:175 Page 7 of 10

Table 5 Clinical outcome in bacteraemic and non-bacteraemic patients

Clinical outcomes Bacteraemic patients Patients without Odds ratio 95% CI p value
(n = 64) bacteraemia (n = 279)
On day 3 after surgery for PIAI
Temperature ≥ 36 °C and ≤ 38 °C, n (%) 25/55 (45.4) 165/243 (67.9) 0.39 (0.21–0.71) < 0.01
WBC count ≥ 5000 and ≤ 15,000/mm3, n (%) 20/55 (57.1) 108/243 (44.4) 0.71 (0.39–1.31) 0.27
SOFA score ≤ 3 points, n (%) 22/55 (40) 104/243 (42.8) 0.89 (0.49–1.62) 0.70
All three criteria, n (%) 7/55 (12.7) 42/243 (17.3) 0.69 (0.29–1.64) 0.54
a
Delta SOFA between day 3 and day 0, median (IQR) 1 (0–4) 2 (0–4) 1.00 (0.91–1.11) 0.82
On day 7 after surgery for PIAI
Temperature ≥ 36 °C and ≤ 38 °C, n (%) 29/44 (65.9) 140/200 (70) 0.82 (0.41–1.65) 0.59
WBC count ≥ 5000 and ≤ 15,000/mm3, n (%) 23/44 (52.3) 84/200 (42) 1.51 (0.78–2.91) 0.21
SOFA score ≤ 3 points, n (%) 22/44 (50) 112/200 (56) 0.78 (0.41–1.51) 0.46
All three criteria, n (%) 12/44 (27.3) 44/200 (22) 1.33 (0.63–2.79) 0.45
a
Delta SOFA between day 7 and day 0, median (IQR) 3.5 (1.2–6) 4 (1–6) 0.98 (0.89–1.11) 0.93
Suspected persistent sepsis, n (%) 36 (58) 144 (53) 1.21 (0.69–2.11) 0.50
Reoperation, n (%) 29 (45) 122 (44) 1.02 (0.59–1.76) 0.81
Total number of reoperations, median (IQR) 1 (1–2) 2 (1–2) 0.88 (0.71–1.10) 0.21
Interval between surgery for PIAI and reoperation, median (IQR) 5 (2.5–7) 5 (2–8) 0.97 (0.87–1.08) 0.42
Main reasons for reoperation
Septic shock/multiple organ failure, n (%) 11/29 (38) 47/122 (39) 0.97 (0.42–2.24) 0.95
Sepsis, n (%) 18/29 (62) 75/122 (62) 1.02 (0.44–2.36) 0.95
Haemorrhage, n (%) 2/29 (7) 9/122 (7) 0.93 (0.18–4.55) 1
Suspicion of bowel ischaemia, n (%) 2/29 (7) 26/122 (21) 0.27 (0.06–1.22) 0.11
Main intraoperative diagnoses at reoperation
Anastomotic leak, n (%) 8/29 (28) 29/122 (24) 1.22 (0.48–3.04) 0.64
Bowel perforation, n (%) 4/29 (14) 25/122 (20) 0.62 (0.19–1.94) 0.60
Ischaemia, n (%) 2/29 (7) 23/122 (19) 0.31 (0.07–1.43) 0.17
Abscess or collections, n (%) 5/29 (17) 14/122 (11) 1.60 (0.53–4.89) 0.37
No identified cause, n (%) 5/29 (17) 10/122 (8) 2.33 (0.73–7.44) 0.16
Medical complications, n (%) 5 (8) 37 (14) 0.55 (0.20–1.47) 0.29
Duration of mechanical ventilation, median (IQR), daysb 8 (4–15) 7 (3–15) 1.45 (0.71–2.95) 0.75
ICU stay, median (IQR), daysb 12 (8–23) 15 (8–25) 0.96 (0.48–1.94) 0.39
Death in ICU, n (%) 26 (41) 82 (29) 1.64 (0.94–2.88) 0.08
Death in hospital, n (%) 26 (41) 88 (32) 1.48 (0.85–2.59) 0.16
CI confidence interval, IQR interquartile range, PIAI postoperative intra-abdominal infection, SOFA Sequential Organ Failure Assessment, WBC white blood cell
a
Per 1 point
b
Variable measured in survivors

case of fungaemia and bacteraemia not targeted by em- This point was confirmed in univariate analysis, while
pirical therapy, could be considered to be signals of other well-known factors, such as emergency surgery
interest. However, the number of cases was too small to and initial severity, were identified as risk factors for
draw any strong conclusions. mortality. Interestingly, the prognosis did not differ in
A poorer prognosis was expected in the cohort of bac- the presence of fungaemia compared with bacteraemia.
teraemic patients. The 41% mortality rate in our study The small sample size of our cohort clearly affected the
confirms the high mortality rate previously reported by validity of these results. Surprisingly, the role of fungae-
De Waele et al. [10]. However, only a trend towards an mia in peritonitis cases has been rarely analysed in the
increased mortality rate was observed in our bacter- literature, while fungaemia is a known factor of poor
aemic patients compared with non-bacteraemic patients. prognosis in many instances [23, 24]. Nevertheless,
Alqarni et al. Critical Care (2018) 22:175 Page 8 of 10

Table 6 Univariate and multivariate analyses of risk factors of intensive care unit (ICU) death in the study population
Univariate analysis Multivariate analysis
Parameters Odds ratio 95% CI p value Odds ratio 95% CI p value
Demographic characteristics and underlying diseases
Age > 62 years 2.45 (1.55–3.90) < 0.001
McCabe Fatal 1.55 (0.96–2.49) 0.08
Immunosuppression or cancer 1.83 (1.15–2.92) 0.01
Urgent initial surgery 2.04 (1.29–3.23) 0.002 2.71 (1.53–4.81) 0.0006
Severity criteria at the time of reoperation
Small bowel perforation 1.57 (0.95–2.59) 0.07
SAPS II > 47 at reoperation 4.26 (2.61–6.94) < 0.001 4.78 (2.71–8.45) < 0.0001
Blood and peritoneal microbiological results
Bacteraemia 1.48 (0.84–2.59) 0.16 1.62 (0.81–3.24) 0.169
Peritoneal samples involving
Enterococci 1.71 (1.08–2.69) 0.019
Non-fermenting Gram-negative Bacteria 2.01 (1.13–3.65) 0.019
Streptococci 0.54 (0.30–0.96) 0.037
Empirical and documented therapies
Empirical imipenem therapy 1.66 (1.01–2.72) 0.044
Empirical vancomycin therapy 1.56 (0.99–2.46) 0.050
Empirical antifungal therapy 1.72 (1.08–2.74) 0.019 2.91 (1.62–5.22) 0.0003
De-escalation documented therapy 0.54 (0.34–0.85) 0.007 0.50 (0.28–0.89) 0.018
Documented therapy using piperacillin/tazobactam 1.74 (1.09–2.78) 0.018
Documented antifungal therapy 1.67 (1.05–2.63) 0.026
The model was adjusted on year of admission
Multivariate logistic regression with stepwise selection: the C-index for the model was 0.81 95% CI 0.76–0.86, the Hosmer-Lemeshow test p value was 0.02
CI confidence interval, SAPS Simplified Acute Physiology Score

fungal infection might play a role in the prognosis, as that discarding these microorganisms may have clinical
indirectly reflected by documented antifungal therapy consequences.
which was a strong criterion of poor prognosis in The duration of anti-infective therapy was not differ-
our cohort. ent between patients with and without bacteraemia. The
The need to treat all isolated microorganisms in these optimal duration of antibiotic therapy for bacteraemia
high-risk ICU patients is a subject of debate. Recently during intra-abdominal infections is unknown. In our
published guidelines have failed to address this issue study, antibiotic therapy for bacteraemic PIAI was ad-
[17, 25, 26]. The usual rule is to target all pathogens ministered for 11 ± 6 days, a duration usually reported in
isolated from both blood and surgical samples. However, the literature. Our data suggest that extended therapy is
the pathogenic role of certain microorganisms such as not required in bacteraemic patients. However, this
staphylococci is debated. Some data suggest a pathogenic study was not tailored to address this issue and any
role of S. aureus [27]. On the contrary, the role of CNS as definitive conclusions would be premature since, in a re-
a contaminant on blood or surgical samples remains cent study analysing the benefit of short-course anti-
much more controversial. In order to limit the risk of con- biotic therapy in PIAI, we reported an increased risk of
tamination, we used strict selection rules for CNS bacter- postoperative bacteraemia and a higher risk of treatment
aemia and we assume that the selected isolates represent failure in the case of short-course therapy [28].
authentic pathogens. By means of these criteria, CNS Several limitations to the present study must be men-
represented only a limited proportion of the organisms tioned, including its retrospective nature. The results of
isolated from bacteraemic patients. Among the 11 bacter- this single-centre analysis may not be directly applicable
aemic patients in whom empirical anti-infective therapy to other institutions. Breakthrough infections and recur-
did not target all bloodstream microorganisms, almost rent bacteraemia could not be assessed due to the short
half of the cases involved CNS. Interestingly, the outcome time window for analysis of bacteraemia. The long study
of these cases was not completely favourable, suggesting duration is also a factor to be taken into account, with
Alqarni et al. Critical Care (2018) 22:175 Page 9 of 10

changing patient characteristics over recent years, with Author details

1
older and more severely ill patients, but the year of in- Département d’Anesthésie-Réanimation, CHU Bichat-Claude Bernard, APHP,
Paris, France. 2King Saud University, Riyadh, Saudi Arabia. 3Université Paris
clusion was adjusted for in the multivariate models. The Diderot, Press Sorbonne Cité, Paris, France. 4Laboratoire de Microbiologie,
quality of infection source control was not evaluated in CHU Bichat-Claude Bernard, APHP, Paris, France. 5INSERM UMR 1137, Paris,
this study, although it is an essential element of the prog- France. 6INSERM UMR 1188, Saint-Denis de la Réunion, France. 7Service de
Chirurgie Générale, CHU Bichat-Claude Bernard, APHP, Paris, France. 8INSERM
nosis of PIAI patients. The number of reoperations, the UMR 1148, Paris, France. 9INSERM UMR 1152, Paris, France.
reasons for these operations, and the surgical findings did
not differ between bacteraemic and non-bacteraemic pa- Received: 11 April 2018 Accepted: 19 June 2018

tients, suggesting that bacteraemia did not play a major

role in the outcome.
References
1. Laupland KB, Zygun DA, Davies HD, Church DL, Louie TJ, Doig CJ.
Conclusion Population-based assessment of intensive care unit-acquired bloodstream
In this single-centre observational study, one-fifth of pa- infections in adults: incidence, risk factors, and associated mortality rate.
tients had bacteraemia during the early management of Crit Care Med. 2002;30:2462–7.
2. Blot S, Vandewoude K, De Bacquer D, Colardyn F. Nosocomial
PIAI. Bacteraemia was not a risk factor for mortality bacteremia caused by antibiotic-resistant gram-negative bacteria in
compared with patients without bacteraemia, and mor- critically ill patients: clinical outcome and length of hospitalization.
bidity criteria did not appear to be modified. None of Clin Infect Dis. 2002;34:1600–6.
3. Vincent JL, Rello J, Marshall J, Silva E, Anzueto A, Martin CD, et al.
the mortality risk factors identified in our study are sus- International study of the prevalence and outcomes of infection in
ceptible to corrective measures designed to improve the intensive care units. JAMA. 2009;302:2323–9.
management of these critically ill patients. Increasing 4. Tabah A, Koulenti D, Laupland K, Misset B, Valles J, Bruzzi de Carvalho F, et
al. Characteristics and determinants of outcome of hospital-acquired
the duration of anti-infective therapy could have been bloodstream infections in intensive care units: the EUROBACT international
proposed to improve the prognosis, although the current cohort study. Intensive Care Med. 2012;38:1930–45.
trend is to reduce the duration of therapy. Early diagno- 5. Boncagni F, Francolini R, Nataloni S, Skrami E, Gesuita R, Donati A, et al.
Epidemiology and clinical outcome of healthcare-associated infections: a
sis and early management of organ failures and their 4-year experience of an Italian ICU. Minerva Anestesiol. 2015;81:765–75.
consequences remain the best rules in the treatment of 6. Riche FC, Dray X, Laisne MJ, Mateo J, Raskine L, Sanson-Le Pors MJ, et al.
these patients. Factors associated with septic shock and mortality in generalized peritonitis:
comparison between community-acquired and postoperative peritonitis.
Crit Care. 2009;13:R99.
Abbreviations 7. Montravers P, Augustin P, Grall N, Desmard M, Allou N, Marmuse JP, et al.
CI: Confidence interval; CNS: Coagulase-negative staphylococci; Characteristics and outcomes of anti-infective de-escalation during health
EAT: Empirical antibiotic therapy; ICU: Intensive care unit; IQR: Interquartile care-associated intra-abdominal infections. Crit Care. 2016;20:83.
range; MDR: Multidrug-resistant; OR: Odds ratio; PIAI: Postoperative
8. Roehrborn A, Thomas L, Potreck O, Ebener C, Ohmann C, Goretzki PE, et al. The
intra-abdominal infection; SAPS: Simplified Acute Physiology Score;
microbiology of postoperative peritonitis. Clin Infect Dis. 2001;33:1513–9.
SOFA: Sequential Organ Failure Assessment
9. Seguin P, Fedun Y, Laviolle B, Nesseler N, Donnio PY, Malledant Y. Risk
factors for multidrug-resistant bacteria in patients with post-operative
Availability of data and materials peritonitis requiring intensive care. J Antimicrob Chemother.
Data and material kept until the end of 2018. 2010;65:342–6.
10. De Waele JJ, Hoste EA, Blot SI. Blood stream infections of abdominal origin
Authors’ contributions in the intensive care unit: characteristics and determinants of death.
Concept and design: PM. Acquisition, analysis or interpretation of data: all Surg Infect. 2008;9:171–7.
authors. Drafting of the manuscript: AA, EK, NG, ATD, PM. Critical revision of 11. Havey TC, Fowler RA, Pinto R, Elligsen M, Daneman N. Duration of antibiotic
the manuscript for important intellectual content: all authors. Statistical therapy for critically ill patients with bloodstream infections: a retrospective
analysis: EK, ATD, PM. Administrative, technical or material support: PM. Study cohort study. Can J Infect Dis Med Microbiol. 2013;24:129–37.
supervision: PM. All authors read and approved the final manuscript. 12. Nelson AN, Justo JA, Bookstaver PB, Kohn J, Albrecht H, Al-Hasan MN.
Optimal duration of antimicrobial therapy for uncomplicated gram-negative
bloodstream infections. Infection. 2017;45:613–20.
Ethics approval and consent to participate
13. McCabe WR, Jackson GG. Gram-negative bacteremia. I Etiology and ecology.
This observational study was conducted according to the terms and
Arch Intern Med. 1962;110:847–55.
regulations of the local institutional review boards (CEERB CHU Bichat, Paris
Diderot University, Paris, agreement 10–008; CNIL 1413211). According to 14. Le Gall JR, Lemeshow S, Saulnier F. A new simplified acute physiology score
French Law, no informed consent was required since this observational (SAPS II) based on a European/North American multicenter study. JAMA.
study did not modify the physician’s treatment decisions. 1993;270:2957–63.
15. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, et al.
The SOFA (sepsis-related organ failure assessment) score to describe organ
Consent for publication dysfunction/failure. On behalf of the working group on Sepsis-related
Not required. problems of the European Society of Intensive Care Medicine. Intensive
Care Med. 1996;22:707–10.
Competing interests 16. Magiorakos AP, Srinivasan A, Carey RB, Carmeli Y, Falagas ME, Giske CG, et
The authors declare that they have no competing interests. al. Multidrug-resistant, extensively drug-resistant and pandrug-resistant
bacteria: an international expert proposal for interim standard definitions for
acquired resistance. Clin Microbiol Infect. 2012;18:268–81.
Publisher’s Note 17. Montravers P, Dupont H, Leone M, Constantin JM, Mertes PM, Laterre PF, et
Springer Nature remains neutral with regard to jurisdictional claims in al. Guidelines for management of intra-abdominal infections. Anaesth Crit
published maps and institutional affiliations. Care Pain Med. 2015;34:117–30.
Alqarni et al. Critical Care (2018) 22:175 Page 10 of 10

18. Dupont H, Bourichon A, Paugam-Burtz C, Mantz J, Desmonts JM. Can yeast

isolation in peritoneal fluid be predicted in intensive care unit patients with
peritonitis? Crit Care Med. 2003;31:752–7.
19. Paugam-Burtz C, Dupont H, Marmuse JP, Chosidow D, Malek L, Desmonts JM,
et al. Daily organ-system failure for diagnosis of persistent intra-abdominal
sepsis after postoperative peritonitis. Intensive Care Med. 2002;28:594–8.
20. Montravers P, Dufour G, Guglielminotti J, Desmard M, Muller C, Houissa H,
et al. Dynamic changes of microbial flora and therapeutic consequences in
persistent peritonitis. Crit Care. 2015;19:70.
21. Harbarth S, Ferriere K, Hugonnet S, Ricou B, Suter P, Pittet D. Epidemiology
and prognostic determinants of bloodstream infections in surgical intensive
care. Arch Surg. 2002;137:1353–9.
22. Weinstein MP, Reller LB. Clinical importance of “breakthrough” bacteremia.
Am J Med. 1984;76:175–80.
23. Dupont H, Paugam-Burtz C, Muller-Serieys C, Fierobe L, Chosidow D,
Marmuse JP, et al. Predictive factors of mortality due to polymicrobial
peritonitis with Candida isolation in peritoneal fluid in critically ill patients.
Arch Surg. 2002;137:1341–7.
24. Montravers P, Perrigault PF, Timsit JF, Mira JP, Lortholary O, Leroy O, et al.
Antifungal therapy for patients with proven or suspected Candida
peritonitis: Amarcand2, a prospective cohort study in French intensive care
units. Clin Microbiol Infect. 2017;23:117. e111–117.e118
25. Mazuski JE, Tessier JM, May AK, Sawyer RG, Nadler EP, Rosengart MR, et al.
The surgical infection society revised guidelines on the management of
intra-abdominal infection. Surg Infect. 2017;18:1–76.
26. Sartelli M, Catena F, Abu-Zidan FM, Ansaloni L, Biffl WL, Boermeester MA, et
al. Management of intra-abdominal infections: recommendations by the
WSES 2016 consensus conference. World J Emerg Surg. 2017;12:22.
27. Fierobe L, Decre D, Muller C, Lucet JC, Marmuse JP, Mantz J, et al.
Methicillin-resistant Staphylococcus aureus as a causative agent of
postoperative intra-abdominal infection: relation to nasal colonization.
Clin Infect Dis. 1999;29:1231–8.
28. Montravers P, Tubach F, Lescot T, Veber B, Esposito-Farese M, Seguin P, et
al. Short-course antibiotic therapy for critically ill patients treated for
postoperative intra-abdominal infection: the DURAPOP randomised clinical
trial. Intensive Care Med. 2018;44:300–10.