Bacterial coinfection is associated with severity of avian influenza A ...

Diagnostic Microbiology and Infectious Disease 84 (2016) 165–169

Contents lists available at ScienceDirect

Diagnostic Microbiology and Infectious Disease journal homepage: www.elsevier.com/locate/diagmicrobio

Bacterial coinfection is associated with severity of avian influenza A (H7N9), and procalcitonin is a useful marker for early diagnosis Meifang Yang, Hainv Gao, Jiajia Chen, Xiaowei Xu, Lingling Tang, Yida Yang, Weifeng Liang, Liang Yu, Jifang Sheng, Lanjuan Li ⁎ State Key Laboratory for Diagnosis and Treatment of Infectious Diseases, The First Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, Hangzhou, China

a r t i c l e

i n f o

Article history: Received 13 January 2015 Received in revised form 1 August 2015 Accepted 24 October 2015 Available online 28 October 2015 Keywords: Bacterial co-infection Influenza A H7N9 Mortality Procalcitonin

a b s t r a c t Patients contracting avian influenza A (H7N9) often develop severe disease. However, information on the contribution of bacterial coinfection to the severity of H7N9 is limited. We retrospectively studied 83 patients with confirmed H7N9 infection from April 2013 to February 2014. The severity of patients with bacterial coinfection and markers for early diagnosis of bacterial coinfection in H7N9 were analyzed. We found Staphylococcus aureus was the most prevalent pathogen. Higher Acute Physiology and Chronic Health Evaluation II score, shock, renal replacement treatment, mechanical ventilation, and extracorporeal membrane oxygenation treatment were more frequently observed in patients with bacterial coinfection. Procalcitonin is more sensitive than C-reactive protein in determining bacterial coinfection in H7N9 patients. In conclusion, H7N9 infection patients with bacterial coinfection had a more severe condition. Elevated procalcitonin is an accurate marker for diagnosing bacterial coinfection in H7N9 patients, thus enabling earlier antibiotic therapy. © 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

2. Methods

A novel avian-origin influenza A (H7N9) virus had emerged and spread among humans in China and led to 435 cases by June 27, 2014, with a mortality rate as high as 36.7% (WHO, nd). Early reports have shown that H7N9 infection could cause progressive pneumonia and multiorgan dysfunction, with 63–76.6% of patients being admitted to an intensive care unit (ICU) (Gao et al., 2013a; Li et al., 2014). Much effort has been devoted to investigate the risk factors for the progression and the fatal outcome of the disease. Our early findings have indicated the presence of a coexisting medical condition as the only independent risk factor for acute respiratory distress syndrome (Gao et al., 2013a). Elevated angiotension II levels are associated with the severity of H7N9-induced disease and may potentially predict patient mortality (Huang et al., 2014). However, initial studies of H7N9 influenza describe few details of patients with bacterial coinfection and the reasonable selection of antibiotic therapy. We conducted a retrospective study that focused on the presence of bacterial coinfection in confirmed cases of H7N9 infection. We aimed to investigate the incidence of bacterial coinfection, the most common pathogen, and an accurate diagnosis marker before obtaining positive cultures for this potential fatal disease and to determine which patients should receive antibiotic therapy at the time of admission.

2.1. Study design and patients

⁎ Corresponding author. Tel.: +86-1-390-651-4210. E-mail address: [email protected] (L. Li).

We performed a retrospective review of medical records of all patients with laboratory-confirmed avian H7N9 influenza infection (by polymerase chain reaction of nasopharyngeal secretions or bronchoalveolar lavage fluid) in a large tertiary hospital, the First Affiliated Hospital, School of Medicine, Zhejiang University, China, from April 2013 to February 2014. The following data were recorded retrospectively: demographic details, microbiologic findings, comorbidities, severity of illness score, the presence of shock, clinical laboratory findings within 24 h of admission, and outcome on hospital day 90. The need for mechanical ventilation, renal replacement therapies, and extracorporeal membrane oxygenation (ECMO) treatment was also recorded. The study was approved by the Ethical Board of the First Affiliated Hospital, School of Medicine, Zhejiang University. Written informed consent was obtained from all patients. 2.2. Definitions A confirmed case was defined as a positive test result for avian H7N9 using reverse transcriptase polymerase chain reaction, as described previously (Gao et al., 2013b). Bacterial coinfection was defined as any bacterial infection presumed by the expert group according to clinical manifestations, with one or more positive cultures or with positive

http://dx.doi.org/10.1016/j.diagmicrobio.2015.10.018 0732-8893/© 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

166

M. Yang et al. / Diagnostic Microbiology and Infectious Disease 84 (2016) 165–169

urine Streptococcus pneumoniae or Legionella pneumophila antigen detection. The cultures should have been obtained from blood, valid sputum, lower respiratory tract (endotracheal or bronchoalveolar lavage) samples, or other normally sterile fluids within the first 72 h of hospitalization. Meanwhile, the urine samples subjected to urine antigen detection should have been positive within 24 h after admission. To determine the severity of illness, Acute Physiology and Chronic Health Evaluation (APACHE) II score (Knaus et al., 1985) was calculated for all patients within 24 h of admission. Shock was defined as that being treated with vasopressors. Mortality was defined as death occurring anytime after admission and up to day 90. 2.3. Method for procalcitonin (PCT) and C-reactive protein (CRP) measurement PCT was tested using the electrochemical luminescence method (Elecsys Brahms PCT, Mannheim, Germany). CRP was measured using immunoturbidimetric assays (Beckman, Carlsbad, CA 92010, USA). 2.4. Statistical analyses Categorical variables are expressed in percentages and frequencies. Continuous variables are expressed means and SD or median. Categorical variables were compared using the chi-square test. Continuous variables were compared using independent-samples t-test or Mann–Whitney U test. Univariate and multivariate logistic regression analyses were performed to predict 90-day mortality. The variables analyzed were age; gender; comorbidities; presence of shock; APACHE II score; serum creatinine; serum creatinine kinase (CK); serum lactate dehydrogenase (LDH); CRP, leukocyte count, platelet count; D-dimer; PCT; bacterial coinfection; and need for mechanical ventilation, renal replacement, and ECMO. The diagnostic accuracy of biomarkers (CRP and PCT) for predicting bacterial coinfection was examined by their receiver operating characteristic curve (ROC), and the area under the curve (AUC), sensitivity, and specificity were also calculated. All tests were 2 tailed, and the significance was set at 5%. Data analysis was performed using SPSS for Windows, version 18.0. 3. Results During the study period, 83 consecutive patients with laboratoryconfirmed avian H7N9 influenza infection who were admitted to the infectious disease department in our hospital were analyzed. The mean age was 57.71 years (57.71 ± 14.60, range, 21–86 years); 41 (49.4%) patients were older than 60 years, and 55 (66.3%) were males. Forty-eight patients (57.8%) had comorbidities, of which hypertension and diabetes were the most common. Two (2.4%) patients were pregnant. The mean APACHE II score at admission was 19.93 ± 8.25. The main demographic and clinical characteristics of patients with and without bacterial coinfection are detailed in Table 1. In our study, 16 (19.3%) patients were diagnosed with bacterial coinfection in the first 72 h of admission based on clinical manifestations, cultures, and urinary antigen detection. However, urine samples of all patients were negative for S. pneumoniae and L. pneumophila antigens. The pathogens that caused bacterial coinfections were Staphylococcus aureus in 4 cases (25%), 3 of which involve methicillinresistant S. aureus (MRSA); Staphylococcus hominis in 3 cases; Staphylococcus epidermidis in 2 cases; Ralstonia mannitolilytica in 2 cases; Acinetobacter baumannii in 2 cases; Pseudomonas aeruginosa in 1 case; Escherichia coli in 1 case; and Burkholderia cepacia in 1 case. Eight patients with bacterial coinfection had bacteremia (S. hominis in 3 cases, S. epidermidis in 2 cases, MRSA in 2 cases, and A. baumannii in 1 case). Three patients (3.6%) had both positive respiratory tract cultures and bacteremia (MRSA in 2 cases and A. baumannii in 1 case). Patients with bacterial coinfection had a more severe condition than those without coinfection and had higher APACHE II score (25.63 ±

Table 1 Demographic and clinical characteristics of 83 avian H7N9 patients with and without bacterial coinfection. Characteristics

Bacterial coinfection (n = 16)

H7N9 only (n = 67)

P value

Age (years), mean Sex (male), n (%) Comorbidities Laboratory finding at admission Serum creatinine (μmol/L) Serum CK (U/L) Serum LDH (U/L) CRP (mg/dL) PCT (ng/mL)

58.25 ± 17.14 8 (50.0%) 9 (56.2%)

57.58 ± 14.07 47 (70.1%) 39 (58.2%)

0.871 0.126 0.973

110.44 ± 91.11 463.81 ± 513.88 783.19 ± 584.47 144.82 ± 84.66 18.80 ± 28.97 11914.44 ± 9789.27

78.99 ± 67.35 0.122 403.06 ± 775.13 0.768 498.13 ± 300.27 0.007 85.10 ± 70.01 0.004 0.58 ± 1.18 0.000 5030.78 ± 5527.12 0.000

6.96 ± 4.43 127.5 ± 82.55 25.63 ± 5.30 13 (81.2%) 15 (93.8%) 10 (62.5%) 8 (50.0%) 29.88 ± 23.13 7 (43.8%) 10 (62.5%)

4.77 ± 3.52 136.07 ± 69.23 18.57 ± 8.27 16 (23.9%) 29 (43.3%) 19 (28.4%) 14 (20.9%) 24.21 ± 19.77 11 (16.4%) 16 (23.9%)

D-Dimer

(μg/L) Leukocyte count (109/L) Platelets counts (109/L) APACHE II score at admission Renal replacement Mechanical ventilation Shock ECMO treatment Hospital stay (days), mean 30-day mortality, n (%) 90-day mortality, n (%)

0.037 0.669 0.002 0.000 0.000 0.010 0.018 0.322 0.017 0.003

5.30 versus 18.57 ± 8.27, P = 0.000) and were more likely to present with shock (62.5% versus 28.4%, P = 0.010); were more likely to need renal replacement treatment (81.2% versus 23.9%, P = 0.000), mechanical ventilation (93.8% versus 43.3%, P = 0.000), and ECMO treatment (50.0% versus 20.9%, P = 0.018); had higher serum levels of LDH, CRP, PCT, D-dimer, and leukocyte counts; and showed higher 90-day hospital mortality (62.5% versus 23.9%, P = 0.003). However, there was no significant trend for longer hospital stay. No differences in age, gender, comorbidities, serum creatinine, CK, and platelets counts were observed (Table 1). PCT and CRP levels were obtained, respectively, in 53 and 83 of the total 83 patients, and 53 patients had both biomarkers measured simultaneously within 24 h after admission. Fig. 1 demonstrates the scatter plot of PCT and CRP values in the 2 groups on admission. Using ROC analysis, AUC to diagnose bacterial coinfection was 0.96 (95% confidence interval [CI]: 0.91–1.00) for PCT compared with 0.68 (95% CI: 0.50–0.87) for CRP (P = 0.005) (Fig. 2). For PCT at a cutoff value of 0.81 μg/L for diagnosis of bacterial coinfection, the sensitivity was 91.7%, and the specificity was 90.2%. All patients received antiviral therapy in our study. The median time from the onset of illness to the initiation of antiviral therapy was 7 days (range, 1–20); 9.6% (8/83) of the patients received antiviral therapy within 48 h after the onset of symptoms. The antiviral regimes include oseltamivir at a daily dose of 150–300 mg in 43 patients, oseltamivir combined with peramivir at a daily dose of 600 mg in 39 patients, and peramivir alone in 1 patient. The mean time from onset to admission was 6.67 (range, 1–15) days. About 91.2% of our patients had received antibiotics before admission, and empiric antibiotic therapy was administered to 58 (69.9%) patients after admission. The antibiotic regimens were beta-lactam monotherapy (in 41 patients, 69.5%), beta-lactam plus vancomycin (in 6 patients, 10.1%), fluoroquinolone monotherapy (in 4 patients, 6.8%), carbapenem monotherapy (in 2 patients, 3.4%), vancomycin monotherapy (in 1 patient, 1.8%), fluoroquinolones plus beta-lactam (in 2 patients, 3.4%), and fosfomycin monotherapy and other combinations (in 2 patients, 3.4%). Empiric antibiotic treatment was inappropriate in 4 (25%) of 16 cases with bacterial coinfection, and 3 patients with inappropriate treatment died. The pathogens associated with inappropriate treatment were A. baumannii in 2 cases, 1 case each of MRSA and E. coli. The all-cause hospital mortality was 28.9% (24/83), and the 90-day mortality was 31.3% (26/83). In the univariate analysis, age; bacterial coinfection; comorbidities; APACHE II score; the presence of shock;


167

Table 2 Significant univariate and multivariate logistic regression analyses of 90-day mortality. Variable

Univariate

Multivariate

OR

95% CI

P value OR

95% CI

P value

D-Dimer

1.047 0.188 0.270 1.209 0.306 0.141 0.248 1.009 1.000

1.008–1.087 0.059–0.599 0.094–0.772 1.107–1.321 0.116–0.808 0.050–0.395 0.089–0.695 1.002–1.015 1.000–1.000

0.017 0.005 0.015 0.000 0.017 0.000 0.008 0.010 0.001

0.959–1.165 0.038–2.128 0.028–2.223 0.952–1.353 0.504–25.958 0.040–1.182 0.026–1.876 0.992–1.014 1.000–1.000

0.265 0.221 0.214 0.158 0.201 0.077 0.166 0.568 0.658

Serum LDH Serum CK

1.003 1.001–1.005 0.004 1.001 1.000–1.002 0.039

Age Bacterial coinfection Comorbidities APACHE II score Shock Renal replacement ECMO CRP

1.057 0.284 0.251 1.135 3.618 0.217 0.220 1.003 1.000

1.004 1.000–1.008 1.001 1.000–1.002

0.079 0.054

regression analysis, the above variables including bacterial coinfection were not associated with increased mortality (Table 2).

4. Discussion

Fig. 1. a, PCT values on admission; b, CRP values on admission.

the need for renal replacement and ECMO treatment; and the higher levels of CRP, D-dimer, serum LDH, and serum CK were significantly associated with 90-day mortality. However, in multivariate logistic

Fig. 2. ROC analysis of PCT (solid line) and CRP (dashed line) for predicting bacterial coinfection at admission, in patients in whom measurements of the 2 biomarkers were obtained simultaneously (n = 53).

In this study, bacterial coinfection diagnosed within 72 h of admission was observed in 19.3% of patients with H7N9, and it was also associated with the severity of the disease. Increased PCT at a cutoff value of 0.81 μg/L can help identify bacterial coinfection in H7N9 influenza patients, thus enabling earlier antibiotic therapy. Although associated with severity and more treatment intensity, bacterial coinfection was not the independent predictor of 90-day mortality in these patients. This is the first report on bacterial coinfection in H7N9 influenza patients. The rate of bacterial coinfection in pH1N1 infection patients showed a great variability in different studies, ranging from 17.5% to 51% in critically ill ICU patients (Cilloniz et al., 2012; Estenssoro et al., 2010; Martin-Loeches et al., 2011; Nguyen et al., 2012; Rice et al., 2012) and from 25% to 55% in fatal cases with autopsy (Anon, 2009; Gill et al., 2010; Mauad et al., 2010). In 2 studies that used a similar definition of bacterial coinfection in pH1N1 infection patients, Rice et al. (2012) reported a rate of 30.3%, which included any adult patient with presumed bacterial pneumonia even if cultures were negative within 72 h of ICU admission. The study by Nguyen et al. (2012) showed a rate as high as 51% in critically ill children based on positive cultures within 72 h of admission. In our study, S. aureus was the most frequently isolated bacteria. Of these isolations, 75% were MRSA. One of them with MRSA bacteremia did not receive appropriate empiric therapy and did not survived. S. aureus pneumonia is a well-recognized complication of influenza. Coinfection with S. aureus, especially with MRSA, has been associated with severe disease and high mortality rate in adults and children infected with season influenza and pH1N1 (Finelli et al., 2008; Kallen et al., 2009; Nguyen et al., 2012). Clinicians managing patients with H7N9 influenza infection should be alert to the possibility of coinfection with S. aureus. The cases of infection by coagulase-negative Staphylococcus, such as S. hominis or S. epidermidis, were considered true infections according to their clinical manifestations and laboratory findings for the following reasons. The novel H7N9 virus causes infection of acute onset and rapid progression. Up to 64% of the patients in our study were admitted to the ICU. Most of the patients with 1 or more blood cultures positive for coagulase-negative Staphylococcus underwent invasive procedures, such as various forms of intubation, before or immediately after admission. The coagulase-negative Staphylococcus is a common pathogen in catheter-related bloodstream infections. Thus, we surmise that these cultures highly likely represented true, hospital-acquired infections. In addition to S. aureus, S. pneumoniae was a frequent bacterial pathogen identified in patients with pH1N1 infection. Unexpectedly, we did not

168


identify S. pneumoniae or other common bacteria such as Haemophilus influenzae and Streptococcus pyogenes in the cultures. About 91.2% of our patients had received antibiotics before admission to our hospital, thus possibly limiting the chances of detecting bacterial coinfection based on a culture test. Thus, the rate of bacterial coinfection in our series could possibly be underestimated. Because bacterial coinfection cannot be easily differentiated from viral infection alone on the basis of the patients' clinical manifestations, radiological results, and routine laboratory results, 71.2% patients received antibiotic therapy within 24 h after admission in the initial 111 cases of H7N9 infection (Gao et al., 2013a), even though bacterial coinfection is considered unlikely. PCT is considered to be a biomarker for bacterial infection in patients with community-acquired pneumonia (CAP) and might be a prognostic marker in those with lower respiratory tract infections (Gilbert, 2011). Some studies have reported that PCT levels can help discriminate isolated viral pneumonia from bacterialand viral-mixed pneumonia in patients during the 2009 A/H1N1 influenza pandemic (Cuquemelle et al., 2011; Ingram et al., 2010; Pfister et al., 2014). In our study, we found PCT levels are more sensitive than CRP levels in discriminating bacterial coinfection in our patients. The optimal cutoff value of PCT for bacterial coinfection was 0.81 μg/L (91.7% sensitivity, 90.2% specificity), with an AUC of 0.96 (0.91–1.00), which was similar to that previously reported for pH1N1 (Cuquemelle et al., 2011; Ingram et al., 2010; Pfister et al., 2014). The all-cause hospital mortality of H7N9 cases in our population was 28.9%, which was higher than that due to seasonal influenza-associated pneumonia (4.4%) (von Baum et al., 2011) and CAP (4.5%) (Chalmers et al., 2011) and that previously reported for adult pH1N1 patients (4.9–23%) (Helferty et al., 2010; Kumar et al., 2009; Rice et al., 2012; Webb et al., 2009; Zhang et al., 2013) but was lower than that due to highly pathogenic avian influenza A (H5N1) viruses (59%) (Patel et al., 2014). Although bacterial coinfection was associated with more severe illness (higher APACHE II score and higher presence of shock); treatment intensity such as a more frequent need for renal replacement treatment, mechanical ventilation, and ECMO treatment; and 90-day mortality in the univariate analysis, it was not associated with an increased mortality in multivariate logistic regression analysis for the 90-day mortality. Several potential limitations in our study need to be addressed. First, this was a retrospective and observational study at a single center, and all patients included in our study were adults. Hence, our results might not be extended to other countries or to children. Second, we defined bacterial coinfection as 1 or more positive cultures obtained from valid samples within the first 72 h of hospitalization. We could not discriminate between community-originated and hospitalacquired infections. Third, a major limitation is that there is no gold standard for the definition of bacterial coinfection in influenza patients. So, the “true” proportion of patients with bacterial coinfection is also difficult to ascertain accurately. Thus, it might decrease the accuracy of bacterial coinfection rate in H7N9 infection patients, with PCT as a predictive marker.

5. Conclusion H7N9 infection patients with bacterial coinfection had a more severe condition. Elevated PCT is an accurate marker for diagnosis of bacterial coinfection in influenza A (H7N9) patients. PCT levels higher than 0.81 μg/L strongly suggest bacterial coinfection in H7N9 influenza patients. We suggest that empiric antibiotic therapy should be given to patients who have bacterial coinfection manifestations and with PCT levels more than 0.81 μg/L. Furthermore, S. aureus, as well as MRSA, should be considered in choosing empiric antibiotic therapy alongside local epidemiological characteristics and drug resistance trends. Further prospective and intervention studies on bacterial coinfection with large sample numbers are needed to clarify the incidence,

risk factors, predictor, and preventive and therapy measures in patients with avian H7N9 virus infection. Financial support This study was supported by grant from Science and Technology Department of Zhejiang Province (2014C03039) and the grant from the National Major S & T Research Projects for the Control and Prevention of Major Infectious Diseases in China (2012ZX10004–206 and 2014ZX10004006). Potential conflicts of interest All authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. Acknowledgments We are indebted to the other clinical and technical staff of the First Affiliated Hospital, College of Medicine, Zhejiang University, for their cooperation in this study. References Bacterial co-infections in lung tissue specimens from fatal cases of 2009 pandemic influenza A (H1N1) - United States, May-August 2009. MMWR Morb Mortal Wkly Rep 2009;58(38):1071–4. Chalmers JD, Taylor JK, Mandal P, Choudhury G, Singanayagam A, Akram AR, et al. Validation of the Infectious Diseases Society of America/American Thoratic Society minor criteria for intensive care unit admission in community-acquired pneumonia patients without major criteria or contraindications to intensive care unit care. Clin Infect Dis 2011;53(6):503–11. Cilloniz C, Ewig S, Menendez R, Ferrer M, Polverino E, Reyes S, et al. Bacterial co-infection with H1N1 infection in patients admitted with community acquired pneumonia. J Infect 2012;65(3):223–30. Cuquemelle E, Soulis F, Villers D, Roche-Campo F, Ara Somohano C, Fartoukh M, et al. Can procalcitonin help identify associated bacterial infection in patients with severe influenza pneumonia? A multicentre study. Intensive Care Med 2011;37(5):796–800. Estenssoro E, Rios FG, Apezteguia C, Reina R, Neira J, Ceraso DH, et al. Pandemic 2009 influenza A in Argentina: a study of 337 patients on mechanical ventilation. Am J Respir Crit Care Med 2010;182(1):41–8. Finelli L, Fiore A, Dhara R, Brammer L, Shay DK, Kamimoto L, et al. Influenza-associated pediatric mortality in the United States: increase of Staphylococcus aureus co-infection. Pediatrics 2008;122(4):805–11. Gao HN, Lu HZ, Cao B, Du B, Shang H, Gan JH, et al. Clinical findings in 111 cases of influenza A (H7N9) virus infection. N Engl J Med 2013a;368(24):2277–85. Gao R, Cao B, Hu Y, Feng Z, Wang D, Hu W, et al. Human infection with a novel avianorigin influenza A (H7N9) virus. N Engl J Med 2013b;368(20):1888–97. Gilbert DN. Procalcitonin as a biomarker in respiratory tract infection. Clin Infect Dis 2011;52(Suppl. 4):S346–50. Gill JR, Sheng ZM, Ely SF, Guinee DG, Beasley MB, Suh J, et al. Pulmonary pathologic findings of fatal 2009 pandemic influenza A/H1N1 viral infections. Arch Pathol Lab Med 2010;134(2):235–43. Helferty M, Vachon J, Tarasuk J, Rodin R, Spika J, Pelletier L. Incidence of hospital admissions and severe outcomes during the first and second waves of pandemic (H1N1) 2009. CMAJ 2010;182(18):1981–7. Huang F, Guo J, Zou Z, Liu J, Cao B, Zhang S, et al. Angiotensin II plasma levels are linked to disease severity and predict fatal outcomes in H7N9-infected patients. Nat Commun 2014;5:3595. Ingram PR, Inglis T, Moxon D, Speers D. Procalcitonin and C-reactive protein in severe 2009 H1N1 influenza infection. Intensive Care Med 2010;36(3):528–32. Kallen AJ, Brunkard J, Moore Z, Budge P, Arnold KE, Fosheim G, et al. Staphylococcus aureus community-acquired pneumonia during the 2006 to 2007 influenza season. Ann Emerg Med 2009;53(3):358–65. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985;13(10):818–29. Kumar A, Zarychanski R, Pinto R, Cook DJ, Marshall J, Lacroix J, et al. Critically ill patients with 2009 influenza A(H1N1) infection in Canada. JAMA 2009;302(17):1872–9. Li Q, Zhou L, Zhou M, Chen Z, Li F, Wu H, et al. Epidemiology of human infections with avian influenza A(H7N9) virus in China. N Engl J Med 2014;370(6):520–32. Martin-Loeches I, Sanchez-Corral A, Diaz E, Granada RM, Zaragoza R, Villavicencio C, et al. Community-acquired respiratory co-infection in critically ill patients with pandemic 2009 influenza A(H1N1) virus. Chest 2011;139(3):555–62. Mauad T, Hajjar LA, Callegari GD, da Silva LF, Schout D, Galas FR, et al. Lung pathology in fatal novel human influenza A (H1N1) infection. Am J Respir Crit Care Med 2010; 181(1):72–9.

M. Yang et al. / Diagnostic Microbiology and Infectious Disease 84 (2016) 165–169 Nguyen T, Kyle UG, Jaimon N, Tcharmtchi MH, Coss-Bu JA, Lam F, et al. Co-infection with Staphylococcus aureus increases risk of severe coagulopathy in critically ill children with influenza A (H1N1) virus infection. Crit Care Med 2012;40(12):3246–50. Patel RB, Mathur MB, Gould M, Uyeki TM, Bhattacharya J, Xiao Y, et al. Demographic and clinical predictors of mortality from highly pathogenic avian influenza A (H5N1) virus infection: CART analysis of international cases. PLoS One 2014;9(3):e91630. Pfister R, Kochanek M, Leygeber T, Brun-Buisson C, Cuquemelle E, Machado MB, et al. Procalcitonin for diagnosis of bacterial pneumonia in critically ill patients during 2009 H1N1 influenza pandemic: a prospective cohort study, systematic review and individual patient data meta-analysis. Crit Care 2014;18(2):R44. Rice TW, Rubinson L, Uyeki TM, Vaughn FL, John BB, Miller III RR, et al. Critical illness from 2009 pandemic influenza A virus and bacterial co-infection in the United States. Crit Care Med 2012;40(5):1487–98.

169

von Baum H, Schweiger B, Welte T, Marre R, Suttorp N, Pletz MW, et al. How deadly is seasonal influenza-associated pneumonia? The German Competence Network for Community-Acquired Pneumonia. Eur Respir J 2011;37(5): 1151–7. Webb SA, Pettila V, Seppelt I, Bellomo R, Bailey M, Cooper DJ, et al. Critical care services and 2009 H1N1 influenza in Australia and New Zealand. N Engl J Med 2009; 361(20):1925–34. WHO risk assessment of human infection with avian influenza A(H7N9) virus. Available at: http://www.who.int/influenza/human_animal_interface/influenza_h7n9/Risk_ Assessment/en/. Zhang PJ, Cao B, Li XL, Liang LR, Yang SG, Gu L, et al. Risk factors for adult death due to 2009 pandemic influenza A (H1N1) virus infection: a 2151 severe and critical cases analysis. Chin Med J (Engl) 2013;126(12):2222–8.