Printer Friendly

Eosinopenia as a diagnostic marker of bloodstream infection in hospitalised paediatric and adult patients: a case-control study.

Bloodstream infection is associated with significant attributable morbidity and mortality. A reliable marker of bloodstream infection may allow timely initiation of antibiotic therapy resulting in improved survival (1,2). An ideal marker of bloodstream infection should be inexpensive, both sensitive and specific, able to be performed quickly and correlate with the severity of the disease. So far, no marker of bloodstream infection available has all these properties.

A reduction in blood eosinophil count, eosinopenia, in response to infection is not a new concept. Eosinopenia was first described in certain infections in 1893 and Bass et al reported a series of papers showing that eosinophil count often reduces in acute inflammation and returns to normal after recovery (35). Since then, eosinopenia has been associated with a variety of clinical conditions including bloodstream infections, viral infections, therapy with corticosteroids and catecholamines, physiological stress, psychiatric conditions such as depression and some allergic diseases (6-9).

Recently, there has been a resurgence of interest in using eosinopenia as a biomarker of infection. Eosinopenia has a strong correlation with leukocytosis in bacterial infection (9) and, using a clinical definition of infection, eosinopenia appears to be superior to a commonly used inflammatory marker - serum C-reactive protein (CRP) concentrations - in distinguishing infection causing systemic inflammatory response from a non-infective cause (10). Because bloodstream infection is a more objective and clearly delineated form of infection, it is the preferred endpoint for assessing the diagnostic utility of eosinopenia. Our recent study has indeed shown that eosinopenia is very common (86%) in critically ill patients with bloodstream infection confirmed by positive blood cultures (11). This study was, however, limited by its small sample size and a lack of a paediatric cohort.

We hypothesised that eosinopenia is a reliable marker of bloodstream infection in both hospitalised paediatric and adult patients, and conducted a case-control study to assess its diagnostic accuracy as a marker of bloodstream infection, and whether eosinopenia could give additional diagnostic information when combined with the usual conventional markers of infection such as serum CRP concentrations and neutrophil counts.

METHODS

Patient population and measurements

This case-control study was assessed by the Hospital Ethics Committee at Royal Perth Hospital (RPH) and Princess Margaret Children's Hospital (PMH) in Western Australia, and was deemed a 'clinical audit'. Western Australia has a population of 2.2 million and Perth, the capital city, has a population of 1.65 million. All tertiary medical services for Western Australia are situated in Perth. RPH is a 580-bed tertiary university teaching hospital with all medical and surgical specialties represented, including heart and lung transplantation. PMH is the only children's hospital in Western Australia.

Consecutive patients with a positive blood culture ('cases') and concurrent randomly selected 'controls' (negative blood culture) between 1 October 2008 and 19 October 2009 at PMH, and between 1 November 2008 and 30 November 2009 at RPH were identified from the Microbiology Laboratory Management Systems Database of both hospitals. The data custodians of the microbiology departments of the two study hospitals independently selected the control patients from the bloodstream infection databases in a random fashion. These two data custodians were not involved in the subsequent data extraction and analysis process in this study. A matched case-control method was not used in this study, because once covariates were matched the significance of the matched covariates could not be analysed in the multivariate analyses.

The medical records of patients with a positive blood culture were reviewed to ensure that the clinical microbiologist and attending clinician both considered the micro-organism in the blood culture as pathological and not a contaminant. All cases and controls had clinical features of systemic infection and suspicion of bloodstream infection. Blood cultures growing coagulase-negative Staphylococcus species were excluded prior to review of medical notes, because this organism frequently represents contamination. The medical records of all controls were also reviewed and patients were excluded if they were treated with antibiotics within one week prior to obtaining the blood culture. Patients who had primary haemotological disorders, immunodeficiency syndromes (e.g. Human Immunodeficiency Virus), as well as those who were treated with chemotherapy or immunosuppressive medications (e.g. corticosteroids, methotrexate, cyclosporin) were also excluded in this study.

All results analysed in this study were performed as part of their clinical management by the treating clinicians. The clinical data analysed included age, gender and mortality data of the patients. The potential markers of bloodstream infection assessed in this study included serum CRP concentrations, total white cell, neutrophil and eosinophil count, all measured within 24 hours of obtaining the blood cultures. Serum CRP concentrations were measured by an immunoenzyme analyser (Hitachi 917, Tokyo, Japan) at RPH and PMH. Eosinophil counts were determined by an automated method with minimal measurable eosinophil counts of 0.01 x [10.sup.9].[l.sup.-1] (10/[mm.sup.3]). Eosinophil counts of less than 0.01 x [10.sup.9.[l.sup.-1] (10/[mm.sup.3]) were not measurable and defined as eosinopenia in this study.

Statistical analyses

Categorical and continuous variables with skewed distributions were analysed by chi-square and Mann-Whitney U tests respectively. Area under the receiver operating characteristic curve (AUROC) was used to assess the predictive value of different inflammatory markers. Because the normal ranges of the total white cell, neutrophil and eosinophil count in the paediatric patients vary significantly between different age cohorts before 18 years of age, the percentage of the upper limit of normal for patients younger than 18 years old was used instead of absolute count in assessing their predictive effect by AUROC. Sensitivity and specificity of eosinopenia and an arbitrarily chosen CRP concentration of >100 mg.[l.sup.-1] as a marker of bloodstream infection were also calculated.

All inflammatory markers were entered into multivariate logistic regression analysis to assess the independent predictive effect of each marker and no variables were removed in the multivariate analyses. A P value of <0.05 was regarded as statistically significant, 95% confidence interval (CI) was used to quantify uncertainty, and all statistical tests in this study were performed with SPSS for Windows (version 13.0 SPSS Inc. IL, USA, 2005).

[FIGURE 1 OMITTED]

RESULTS

Among a total of 502 adult patients with a positive blood culture and 551 adult patients with infection but without a positive blood culture, 157 and 195 patients were selected as 'cases' and 'controls', respectively. Among a total of 463 paediatric patients with a positive blood culture and 403 paediatric patients with infection but without a positive blood culture, 85 and 94 paediatric patients were considered as 'cases' and 'controls' respectively (Figure 1).

A total of 165 and 92 episodes of bloodstream infections among the 157 adult and 85 paediatric 'cases' were identified during the study period, respectively. Eight adult and seven paediatric patients had more than one organism in the same blood culture specimen when the diagnosis of bloodstream infection was made. Seventy of the 165 bloodstream infections in the adult patients (42.4%) were due to Gram-positive organisms, 91 (55.2%) were due to Gram-negative organisms and four (2.4%) were due to fungal organisms. Among the 92 bloodstream infections in the paediatric patients, 60 (65.2%) were due to Gram-positive organisms, 29 (31.5%) were due to Gram-negative organisms and three (3.3%) were due to fungal organisms.

The median eosinophil count in the adult patients with bloodstream infection was significantly lower than those of the controls (0.07 vs 0.18 x [10.sup.9].[l.sup.-1] [70 vs 180 /[mm.sup.3]], respectively, P=0.001). Eosinopenia, unmeasurable eosinophil count, was significantly more common among the cases than the controls (46.5% vs 21.5%, respectively, P=0.001). The CRP concentrations, total white cell and neutrophil count were also significantly higher among the cases than the controls (Table 1).

In the univariable analyses, the total white cell count (AUROC 0.569, 95% CI 0.506 to 0.632), neutrophil count (AUROC 0.615, 95% CI 0.554 to 0.677), eosinophil count (AUROC 0.349, 95% CI 0.288 to 0.411), and the CRP concentrations (AUROC 0.680, 95% CI 0.621 to 0.739) all had some degree of predictive ability as a marker of bloodstream infection in adult patients (Table 2 and Figure 2). The specificity of eosinopenia to predict bloodstream infection in adult patients was reasonable (79%, 95% CI 74 to 82), but its sensitivity was low (47%, 95% CI 41 to 52). CRP >100 mg.[l.sup.-1] was more sensitive but less specific than eosinopenia as a marker of bloodstream infection in adult patients (sensitivity: 62%, 95% CI 57 to 67 vs 47%, 95% 41 to 52 and specificity: 66%, 95% CI 61 to 71 vs 79%, 95% 74 to 82).

In contrast to the adult patients, eosinophil count and eosinopenia were not significantly different between paediatric cases and controls (Table 3). CRP concentrations and neutrophil counts were however, significantly different between the cases and controls. In fact, CRP concentrations (AUROC 0.748, 95% CI 0.675 to 0.820, P=0.001) had the best predictive ability when used alone in paediatric patients (Table 3 and Figure 2). Using CRP concentration >100 mg.[l.sup.-1] as a predictor, the specificity of CRP to predict bloodstream infection in paediatric patients was very high (93%, 95% CI 89 to 96), but its sensitivity was very low (31%, 95% CI 24 to 38). Eosinophil count has very little overall predictive ability (AUROC 0.448, 95% CI 0.363 to 0.533, P=0.237) and the sensitivity (54%, 95% CI 47 to 61) and specificity (56%, 95% CI 49 to 63) of eosinopenia to predict bloodstream infection in paediatric patients were both low.

Multivariate analyses confirmed that the CRP concentrations, total white cell and neutrophil counts were independent predictors of bloodstream infection in both paediatric and adult patients (Table 4), but not eosinophil count.

DISCUSSION

Our results showed that eosinopenia was common (46.5%) and had a reasonable specificity (79%) as a marker of bloodstream infections in adult patients. Eosinopenia was, however, not sensitive and its overall predictive ability was not high as a marker of bloodstream infection in adult patients. The use of eosinopenia as a marker of bloodstream infection was very limited in paediatric patients. Neutrophil counts and CRP concentrations both appeared to be better markers of bloodstream infection than eosinopenia in paediatric patients.

[FIGURE 2 OMITTED]

Although eosinophils only account for a very small proportion of the peripheral white blood cells, their production is tightly regulated by interleukin-3, interleukin-5 and granulocyte-macrophage colonystimulating factor (3,10,12). Without these three cytokines, eosinophils survive for less than 48 hours (12). These three cytokines are however, not significantly activated in patients with sepsis (13) and as such, it is believed to be the main mechanism resulting in eosinopenia in patients with severe sepsis and bloodstream infections (3,14,15). Our results confirmed that eosinopenia is common in adult patients with bloodstream infections (14,15). Although eosinopenia has a reasonable specificity (79%) as a marker of bloodstream infection in adult patients, its sensitivity is low. Furthermore, eosinopenia does not appear to be useful when combined with CRP concentrations, total white cell and neutrophil counts (Table 4). These results suggest that the presence of eosinopenia can be considered as an inexpensive warning test for bloodstream infections in adult patients so that further investigations can be initiated to exclude bloodstream infection. An absence of eosinopenia is, however, not sensitive enough to exclude bloodstream infection in hospitalised adult patients.

Although earlier studies suggested that eosinopenia could also be useful as a marker of infection in paediatric patients (16,17), we could not confirm the findings of these studies. Furthermore, we found that CRP concentration and neutrophil count were better markers of bloodstream infection than eosinopenia. There are a few possible explanations for these results. First, we have included very young paediatric patients with bloodstream infection (median age two years old). Very young patients, especially neonates, may have an immature immune response and subtle differences in the activation of cytokines in response to bloodstream infection (18). Second, the sample size of our study was bigger than the previous studies and we have also excluded the confounding effect of antibiotic therapy in the control group. Third, our patients were sicker than the patients reported in the previous reports (16,17). Bloodstream infection is a more objective and clearly delineated endpoint of sepsis than clinical infection. It is possible that bloodstream infection induces a more significant generalised systemic inflammation than localised infection (13). Fourth, children generally have a shorter prodrome than adults and thus, the expected drop in eosinophil count may not have yet occurred. As such, it is possible that the relative ability of CRP, neutrophil counts and eosinopenia in differentiating infections is different, depending on the severity of systemic inflammation and amount of chemotactic factors released into the systemic circulation.

This study has limitations. First, although the sample size of this observational study was reasonably large, it may still be underpowered to demonstrate the potential predictive ability of eosinopenia as a marker of bloodstream infection, especially in paediatric patients. Second, we only had a small number of patients with fungaemia (n=7). The ability of eosinopenia to serve as a better marker of fungaemia remains uncertain, but this merits further investigation by a large study. Third, we have not included other biomarkers. We are not aware of any studies that have compared eosinopenia with procalcitonin, which may be a better marker of sepsis than CRP (19,20). Fourth, we did not collect clinical data other than age, gender and mortality of the study patients. Clinical data such as symptoms of rigor, body temperature, chest X-ray changes and comorbidities may provide additional information on the characteristics of the cases and controls, but including a substantial number of covariates in the analyses would have exceeded the power of the study. Finally, we have excluded patients who have haematological diseases and immunosuppressive therapy, and tropical infectious diseases occur rarely in our study centres. As such, our findings are not generalisable to these patients or to centres with a high prevalence of tropical infectious diseases.

Our study used a cut-off of less than 0.01 x [10.sup.9]/l or 10 cells/[mm.sup.3] to determine eosinopenia. A more sensitive measurement may improve the diagnostic utility. However, our results suggest that when eosinopenia was modelled as a continuous variable, its diagnostic utility as a marker of bloodstream infection was limited as evidenced by its area under the receiver operating characteristic curve.

In summary, the presence of eosinopenia can be considered as an inexpensive warning test for bloodstream infection in hospitalised adult patients so that further blood tests and investigations can be initiated to exclude bloodstream infection. However, an absence of eosinopenia does not exclude bloodstream infection in hospitalised adult patients. The use of eosinopenia, as a marker of bloodstream infection, is very limited in hospitalised paediatric patients. Neutrophil counts and CRP concentrations are better markers of bloodstream infection than eosinopenia in this group.

REFERENCES

(1.) Valles J, Rello J, Ochagavia A, Garnacho J, Alcala MA. Community-acquired bloodstream infection in critically ill adult patients: impact of shock and inappropriate antibiotic therapy on survival. Chest 2003; 123:1615-1624.

(2.) Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Intensive Care Med 2008; 34:17-60.

(3.) Bass DA, Gonwa TA, Szejda P, Cousart MS, DeChatelet LR, McCall CE. Eosinopenia of acute infection: production of eosinopenia by chemotactic factors of acute inflammation. J Clin Invest 1980; 65:1265-1271.

(4.) Bass DA. Behavior of eosinophil leukocytes in acute inflammation. II. Eosinophil dynamics during acute inflammation. J Clin Invest 1975; 56:870-879.

(5.) Bass DA. Behavior of eosinophil leukocytes in acute inflammation. I. Lack of dependence on adrenal function. J Clin Invest 1975; 55:1229-1236.

(6.) Juhlin L, Michaelsson G. A new syndrome characterised by absence of eosinophils and basophils. Lancet 1977; 1:1233-1235.

(7.) Freeman G. Syndromes associated with eosinopenia. Allergy 1998; 53:331-333.

(8.) Gil H, Magy N, Mauny F, Dupond JL. Value of eosinopenia in inflammatory disorders: an "old" marker revisited. Rev Med Interne 2003; 24:431-435.

(9.) Mann A, Lehmann H. The eosinophil level in psychiatric conditions. Can Med Assoc J 1952; 66:52-58.

(10.) Abidi K, Khoudri I, Belayachi J, Madani N, Zekraoui A, Zeggwagh AA et al. Eosinopenia is a reliable marker of sepsis on admission to medical intensive care units. Crit Care 2008; 12:R59.

(11.) Ho KM, Towler SC. A comparison of eosinopenia and C-reactive protein as a marker of bloodstream infections in critically ill patients: a case control study. Anaesth Intensive Care 2009; 37:450-456.

(12.) Hogan SP, Rosenberg HF, Moqbel R, Phipps S, Foster PS, Lacy P et al. Eosinophils: biological properties and role in health and disease. Clin Exp Allergy 2008; 38:709-750.

(13.) Cavaillon JM, Adib-Conquy M, Fitting C, Adrie C, Payen D. Cytokine cascade in sepsis. Scand J Infect Dis 2003; 35:535-544.

(14.) Raz R, Ben-Israel Y, Gronich D, Granot E, Colodner R, Visotzky I. Usefulness of blood cultures in the management of febrile patients in long-term care facilities. Eur J Clin Microbiol Infect Dis 2005; 24:745-748.

(15.) Lipkin WI. Eosinophil counts in bacteremia. Arch Intern Med 1979; 139:490-491.

(16.) Montesanti M, Testa G, Biagi C, Bartolini F. Pattern of circulating eosinophils in allergic children suffering from infectious disease. Minerva Pediatr 1997; 49:187-191.

(17.) Montesanti M, Testa G, Biagi C, Bartolini F. Trend of circulating eosinophils in healthy children and children suffering from infectious diseases. A retrospective study. Minerva Pediatr 1997; 49:179-186.

(18.) Farquhar J. The evaluation of the eosinopenic response to corticotrophin and cortisone in the newborn infant. Arch Dis Child 1955; 30:133-140.

(19.) Ruiz-Alvarez MJ, Garcia-Valdecasas S, De Pablo R, Sanchez Garcia M, Coca C, Groeneveld TW et al. Diagnostic efficacy and prognostic value of serum procalcitonin concentration in patients with suspected sepsis. J Int Care Med 2009; 24:63-71.

(20.) Naoki A, Seitaro F, Shigeatu E, Isao S, Kazuhiro K, Yasuhiro Y et al. Multicenter prospective study of procalcitonin as an indicator of sepsis. J Infect Chemother 2005; 11:152-159.

B. A. WIBROW *, K. M. HO ([dagger]), J. P. FLEXMAN ([double dagger]), A. D. KEIL ([section]), D. L. KOHRS **

Royal Perth Hospital and Princess Margaret Hospital for Children, Perth, Western Australia, Australia

* M.B., B.S., Registrar, Royal Perth Hospital.

([dagger]) Ph.D., F.C.I.C.M., Associate Professor, Intensivist, Royal Perth Hospital.

([double dagger]) M.B., B.S., Ph.D., B.Sc., F.R.C.P.A., Head of Department, Department of Microbiology and Infectious Diseases, Royal Perth Hospital, PathWest Laboratory Medicine and Clinical Professor, Department of Microbiology and Immunology and Pathology and Laboratory Medicine, University of Western Australia.

([section]) M.B., B.S., F.R.C.P.A., Microbiologist, Department of Microbiology, Pathwest Laboratory Medicine, Princess Margaret Hospital for Children.

** M.B., Ch.B., F.A.C.E.M., Emergency Physician, Royal Perth Hospital; Clinical Lecturer University of Western Australia and Adjunct Senior Clinical Lecturer, University of Notre Dame.

Address for correspondence: Dr B. A. Wibrow, 38 Chrysostom St, North Beach, WA 6020.

Accepted for publication on October 20, 2010.
TABLE 1
Difference in characteristics between bacteraemic and randomly
selected blood culture negative controls among all adult hospitalised
patients by univariable analyses

Variable Bacteraemia Controls P
 (n=157) (n=195) value *

Age, y (SD) 62.2 (20.7) 56.0 (22.7) 0.015
[median, IQR] [67, 46-79] [56, 34-78]

Male, n (%) 90 (57.3) 121 (62.1) 0.383

CRP, mg/l (SD) 164.5 (122) 99.2 (100) 0.001
[median, IQR] [130, 63-240] [60, 27-130]

Total white cell 13.8 (6.2) 12.2 (5.8) 0.005
count, x [12.7,9.9-17.0] [11.5, 8.0-15.2]
[10.sup.9]/l (SD)
[median, IQR]

Eosinophil count, 0.07 (0.13) 0.18 (0.94) 0.001
x [10.sup.9]/l (SD) [0.01, 0-0.07] [0.05, 0.01-0.14]
[median, IQR]

Eosinopenia 73 (46.5) 42 (21.5) 0.001
(undetectable
count), n (%)

Neutrophil count, 12.1 (5.8) 9.8 (5.5) 0.001
x [10.sup.9]/l (SD) [10.9, 8.2-14.9] [8.7, 5.9-12.1]
[median, IQR]

Hospital mortality, 33 (21.0) 31 (15.9) 0.266
n (%)

IQR=interquartile range, CRP=C-reactive protein. * P values by
chi-squared or Mann-Whitney U test.

TABLE 2
Difference in characteristics between bacteraemic and randomly
selected blood culture negative controls among all paediatric
hospitalised patients by univariable analyses

Variable Bacteraemia Controls
 (n=85) (n=94)

Age, y (SD) 4.9 (5.5) 3.3 (4.1)
[median, IQR] [2, 0.5-10] [1.5, 0.2-5.0]

CRP, mg/l (SD) 83.2 (97) 27 (48)
[median, IQR] [43, 11-126] [7, 4-26]

Total white cell count, 13.2 (8.3) 12.7 (5.6)
x [10.sup.9]/l (SD) [11, 7-18] [11, 9-16]
[median, IQR]

Total white cell count as 80 (50) 74 (37)
% of maximum normal
count for the age

Eosinophil count x 0.18 (0.29) 0.22 (0.27)
[10.sup.9]/l (SD) [0.09, 0-0.20] [0.1, 0-0.34]
[median, IQR]

Eosinophil count as % 24 (42) 26 (33)
of maximum normal
count for the age

Eosinopenia 46 (54) 41 (44)
(undetectable count),
n (%)

Neutrophil count, x 9.7 (7.5) 7.7 (5.5)
[10.sup.9]/l [8, 4-13] [6, 5.3-11]
(SD) [median, IQR]

Neutrophil count as % 112 (89) 88 (68)
of maximum normal
count for the age

Variable P value *

Age, y (SD) 0.129
[median, IQR]

CRP, mg/l (SD) 0.001
[median, IQR]

Total white cell count, 0.529
x [10.sup.9]/l (SD)
[median, IQR]

Total white cell count as 0.399
% of maximum normal
count for the age

Eosinophil count x 0.203
[10.sup.9]/l (SD)
[median, IQR]

Eosinophil count as % 0.751
of maximum normal
count for the age

Eosinopenia 0.180
(undetectable count),
n (%)

Neutrophil count, x 0.107
[10.sup.9]/l
(SD) [median, IQR]

Neutrophil count as % 0.045
of maximum normal
count for the age

IQR=interquartile range. CRP=C-reactive protein. * P values by
chi-squared or Mann-Whitney U test.

TABLE 3
Area under the receiver operating characteristic curve of C-reactive
protein, eosinophil, neutrophil and total white cell count as a marker
of bacteraemia by univariable analysis for adult (>18 years old) and
paediatric patients (0 to 18 years old)

Variable AUROC curve P value
 (95% CI)

Adult patients

White cell count 0.569 (0.506-0.632) 0.034

Neutrophil count 0.615 (0.554-0.677) 0.001

Eosinophil count 0.349 (0.288-0.411) 0.001

C-reactive protein (mg.[l.sup.-1]) 0.680 (0.621-0.739) 0.001

Paediatric patients

White cell count (as % of 0.511 (0.424-0.598) 0.797
maximum normal count for the
age of the population)

Neutrophil count (as % of 0.582 (0.497-0.666) 0.061
maximum normal count for the
age of the population)

Eosinophil count (as % of 0.448 (0.363-0.533) 0.237
maximum normal count for the
age of the population)

C-reactive protein (mg.[l.sup.-1]) 0.748 (0.675-0.820) 0.001

AUROC=area under the receiver operating characteristic curve,
CI=confidence interval.

TABLE 4
Multivariate analysis showing the independent effect of C-reactive
protein concentrations, eosinophil, neutrophil and total white cell
count as a marker of bacteraemia in (a) adult and (b) paediatric
hospitalised patients

Variable Wald Odds ratio P value
 statistics (95% CI)

Adult patients

Total WCC 18.43 0.593 <0.001
(per [10.sup.9]/l increment) (0.467-0.753)

Eosinophil count 0.19 0.852 0.659
(per [10.sup.9]/l increment) (0.419-1.733)

Neutrophil count 20.56 1.801 <0.001
(per [10.sup.9]/l increment) (1.396-2.322)

CRP 14.34 1.041 <0.001
(per 10 mg/l increment) (1.021-1.061)

Paediatric patients

Total WCC 5.76 0.843 0.016
(per [10.sup.9]/l increment) (0.734-0.969)

Eosinophil count 0.52 1.588 0.472
(per [10.sup.9]/l increment) (0.450-5.596)

Neutrophil count 6.04 1.211 0.014
(per [10.sup.9]/l increment) (1.040-1.411

CRP 9.67 1.041 0.002
(per 10 mg/l increment) (1.021-1.061)

CI=confidence interval, WCC=white cell count, CRP=C-reactive
protein. For the adult patients: Hosmer-Lemeshow chi-square
statistic and Nagelkerke R2 of the model was 4.80 (P=0.779) and
0.214, respectively. For the paediatric patients: Hosmer-Lemeshow
chi-square statistic and Nagelkerke [R.sup.2] of the model was 3.82
(P=0.873) and 0.217, respectively.
COPYRIGHT 2011 Australian Society of Anaesthetists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Wibrow, B.A.; Ho, K.M.; Flexman, J.P.; Keil, A.D.; Kohrs, D.L.
Publication:Anaesthesia and Intensive Care
Article Type:Report
Geographic Code:8AUST
Date:Mar 1, 2011
Words:4132
Previous Article:Non-invasive ventilation after extubation in patients with chronic obstructive airways disease: a randomised controlled trial.
Next Article:Antimicrobial utilisation in 37 Australian and New Zealand intensive care units.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters