Volume 11 Supplement 4

Sepsis 2007

Open Access

Gender, infection, critical illness, and death: superior survival for female trauma patients admitted to the intensive care unit treated for infection in a dedicated, prospective, multicenter study

  • Brian Swenson1,
  • Hugo Bonatti1,
  • Lesly Dossett1,
  • Addison May1 and
  • Robert Sawyer1
Critical Care200711(Suppl 4):P52

https://doi.org/10.1186/cc6031

Published: 26 September 2007

Background

Multiple studies have suggested gender-based differences exist in infection incidence and outcome in critically ill patients, but these relationships have not been clearly defined. We present the largest prospective multicenter study designed to measure the impact of gender on infection incidence and outcome in critically ill surgical and trauma patients to date.

Methods

Patients admitted for at least 3 days to surgical or trauma intensive care units (ICUs) at two academic tertiary care centers from October 2001 to May 2006 were prospectively followed for the occurrence of infection and clinical outcomes. Logistic regressions were performed to assess independent risk of gender on the acquisition of an ICU infection and inhouse mortality in patients with ICU-acquired infections. Demographic and clinical variables that were significant by univariate analyses were included in the models.

Results

The final cohort followed 2,291 patients (trauma n = 1,407, nontrauma n = 884) with 383 (27.2%) female trauma patients and 413 (46.7%) female nontrauma patients. The proportion of patients with at least one ICU-acquired infection was 47.6% for trauma males and 43.1% for trauma females (P = 0.13); and was 44.2% for nontrauma male patients and 45.8% for nontrauma female patients (P = 0.63). The inhospital mortality of patients with at least one ICU-acquired infection was 12.5% for trauma males and 7.3% for trauma females (P = 0.06); and was 30.8% for nontrauma male patients and 28.0% for nontrauma female patients (P = 0.55). Logistic regression analysis did not show gender to be a significant variable in the acquisition of an ICU infection. The results of a logistic regression analysis of predictors of inhospital mortality in patients with ICU-acquired infections are presented in Table 1. Only variables significant by univariate analysis were entered into logistic models. Empty cells represent variables that were significant for only one group (for example – only trauma, but not nontrauma).
Table 1

Logistic regression analysis assessing predictors of mortality in patients with ICU-acquired infections

 

Infected trauma patients (n = 652)

Infected nontrauma patients (n = 397)

Variable

Odds ratio

95% confidence interval

P value

Odds ratio

95% confidence interval

P value

Female gender

0.45

0.22–0.95

0.036

1.15

0.696–1.911

0.58

Age, per year

1.02

0.10–1.04

0.087

1.04

1.019–1.062

<0.001

White race

1.00

-

 

1.00

-

-

Black race

0.62

0.24–1.62

0.33

1.08

0.44–2.59

0.87

Other race

1.03

0.29–3.67

0.97

3.61

1.08–12.1

0.037

Any comorbidity

0.89

0.40–2.00

0.77

0.76

0.29–1.98

0.57

Chronic corticosteroid use

0.48

0.05–5.0

0.54

1.35

0.66–2.74

0.41

Hepatic insufficiency

4.25

1.33–13.6

0.015

1.61

0.78–3.33

0.20

Cardiac disease

1.18

0.49–2.82

0.72

   

Baseline serum creatinine ≥2.0 (units)

7.46

1.11–50.1

0.039

   

Malignancy

0.17

0.018–1.56

0.12

1.39

0.80–2.44

0.24

Body mass index, per kg/m2

   

0.98

0.96–1.01

0.24

Admission from home

   

1.03

0.41–2.60

0.95

McCabe score at ICU admission, per point

1.21

0.60–2.44

0.59

1.35

0.91–2.01

0.14

APACHE II score at ICU admission, per point

1.07

1.02–1.12

0.0081

1.08

1.04–1.12

<0.001

WHO functional status at ICU admission, per point

1.46

0.89–2.39

0.14

   

Multiple Organ Dysfunction score at ICU admission, per point

1.12

1.04–1.22

0.0040

   

Probability of trauma survival

0.49

0.23–1.02

0.055

   

ICU length of stay, per day

1.02

0.97–1.07

0.42

   

Hospital length of stay, per day

0.95

0.91–0.98

0.032

   

Bold data are statistically significant.

Conclusion

Gender does not appear to play a role in acquisition of an ICU infection in either trauma or nontrauma patients. In contrast, female gender appears to provide a strong survival advantage in trauma patients with ICU-acquired infections. This phenomenon was not observed in the nontrauma group. Further investigation into the hormonal and cytokine differences between the genders in critically ill trauma patients and their response to infection may prove beneficial.

Authors’ Affiliations

(1)
Department of Surgery, University of Virginia Health System

Copyright

© BioMed Central Ltd 2007

Advertisement