Skip to main content

Table 2 Results of statistical analyses at the ICU level

From: Impact of availability of guidelines and active surveillance in reducing the incidence of ventilator-associated pneumonia in Europe and worldwide

  (1) (2) (3) (4) (5) (6) (7)
  Logistic regression VAP measure 1 Logistic regression VAP measure 2 Logistic regression VAP measure 3 Logistic regression VAP measure 4 Logistic regression VAP measure 5 Logistic regression VAP measure 6 Linear regression VAP measures 1-6
Availability of VAP guidelines on ICU level 1.803 2.721*** 2.137*** 1.336* 1.456* 2.847*** 0.573***
Existence of VAP surveillance system on ICU level 1.722 1.320* 1.434** 1.001 1.403* 1.908*** 0.244***
Years of experience in critical care 1.037 1.004 0.992 1.018* 0.996 1.032*** 0.00594*
Female sex 0.874 0.919 1.303 0.677** 0.987 1.095 -0.0329
Number of beds in ICU 0.987 0.996 1.022* 1.013 0.992 1.009 0.00316
ES 1 1 1 1 1 1 0
AR   2.747** 0.189*** 3.179 0.265** 0.617 -0.0933
AT 0.681 1.687* 0.0842*** 1.328 3.800* 1.421 -0.147
AU   2.333 0.0592*** 0.173***   3.397 -0.222
BE   1.685 0.693 0.0783*** 2.809 2.948 -0.126
BR   3.506** 0.339 1.367 0.119*** 0.267** -0.293
CH 0.720 3.963*** 0.204** 0.531 0.517 3.690 0.146
CO 0.669 1.924 0.343 0.582 0.254** 6.657 -0.00351
DE 0.851 1.982* 0.0889*** 0.885 5.765 1.061 -0.108
FR 0.704 0.426*** 0.0958*** 0.190*** 0.803 4.085*** -0.598***
GB 1.329 6.824*** 0.563 0.0735*** 0.531 1.329 -0.0721
GR 0.694 1.062 0.462 0.168*** 0.101*** 0.955 -0.704**
IN   30.52*** 0.490 0.356* 0.314* 4.880* 0.382***
IT 0.222* 0.908 0.291*** 0.0844*** 0.189*** 0.491** -0.953***
MX 0.650 1.277 0.0506*** 0.584 0.734 1.027 -0.509**
NL 0.108** 2.493 0.0299*** 0.152*** 0.435 1.482 -0.818**
PE 0.506 1.846 0.00364*** 0.839 0.313* 0.198** -1.168***
PT   1.024 0.301** 0.193*** 0.278** 1.164 -0.425**
Constant        4.309***
N 1270 1668 1676 1670 1621 1641 1730
  1. VAP prevention measure 1: Head of bed elevation; VAP prevention measure 2: daily sedation and weaning protocol; VAP prevention measure 3: Oral care with chlorhexidine; VAP prevention measure 4: No ventilator circuit change unless indicated; VAP prevention measure 5: Cuff pressure control at lease every 24 hrs; VAP prevention measure 6: Strict hand hygiene using alcohol. In column (1) to (6), odds ratios of the logistic regression models are shown; Column (7) shows non-exponentiated regression coefficients of the linear regression model that was fitted by the least squares approach. In the regression analyses, all 77 countries were included as categorical variables. For lack of space, however, only those for countries with most responses (n > 20) are shown in the present table. The clustering within pseudo ICUs was taken into account by using robust standard errors (application of the cluster option in Stata 12). Country appreviations: ES = SPAIN; FR = FRANCE; IT = ITALY; AT = AUSTRIA; GB = UNITED KINGDOM; DE = GERMANY; IN = INDIA; PT = PORTUGAL; AR = ARGENTINA; BE = BELGIUM; CO = COLOMBIA; MX = MEXICO; NL = NETHERLANDS; CH = SWITZERLAND; AU = AUSTRALIA; GR = GREECE; PE = PERU; BR = BRAZIL.
  2. * p < 0.05, ** p < 0.01, *** p < 0.001.