/v1/supplement/title

We’re sorry, something doesn't seem to be working properly.

Please try refreshing the page. If that doesn't work, please contact us so we can address the problem.

Nosocomial pneumonia and bacteraemia in the Belgian intensive care units (ICU) network: epidemiology and risk factors

  • G Hanique1,
  • O Ronveaux2,
  • B Jans2,
  • C Eeckman3 and
  • R Mertens2
Critical Care19971(Suppl 1):P043

DOI: 10.1186/cc49

Published: 1 March 1997

Introduction

Surveillance of nosocomial infections (NI) with feedback of own results to healthcare personnel has repeatedly been shown to be an efficacious measure, that can significantly contribute to the prevention of these infections. Moreover, multicentric surveillance can give an added value to local results by offering aggregate results as a basis tor comparison. However, one of the prerequisites for meaningful comparisons to be made is that the results be comparable in terms of case mix, intrinsic patient risk and exposure to high-risk devices and treatments.

Objectives

The aim of this study was to measure the incidence of nosocomial pneumonia (PN) and blood stream infection (BSI) in Belgian ICUs and to evaluate the influence of the different risk factors included in the surveillance on these NIs.

Materials and methods

In January 1996, a new voluntary nationwide surveillance system was initiated jointly by the Belgian Intensive Care Society (SIZ) and the Institute of Hygiene and Epidemiology (IHE). Between January and June, 8475 patients from 64 different ICU units were observed during a total of 77 ICU-trimesters. Data included patient characteristics at entry (including type of admission. SAPS II score and underlying disease), day-by-day exposure to high-risk devices and treatments (including ventilation and central lines), and vital status at discharge.

PN and BSI were definitions based upon HELICS criteria. For each episode diagnostic criteria were registered in detail, including microbiological evidence. Multivariate logistic regression was used to determine the significant risk factors (RF) and their associated odds ratios (OR).

Results

In total, 547 PN and 200 BSI were observed (incidence: 6.5% and 2.4%; first episodes only). Median SAPS II score was 28 (PN: 42, BSI: 41.5); median length of stay was 4 days (PN: 14, BSI: 15) and global death rate was 9.8% (PN: 30.5%, BSI: 35.5%). Significant RF at admission (P < 0.01) and their OR are given in the table.

During stay, crude IN risk is evidently linked to duration of stay, PN risk to number of ventilation days, BSI risk to central catheter days and incurrent pneumonia. However, the complex nature of the confounding by duration of stay, (early) mortality or discharge requires more complex models than this first approach by logistic regression.

A logistic model was also used to predict mortality risk for the whole group and for the PN and BSI subgroups.

Conclusions

Our results are in accordance with those currently found in the literature. Even if the risk factors are not exactly the same, they represented the two classical risk poles: severity of illness and the exogenous exposition to specific devices. More research is needed to see how risk assessment can be simplified, whilst keeping its value for risk adjustment with the aim of comparing performance between individual units.

Table

Risk factors for pneumonia

OR

Risk factors for bacteraemia

OR

SAPS II score ≥ 20 and < 30

2.3

SAPS II score ≥ 30 and < 50

5.1

SAPS II score ≥ 30 and < 50

4.4

SAPS II score ≥ 50

8.4

SAPS II score ≥ 50

11.7

Infection at entry (any site)

1.3

Prior thoracic surgery

3.5

  

Unscheduled surgical admission

1.6

  

Authors’ Affiliations

(1)
Clinique Universitaires St Lue
(2)
Institute of Hygiene and Epidemiology, Ministry of Social Affairs and Public Health
(3)
Universitair Ziekenhuis Gent

Copyright

© Current Science Ltd 1997

Advertisement