Objectives: To study time trends in overall and pathogen-specific incidence rates in the national, hospitalwide surveillance of nosocomial bloodstream infections (BSI) in Belgian hospitals from 1992 to 2001. Methods: Since October 1992 the Scientific Institute of Public Health (IPH) in Brussels invites quarterly all Belgian hospitals to participate on a voluntary basis to the national, hospital-wide, nosocomial BSI surveillance. Surveillance data are transmitted by the hospitals to the IPH for analysis, national benchmarking and feedback. Data from October 1992 until June 2001 were included in the analysis. BSIs were defined as nosocomial when they occurred after 48 hours of hospital stay. Analysis for time trends was carried out using linear regression for repeated (min. 3 participation’s) observations. Trends for year were examined adjusting for hospital size and frequency of blood culturing.
Population: During the study period 148 hospitals (approximately 80% of all Belgian hospitals) participated at least once to the surveillance. 73 hospitals participated at least three times (including recent participation) and were eligible for trend analysis. Thus, a total of 991 surveillance-quarters (mean of 13.6 quarters by hospital) and 20491 nosocomial BSI episodes were included.
Results: The overall incidence rate was 7.6 BSI episodes per 10000 patient-days (pooled mean 7.2, median: 6.5) and increased from 6.0 in 1992-1994 to 8.0 in 2000-2001. However this trend was accompanied by an increase of blood culturing from 23.4 blood culture sets/1000 patient-days in 1992- 1994 to 31.2 in 2000-2001 (correlation coefficient=0.50, p= 500 beds) had a higher BSI-incidence rate (9.4/10000 p-days) than medium size (250-499; 6.8/10000 ptdays) and small hospitals (<250 beds; 5.9/10000p-days). Coagulase-negative staphylococci were the most frequently isolated micro-organisms (19.9%), followed by E. Coli (14.2%), S. aureus (13.6%), Streptococcus sp. (6.4%, with 2.1% S. pneumoniae), Enterobacter sp. (5.9%), Enterococcus sp. (5.6%), Candida sp. (5.4%), Klebsiella sp. (5.2%), Pseudomonas sp. (5.0%), and Serratia sp. (2.5%).
Conclusions: The increase in the hospital-wide incidence of BSI in Belgian hospitals from 1992 to 1999 may, at least in part, be explained by an increasing intensity of surveillance, as demonstrated by a pronounced increase in the frequency of blood culturing during the same period.