Magnitude of bacteraemia predicts one-year mortality

Kim Gradel, Henrik Schønheyder, Mette Søgaard, Claus Dethlefsen, Henrik Nielsen

Research output: Contribution to conference without publisher/journalPosterResearch

Abstract

Objectives: All hospitals in our region use the BacT/Alert® blood culture (BC) system with a 3-bottle BC set for adults. We hypothesized that the magnitude of bacteremia (i.e., number of positive bottles in the initial BC set) predicted one-year mortality. Methods In a population-based study we analyzed all patients with monomicrobial bacteremia in North Jutland County, Denmark, 1996-2004. Data from the County Bacteremia Registry were linked to the Hospital Discharge Registry (comorbidity) and the Danish Civil Registration System (vital status, including date of emigration or death). Patients with a BC index of 1 (i.e., one positive bottle) were chosen as the reference group. We computed Kaplan-Meier curves and performed Cox regression analyses to estimate mortality rate ratios (MRRs) with 95 % confidence intervals [CIs] 30 and 365 days after the initial BC sampling date, first in crude analyses, second in analyses adjusted for age, comorbidity, acquisition of infection (community, nosocomial, orhealth-care related), and incident or recurrent episode. In addition we stratified the analyses on acquisition of infection and pathogen group. Results A total of 6955 patients had 8152 episodes of monomicrobial bacteremia, among which one-year follow-up was possible for 8108 (99.5 %). Of these, 2539, 1511, and 4058 episodes had a BC index of 1, 2, and 3, respectively. In crude analyses, 30-day MRRs were 0.98 [0.85-1.14] and 1.20 [1.07-1.34] for BC indices 2 and 3, respectively, and similar MRRs were found after 1 year (0.99 [0.89-1.09] and 1.12 [1.04-1.21]). All estimates remained unchanged in the adjusted analyses. Results for community-acquired and health-care related bacteremia episodes were consistent with the non-stratified results, whereas all MRRs for nosocomial infections were close to 1. A BC index of 3 had the strongest long-term prognostic impact in pneumococcal bacteremia (n=855) (adjusted MRR 1.60 [1.09-2.34]). In patients with community-acquired or health-care related bacteremia, high magnitude of bacteremia (i.e., a BC index of 3) predicted increased  30-day as well as 365-day mortality.

Original languageEnglish
Publication date2007
Number of pages1
Publication statusPublished - 2007
EventEuropean Congress of Clinical Microbiology and Infectious Diseases - Munich, Germany
Duration: 31 Mar 20073 Apr 2007
Conference number: 17

Conference

ConferenceEuropean Congress of Clinical Microbiology and Infectious Diseases
Number17
CountryGermany
CityMunich
Period31/03/200703/04/2007

Fingerprint

Bacteremia
Mortality
Registries
Comorbidity
Blood Culture
Community Health Services
Community Hospital
Emigration and Immigration
Denmark
Cross Infection
Infection
Regression Analysis
Confidence Intervals
Delivery of Health Care

Cite this

Gradel, K., Schønheyder, H., Søgaard, M., Dethlefsen, C., & Nielsen, H. (2007). Magnitude of bacteraemia predicts one-year mortality. Poster session presented at European Congress of Clinical Microbiology and Infectious Diseases, Munich, Germany.
Gradel, Kim ; Schønheyder, Henrik ; Søgaard, Mette ; Dethlefsen, Claus ; Nielsen, Henrik. / Magnitude of bacteraemia predicts one-year mortality. Poster session presented at European Congress of Clinical Microbiology and Infectious Diseases, Munich, Germany.1 p.
@conference{c77cc8203c3111dcb148000ea68e967b,
title = "Magnitude of bacteraemia predicts one-year mortality",
abstract = "Objectives: All hospitals in our region use the BacT/Alert{\circledR} blood culture (BC) system with a 3-bottle BC set for adults. We hypothesized that the magnitude of bacteremia (i.e., number of positive bottles in the initial BC set) predicted one-year mortality. Methods In a population-based study we analyzed all patients with monomicrobial bacteremia in North Jutland County, Denmark, 1996-2004. Data from the County Bacteremia Registry were linked to the Hospital Discharge Registry (comorbidity) and the Danish Civil Registration System (vital status, including date of emigration or death). Patients with a BC index of 1 (i.e., one positive bottle) were chosen as the reference group. We computed Kaplan-Meier curves and performed Cox regression analyses to estimate mortality rate ratios (MRRs) with 95 {\%} confidence intervals [CIs] 30 and 365 days after the initial BC sampling date, first in crude analyses, second in analyses adjusted for age, comorbidity, acquisition of infection (community, nosocomial, orhealth-care related), and incident or recurrent episode. In addition we stratified the analyses on acquisition of infection and pathogen group. Results A total of 6955 patients had 8152 episodes of monomicrobial bacteremia, among which one-year follow-up was possible for 8108 (99.5 {\%}). Of these, 2539, 1511, and 4058 episodes had a BC index of 1, 2, and 3, respectively. In crude analyses, 30-day MRRs were 0.98 [0.85-1.14] and 1.20 [1.07-1.34] for BC indices 2 and 3, respectively, and similar MRRs were found after 1 year (0.99 [0.89-1.09] and 1.12 [1.04-1.21]). All estimates remained unchanged in the adjusted analyses. Results for community-acquired and health-care related bacteremia episodes were consistent with the non-stratified results, whereas all MRRs for nosocomial infections were close to 1. A BC index of 3 had the strongest long-term prognostic impact in pneumococcal bacteremia (n=855) (adjusted MRR 1.60 [1.09-2.34]). In patients with community-acquired or health-care related bacteremia, high magnitude of bacteremia (i.e., a BC index of 3) predicted increased  30-day as well as 365-day mortality.",
author = "Kim Gradel and Henrik Sch{\o}nheyder and Mette S{\o}gaard and Claus Dethlefsen and Henrik Nielsen",
note = "Titel p{\aa} proceedings: Proceedings of 17th ECCMID Oversat titel p{\aa} proceedings: Proceedings of 17th ECCMID; null ; Conference date: 31-03-2007 Through 03-04-2007",
year = "2007",
language = "English",

}

Gradel, K, Schønheyder, H, Søgaard, M, Dethlefsen, C & Nielsen, H 2007, 'Magnitude of bacteraemia predicts one-year mortality', Munich, Germany, 31/03/2007 - 03/04/2007, .

Magnitude of bacteraemia predicts one-year mortality. / Gradel, Kim; Schønheyder, Henrik; Søgaard, Mette; Dethlefsen, Claus; Nielsen, Henrik.

2007. Poster session presented at European Congress of Clinical Microbiology and Infectious Diseases, Munich, Germany.

Research output: Contribution to conference without publisher/journalPosterResearch

TY - CONF

T1 - Magnitude of bacteraemia predicts one-year mortality

AU - Gradel, Kim

AU - Schønheyder, Henrik

AU - Søgaard, Mette

AU - Dethlefsen, Claus

AU - Nielsen, Henrik

N1 - Titel på proceedings: Proceedings of 17th ECCMID Oversat titel på proceedings: Proceedings of 17th ECCMID

PY - 2007

Y1 - 2007

N2 - Objectives: All hospitals in our region use the BacT/Alert® blood culture (BC) system with a 3-bottle BC set for adults. We hypothesized that the magnitude of bacteremia (i.e., number of positive bottles in the initial BC set) predicted one-year mortality. Methods In a population-based study we analyzed all patients with monomicrobial bacteremia in North Jutland County, Denmark, 1996-2004. Data from the County Bacteremia Registry were linked to the Hospital Discharge Registry (comorbidity) and the Danish Civil Registration System (vital status, including date of emigration or death). Patients with a BC index of 1 (i.e., one positive bottle) were chosen as the reference group. We computed Kaplan-Meier curves and performed Cox regression analyses to estimate mortality rate ratios (MRRs) with 95 % confidence intervals [CIs] 30 and 365 days after the initial BC sampling date, first in crude analyses, second in analyses adjusted for age, comorbidity, acquisition of infection (community, nosocomial, orhealth-care related), and incident or recurrent episode. In addition we stratified the analyses on acquisition of infection and pathogen group. Results A total of 6955 patients had 8152 episodes of monomicrobial bacteremia, among which one-year follow-up was possible for 8108 (99.5 %). Of these, 2539, 1511, and 4058 episodes had a BC index of 1, 2, and 3, respectively. In crude analyses, 30-day MRRs were 0.98 [0.85-1.14] and 1.20 [1.07-1.34] for BC indices 2 and 3, respectively, and similar MRRs were found after 1 year (0.99 [0.89-1.09] and 1.12 [1.04-1.21]). All estimates remained unchanged in the adjusted analyses. Results for community-acquired and health-care related bacteremia episodes were consistent with the non-stratified results, whereas all MRRs for nosocomial infections were close to 1. A BC index of 3 had the strongest long-term prognostic impact in pneumococcal bacteremia (n=855) (adjusted MRR 1.60 [1.09-2.34]). In patients with community-acquired or health-care related bacteremia, high magnitude of bacteremia (i.e., a BC index of 3) predicted increased  30-day as well as 365-day mortality.

AB - Objectives: All hospitals in our region use the BacT/Alert® blood culture (BC) system with a 3-bottle BC set for adults. We hypothesized that the magnitude of bacteremia (i.e., number of positive bottles in the initial BC set) predicted one-year mortality. Methods In a population-based study we analyzed all patients with monomicrobial bacteremia in North Jutland County, Denmark, 1996-2004. Data from the County Bacteremia Registry were linked to the Hospital Discharge Registry (comorbidity) and the Danish Civil Registration System (vital status, including date of emigration or death). Patients with a BC index of 1 (i.e., one positive bottle) were chosen as the reference group. We computed Kaplan-Meier curves and performed Cox regression analyses to estimate mortality rate ratios (MRRs) with 95 % confidence intervals [CIs] 30 and 365 days after the initial BC sampling date, first in crude analyses, second in analyses adjusted for age, comorbidity, acquisition of infection (community, nosocomial, orhealth-care related), and incident or recurrent episode. In addition we stratified the analyses on acquisition of infection and pathogen group. Results A total of 6955 patients had 8152 episodes of monomicrobial bacteremia, among which one-year follow-up was possible for 8108 (99.5 %). Of these, 2539, 1511, and 4058 episodes had a BC index of 1, 2, and 3, respectively. In crude analyses, 30-day MRRs were 0.98 [0.85-1.14] and 1.20 [1.07-1.34] for BC indices 2 and 3, respectively, and similar MRRs were found after 1 year (0.99 [0.89-1.09] and 1.12 [1.04-1.21]). All estimates remained unchanged in the adjusted analyses. Results for community-acquired and health-care related bacteremia episodes were consistent with the non-stratified results, whereas all MRRs for nosocomial infections were close to 1. A BC index of 3 had the strongest long-term prognostic impact in pneumococcal bacteremia (n=855) (adjusted MRR 1.60 [1.09-2.34]). In patients with community-acquired or health-care related bacteremia, high magnitude of bacteremia (i.e., a BC index of 3) predicted increased  30-day as well as 365-day mortality.

M3 - Poster

ER -

Gradel K, Schønheyder H, Søgaard M, Dethlefsen C, Nielsen H. Magnitude of bacteraemia predicts one-year mortality. 2007. Poster session presented at European Congress of Clinical Microbiology and Infectious Diseases, Munich, Germany.