1 25th ECCMID - 2007, Munich, Germany Magnitude of bacteremia predicts one-year mortality Kim O. Gradel 1, Henrik C. Schønheyder 1,2, Mette Søgaard 1,

25th ECCMID - 2007, Munich, Germany 1

Magnitude of bacteremia predicts one-year mortality

Kim O. Gradel1, Henrik C. Schønheyder1,2, Mette Søgaard1, Claus Dethlefsen1, Henrik Nielsen1

1Aalborg Hospital, Aarhus University Hospital, Denmark2Dartmouth-Hitchcock Medical Center, Lebanon, NH, USA


Prognostic bacteremia studies with magnitude of bacteremia

8 studies:– Old (range 1952-1995)– Few patients (range 20-464)– Often selected patient groups and bacteria– No multivariable analyses used– Only in-hospital mortality reported– Prognosis generally worsened with higher

magnitude of bacteremia


Study type

• One-year follow-up study

• Population based

• Registry study:– Registries linked via the personal

identification number (unique number given to all Danish residents)


Setting

• North Jutland County, Denmark

• Background population: ~500,000 residents (~9% of Denmark’s population)

• All acutely ill patients admitted to the nearest hospital in their county

• Centralized clinical microbiology service


Registries

• North Jutland Bacteremia Registry– Microbiological and clinical data

• Hospital Discharge Registry– Discharge diagnoses, coded by physicians

• Civil Registration System– Vital status, incl. date of emigration and

death


Inclusion criteria

• Study period: 1996-2004

• Adults (16 years and older)

• Mono-microbial bacteremias

• First-time bacteremias


Blood culture system, I

• BacT/Alert®

• Automatic colometric detection of CO2 produced by growing micro-organisms

• Incubation up till 7 days

• ~30 mL blood in each specimen, distributed equally in 3 bottles (2 aerobic, 1 anaerobic)


Blood culture system, II


Blood culture system, III


Aims

1. Is ”Magnitude of bacteremia” (bacterial growth in 1, 2, or 3 blood culture bottles) associated with short-term and long-term all-cause mortality?

2. Is this consistent within strata:– Age– Comorbidity– Speciality (medical, surgical)


Variables

• Primary exposure: 1, 2, or 3 bacteria positive blood culture bottles in the patient’s initial blood culture (BC) set (BC index 1 [ref.], 2, or 3)

• Outcome: All-cause mortality

• Covariates:– Age– Gender– Charlson comorbidity score– Speciality– Acquisition of infection – Focus


6Tumor, metastaticAIDS

3Moderate/severe liver disease

2HemiplegiaDiabetes (type 1 and 2), with complicationsAny tumorLeukaemiaLymphoma

1Myocardial infarctionCongestive heart failurePeripheral vascular diseaseCerebrovascular diseaseDementiaChronic pulmonary diseaseConnective tissue diseaseUlcer diseaseMild liver diseaseDiabetes (type 1 and 2), without complications

ScoreDisease category

Charlson comorbidity index

Overall score: 1Overall score: 0Overall score: 2Overall score: 4


Methods

• BC index distribution, overall and stratified

• Kaplan-Meier survival curve estimates, overall and stratified

• Cumulative mortality, standardized to the age and comorbidity distribution in the BC index 1 group

• Cox-regression analyses:– Crude– Adjusted, overall – Adjusted, stratified


Study population

• 6406 patients

• 53.6% males

• Median age: 72 years


BC index distribution

• The overall BC index distribution was U-shaped:– BC index 1: 1993 (31.1%)– BC index 2: 1174 (18.3%)– BC index 3: 3239 (50.6%)

• Generally few deviations from the overall BC index distribution within strata

• 92.5% were facultatively anaerobic

0

10

20

30

40

50

60

1 2 3

BC index

% B

C bo

ttles


Kaplan-Meier mortality curves, all

BC index 3BC index 2BC index 1

0

10

20

30

40

50

% m

orta

lity

07 30 365Days after first-time bacteremia


Kaplan-Meier mortality curves, medical patients (n = 3842)


0

10

20

30

40

50

% m

orta

lity



Kaplan-Meier mortality curves, surgical patients (n = 1691)


0

10

20

30

40

50

% m

orta

lity



365-day cumulative mortality, standardized to age or

comorbidity distribution as in the BC index 1 group

All 16-64 65-80 >80 0 1-2 >20

10

20

30

40

50

60

70 BC index 1BC index 2BC index 3

Age (years) Charlson score

% m

ort

alit

y


Cox-regression, overall results

1.4 (1.2-1.7)

1.3 (1.1-1.6)

1.3 (1.1-1.6)

0-7 days

Mortality rate ratios (95% CIs) for BC index 3 (BC index 1 = reference)

1.5 (1.2-1.8)

1.4 (1.1-1.7)

1.4 (1.1-1.7)

8-30 Days 31-365 days

1.1 (1.0-1.3)

BC index, age, gender, comorbidity, speciality, acquisition of infection, focus

1.1 (0.9-1.2)BC index, age, comorbidity

1.1 (0.9-1.2)BC index (crude)

Covariates


Cox-regression, age groups

1.1 (0.9-1.4)1.4 (1.0-1.9)1.1 (0.8-1.4)>80

1.0 (0.8-1.2)1.6 (1.2-2.2)1.4 (1.1-1.9)65-80

1.3 (1.0-1.7)1.0 (0.6-1.5)1.5 (1.1-2.2)16-64

31-365 days8-30 Days0-7 days


Age, years


Cox-regression, comorbidity

1.0 (0.8-1.2)1.1 (0.7-1.6)1.1 (0.8-1.5)0

1.0 (0.8-1.3)1.7 (1.2-2.5)1.5 (1.1-2.0)>2

1.2 (1.0-1.5)1.4 (1.0-1.8)1.4 (1.1-1.8)1-2

31-365 days8-30 Days0-7 days


Charlson score


Strengths of the study

• Population-based

• High statistical precision

• Complete follow-up


Limitations of the study

• “Magnitude of bacteremia” assessed by indirect (i.e., semi-quantitative) method

• Secondary data sources

• No information on severity of bacteremia (e.g., as described in the APACHE score system) or blood biochemistry parameters


Conclusions, I

• In general, BC index 3 patients had ~4% higher one-year mortality compared to BC index 1 patients

• For medical patients, BC index 3 was a marker of short-term mortality (~0-30 days)

• For surgical patients, BC index 3 was a marker of long-term mortality (beyond ~30 days)


Conclusions, II

• BC index was also associated with the cumulative mortality, but baseline patient characteristics (age, comorbidity) were more important

• The overall trends differed between some strata.


Thank you!Danke!

Tak!

1 25th ECCMID - 2007, Munich, Germany Magnitude of bacteremia predicts one-year mortality Kim O. Gradel 1, Henrik C. Schønheyder 1,2, Mette Søgaard 1,

Documents