Table 1

Comparison of epidemiologic studies of severe sepsis

Methods

Results for severe sepsis in the ICU

Reference

Year

Region

Design

Case identification

Time frame

Sampling frame

No. of patients screened

No. of cases identified

Age (years; mean)

Sex (% male)

% of ICU ads

Pop. incid.

Hospital mortality (%)

[22]

1995

Italy

Prospective

Consensus criteria

April 1993 – March 1994

First 3 cases each month in 99 ICUs in northern Italy

1101

128

11.6

52.2^a

[8]

1995

France

Prospective

Consensus criteria

January – February 1993

All cases in 170 medical ICUs

11,828

742

61.4

6.3

[23]

1995

USA

Prospective

Consensus criteria

August 1992 – April 1993

All cases in 3 ICUs and 3 floors in one hospital

3708^b

467

55.1^b

60^b

12.6^b

20^b

[9]

1997

USA

Prospective

Consensus criteria

January 1993 – April 1994

All ICU patients and all floor patients with blood cultures at 8 hospitals

9763

990

59^b(median)

56^a

10.1

34^b(28-day)

[12]

2001

USA

Retrospective

ICD-9-CM codes representing Consensus criteria

1995

All cases at all hospitals (n = 936) in 7 US states

880,473

98,613

63.8^b

49.6^b

11.2

1.53/1000

34.1

[14]

2003

Retrospective

Consensus criteria

December 1995 – February 2000

All cases on day 1 in 91 ICUs in national registry

56,673

15,362

65 (median)

54.3

27.1

0.51/1000

47.3

[24]

2003

USA

Retrospective

Consensus criteria

1998–1999

All cases on day 1 in 50 ICUs in Project IMPACT

21,480

2434

63.6

11.3

36.3

[25]

2003

Europe, Canada, Israel

Prospective

Consensus criteria

May 1997 – May 1998

All cases in 28 ICUs

14,364

2124

14.8

19.6–49.3 in subgroups

[13]

2003

USA

Retrospective

ICD-9-CM code for septicemia only

1979–2000

NHDS, a 1% subset of all US hospital admissions

750 M^b

10.3 M^b

57.4–60.8^b(1979–2000)

49.6–46.8^b(1979–2000)

27.8–17.9^b

[10]

2004

France

Prospective

Consensus criteria

2 weeks in November – December 2001

All cases in 206 ICUs

3738

546

65 (median)

66.9

14.6

41.9 at 2 2 months

[11]

2004

Australia/NZ

Prospective

Consensus criteria

May – August 1999

All cases in 23 ICUs

5878

691

60.7

56.9

11.8

0.77/1000

37.5

[7]

2004

Netherlands

Prospective

Consensus criteria

24 hours in December 2001

All cases in 47 ICUs

455

134

0.54/1000

[6]

2004

Norway

Retrospective

ICD-10-CM codes representing Consensus criteria

1999

All cases in all Norwegian hospitals

57.9

0.47/1000

[5]

2004

Brazil

Prospective

Consensus criteria

8 months in May 2001 – January 2002

5 ICUs

1383

241

66.2 median, all septic patients

17.4

46.9

^aMortality rate was reported only for patients admitted with severe sepsis to the intensive care unit (ICU) but not for those who developed severe sepsis later. ^bResults are for all patients (both ICU and non-ICU) with severe sepsis in the study. ads, admissions; Pop. incid., population incodence; ICD-9-CM, International Classification of Diseases, 9th Revision, Clinical Modification; Project IMPACT, A proprietary dataset of Cerner Corporation, which contains organisational, patient-care and outcome data for over 100 ICUs in US and Latin America; NHDS, US National Hospital Discharge Survey.

Linde-Zwirble and Angus Critical Care 2004 8:222 doi:10.1186/cc2917