Format

Send to

Choose Destination
Intensive Care Med. 2006 Nov;32(11):1706-12. Epub 2006 Sep 7.

A systematic review of the quality of genetic association studies in human sepsis.

Author information

1
Department of Anaesthesia, University of Newcastle upon Tyne, Royal Victoria Infirmary, NE1 4 lP, Newcastle upon Tyne, UK.

Abstract

OBJECTIVE:

Epidemiological studies demonstrate that inherited factors play a major role in the development and prognosis of sepsis. However, genetic association studies in sepsis have produced contradictory evidence of an effect from individual polymorphisms. Major methodological flaws have been reported in a number of genetic association studies in non-septic populations, relating to problems with experimental design, statistical analysis, study size, power and replication. We hypothesised that genetic association studies investigating sepsis suffer from similar problems, and that this explains the lack of consistent evidence for an effect from polymorphisms.

DESIGN:

A systematic review was conducted of published genetic association studies in sepsis from 1996-2005 using a newly devised scoring system for study quality and rigour. A Bayesian statistical analysis was also carried out to assess the false-positive report probability of identified studies.

MEASUREMENTS AND RESULTS:

Study quality was assessed using a 10-point scoring system designed from published reporting guidelines. The majority of studies were of low to intermediate quality, with deficiencies in control group selection, genetic assay technique, study blinding, statistical interpretation, study replication, study size and power. Bayesian analysis indicated that many of the studies reporting a positive association between a genetic polymorphism and sepsis were likely to represent false-positive associations.

CONCLUSIONS:

The quality and size of genetic association studies in septic patients needs to improve if advances in identifying genetic effects in sepsis are to occur. Investigators should, as a minimum, follow recommended guidelines when designing studies.

PMID:
16957907
DOI:
10.1007/s00134-006-0327-y
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Springer
Loading ...
Support Center