Question 8: Is there evidence of statistical heterogeneity in Figure 2?
No, there is little evidence of statistical heterogeneity.
There is no pair of confidence intervals which does not overlap, so it would be quite feasible for there to be a common odds ratio with which all the studies would be consistent. The test for heterogenity has P=0.3, which is nowhere near significant. Hence there is little evidence of heterogeneity.
However, the numbers of patients re-intubated is quite small, zero in two groups, and there is not a lot of information. As there is clear clinical heterogeneity, we might want to be cautious here and treat the studies as heterogenous anyway.
Back to question.
Back to Systematic Reviews index.
To Martin Bland's home page.
This page maintained by Martin Bland.
Last updated: 18 February, 2009.
Back to top.