Suggested answer to exercise: Education for atopic dermatitis, 3

The main outcome measure was severity of eczema as a score on an atopic dermatitis scale. The abstract contained the following:

‘Significant improvements in severity of eczema and subjective severity were seen in all intervention groups compared with control groups (total score for severity: age 3 months to 7 years –17.5, 95% confidence intervals –19.6 to –15.3 v –12.2, –14.3 to –10.1 . . .)’ These numbers are the mean change in severity score, before minus after, over 12 months.

Question 2: We are told that there was a significant improvements in severity of eczema and in the intervention group compared with the control group. Could we deduce this from the confidence intervals presented, without further calculation?

Suggested answer:

The confidence intervals do not overlap, and so we could conclude from them that the two groups are significantly different. Our intervals estimates do not include any of the same possible values for the mean changes in scores.

The difference would be significant if the confidence interval for the first group did not contain the population value for the second. As the confidence interval for the second group does not overlap with the first, the population value for the second group being in the first confidence interval would not be consistent with the data.

Back to Exercise: Education for atopic dermatitis.

To Clinical Biostatistics index.

To IAPT index.

To Martin Bland's M.Sc. index.

To Martin Bland's home page.

This page maintained by Martin Bland.
Last updated: 22 January, 2008.

Back to top.