Menu ENN Search
Language: English Français

Result with in 95% CI >100?

This question was posted the Assessment forum area and has 2 replies. You can also reply via email – be sure to leave the subject unchanged.

» Post a reply

Anonymous 534

Project coordinator

Normal user

25 Jan 2011, 02:18

Hi,
I found a result with 96% coverage of Vitamin A 95% CI (91.4 - 101.8). Is this correct? Meaning can one have result >100 within the interval?

Thanks

André BRIEND

Frequent user

25 Jan 2011, 17:06

Dear anonymous 454

Formulas commonly used to calculate CI are based on approximations. They are often very poor at calculating CI for percentages close to 0% or 100%. Of course, CI cannot be > 100%.

Mark Myatt

Consultant Epideomiologist

Frequent user

25 Jan 2011, 17:24

Is this is a survey result rather a result worked out from workload returns.

If it is a survey result then this is probably due to you (or your software) using an approximate method to calculate the 95% CI. Most approximate methods break down at low and high proportions. You need to use an exact method or, lazily, report the upper confidence limits as 99.9% (you know it cannot be 100% because you have some uncovered cases). A similar question and answers can be found at:

http://www.en-net.org.uk/question/290.aspx

If it is a workload return then either the number of doses delivered or the estimate of the size of the target population is wrong. It is not uncommon for coverage estimates from workload returns to be way off.

I hope this helps.

Back to top

» Post a reply