Hello Dr. Amira,

Meanchange= m2-m1

SDchange= sqrt(SD12 + SD22 - (2*corr*SD1*SD2))

- m1,SD1: pre

- m2, SD2: post

- corr: correlation coefficient, a value that describes how similar the baseline and final measurements were across participants.

corr= (SD12 + SD22 - SD change2) / (2*SD1*SD2)

As you see, corr is calculated using the SD1, SD2, and SDchange. Since you do not have the SDchange, you may consider imputing this value (corr) from another similar study (study that provides SDchange).

BW

Hello Dr. Amira,

Meanchange= m2-m1

SDchange= sqrt(SD12 + SD22 - (2*corr*SD1*SD2))

- m1,SD1: pre

- m2, SD2: post

- corr: correlation coefficient, a value that describes how similar the baseline and final measurements were across participants.

corr= (SD12 + SD22 - SD change2) / (2*SD1*SD2)

As you see, corr is calculated using the SD1, SD2, and SDchange. Since you do not have the SDchange, you may consider imputing this value (corr) from another similar study (study that provides SDchange).

BW

thanks dr Ahmed for your answer

please what is the refrence to this method ?

Hello Dr,

I have made substantial changes to the answer above, please check.

These equations are adapted from Cochrane handbook, Chapter 16.

Kind regards,

Find a more comprehensive explanation in this URL

http://handbook.cochrane.org/chapter_16/16_1_3_2_imputing_standard_deviations_for_changes_from_baseline.htm

Do you mean pre-post treatment or intervention?

How to calculate mean change by using baseline and follow up data????