I am doing a meta-analysis of 18 individual studies using stata 17.0
- The effect size uses the odds ratio and 95%CI calculated from individual analyses.
- The method is to use a random effects model.
- Study sample sizes range from 40 to 1,000,000 participants. Therefore, I want to include the sample size in the weights.
The variables are:
ES = odds ratio
LCI= Low confidence interval
UCI= Upper confidence interval
ID= study ID
PATICIPANTSIZE= Study sample sizes
Q1.chether the command below is correct.
"metan ES LCI UCI, wgt( PATICIPANTSIZE ) label(namevar=ID) forestplot(null(1))"
Q2. or would you please recommend what command is appropriate to use for this analysis?
Q3. Lastly, I am wonder,
if the I2 is above 90%, What statement should I perform command for further analysis?
- The effect size uses the odds ratio and 95%CI calculated from individual analyses.
- The method is to use a random effects model.
- Study sample sizes range from 40 to 1,000,000 participants. Therefore, I want to include the sample size in the weights.
The variables are:
ES = odds ratio
LCI= Low confidence interval
UCI= Upper confidence interval
ID= study ID
PATICIPANTSIZE= Study sample sizes
Q1.chether the command below is correct.
"metan ES LCI UCI, wgt( PATICIPANTSIZE ) label(namevar=ID) forestplot(null(1))"
Q2. or would you please recommend what command is appropriate to use for this analysis?
Q3. Lastly, I am wonder,
if the I2 is above 90%, What statement should I perform command for further analysis?
Comment