Poisson confidence interval

[Mich Prov]

Hi Statalists, the command "strate diabete, per(100)" after stset allows to calculate incidence rate and 95 5 CI but i think that they are not calculated from Poisson method... how can i obtain Poisson confidence intervals ??

[Marcos Almeida]

Hello Mich,

About -strate - and the method for calculating the CIs, as informed in the Manual (http://www.stata.com/manuals13/ststrate.pdf),

strate tabulates the rate, formed from the number of failures divided by the person-time, by different levels of one or more categorical explanatory variables specified by varlist. Confidence intervals for the rate are also given. By default, the confidence intervals are calculated using the quadratic approximation to the Poisson log likelihood for the log-rate parameter. However, whenever the Poisson assumption is questionable, jackknife confidence intervals can also be calculated.

Hopefully that helps.

Best,

Marcos

[Mich Prov]

yes... i want to clarify what manual tell us :
"Confidence
intervals for the rate are also given. By default, the confidence intervals are calculated using the
quadratic approximation to the Poisson log likelihood for the log-rate parameter".
Thank you so much doctor Almeida!