Announcing power tworates_zhu: Stata module to calculate sample size or power for a two-sample test of rates (negative binomial regression)

[Mark Chatfield]

Dear Statalisters,

I have just posted my program -power tworates_zhu- on SSC. It implements the equations in Zhu and Lakkis (2014), which is useful if, like me, you don’t have access to PASS Sample Size Software.

ssc install power_tworates_zhu

Zhu, H. and Lakkis, H. 2014. Sample Size Calculation for Comparing Two Negative Binomial Rates. Statistics in Medicine, Volume 33, Pages 376-387.
https://onlinelibrary.wiley.com/doi/....1002/sim.5947

PASS Sample Size Software, Chapter 438.
https://ncss-wpengine.netdna-ssl.com...mial_Rates.pdf

Best wishes, Mark

[Jens Eg Noergaard]

Dear Mark,

Thank you for this wonderful tool to make sample size calculations that accounts for overdispersion. I am not a statistician so it would be great if you could tell me exactly how to determine the factor to report in the overdispersion(x) command. I research fall prevention, and in a previous trial I tested for overdispersion as: r^2 / mean. The calculation was like this: (1.7)^2 / 0.99 = 2.96. Is this the value I should I use in future sample size calculation of fall preventive trials?

Hope to gain some clarity on this.

Best regards

Jens