Sample size calculation for cluster randomized tria

[Henri]

Hi Dear,

Can you help with codes to compute sample size for a two-group cluster randomized trial.

The ICC is assumed to be 0.05. 

Proportion of outcomes in the control group is 35% and in intervention group is 10%. 

Equal size is assumed per cluster.

Two-tailed alpha: 5%

Power: 80%.

if you can add codes that can help having a power plot according to cluster size that would be great!

Thank you a lot

H. 

[Reeza]

What's ICC here?

In general, you want PROC POWER, but how you plan to do the analysis and type of data also affect sample size calculations, ie percentages vs mean have different code.

https://support.sas.com/resources/papers/proceedings/proceedings/sugi29/211-29.pdf

There is a lot out there on PROC POWER if you search on Lexjansen.com or Google

Recommended search: "Proc power tutorial site:lexjansen.com" and consider adding clinical trial if you get too many hits. 

I'll move your question to the Statistics Forum as well. 

---

@Henri wrote:

Hi Dear,

Can you help with codes to compute sample size for a two-group cluster randomized trial.

The ICC is assumed to be 0.05. 

Proportion of outcomes in the control group is 35% and in intervention group is 10%. 

Equal size is assumed per cluster.

Two-tailed alpha: 5%

Power: 80%.

if you can add codes that can help having a power plot according to cluster size that would be great!

Thank you a lot

H. 

---