(Total Views: 583)
Posted On: 01/13/2021 11:46:38 AM
Post# of 145254
![](/m/assets/46931549/no_avatar_available_thumb.jpg)
Re: Borel Fields #72434
https://www.medcalc.org/calc/relative_risk.php
They get slightly different p-values for relative risk, maybe using one-sided test?
Also, you are only using N=393. Since N=394, would add one more in LL group.
For last example with a=52, b=211, c=38, and d=93, get RR=0.68 (32% relative reduction in risk of death, from 29% to 19.8%), with 95% CI from 0.47 to 0.98 (2% to 53% relative reduction).
They use 95%CI. Can subsitute Z value for other %CI if desired; for 95%, Z=1.96.
That's about what I expect, for LL use to reduce risk of death from 30% to 20%, on top of standard of care (steroids, remdesivir, other like anti-IL-6). Maybe better?
They get slightly different p-values for relative risk, maybe using one-sided test?
Also, you are only using N=393. Since N=394, would add one more in LL group.
For last example with a=52, b=211, c=38, and d=93, get RR=0.68 (32% relative reduction in risk of death, from 29% to 19.8%), with 95% CI from 0.47 to 0.98 (2% to 53% relative reduction).
They use 95%CI. Can subsitute Z value for other %CI if desired; for 95%, Z=1.96.
That's about what I expect, for LL use to reduce risk of death from 30% to 20%, on top of standard of care (steroids, remdesivir, other like anti-IL-6). Maybe better?
![](/m/images/thumb-up.png)
![](/m/images/thumb-down.png)