. Author manuscript; available in PMC: 2013 Oct 1.

Published in final edited form as: Kidney Int. 2013 Jan 16;83(4):749–756. doi: 10.1038/ki.2012.428

Table 4.

Sample size and Classification Power for the NEPTUNE Study

	Prevalence = 0.6				Prevalence = 0.3
		r_m = 5	r_C = 10			r_m = 5	r_C = 10

δ_m/σ_m	PCC^*	PCC			PCC^*	PCC

		n=100	n=150	n=200		n=100	n=150	n=200
0.2	0.788	0.638	0.667	0.687	0.812	0.670	0.675	0.690
0.3	0.881	0.776	0.803	0.818	0.892	0.779	0.807	0.822

		r_m = 50	r_C = 10			r_m = 50	r_C = 10

δ_m/σ_m	PCC^*	PCC			PCC^*	PCC

		n=100	n=150	n=200		n=100	n=150	n=200
0.2	0.941	0.639	0.695	0.753	0.948	0.694	0.697	0.754
0.3	0.990	0.866	0.938	0.964	0.991	0.867	0.939	0.964

“Sample size and power determination are based on a classification model for the primary endpoints with significant predictors. PCC* is the maximal classification power that would be obtained under r_m genetic predictors (of 20,000 genes) and r_c clinical predictors (of 75 clinical parameters) with respect to the standardized effect-sizes, δ_m/σ_m, where δ_m and σ_m are the effect size and standard deviation of an important biomarker, respectively. In parallel, PCC is the achievable power in the actual study with n patients enrolled in the FSGS, MCD, or MN subcohorts in NEPTUNE.”