Additive Hazard Regression for the Analysis of Clustered Survival Data from Case-Cohort Studies

Table 1

Simulation results for the estimation of .

Design		Method
			Bias	SE	SEE	BSE	CP	Bias	SE	SEE	BSE	CP
A	0.5	SC	0.0010	0.0568	0.0574	0.0500	0.948	−0.0021	0.0677	0.0635	0.0678	0.944
		WC	−0.0026	0.0526	0.0576	0.0550	0.969	−0.0023	0.0603	0.0646	0.0630	0.966
		T	−0.0021	0.0533	0.0525	0.0496	0.953	−0.0015	0.0594	0.0597	0.0642	0.956
	2	SC	0.0015	0.0573	0.0569	0.0506	0.948	−0.0020	0.0668	0.0656	0.0672	0.942
		WC	−0.0031	0.0531	0.0571	0.0560	0.953	−0.0031	0.0625	0.0648	0.0602	0.953
		T	−0.0016	0.0529	0.0536	0.0562	0.942	−0.0010	0.0584	0.0603	0.0618	0.947
B	0.5	SC	−0.0035	0.0578	0.0569	0.0460	0.937	−0.0012	0.0683	0.0644	0.0766	0.920
		WC	0.0023	0.0593	0.0612	0.0609	0.943	0.0004	0.0642	0.0640	0.0635	0.938
		T	−0.0030	0.0534	0.0511	0.0605	0.934	−0.0039	0.0629	0.0613	0.0614	0.936
	2	SC	−0.0012	0.0568	0.0573	0.0589	0.941	−0.0009	0.0679	0.0654	0.0685	0.920
		WC	0.0027	0.0578	0.0581	0.0513	0.939	0.0015	0.0653	0.0637	0.0591	0.938
		T	−0.0019	0.0534	0.0526	0.0480	0.940	−0.0026	0.0629	0.0613	0.0557	0.936
C	0.5	SC	−0.0009	0.0576	0.0570	0.0551	0.948	−0.0019	0.0685	0.0642	0.0651	0.929
		WC	−0.0032	0.0543	0.0571	0.0571	0.958	−0.0039	0.0605	0.0650	0.0632	0.963
		T	−0.0015	0.0540	0.0518	0.0481	0.941	−0.0047	0.0605	0.0591	0.0679	0.948
	2	SC	−0.0021	0.0576	0.0570	0.0515	0.948	−0.0024	0.0679	0.0653	0.0614	0.934
		WC	−0.0017	0.0547	0.0562	0.0565	0.942	−0.0026	0.0614	0.0647	0.0644	0.956
		T	−0.0010	0.0549	0.0523	0.0510	0.937	−0.0038	0.0587	0.0604	0.0613	0.943

Estimate of from five methods: SC = estimating using the subcohort ; WC = estimating using the whole cohort, ; T = using true value, .