A Bayesian Adaptive Design in Cancer Phase I Trials Using Dose Combinations in the Presence of a Baseline Covariate

Table 1

Operating characteristics summarizing trial safety for designs using a covariate (WC), ignoring the covariate (IC), and parallel trials (PT) considering continuous dose combinations, .

Scenario	Design	Average % DLTs
		(% Trials: DLT rate ; % Trials: DLT rate )
		Overall	Z = 0	Z = 1
(a)	WC	30.42 (7.1; 0.5)	21.68 (56.6; 0.0)	39.17 (0.4; 31.5)
	IC	30.10 (8.6; 0.4)	14.05 (86.7; 0.1)	46.15 (0.0; 65.8)
	PT	-	21.95 (56.6; 0.0)	29.55 (10.9; 1.8)
(b)	WC	24.34 (45.4; 0.0)	19.47 (73.8; 0.1)	29.21 (13.4; 1.8)
	IC	25.72 (34.6; 0.0)	19.07 (68.1; 0.4)	32.37 (10.2; 11.7)
	PT	-	16.87 (88.3; 0.0)	19.04 (78.0; 0.0)
(c)	WC	24.15 (47.7; 0.0)	13.98 (93.6; 0.0)	34.32 (2.7; 11.0)
	IC	24.90 (41.6; 0.0)	8.60 (97.3; 0.0)	41.21 (0.8; 42.6)
	PT	-	14.78 (96.5; 0.0)	22.21 (55.2; 0.1)
(d)	WC	16.84 (98.0; 0.0)	5.63 (100.0; 0.0)	28.04 (15.4; 0.0)
	IC	21.55 (73.3; 0.0)	2.89 (100.0; 0.0)	40.22 (1.0; 36.8)
	PT	-	6.43 (100.0; 0.0)	14.54 (97.6; 0.0)

Scenario : (a) , (b) , (c) , and (d) .