Table 4.

Simulation Scenario: Parameters based on the observed EG arm from the growth hormone study.

φi = P (S = i)		OMVN		MVN
φ1	0.3	${\beta }_{2,0}^{(2)}$	−28	${\beta }_0^{(\ge 2)}$	15
φ2	0.2	${\beta }_{2,1}^{(2)}$	2	${\beta }_1^{(\ge 2)}$	1
φ3	0.5	${\beta }_{2,0}^{(3)}$	18	${\beta }_0^{(\ge 3)}$	−5
Baseline		${\beta }_{2,1}^{(3)}$	1	${\beta }_1^{(\ge 3)}$	0.5
${\mu }_1^{(1)}$	58	${\beta }_{3,0}^{(3)}$	−5	${\beta }_2^{(\ge 3)}$	0.7
${\mu }_1^{(2)}$	56	${\beta }_{3,1}^{(3)}$	0.5	${\sigma }_{2\mid 2-}^{(\ge 2)}$	4
${\mu }_1^{(3)}$	78	${\beta }_{3,2}^{(3)}$	0.7	${\sigma }_{3\mid 3-}^{(\ge 3)}$	2.3
${\sigma }_1^{(1)}$	12.3	${\sigma }_{2\mid 2-}^{(2)}$	9
${\sigma }_1^{(2)}$	16	${\sigma }_{2\mid 2-}^{(3)}$	6.3
${\sigma }_1^{(3)}$	4	${\sigma }_{3\mid 3-}^{(3)}$	4