Table 4.

Results of MCMC Simulation, with Unconditional Means and Correlations between Parameters

		Posterior Mean and Variance		Correlations
Parameter	Description	Mean	Variance	α	${\mathit{\text{d}}}_{\mathbf{2}}^{\mathit{\text{M}}}$	${\mathit{\text{d}}}_{\mathbf{3}}^{\mathit{\text{M}}}$	${\mathit{\text{d}}}_{\mathbf{4}}^{\mathit{\text{M}}}$	PB	dS	pL	qM	qS
α	Baseline mortality (logit scale)	−1.07	0.11	1	0	0	0	0	0	0	0	0
$d_2^M$	Effect of albumin on mortality	−0.31	0.51	0	1	0.64	0.44	0	0.25	0	0	0
$d_3^M$	Effect of saline on mortality	−2.34	0.64	0	0.64	1	0.51	0	0.27	0	0	0
$d_4^M$	Effect of gelofusine on mortality	−0.19	1.81	0	0.44	0.51	1	0	0.27	0	0	0
PB	Baseline NS	0.10	0.04	0	0	0	0	1	0	0	0	0
dS	Change to NS risk in “saved” patients	1.88	4.52	0	0.25	0.27	0.27	0	1	0	0	0
pL	Probability that short-term NS proves permanent	0.50	0.08	0	0	0	0	0	0	1	0	0
qM	QALYs gained by those who survive and are NS-free	19.99	24.97	0	0	0	0	0	0	0	1	0
qS	QALY loss from NS	5.38	7.94	0	0	0	0	0	0	0	0	1

Note: MCMC = Markov chain Monte Carlo; NS = neurological sequelae; QALY = quality-adjusted life-year.