Results and lessons learnt from a randomized controlled trial: prophylactic treatment of vestibular migraine with metoprolol (PROVEMIG)

Table 3 Key secondary outcome results (least square mean change difference or odds ratio)^a

Secondary End points	N^b	Placebo	Metoprolol
DHI mean total score	91
LS mean change^c (95% CI)		0.159 (−0.252 to 0.570)	0.080 (−0.310 to 0.470)
Difference vs. placebo (95% CI)		−0.079 (−0.360 to 0.201)
P value		0.577
Pursuit eye movement	92
Patients achieving response^d, n (%)		5 (11.6)	8 (16.3)
OR (95% CI)		0.132 (0.045–0.305)	0.195 (0.092–0.416)
Difference vs. placebo, OR (95% CI)		1.483 (0.454–5.277)
P value		0.520
SVV	90
Patients achieving response^d, n (%)		4 (9.5)	2 (4.2)
OR (95% CI)		0.105 (0.032–0.262)	0.043 (0.012–0.179)
Difference vs. placebo, OR (95% CI)		0.413 (0.055–2.235)
P value		0.322

CI confidence interval
^a Least square (LS) mean change difference (estimates derived by complete case analysis of covariance for absolute change scores) or odds ratio (OR; estimates derived by logistic regression (unadjusted)); analysis of absolute change from baseline at month 6 for Dizziness Handicap Inventory (DHI) by analysis of covariance; for change state from baseline at month 6 in eye movement and subjective visual vertical (SVV) by logistic regression for the full analysis set (FAS) sample
^b Numbers of patients with non-missing observations for both baseline and 6-month visit (FAS population: n = 127; placebo n = 63; metoprolol n = 64)
^c Change score means difference between post-intervention score at month 6 versus baseline score; see Table 1 for description of DHI score ranges
^d Logistic regression for the change state in smooth pursuit eye movement or SVV between baseline and month 6 (1, change from abnormal to normal; 0, otherwise); pursuit eye movement — treatment response means change from ‘saccadic’ to ‘smooth’

ISSN: 1745-6215