Impact of methodological choices in comparative effectiveness studies: application in natalizumab versus fingolimod comparison among patients with multiple sclerosis

Table 1 Overview of the analytical approaches used in the present work according to the outcomes

Outcome	PS method	Model
Counts of relapses^a	Weighting^b	Weighted negative binomial model of disease outcomes by treatment
Counts of relapses^a	Matching^c	Generalized estimating equations with negative binomial distribution and cluster for treatment status
Time to first relapse^a	Weighting^b	Weighted Cox model of disease outcomes by treatment
Time to first relapse^a	Matching^c	Frailty Cox model of disease outcomes by treatment
Time to first EDSS worsening^a	Weighting^b	Weighted Cox model of disease outcomes by treatment adjusted for visit density
Time to first EDSS worsening^a	Matching^c	Frailty Cox model of disease outcomes by treatment adjusted for visit density
Time to first EDSS improvement^a	Weighting^b	Weighted Cox model of disease outcomes by treatment adjusted for visit density
Time to first EDSS improvement^a	Matching^c	Frailty Cox model of disease outcomes by treatment adjusted for visit density

^aAnalyses were conducted in intention-to-treat, on treatment and pairwise-censoring (matching only) frameworks
^bTwo type of weights were considered (inverse probability weighting and weighting by the odds)
^cThree values of calipers were considered (0.02, 0.1, 0.2)

ISSN: 1471-2288