Multiple Sclerosis Journal - October 2017 - 1489

A Manouchehrinia, O Beiki et al.
Materials and methods
Patient population
Incidence and prevalence of MS in Sweden are estimated to be around 10.2 and 189 per 100,000, respectively, which is among the highest in the world.3,4 In
this work, we analysed data from patients registered
in the Swedish MS Register (SMSreg).5 SMSreg is a
national quality register containing clinical and demographic data on about 17,600 MS patients (at the time
of this study). The SMSreg is designed to assure quality health care for patients with MS and is used in
almost all 64 Swedish neurology clinics. The register
covers around 80% of all prevalent cases of MS in
Sweden.6
Study design
We conducted a cohort analyses measuring the time
from date of birth, disease onset and diagnosis to the
date of several major disability and clinical course
milestones and mortality.
Outcome measures
EDSS score milestones 3.0, 4.0 and 6.0. The EDSS
score is the most-used method of quantifying disability in MS. The scale ranges from 0 to 10 in 0.5 unit
increments. Scoring is performed by a neurologist
and is based on the examination of eight functional
systems.7 Age at and time from onset and diagnosis of
MS to sustained EDSS score milestones 3.0 (moderate disability but no impairment of walking), 4.0 (significant disability but able to walk without aid or rest
for 500 m) and 6.0 (requires unilateral assistance to
walk about 100 m with or without resting) were our
milestones of interest with regard to EDSS score. An
EDSS score was defined as sustained if there has been
at least one more recorded EDSS score and no proceeding score was more than 0.5 score lower.
Onset of progressive MS. Year of transition to SPMS
is estimated and recorded in SMSreg by a neurologist.
Time from birth and disease onset to onset of SPMS
(mid-year of registered year in SMSreg) were estimated using the Kaplan-Meier method.
All-cause mortality. Living status and date of death
(if deceased) are yearly updated in SMSreg using statistics from the Swedish death register with total population coverage. We calculated the Kaplan-Meier
estimates of age at death and years from disease onset
to death. We also calculated age and sex standardised
mortality ratios (SMRs) according to the 2010 mortality rates of the Swedish general population.
journals.sagepub.com/home/msj

This study was approved by the Swedish regional
ethical review board in Stockholm (EPN).
Statistical analysis
Data were summarised using descriptive statistics.
The Kaplan-Meier method was used to estimate the
median time from date of birth, disease onset and
diagnosis to the outcome of interest. Log-rank tests
and flexible parametric models8 were used to compare the median times and hazards between sexes and
patient groups based on MS phenotype at onset (bout
onset vs progressive onset). We also investigated the
geographical differences in the risk of reaching disability milestones between north (Norrland), middle
(Svealand) and south of Sweden (Götaland).
To examine the impact of missing outcome dates on
the estimations, we performed the sensitivity analysis
used by Tremlett and colleagues.9 Shortly, when time
of EDSS score 6.0 was unknown, best-, worst- and
middle-case scenarios assumed that patients have
reached EDSS 6.0 at the time of first clinic visit, disease onset and time between disease onset and first
clinic visit, respectively.
We estimated the average treatment effect in treated
(ATET) on mean age at EDSS score 6.0 and conversion to SPMS using inverse-probability weighting
with regression adjustment.10,11 Outcome adjustment
was made for sex, date of MS onset and date of birth.
We model assignment to the treatment as a function of
first recorded EDSS score, age at the first EDSS score,
onset age and date of onset. Patients were categorised
into two treatment groups. Untreated was defined as
having been exposed to a first-line DMT (interferon
beta 1a, interferon beta 1b and glatiramer acetate) for
less than 6 months and treated was defined as at least
6 months of exposure to a first-line DMT. Neither of
the groups had exposure to a second-line DMT. ATET
gave us the estimated average years gained by the
means of treatment to EDSS score 6.0 and conversion
to SPMS. Since ATET is obtained from a treated population, the conditional independence and the sufficient
overlap assumptions are less restrictive.
Except in the estimation of time to death, we only
included patients with at least three neurology clinic
visits to ensure sufficient specialist exposure and
robust estimations. A sub-analysis of time to EDSS
score 6.0 with no restriction was performed to ensure
that no selection bias has been introduced to the main
analyses by only including patients with minimum
three clinic visits. Statistical analyses were performed
using Stata 14.2.12
1489

https://journals.sagepub.com/home/msj

Table of Contents for the Digital Edition of Multiple Sclerosis Journal - October 2017