Multiple Sclerosis Journal - October 2017 - 1513

JMA Wijnands, E Kingwell et al.
Table 4. Adjusted rate ratios (aRR) for antimicrobial prescriptions in people with multiple sclerosis as compared with controls.

All antimicrobial prescriptions
MS population versus controls
MS population versus controls
stratified by sex
Males
Females
MS population versus controls
stratified by age (years)
18-39
40-49
50-59
60+
Three most common
antimicrobial prescriptions
Antibiotics
Antivirals
Antimycotics

Number of prescriptions per
10,000 person-years

Model 1

Model 2

Model 3

MS population

Controls

RR (95%)

aRR (95%)

aRR (95%)

12,647

8248

1.55 (1.48-1.62)

1.58 (1.50-1.65)

1.57 (1.49-1.65)

9522
13,667

6051
8991

1.66 (1.51-1.83)
1.52 (1.44-1.60)

1.71 (1.54-1.90)
1.54 (1.46-1.63)

1.72 (1.54-1.92)
1.54 (1.45-1.62)

11,581
11,896
13,113
14,255

8937
7785
7776
8849

1.32 (1.20-1.45)
1.55 (1.46-1.66)
1.67 (1.56-1.79)
1.66 (1.53-1.80)

1.32 (1.21-1.46)
1.55 (1.46-1.65)
1.66 (1.55-1.78)
1.64 (1.51-1.78)

1.32 (1.20-1.46)
1.54 (1.45-1.64)
1.62 (1.52-1.73)
1.66 (1.52-1.82)

10,858
1228
522

7123
784
309

1.54 (1.48-1.61)
1.52 (1.15-1.99)
1.67 (1.46-1.91)

1.57 (1.50-1.64)
1.49 (1.11-2.00)
1.62 (1.41-1.87)

1.56 (1.49-1.63)
1.51 (1.12-2.04)
1.61 (1.39-1.85)

RR: rate ratio; MS: multiple sclerosis; SES: socioeconomic status.
Model 1: crude (people with MS and controls matched by sex, year of birth, and place of residence; no additional adjustments made); Model 2: as for Model 1,
plus adjustment for sex (except for the sex-specific models), age (continuous, except for the age-specific models), year of index date (1996-2001, 2002-2007,
and 2008-2013), SES (quintiles), immunotherapy use not specific to MS (yes, no); Model 3: as for model 2, plus adjustment for the number of comorbidities
(none, 1, and ⩾2).

the effect was greater among men than women with
MS (relative to men and women in the matched general population). This finding was concordant with
that of a previous study that identified an almost fivefold increased risk for a first hospitalization due to an
infection among men with MS (relative to men without MS) as compared to a fourfold increased risk
among women with MS (relative to women without
MS).3 We showed that these sex-related differences
could be explained in part by the disproportionately
higher risk for urinary system infections in men with
MS. Men with MS are often reported as having a
poorer outcome in terms of disease progression,34 possibly increasing their risk for infections. Alternatively,
the attenuated association in women might partly be
explained by higher health services utilization by
women in general, decreasing the difference in rates
between women with and without MS.35
To the best of our knowledge, this is one of the first
studies to examine infection-related medical encounters in both the inpatient and outpatient care settings
in the MS population. Study strengths included access
to a large population-based dataset, a validated algorithm to identify people with MS, and a sophisticated

journals.sagepub.com/home/msj

statistical approach in which we considered all medial
encounters rather than just the first event. We carefully adjusted for potential confounders, although as
for all observational studies, we cannot rule out a role
of other, unmeasured, explanatory factors such as
diet, exercise, or smoking. Our study was designed to
examine infection-related medical encounters in the
general MS population, including those taking
MS-specific immunomodulatory drugs. When we
only considered the time prior to initiation of an
MS-specific immunomodulatory drug in a complementary analysis, the direction and interpretation of
findings did not change. It is possible that people with
MS are monitored more closely for infections than
their peers without MS. However, we attempted to
minimize this potential surveillance bias by only considering the primary diagnostic code, and therefore
the main reason of each medical encounter. In addition, after combining infection-related encounters
within a 2-week period and treating them as a single
event, we still found significantly higher health care
utilization in people with MS as compared to their
matched controls. However, the magnitude of the
association between MS and physician visits did
decrease, which suggests that people with MS might

1513
https://journals.sagepub.com/home/msj
Table of Contents for the Digital Edition of Multiple Sclerosis Journal - October 2017
