Multiple Sclerosis Journal - October 2017 - 1508

Multiple Sclerosis Journal 23(11)
Analyses
We examined the associations between MS and number of infection-related physician visits, and MS and
number of antimicrobials prescriptions by negative
binomial regression models. As hospital admissions
are less common, we assessed the association between
MS and infection-related hospitalizations (absence vs
presence) by logistic regression models. The models
were fitted via generalized estimating equations with
an exchangeable working correlation structure, and
adjusted for sex, current age (continuous), index year
(1996-2001, 2002-2007, and 2008-2013), SES
(quintiles), immunotherapy use (not specific for MS),
and comorbidity (none, 1, and ⩾2). The logarithm of
residency time was included as an offset in the negative binomial models. Subgroup analyses included
sex and age group (18-39, 40-49, 50-59, and 60+
years). Findings were reported as adjusted incidence
rate ratios (aRRs) or adjusted odds ratios (aORs).
Similar approaches were taken for the five most common infection groups and drug classes, except for
hospitalizations, where due to low frequencies we
assessed the number of admissions over the entire
study period using negative binomial regression.
Prompted by the observation of sex differences in the
primary analyses, sex differences were also explored
across these outcomes.
Complementary analyses included (1) censoring follow-up time at initiation of a MS-specific immunomodulatory drug (Supplementary Appendix e-1,
Table e-1a) and (2) combining infection-related
encounters occurring within each 2-week period into
one to minimize the impact of repeated encounters for
the same infection. Analyses were performed separately for hospital, physician, and prescription data
using the same approach as the main analyses.
Statistical analyses were performed using IBM SPSS
Statistics Version 22 (IBM, Armonk, NY, USA) and
SAS, Version 9.4 (SAS Institute, Cary, NC, USA).
Results
Between 1996 and 2013, we identified 7179 persons
with MS and 35,837 matched controls from the general population. For both populations, the mean age at
the index date was 45.4 years and 73.8% were women
(Table 1).
Infection-related hospitalizations
People with MS were more than twice as likely to
be hospitalized for infections (aOR: 2.39; 95%
1508

confidence interval (CI): 2.16-2.65) as the matched
controls were (Table 2; model 3). The difference in
hospitalization risk between people with and without
MS was greater among men (aOR: 3.03; 95% CI:
2.50-3.68) than among women (aOR: 2.19; 95% CI:
1.95-2.46). Infection-related hospitalizations were
more frequent in MS across all age groups, rising
from a 1.5-fold increased risk in those under 40 years
of age, to a 2.7-fold increased risk in those aged
⩾60 years. Infections that prompted the highest number of hospitalizations in the total cohort were infections of the urinary system (32/10,000 person-years
(PY)), pneumonia (17/10,000 PY), intestinal infections (11/10,000 PY), infections of the skin and subcutaneous tissue (8/10,000 PY), and sepsis (8/10,000
PY). People with MS had more hospitalizations for
each of these types of infections, with the greatest difference being for infections of the urinary system
(aRR: 3.26; 95% CI: 3.10-3.43) (Table 2; model 3).
The post hoc analyses exploring sex-related differences across the five most common infection groups
showed that men with MS generally had a higher relative risk than women. This was particularly evident
for infections of the urinary system (the aRR for men
with MS relative to men without MS was 6.43 (95%
CI: 4.57-9.06); the equivalent aRR for women was
2.74 (95% CI: 2.26-3.31); Figure 1).
Infection-related physician visits
People with MS had 41% (aRR: 1.41; 95% CI: 1.36-
1.47) more infection-related physician visits than the
matched controls (Table 3; model 3). This difference
was greater among men (aRR: 1.69; 95% CI: 1.52-
1.87) than women (aRR: 1.35; 95% CI: 1.29-1.40).
Infection-related physician visits were elevated in MS
across all age groups, rising from 25% more visits in
those under 40 years of age, to 63% in those aged
⩾60 years. Infections that prompted the highest number of physician visits in the total cohort were upper
respiratory tract infections (2210/10,000 PY), infections of the urinary system (978/10,000 PY), bronchitis and bronchiolitis (892/10,000 PY), infections of
the skin and subcutaneous tissue (869/10,000 PY),
and pneumonia (367/10,000 PY). People with MS
had more physician visits for all of these types of
infections except for upper respiratory tract infections, for which no significant difference was found.
The greatest differences between people with and
without MS were for infections of the urinary system
(aRR: 2.20; 95% CI: 2.02-2.38) and pneumonia
(aRR: 2.05; 95% CI: 1.75-2.40) (Table 3; model 3),
with men exhibiting a greater burden than women
when compared to men and women in the matched
populations, respectively (Figure 1).
journals.sagepub.com/home/msj
https://journals.sagepub.com/home/msj
Table of Contents for the Digital Edition of Multiple Sclerosis Journal - October 2017
