Content uploaded by Patrick Sh
Author content
All content in this area was uploaded by Patrick Sh on Aug 23, 2017
Content may be subject to copyright.
Please
cite
this
article
in
press
as:
Saunders-Hastings,
P.,
et
al.,
Effectiveness
of
personal
protective
measures
in
reducing
pandemic
influenza
transmission:
A
systematic
review
and
meta-analysis.
Epidemics
(2017),
http://dx.doi.org/10.1016/j.epidem.2017.04.003
ARTICLE IN PRESS
G Model
EPIDEM-263;
No.
of
Pages
20
Epidemics
xxx
(2017)
xxx–xxx
Contents lists available at ScienceDirect
Epidemics
journal homepage: www.elsevier.com/locate/epidemics
Review
Effectiveness
of
personal
protective
measures
in
reducing
pandemic
influenza
transmission:
A
systematic
review
and
meta-analysis
Patrick
Saunders-Hastingsa,∗,
James
A.G.
Crispoa,b,
Lindsey
Sikorac,
Daniel
Krewskia
aUniversity
of
Ottawa,
McLaughlin
Centre
for
Population
Health
Risk
Assessment,
850
Peter
Morand
Crescent,
Ottawa,
ON,
K1G
5Z3,
Canada
bUniversity
of
Pennsylvania,
Department
of
Neurology,
Philadelphia,
PA,
United
States
cUniversity
of
Ottawa,
Health
Sciences
Library,451
Smyth
Road,
Ottawa,
ON,
Canada
a
r
t
i
c
l
e
i
n
f
o
Article
history:
Received
28
December
2016
Received
in
revised
form
20
March
2017
Accepted
24
April
2017
Available
online
xxx
Keywords:
Pandemic
influenza
Systematic
review
Hand
hygiene
Facemask
Cough
etiquette
a
b
s
t
r
a
c
t
The
goal
of
this
review
was
to
examine
the
effectiveness
of
personal
protective
measures
in
preventing
pandemic
influenza
transmission
in
human
populations.
We
collected
primary
studies
from
Medline,
Embase,
PubMed,
Cochrane
Library,
CINAHL
and
grey
literature.
Where
appropriate,
random
effects
meta-analyses
were
conducted
using
inverse
variance
statistical
calculations.
Meta-analyses
suggest
that
regular
hand
hygiene
provided
a
significant
protective
effect
(OR
=
0.62;
95%
CI
0.52–0.73;
I2=
0%),
and
facemask
use
provided
a
non-significant
protective
effect
(OR
=
0.53;
95%
CI
0.16–1.71;
I2=
48%)
against
2009
pandemic
influenza
infection.
These
interventions
may
therefore
be
effective
at
limiting
transmission
during
future
pandemics.
PROSPERO
Registration:
42016039896.
©
2017
The
Authors.
Published
by
Elsevier
B.V.
This
is
an
open
access
article
under
the
CC
BY
license
(http://creativecommons.org/licenses/by/4.0/).
Contents
1.
Introduction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
2.
Methods
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
2.1.
Overview
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
2.2.
Search
strategy
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
2.3.
Eligibility
criteria
and
study
inclusion
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
2.4.
Data
extraction
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
2.5.
Quality
assessment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
2.6.
Data
synthesis
and
analysis
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
3.
Results.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.00
3.1.
Included
studies
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
3.2.
Quality
assessment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
3.3.
Hand
hygiene
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
3.4.
Facemask
use
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
4.
Discussion
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
5.
Conclusion
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
Conflicts
of
interest
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
Author
contributions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
Acknowledgements.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.00
Appendix
A.
PRISMA
2009
checklist
for
systematic
review
and
meta-analysis
of
personal
protective
measure
effectiveness
(Moher
et
al.,
2009).
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.00
Appendix
B.
Search
strategy
for
relevant
databases
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
Appendix
C.
Articles
excluded
during
full
review
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
Appendix
D.
Summary
of
individual
study
results
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
References
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
00
∗Corresponding
author
at:
55
Metcalfe
Street,
Suite
700,
Ottawa,
ON,
K1P
6L5,
Canada.
E-mail
address:
psaun076@uottawa.ca
(P.
Saunders-Hastings).
http://dx.doi.org/10.1016/j.epidem.2017.04.003
1755-4365/©
2017
The
Authors.
Published
by
Elsevier
B.V.
This
is
an
open
access
article
under
the
CC
BY
license
(http://creativecommons.org/licenses/by/4.0/).
Please
cite
this
article
in
press
as:
Saunders-Hastings,
P.,
et
al.,
Effectiveness
of
personal
protective
measures
in
reducing
pandemic
influenza
transmission:
A
systematic
review
and
meta-analysis.
Epidemics
(2017),
http://dx.doi.org/10.1016/j.epidem.2017.04.003
ARTICLE IN PRESS
G Model
EPIDEM-263;
No.
of
Pages
20
2
P.
Saunders-Hastings
et
al.
/
Epidemics
xxx
(2017)
xxx–xxx
1.
Introduction
Influenza
pandemics
may
arise
from
antigenic
shifts,
when
reas-
sortment
between
different
viral
strains
results
in
the
emergence
of
a
novel
influenza
virus
to
which
most
individuals
are
immunolog-
ically
naïve
(Zambon,
1999).
If
this
new
pathogen
causes
clinical
illness
in
humans
and
is
able
to
transmit
effectively
between
humans,
a
global
pandemic
may
occur.
This
has
happened
four
times
in
the
past
one
hundred
years:
the
1918
Spanish
flu,
the
1957
Asian
flu,
the
1968
Hong
Kong
flu
and
the
2009
Swine
flu.
Together
these
events
have
resulted
in
millions
of
cases
of
illness,
hospitalization
and
death,
as
well
as
a
significant
social
and
eco-
nomic
burden
(Humphries,
2013;
Henderson
et
al.,
2009;
Guan
et
al.,
2010;
Simonsen
et
al.,
2013).
The
Spanish
flu
demonstrates
the
catastrophic
potential
of
such
events,
having
caused
between
20
and
50
million
deaths
globally
(Humphries,
2013;
Jordan,
1927;
Patterson
and
Pyle,
1991;
Johnson
and
Mueller,
2002).
Advances
in
medicine
and
public
health
render
such
dramatic
consequences
unlikely
today
(Saunders-Hastings
and
Krewski,
2016).
The
emer-
gence
of
antivirals,
vaccines
and
mechanical
ventilators
should
help
protect
from
such
a
catastrophic
pandemic
in
the
future,
and
pandemic-attributable
mortality
has
decreased
in
the
three
influenza
pandemics
since
the
Spanish
flu9.
However,
the
unpre-
dictable
nature
of
influenza
pandemics,
coupled
with
increasing
opportunities
for
viral
reassortment,
necessitate
further
studies
of
appropriate
mechanisms
to
respond
to
such
events
and
mitigate
their
impact.
The
irregular
cycle
of
influenza
pandemics
makes
them
diffi-
cult
to
study,
with
most
of
the
available,
scientifically
rigorous
data
deriving
from
the
recent
2009
pandemic.
This
is
problematic
given
that
the
2009
pandemic
strain
—
A(H1N1)pdm09
—
is
known
to
have
been
quite
mild,
with
hospitalization
and
death
rates
similar
to
recent
seasonal
influenza
(Henderson
et
al.,
2009).
The
disease
characteristics
of
future
pandemics
may
differ
substantially
from
those
in
the
past.
Little
is
known
about
the
effectiveness
of
a
suite
of
potential
interventions
to
interrupt
pandemic
influenza
infec-
tion.
This
is
especially
true
of
non-pharmaceutical
measures
such
as
social
distancing
(school
closure,
patient
quarantine)
and
personal
protective
measures
(PPMs).
Pharmaceutical
measures
such
as
pan-
demic
vaccination
are
effective
(Breteler
et
al.,
2013;
Demicheli
et
al.,
2014;
Manzoli
et
al.,
2009;
Yin
et
al.,
2012;
Saunders-Hastings
et
al.,
2016),
but
may
not
be
available
in
the
early
stages
of
a
pan-
demic
influenza
outbreak
(Madhav,
2013;
Longini
et
al.,
2004).
Social
distancing
policies,
meanwhile,
are
of
uncertain
effective-
ness,
and
are
often
expensive,
unpopular
and
difficult
to
implement
(Isfeld-Kiely
and
Moghadas,
2014;
Perez
Velasco
et
al.,
2012;
Borse
et
al.,
2011).
Consequently,
patient
quarantine
has
not
been
broadly
implemented
since
the
1918
pandemic
(Markel
et
al.,
2006),
while
uncertainty
regarding
the
effectiveness
of
school
closure
has
lim-
ited
its
implementation
over
the
course
of
the
past
three
pandemics
(Trotter
et
al.,
1959;
Earn
et
al.,
2012).
However,
PPMs
such
as
respiratory
etiquette,
hand
hygiene
and
the
use
of
facemasks
are
inexpensive
and
easy
to
implement,
and
are
commonly
recom-
mended
and
undertaken
during
influenza
outbreaks
(Aledort
et
al.,
2007;
Aiello
et
al.,
2010a;
van
der
Weerd
et
al.,
2011).
The
primary
objective
of
this
review
and
meta-analysis
is
to
quantify
the
effectiveness
of
PPMs
in
reducing
the
risk
of
human-
to-human
pandemic
influenza
infection.
A
secondary
objective
is
to
assess
the
relative
effectiveness
of
these
interventions.
This
is
cur-
rently
an
important
knowledge
gap:
a
search
for
existing
systematic
reviews
evaluating
pandemic
influenza
interventions
found
only
a
single
systematic
review
(Wong
et
al.,
2014)
on
the
effective-
ness
of
PPMs
in
preventing
pandemic
influenza
infection;
focusing
specifically
on
hand
hygiene
in
the
community,
the
review
found
only
one
study
that
was
conducted
during
an
influenza
pandemic
(Suess
et
al.,
2011).
However,
the
authors
only
included
randomized
control
trials
(RCTs)
in
their
analysis,
potentially
missing
impor-
tant
insights
from
observational
studies.
Given
the
important
role
PPMs
may
play
in
the
early
stages
of
a
future
pandemic,
this
review
provides
an
important
and
timely
assessment
of
the
state
of
PPM
literature
and,
where
possible,
quantification
of
pooled
estimates
of
PPM
effectiveness
in
interrupting
pandemic
influenza
transmis-
sion.
2.
Methods
2.1.
Overview
Preferred
Reporting
Items
for
Systematic
Reviews
and
Meta-
Analyses
(PRISMA)
guidelines
were
used
to
guide
the
development
of
the
systematic
review
methodology
(Appendix
A)
(Moher
et
al.,
2009).
A
protocol
was
developed
a
priori
a
registered
in
the
National
Institute
for
Health
Research
International
Prospective
Register
of
Systematic
Reviews
(PROSPERO)
(Saunders-Hastings
et
al.,
2016).
2.2.
Search
strategy
The
literature
search
was
conducted
by
PSH
on
June
30,
2016,
with
no
language
or
date
restrictions.
Searches
were
conducted
across
five
databases:
PubMed
(all
dates),
Medline
via
Ovid
(1946-
June
30,
2016),
Embase
via
Ovid
(1947-June
30,
2016),
Cochrane
Library
via
Ovid
(all
dates)
and
the
Cumulative
Index
to
Nursing
and
Allied
Health
(CINAHL;
all
dates).
Database-specific
variations
are
included
in
Appendix
B.
To
supplement
these
searches,
researchers
conducted
searches
of
the
reference
lists
of
included
studies,
and
of
the
grey
literature
using
Google
Scholar.
2.3.
Eligibility
criteria
and
study
inclusion
In
this
review,
investigators
sought
to
assess
the
effectiveness
of
commonly
recommended
PPMs
in
reducing
the
risk
of
pan-
demic
influenza
infection
in
humans.
Personal
protective
measures
included
any
form
of
hand
hygiene,
use
of
facemasks
or
respiratory
etiquette
(covering
mouth
during
coughing
and
sneezing).
Inter-
ventions
more
commonly
recommended
for
healthcare
staff,
and
less
likely
to
be
implemented
in
community
settings,
were
not
con-
sidered.
These
interventions
include
the
use
of
goggles,
gowns
and
gloves
to
prevent
influenza
transmission.
Table
1
lists
the
relevant
eligibility
criteria
developed
a
priori
and
applied
throughout
the
screening
process.
All
citations
were
imported
into
the
web-based
system-
atic
review
software
DistillerSR
(Evidence
Partners
Incorporated,
Ottawa,
Canada).
Following
deduplication,
two
independent,
blind
reviewers
conducted
title
and
abstract
screening
using
a
pilot-
tested
DistillerSR
screening
form
that
reflected
the
eligibility
criteria.
An
assenting
response
from
at
least
one
reviewer
resulted
in
article
inclusion
for
full
review,
where
articles
were
again
sub-
jected
to
blind
review
by
two
independent
reviewers
using
a
piloted
DistillerSR
form.
At
this
stage,
disagreements
were
resolved
by
con-
sensus;
third-party
arbitration
was
implemented
as
necessary.
2.4.
Data
extraction
Two
independent
reviewers
(PSH,
JC)
extracted
data
from
included
studies
using
an
adapted
data
collection
form
developed
by
The
Cochrane
Collaboration
(Anon.,
2013).
The
form
collected
information
on
study
population(s),
methods,
intervention(s),
out-
come
measure(s)
and
results.
Please
cite
this
article
in
press
as:
Saunders-Hastings,
P.,
et
al.,
Effectiveness
of
personal
protective
measures
in
reducing
pandemic
influenza
transmission:
A
systematic
review
and
meta-analysis.
Epidemics
(2017),
http://dx.doi.org/10.1016/j.epidem.2017.04.003
ARTICLE IN PRESS
G Model
EPIDEM-263;
No.
of
Pages
20
P.
Saunders-Hastings
et
al.
/
Epidemics
xxx
(2017)
xxx–xxx
3
Table
1
Eligibility
criteria
for
assessed
studies.
Category
Inclusion
Criteria
Exclusion
Criteria
Population
Humans
exposed
to
a
pandemic
influenza
In
vivo
and
in
vitro
laboratory
studies,
and
non-human
species
Intervention(s)
Any
form
of
hand
hygiene,
respiratory
etiquette,
or
the
wearing
of
facemasks
No
intervention
or
other
interventions,
including
vaccination,
antiviral
use,
school
or
work
closure,
or
contact
reduction
Comparison(s)
The
impact
or
effectiveness
of
an
intervention
of
interest
compared
to
no
intervention
or
other
forms
of
intervention.
No
comparisons
made
Outcome(s)
Quantified
change
in
risk
of
pandemic
influenza
transmission
No
quantified
impact
Study
type Randomized
trials,
case-control,
cohort,
and
cross-sectional
studies
Predictive
mathematical
modeling
studies;
case
study
and
case
series
reports;
case-crossover,
crossover,
before-after,
and
ecologic
studies;
expert
opinion
and
editorials;
systematic
reviews
and
meta-analyses
2.5.
Quality
assessment
Quality
assessment
was
conducted
for
included
studies
during
the
data
extraction
process.
Tables
assessing
the
risk
of
bias
for
randomized
trials,
case-control
and
cohort
studies
were
included
in
the
data
extraction
form.
Quality
assessment
of
randomized
tri-
als
adopted
the
Cochrane
Collaboration’s
tool
for
assessing
risk
of
bias
(Higgins
et
al.,
2011).
Quality
appraisal
for
case-control
and
cohort
studies
was
done
using
the
Newcastle-Ottawa
scale
(Wells
et
al.,
2014).
Both
tools
have
been
externally
validated
and
are
com-
monly
used
in
reviews
of
public
health
interventions
(Higgins
et
al.,
2011;
Stang,
2010).
Disagreements
regarding
potential
sources
of
bias
were
discussed
and
resolved
by
consensus.
The
quality
of
cross-
sectional
studies
was
not
assessed,
as
the
risk
of
bias
in
these
studies
was
deemed
to
be
innately
high,
and
cross-sectional
studies
were
excluded
from
data
pooling.
Study
quality
was
not
used
as
an
eli-
gibility
criterion;
rather,
potential
sources
of
bias
are
noted
in
the
discussion
of
the
implications
of
review
findings.
2.6.
Data
synthesis
and
analysis
Data
from
included
studies
were
grouped
based
on
categories
of
interventions
and
outcome
measures.
Synthesis
proceeded
in
a
two-stage
process.
Comparable
data
and
risk
estimates
from
studies
with
similar
interventions
and
outcome
measures
were
imported
into
Revman
5.3.5
(http://tech.cochrane.org/revman).
Where
possible,
we
used
estimates
of
effect
from
models
adjusted
for
the
maximum
number
of
covariates.
Inverse
variance
weight-
ing
and
random-effects
modeling
was
used
in
all
meta-analyses.
The
presence
of
significant
data
heterogeneity,
as
measured
using
the
I2statistic
(I2>
50%
was
considered
to
represent
possible
sub-
stantial
uncertainty,
as
conveyed
in
the
Cochrane
Handbook
(Anon.,
2011),
did
not
preclude
data
pooling,
but
was
noted
in
the
discus-
sion
of
results.
In
instances
where
only
a
single
study
was
identified,
intervention
results
were
excluded
from
meta-analyses
and
were
qualitatively
described.
Data
permitting,
subgroup
analyses
by
age
were
conducted.
Risk
of
publication
bias
was
assessed
via
funnel
plot
inspection
and
Egger
regression
test.
3.
Results
A
total
of
2660
citations
were
retrieved
from
our
search
of
the
literature.
Following
elimination
of
duplicates,
1671
citations
were
retained
for
title
and
abstract
review.
Of
these,
52
were
retained
for
full
review,
along
with
a
single
study
identified
through
additional
searches.
Overall,
16
studies
met
the
inclusion
criteria
for
our
study
and
were
subject
to
data
extraction.
The
study
selection
process
is
summarized
in
Fig.
1,
while
a
justification
for
articles
excluded
during
full
review
is
included
in
Appendix
C.
3.1.
Included
studies
Of
the
16
studies
included,
15
(Cheng
et
al.,
2010;
Deng
et
al.,
2010;
Li
et
al.,
2011;
Liu
et
al.,
2011;
Toyokawa
et
al.,
2011;
Godoy
et
al.,
2012;
Kim
et
al.,
2012;
Suess
et
al.,
2012;
Zhang
et
al.,
2012;
Kuster
et
al.,
2013;
Remschmidt
et
al.,
2013;
Azor-
Martinez
et
al.,
2014;
Merk
et
al.,
2014;
Torner
et
al.,
2015;
Zhang
et
al.,
2013)
quantitatively
described
at
least
one
measure
of
hand
hygiene
effectiveness,
while
eight
(Cheng
et
al.,
2010;
Deng
et
al.,
2010;
Toyokawa
et
al.,
2011;
Kim
et
al.,
2012;
Suess
et
al.,
2012;
Zhang
et
al.,
2012;
Kuster
et
al.,
2013;
Jaeger
et
al.,
2011)
mea-
sured
the
effectiveness
of
facemask
use;
no
studies
evaluating
the
impact
of
any
form
of
respiratory
etiquette,
such
as
covering
one’s
mouth
during
coughing
or
sneezing,
were
identified.
All
studies
derived
from
the
A(H1N1)pdm09
pandemic.
Six
unique
combina-
tions
of
intervention
group
and
outcome
measures
were
pooled
for
meta-analysis;
two
of
these
had
significant
data
heterogeneity.
There
were
insufficient
data
to
conduct
meta-analyses
based
on
age
group.
The
characteristics
of
individual
studies
are
summarized
in
Table
2,
organized
by
study
and
intervention
types,
while
results
from
individual
studies
are
included
in
Appendix
D.
3.2.
Quality
assessment
Inter-rater
agreement
of
the
quality
of
included
studies
was
strong.
The
results
of
this
assessment
are
included
in
Tables
3–5
,
which
highlights
potential
sources
of
bias
across
different
study
types.
These
are
incorporated
into
the
discussion
of
the
implica-
tions
of
findings
of