A cohort study of coagulase negative staphylococcal mastitis in selected dairy herds in Prince Edward Island

Abstract

The epidemiology and importance of coagulase negative staphylococcal (CNS) mastitis in Prince Edward Island had not been documented. To investigate this, a cohort of 84 cows at seven farms were quarter sampled eight times over a lactation, commencing with samples taken prior to drying off in the previous lactation. Thirteen species of CNS were isolated. The quarter prevalence of CNS mastitis varied from 4.8% to 6.4% in the first five months of lactation and increased to 14.2 to 16.6% in the last four months of lactation. The geometric mean somatic cell counts (SCC) for quarters infected with CNS and uninfected quarters were 90 x 10(3) and 64 x 10(3) respectively (difference significant at p > 0.005). The two month new infection risk of CNS was 9.0% while the two month elimination risk was 74.4%. Infection with CNS did not alter the risk of subsequent infection with Staphylococcus aureus. The results from this project support the classification of CNS as a minor pathogen in mastitis control programs.

Full text

A
Cohort
Study
of
Coagulase
Negative
Staphylococcal
Mastitis
in
Selected
Dairy
Herds
in
Prince
Edward
Island
T.
Jeffrey
Davidson,
Ian
R.
Dohoo,
Alan
W.
Donald,
Harry
Hariharan
and
Kathy
Collins
ABSTRACT
The
epidemiology
and
importance
of
coagulase
negative
staphylococcal
(CNS)
mastitis
in
Prince
Edward
Island
had
not
been
documented.
To
investigate
this,
a
cohort
of
84
cows
at
seven
farms
were
quarter
sampled
eight
times
over
a
lactation,
commenc-
ing
with
samples
taken
prior
to
drying
off
in
the
previous
lactation.
Thirteen
species
of
CNS
were
isolated.
The
quarter
prevalence
of
CNS
mastitis
varied
from
4.8%
to
6.4%
in
the
first
five
months
of
lactation
and
increased
to
14.2
to
16.6%
in
the
last
four
months
of
lactation.
The
geometric
mean
somatic
cell
counts
(SCC)
for
quarters
infected
with
CNS
and
uninfected
quarters
were
90
x
103
and
64
x
103
respectively
(difference
significant
at
p
>
0.005).
The
two
month
new
infection
risk
of
CNS
was
9.0%
while
the
two
month
elimination
risk
was
74.4%.
Infection
with
CNS
did
not
alter
the
risk
of
subsequent
infection
with
Staphylococcus
aureus.
The
results
from
this
project
support
the
classification
of
CNS
as
a
minor
pathogen
in
mastitis
control
programs.
RESUME
Aucun
rapport
sur
l'epidemiologie
et
l'importance
des
staphylocoques
coagulase
negatif
(SCN)
sur
l'Ile-du-
Prince-Edouard
n'a
ete
publiee
a
ce
jour.
Une
etude
fut
entreprise
sur
une
cohorte
de
84
vaches
riparties
sur
7
fermes.
Huit
echantillons
de
lait
de
chaque
quartier
furent
pris
sur
une
periode
d'une
lactation.
L'echantfllon-
nage
a
debute
avant
le
tarissement
de
la
lactation
precedente.
Treize
especes
de
SCN
furent
isolees.
La
prevalence
par
quartier
de
mammite
causee
par
des
SCN
variait
de
4,8
A
6,4%
durant
les
cinq
premiers
mois
de
lactation
et
a
augmente
jusqu'A
14,2%
a
16,6!7o
durant
les
quatre
derniers
mois
de
lac-
tation.
La
moyenne
geometrique
du
comptage
des
cellules
somatiques
(CCS)
etait
de
90
x
103
pour
les
quartiers
infectes
avec
des
SCN
et
de
64
x
103
pour
les
quartiers
non
infectes
(diffe-
rence
significative
a
p
<
0,005).
Le
risque
de
nouvelle
infection
au
SCN
pour
une
periode
de
2
mois
etait
de
9,0%
et
le
risque
d'elimination
pour
une
periode
de
deux mois
etait
de
74,4%.
L'infection
avec
un
SCN
n'a
pas
modifle
le
risque
d'une
infection
subsequente
avec
Staphylococcus
aureus.
Les
risultats
de
cette
etude
sont
en
accord
avec
la
classification
des
SCN
comme
pathogenes
mineurs
dans
les
programmes
de
contr8le
de
la
mam-
mite.
(Traduit
par
Dr
Emile
Bouchard)
INTRODUCTION
Coagulase
negative
staphylococci
(CNS)
are
the
organisms
most
fre-
quently
isolated
from
milk
samples
of
lactating
cows
(1,2).
In
the
past,
little
was
known
about
the
epidemiology
of
CNS
intramammary
infections
(IMI)
because
CNS
were
classified
as
minor
pathogens
and
considered
insignificant
in
mastitis
control
programs
(3-11).
In
Prince
Edward
Island
(P.E.I.),
there
appeared
to
be
a
high
prevalence
of
CNS
in
a
number
of
dairy
herds.
This
study
examined
selected
dynamics
of
this
organism
in
these
herds.
In
the
United
States,
there
are
discrepancies
in
reports
describing
the
prevalence
of
CNS
mastitis
throughout
lactation
of
dairy
cows.
Various
stud-
ies
have
reported
that
the
prevalence
increases
(6),
decreases
(8,12)
and
remains
constant
(7)
throughout
lacta-
tion.
The
spontaneous
elimination
rate
of
CNS
during
lactation
also
varied
from
15.5%
to
407o
in
studies
in
the
United
States
and
France
(6,13,14).
The
protective
effect
of
CNS
on
the
rate
of
infection
of
major
pathogens
has
been
studied
with
varying
results
(4,5,9,10,16-19).
Coagulase
negative
staphylococci
have
been
shown
to
increase
the
somatic
cell
count
(SCC)
in
infected
quarters
in
most
studies
(7,10,15,20).
The
objectives
of
this
study
were
to
determine
the
new
infection
and
spon-
taneous
elimination
rates
of
CNS
throughout
lactation,
to
determine
the
influence
of
CNS
infection
on
subse-
quent
infections
with
Staphylococcus
aureus
and
to
investigate
the
influence
of
CNS
infection
on
SCC
in
selected
dairy
herds
in
P.E.I.
MATERIALS
AND
METHODS
SELECTION
OF
HERDS
AND
COWS
Seven
dairy
herds
within
a
30
km
radius
of
Charlottetown,
P.E.I.
were
selected
to
participate
in
this
study.
Herds
were
selected
on
the
basis
of
an
apparently
elevated
prevalence
of
CNS
mastitis,
a
willingness
to
participate
in
the
study
and
an
established
profes-
sional
relationship
with
the
principal
investigator.
In
these
herds,
84
cows
were
due
to
freshen
between
May
and
August
1988.
Heifers
were
not
included
because
they
would
be
unable
to
pro-
vide
all
the
samples
required
(i.e.
predrying/and
drying
off
samples).
All
of
the
herds
in
this
study
prac-
ticed
postmilking
teat
dipping
and
dry
cow
therapy.
Dry
cow
therapy
was
Can
J
Vet
Res
1992;
56:
275-280
Atlantic
Veterinary
College,
University
of
Prince
Edward
Island,
550
University
Avenue,
Charlottetown,
Prince
Edward
Island
CIA
4P3.
Sources
of
support:
Technology
Development
Program,
Agriculture
Canada,
Tuco
Products
Company
and
Ayerst
Laboratories.
Submitted
June
18,
1991.
275

Fig.
1.
Flowchart
used
to
identify
organisms
isolated
from
milk
samples.
cloxacillin
or
a
novobiocin-penicillin
combination.
There
was
no
history
of
Strepto-
coccus
agalactiae
on
these
farms
for
at
least
five
years.
At
the
beginning
of
the
study,
bulk
tank
samples
from
all
the
farms
were
taken
weekly
for
three
weeks.
All
these
samples
were
negative
for
S.
agalactiae.
The
herds
were
assumed
to
be
free
of
this
organism
for
the
duration
of
the
study.
SAMPLING
SCHEDULE
Each
cow
was
scheduled
to
be
quarter
sampled
eight
times
during
the
study.
Samples
were:
predry
sample
(SI),
taken
one
week
before
drying
off;
dry
sample
(S2),
taken
at
the
time
of
drying
off;
fresh
sample
(S3),
taken
within
one
week
after
freshening;
post-
fresh
sample
(S4),
taken
between
one
and
two
weeks
after
freshening;
two
to
three
month
sample
(S5),
taken
two
to
three
months
after
freshening;
four
to
five
month
sample
(S6),
taken
four
to
five
months
after
freshening;
six
to
seven
month
sample
(S7),
taken
six
to
seven
months
after
freshening
and
the
eight
to
nine
month
sample
(S8),
taken
eight
to
nine
months
after
freshening.
COLLECTION
AND
PROCESSING
OF
SAMPLES
For
Sl
and
S2,
the
cows
were
mistakenly
milked
before
collection
of
the
samples.
Immediately
after
the
removal
of
the
milking
equipment,
the
teat
ends
were
scrubbed
with
alcohol
and
quarter
samples
were
obtained
in
an
aseptic
manner.
For
S3
to
S8,
fore-
milk
was
used
for
sampling
purposes.
After
the
producer
had
washed
and
dried
the
teats,
the
teat
ends
were
scrubbed
with
alcohol
and
four
streams
of
foremilk
were
discarded
before
collection
of
the
samples
in
an
aseptic
manner.
Samples
were
taken
directly
from
the
farm
to
the
Diagnostic
Micro-
biology
Laboratory
at
the
Atlantic
Veterinary
College.
Platinum
loops
were
used
to
transfer
0.01
mL
of
milk
aseptically
from
the
milk
vials
to
a
blood
agar
plate
(Columbia
agar
base,
5%
defibrinated
sheeps'
blood).
This
plate
was
then
incubated
at
35°C
for
24
h.
After
the
transfer
of
the
milk
to
the
blood
agar
plates,
the
milk
vials
were
sent
to
the
P.E.I.
Provincial
Dairy
Laboratory
for
somatic
cell
count
determination
by
the
Fossomatic
Somatic
Cell
Counter:
Model
215
(A/SN.
Foss
Electric,
69,
Slangerupgade
Dk-3400
Hillerod,
Denmark).
If
a
delay
in
this
transfer
was
anticipated,
the
samples
were
kept
in
a
refrigerator.
The
maximum
delay
was
24
h.
Figure
1
shows
the
flowchart
used
to
identify
the
various
organisms.
A
Gram
stain
was
performed
on
each
sample.
Gram-negative
rods
and
gram-positive
rods
were
identified
as
such
and
not
classified
further.
A
catalase
test
was
performed
on
gram
+
ve
cocci
to
dif-
ferentiate
catalase
negative
Streptococ-
cus
sp.
from
catalase
positive
Staphyl-
ococcus
and
Micrococcus.
Streptococci
were
not
further
identified.
The
tube
coagulase
test
was
performed
on
all
catalase
positive
cocci.
Coagulase
posi-
tive
cocci
were
classified
as
S.
aureus.
Coagulase
negative
cocci
were
classi-
fied
as
either
micrococci
or
coagulase
negative
staphylococci
by
colony
char-
acteristics
and
the
glucose
fermenta-
tion
test.
Number
of
colonies
of
all
organisms
were
recorded.
SPECIES
IDENTIFICATION
Between
July
18
and
August
2,
1988
all
isolates
of
staphylococci
were
saved
for
species
identification.
These
cul-
tures
were
frozen
on
trypticase
soy
agar
(TSA)
slants
and
analyzed
in
January
1989
using
Staph
Ident
(API
Laboratory
Products
Ltd.,
St-Laurent,
Quebec)
and
repetition
of
the
tube
coagulase
test.
In
February
and
March
1989,
eight
coagulase
+
ve
staphylo-
cocci
samples
and
100
coagulase
-
ve
staphylococci
samples
were
tested
in
a
similar
manner.
CLASSIFICATION
OF
CULTURE
RESULTS
The
method
of
determining
the
number
of
colonies
needed
to
establish
a
CNS
infection
was
to
compare
the
colony
count
on
S1
and
S2
samples
(2
1
colony,
2
2
colonies,
.
5
colo-
nies)
to
a
standard.
The
standard
was
276
Gram
Stain
-yE
+v
G
(-)
ro
Morphology
r
cocci
Catalase
Strep
sp
Cog
Colony
characteristics
and
glucose
fermentation
test
micrococci
CNS
S.
aureus

defined
as
positive
culture
results
for
CNS
in
both
SI
and
S2.
The
CNS
>2
colonies
had
the
most
acceptable
combination
of
sensitivity,
specificity
and
overall
correct
classification
and
was
selected
as
the
criterion
for
a
posi-
tive
culture
(Table
I).
Based
on
previ-
ously
published
criteria,
a
minimum
of
two
colonies
and
five
colonies
were
required
for
positive
infection
status
for
S.
aureus
and
gram
-
ve
rods
respectively
(21).
Culture
results
were
classified
as
fol-
lows.
A
culture
was
classified
as
coagu-
lase
negative
staphylococcus-pure
(CNSP)
if
it
had
at
least
two
colonies
of
CNS
with
no
other
growth.
A
cul-
ture
was
classified
as
coagulase
nega-
tive
staphylococcus-all
(CNSA)
if
it
had
at
least
two
colonies
of
CNS
with
no
more
than
one
other
colony
type
isolated.
A
culture
was
classified
as
S.
aureus
(SA2)
if
it
had
at
least
two
colonies
of
S.
aureus.
A
culture
was
classified
as
gram
-
ve
rods
(GN5)
if
it
had
at
least
five
colonies
of
gram
-
ve
rods.
The
prevalence
of
strepto-
cocci
(nonagalactiae)
was
very
low,
therefore
these
organisms
and
gram
+
ve
rods
were
classified
as
contami-
nants
and
samples
yielding
only
these
organisms
were
recorded
as
having
no
significant
growth
(NSG).
If
three
or
more
colony
types
were
isolated
from
a
sample,
the
sample
was
considered
to
be
contaminated
(CON).
However,
if
S.
aureus
was
isolated,
then
it
was
considered
a
SA2
positive
culture
even
if
three
different
colony
types
were
isolated
(22).
STATISTICAL
ANALYSIS
Data
were
entered
and
stored
in
dBASE
III
Plus
(23),
a
database
management
software
package.
Statis-
tical
analyses
were
conducted
using
Minitab
Version
7.1
(24)
on
a
main-
frame
computer
(VAX).
The
chi-square
statistic
was
used
to
determine
significant
differences
between
the
prevalences,
by
stage
of
lactation,
of
CNSP
and
CNSA.
This
statistic
was
also
used
to
determine
the
significance
of
CNS
infection
on
sub-
sequent
infections
with
S.
aureus.
Multiple
linear
regression
analyses
were
conducted
to
determine
if
CNSP
and
CNSA
had
a
significant
effect
on
the
natural
log
of
the
SCC
(lnSCC).
The
new
infection
and
spontaneous
elimination
risks
between
consecutive
samples
(approximately
two
months
apart)
were
calculated
for
CNSA,
CNSP
and
SA2.
The
numerator
for
the
new
infection
risk
was
the
percent-
age
of
uninfected
quarters
in
the
initial
sample
that
were
infected
in
the
sub-
sequent
sample.
The
denominator
was
the
percentage
of
uninfected
quarters
in
the
initial
sample.
The
numerator
of
the
elimination
risk
was
the
percentage
of
infected
quarters
in
the
initial
sam-
ple
that
spontaneously
eliminated
the
infection
before
the
subsequent
sam-
ple.
The
denominator
was
the
percent-
age
of
infected
quarters
in
the
initial
sample.
All
proportions
were
expressed
as
percentages
and
were
"two
month
risks"
(i.e.
the
percentage
of
new
cases
over
a
two
month
period
among
quarters
at
risk).
The
effect
of
a
CNS
infection
in
a
quarter
on
the
subsequent
infection
rate
of
SA2
in
that
quarter
was
calcu-
lated.
The
numerator
was
the
percent-
age
of
quarters
infected
with
CNS
in
the
initial
sample
that
became
infected
with
SA2
in
the
subsequent
sample.
The
denominator
was
the
percentage
of
quarters
infected
with
CNS
and
uninfected
with
SA2
in
the
initial
sample.
RESULTS
All
eight
samples
were
collected
and
processed
on
56
of
the
84
cows
in
the
cohort.
Of
the
remaining
28
cows
with
missing
samples,
15
missed
one
sample
and
eight
missed
two
samples.
The
main
reason
for
missing
data
was
the
lack
of
opportunity
to
collect
S1
and
S8
before
the
cows
were
dried
off.
A
total
of
64
isolates
of
staphylo-
cocci
were
saved
in
July
and
August
1988
for
species
identification.
Of
the
24
initially
identified
to
be
coagulase
+
ve,
17
were
subsequently
identified
as
coagulase
-ve
and
not
S.
aureus.
The
40
cultures
initially
identified
as
CNS
were
confirmed
to
be
CNS.
Samples
were
also
saved
in
February
and
March
1989
for
species
identifica-
tion.
The
eight
initially
identified
as
coagulase
+
ve
were
S.
aureus,
and
the
100
initially
identified
as
coagulase
-
ve
were
CNS.
The
species
and
num-
ber
of
isolations
of
CNS
confirmed
in
the
above
samples
are
in
Table
II.
The
quarter
prevalence
of
organisms
by
sample
is
given
in
Table
III.
The
prevalences
of
CNSP
and
CNSA
were
TABLE
I.
Sensitivities,
specificities
and
correct
classification
of
CNS
infections
in
an
84
cow
study
in
P.E.I.
No.
of
CNS
colonies
-1
<2
>5
Sensitivity
100%
78%
53%
Specificity
75%
83%
91%
Correct
classification
81%
84% 87%
The
standard
was
defined
as
positive
culture
results
for
CNS
in
both
SI
and
S2
Three
criteria
were
compared
to
the
standard;
>
1
colonies
of
CNS,
>
2
colonies
of
CNS
and
>5
colonies
of
CNS
For
each
criteria,
all
the
quarters
in
SI
and
S2
were
compared
to
the
standard
in
the
corre-
sponding
quarter.
The
sensitivity,
specificity
and
correct
classification
was
computed
for
each
criteria
TABLE
H.
Specation
by
Staph
Ident
of
a
subset
of
CNS
isolates
from
an
84
cow
study
in
P.E.I.
Staphylococcus
hominis
S.
sciuri
S.
xylosus
X.
epidermidis
S.
warneri
S.
capitis
S.
saprophytius
S.
hemolyticus
S.
simulans
S.
cohnii
S.
intermedius
S.
warnei
S.
hyicus
Total
number
of
CNS
speciated
25
24
18
17
16
16
11
10
6
5
4
4
1
157
relatively
constant
throughout
the
first
five
months
of
lactation,
ranging
from
4.8
%
and
13.1%
for
S3
to
6.4%
and
16.7%
for
S6
for
CNSP
and
CNSA
respectively.
The
prevalences
increased
in
the
sixth
to
ninth
month
of
lactation
to
16.6%
and
25.3
!7
for
S7
and
14.2%7o
and
27.4%
for
S8
for
CNSP
and
CNSA
respectively.
The
S1
and
S2
had
prevalences
of
6.5%
and
7.8%,
and
17.5%
and
20.0%
respectively,
for
CNSP
and
CNSA
respectively.
There
were
significant
differences
in
the
prevalence
of
CNSP
by
stage
of
lacta-
tion
(p
<
0.001)
and
CNSA
by
the
stage
of
lactation
(p<0.001).
The
mean
SCC
for
S5
to
S8
are
pre-
sented
in
Table
I.
Somatic
cell
counts
were
not
used
for
S1
and
S2
because
samples
were
taken
after
the
quarters
were
milked.
These
stripping
samples
would
not
yield
reliable
estimates
of
the
SCC
(25).
Counts
for
S3
and
S4
were
also
eliminated
as
these
samples
were
taken
within
two
weeks
of
freshening,
thus
rendering
them
invalid
277

TABLE
III.
Prevalence
of
organisms
(quarter
basis)
in
an
84
cow
cohort
study
in
P.E.I.
Sample
SI
s2
S3
S4
S5
S6
S7
s8
(Predry)
(Dry)
(Fresh)
(Postfresh)
(2-3
m)
(4-5
m)
(6-7
m)
(8-9
m)
No.
of
samples
Observation
308
210
336 332
320
312
308
288
CNSa
pure
(CNSP)
6.5%
7.8%
4.8%
4.2%
4.7%
6.4%
16.6%
14.2%
CNS
all
(CNSA)
17.5%
20.0%
13.1%
9.9%
10.9% 16.7%
25.3%
27.4%
S.
aureus
(SA2)
NA NA
NA
NA
3.8%
5.8%
3.6%
2.8%
Gram
-ve
rod
(GN5)
1.3%
1.3%
0.6%
0.9%
0.3%
0.3%
0.0%
0.3%
Contaminated
7.8% 8.1%
9.2% 8.7%
14.7%
3.2% 7.8%
8.0%
No
growth
66.9%
63.4%
74.1%
78.6%
69.7%
75.0%
63.6%
61.1%
Arithmetic
mean
SCCb
(x
1C3)
NA
NA
NA
NA
136.8
175.1
217.2
390.3
Geometric
mean
SCC
(x
103)
NA
NA
NA
NA
31.8
45.8
79.9
126.6
aCNS
=
coagulase
negative
staphylococci
bSCC
=
somatic
cell
count
NA
=
not
applicable
(22).
The
mean
SCC
increased
from
S5
to
S8
arithmetically
(136.8
x
103
to
390.3
x
103)
and
geometrically
(31.8
x
103
to
126.6
x
103).
The
mean
SCC
by
organism
is
given
in
Table
IV.
The
overall
arithmetic
mean
SCC
for
CNSP
and
CNSA
were
291.2
x
103
and
251.0
x
103
respec-
tively.
Mean
SCC
for
CNSP
and
CNSA
were
both
higher
than
mean
SCC
for
NSG.
The
two
month
new
infection
and
elimination
risks
are
given
in
Table
IV.
The
infection
risk
was
greater
between
S6
to
S7
(CNSP-14.7%,
CNSA-22.5)
and
S7
to
S8
(CNSP-11.5%,
CNSA-
24.0%)
than
between
S4
to
S5
(CNSP-
4.0%o,
CNSA-9.3%)
and
S5-S6
(CNSP-6.7%,
CNSA-14.4%).
The
over-
all
two
month
infection
risk
for
CNSP
and
CNSA
were
9.0%
and
17.0Gb
respectively.
The
two
month
elimination
risks
ranged
from
52.4%
to
100Gb
for
CNSP
and
60.001o
to
72.4%
for
CNSA.
The
overall
two
month
elimination
risks
for
CNSP
and
CNSA
were
74.4%
and
64.7%
respectively.
The
overall
rate
of
new
infection
with
SA2
was
2.2%.
The
new
infection
rates
for
SA2
were
2.7%
and
1.3%
for
quarters
previously
infected
with
CNSP
and
CNSA
respectively.
The
new
infection
rates
for
SA2
were
2.0%o
and
2.50%
for
quarters
previously
not
infected
with
CNSP
and
CNSA
respec-
tively.
There
were
no
significant
dif-
ferences
between
these
rates.
The
natural
log
of
the
SCC
(lnSCC)
was
regressed
on
CNSP
and
CNSA
separately
to
determine
if
CNSP
or
CNSA
had
a
significant
influence
on
SCC.
Cow,
sample
number,
lactation
number,
farm
of
origin,
SA2
infection
and
GN5
infection
were
controlled
in
the
regression.
Coagulase
negative
staphyloccus-pure
and
CNSA
had
a
significant
positive
influence
on
lnSCC.
The
coefficients
for
CNSP
and
CNSA
in
the
regression
equation
were
0.2721
(p=0.024)
and
0.1813
(p=
0.049)
respectively.
Overall
r2
values
were
29.601o
and
29.5%
for
models
con-
taining
CNSP
and
CNSA
respectively.
DISCUSSION
There
are
no
standard
criteria
for
the
diagnosis
of
CNS
infections
in
the
bovine
udder.
For
this
reason
CNS
was
evaluated
by
two
criteria,
CNSP
and
CNSA.
Most
analyses
were
based
on
results
from
a
single
sample
so
it
was
necessary
to
establish
a
minimum
number
of
colonies
required
for
a
diagnosis
of
CNS.
Based
on
results
from
one
set
of
paired
samples,
two
colonies
provided
the
best
combination
of
sensitivity,
specificity
and
overall
correct
classification.
The
reason
for
the
misclassification
of
organisms
in
the
samples
saved
for
speciation
in
July
and
August
1988
was
not
determined.
It
was
concluded
that
identification
of
S.
aureus
was
not
reliable
during
the
summer
of
1988.
Therefore
S.
aureus
data
were
not
used
in
the
analysis
of
data
from
S1,
S2,
S3
and
S4.
During
this
period,
CNS
pre-
valence
was
probably
underestimated
by
approximately
12%
due
to
misclas-
sification
of
CNS
as
S.
aureus.
The
species
and
number
of
isola-
tions
of
CNS
may
affect
the
manner
in
which
CNS
affects
mastitis
control
in
a
herd.
The
overall
prevalences
of
CNSP
(8.0%)
and
CNSA
(17.40/o)
were
similar
to
values
reported
from
other
studies.
In
a
clinical
trial
involving
16
herds
in
Vermont,
the
prevalences
of
CNS
in
teat
dipped
and
nonteat
dipped
quarters
were
7.2%
and
11.0%
respectively
(3).
The
criterion
for
infec-
tion
was
positive
duplicate
foremilk
samples.
Another
study
in
Britain
involving
30
herds,
found
a
13.4%
prevalence
of
CNS
in
herds
not
prac-
tising
teat
dipping
and
dry
cow
treat-
ment,
and
a
6.3
%
prevalence
in
the
same
herds
after
three
years
of
teat
dipping
and
dry
cow
therapy
(26).
The
criterion
for
infection
was
two
con-
secutive
isolations
of
CNS
from
samples
taken
one
week
apart.
All
the
herds
in
the
current
study
practised
teat
dipping
and
dry
cow
treatment.
A
five
year
study
of
several
herds
in
Sweden
found
7.6%
of
quarters
infected
with
a
pure
culture
of
CNS
(27).
The
Uni-
versity
of
Kentucky
dairy
herd
had
a
CNS
quarter
prevalence
of
14.3%7o
(7).
Criteria
for
infection
in
this
study
were
not
given,
making
the
findings
difficult
to
interpret.
Comparison
of
results
among
studies
can
be
difficult
because
of
the
lack
of
common
criteria
for
infection
status
(13).
The
overall
two
month
new
infection
risks
for
CNSP
and
CNSA
were
9.01o
and
17.0%
respectively.
The
new
infec-
tion
risk
increased
during
lactation.
Other
studies
have
shown
infection
risks
to
remain
fairly
constant
during
lactation.
In
the
results
from
the
two
University
of
Wisconsin
herds,
infec-
tion
risks
were
constant
but
no
values
were
presented
(6).
One
previous
study
involving
24
heifers in
Iowa,
showed
the
one
month
infection
risk
to
be
278

fairly
constant
throughout
lactation
at
approximately
12%
(14).
The
overall
two
month
elimination
risks
for
CNSP
and
CNSA
of
74.4%
and
64.7/o
respectively
during
lacta-
tion
were
considerably
higher
than
in
other
studies.
In
the
University
of
Wisconsin
study,
where
the
criterion
for
infection
rate
was
two
cultures
positive
for
CNS
one
week
apart
and
the
criterion
for
elimination
was
two
cultures
negative
for
CNS
one
week
apart,
the
elimination
risk
over
the
lac-
tation
was
15.5%
(6).
This
low
rate
may
have
been
due
to
the
strict
criteria
for
elimination
of
CNS.
In
another
study
involving
two
commercial
herds
in
France,
the
three
week
elimination
risk
was
34%
(13).
The
criterion
for
elimination
risk
was
similar
to
the
Uni-
versity
of
Wisconsin
study
except
that
sampling
took
place
every
three
weeks.
In
this
study,
foremilk
samples
were
taken
without
discarding
the
first
strip-
pings.
These
strippings
are
reported
to
contain
more
CNS
than
the
rest
of
the
milk
(28).
The
species
composition
in
first
strippings
compared
to
other
samples
and
the
frequency
of
sampling
may
account
for
the
different
elimina-
tion
risks.
In
the
study
involving
20
heifers
in
Iowa,
the
one
month
elimination
risk
was
407o
(14).
The
criterion
for
new
infections
was
the
repetition
of
the
isolation
of
CNS
after
it
was
first
isolated,
and
an
elevated
SCC
count.
The
study
did
not
define
the
criterion
for
elimination.
The
species
composition
of
CNS
in
the
various
studies
probably
vary
(1),
which
could
account
for
the
different
persistences
(14).
Different
persistences
would
account
for
the
differences
in
elimination
risks
(7,13).
This
study
did
not
demonstrate
an
influence
of
CNS
infections
in
protect-
ing
the
quarter
from
S.
aureus
infec-
tions.
The
failure
to
demonstrate
this
may
have
been
due
to
the
small
num-
ber
of
quarters
previously
infected
with
CNS
which
subsequently
became
infected
with
S.
aureus.
The
evidence
from
the
literature
seems
equivocal.
There
are
numerous
studies
showing
a
protective
effect
from
infection
with
major
pathogens
by
CNS.
A
study
in
1966
involving
a
herd
in
Great
Britain
suggested
that
mildly
pathogenic
CNS
might
protect
the
udder
from
infection
by
more
pathogenic
bacteria
(4).
In
a
controlled
laboratory
experiment
in
TABLE
IV.
The
arthmetic
and
geometric
mean
SCC
by
organisms
from
an
84
cow
(1228
samples)
cohort
study
in
P.E.I.
Arithmetic
mean
Geometric
mean
SCC
SCC
x103
x103
S5-S8
S5-S8
CNSb
pure
(CNSP)
291
90
CNS
all
(CNSA)
251
75
Staphylococcus
aureus
(SA2)
920
375
Gram
-
ve
rods
(GN5)
44
43
No
growth
186
64
aSCC
=
somatic
cell
count
bCNS
=
coagulase
negative
staphylococci
TABLE
V.
The
two
month
new
infection
risk
and
elimination
risk
for
320
quarters
in
a
cohort
study
in
P.E.I.
Sample
S4-S5
S5-
S6
S6-S7
S7-S8
Coagulase
negative
staph-pure
(CNSP)
Both
-ve
(o)
91.7
89.1
79.9
75.0
New
infection
(%)
3.8
6.4
13.8
9.7
Elimination
(¾)
3.5
4.5
3.3
11.5
Both
+ve
(%)
1.0
0
3.0
3.8
Two
month
infection
risk
(!)
4.0
6.7 14.7
11.5
Two
month
elimination
risk
(No)
77.8 100.0
52.4
75.2
Coagulase
negative
staph-all
(CNSA)
Both
-ve
(¾)
81.3
76.3
64.8
57.3
New
infection
(¾)
8.3
12.8
18.8
18.1
Elimination
(%)
7.6
7.4
9.9
15.3
Both
+ve
(%)
2.9
3.5
6.6
9.0
Two
month
infection
risk
(o)
9.3
14.4
22.5
24.0
Two
month
elimination
risk
(¾)
72.4
67.9
60.0
63.0
S.
aureus
(SA2)
Both
-
ve
(¾)
95.6
92.6
92.8 93.8
New
infection
(¾)
2.5
3.5
1.3
2.4
Elimination
(¾)
0.6
1.9
3.6
2.4
Both
+ve
(%)
1.6
1.9
2.3
1.4
Two
month
infection
risk
(%)
2.5
3.6
1.4
2.5
Two
month
elimination
risk
(¾)
27.3
50.0
61.0
63.2
S4
=
1-2
weeks
postpartum
S5
=
2-3
months
postpartum
S6
=
4-5
months
postpartum
S7
=
6-7
months
postpartum
S8
=
8-9
months
postpartum
Britain,
quarters
infected
with
S.
epi-
dermidis
were
less
susceptible
to
infec-
tion
with
Escherichia
coli
or
S.
agalac-
tiae
(9,15).
The
authors
explained
this
as
being
due
to
an
increased
SCC.
In
contrast,
a
study
based
on
a
single
Ohio
dairy
herd
showed
that
the
rate
of
environmental
streptococcal
infec-
tions
was
greater
in
quarters
infected
with
CNS
(5).
Rates
of
coliform
infec-
tions
were
neither
enhanced
nor
reduced
in
quarters
infected
with
CNS.
Results
from
trials
in
France
and
Sweden
have
shown
a
protective
effect
of
CNS
against
experimental
infections
with
S.
aureus
(10,16-19).
The
mech-
anism
of
this
protection
in
one
study
*was
stated
as
being
due
mainly
to
an
increase
in
SCC
caused
by
CNS
(10).
A
study
at
the
Hill
Farm
Research
Sta-
tion
dairy
farm
in
Louisiana
suggested
that
CNS
afforded
little
resistance
to
S.
aureus
infections
(29).
Infection
with
CNSP
and
CNSA
had
a
significant
positive
influence
on
lnSCC.
Other
studies
concur
with
this
finding.
The
natural
log
of
the
SCC
was
approximately
two-times
higher
in
CNS
infected
quarters
than
in
uninfected
quarters
in
studies
done
in
Ohio,
Kentucky,
Sweden
and
Wisconsin
(5,7,10,20).
The
natural
log
of
the
SCC
279

was
58
x
103,
216
x
103
and
67
x
103
for
uninfected
quarters,
and
105
X
103,
500
x
103
and
122
x
103
for
infected
quarters
respectively
(7,10,20).
In
a
seventeen
cow
study
in
England,
S.
epidermidis
caused
lnSCC
to
increase
from
three
to
ten
times
that
of
uninfected
quarters
(15).
In
another
English
herd,
most
species
of
CNS
did
not
cause
an
increase
in
lnSCC
(4).
Infection
with
gram-negative
rods
had
a
negative
influence
on
lnSCC.
This
was
not
significant,
probably
due
to
the
small
number
of
gram-negative
rods
isolated.
Coagulase
negative
staphylococcus-
pure
had
a
greater
effect
on
lnSCC
than
CNSA.
Predicted
SCC
for
a
quarter
uninfected
with
CNS,
infected
with
CNSA
and
infected
with
CNSP
were
87
x
103,
102
x
103
and
115
x
103
respectively
in
a
midlactation
cow.
This
study
suggests
that
on
a
cow
basis,
CNS
infections
are
relatively
unimportant,
however
if
the
prevalence
of
infection
is
high
in
a
herd,
a
signif-
icant
increase
in
bulk
tank
SCC
could
result.
REFERENCES
1.
DEVRIESE
LA,
DEKEYSER
H.
Prevalence
of
different
species
of
coagulate-negative
staphylococci
on
teats
and
in
milk
samples
from
dairy
cows.
J
Dairy
Res
1980;
47:
155-158.
2.
SCHALM
OW,
LASMANIS
J.
Distribution
of
micrococci
and
other
bacteria
in
milk
samples
from
a
single
dairy
herd
after
twelve
years
of
mastitis
control.
Am
J
Vet
Res
1957;
18:
778.
3.
HOGAN
JS,
WHITE
DG,
PANKEY
JW.
Effects
of
teat
dipping
on
intramammary
infections
by
staphylococci
other
than
Staphylococcus
aureus.
J
Dairy
Sci
1987;
70:
873-879.
4.
EDWARDS
SJ,
JONES
GW.
The
distribu-
tion
and
characters
of
coagulase-negative
staphylococci
of
the
bovine
udder.
J
Dairy
Res
1966;
33:
261-270.
5.
HOGAN
JS,
SMITH
KL,
TODHUNTER
DA,
SCHOENBERGER
PS.
Rate
of
envi-
280
ronmental
mastitis
in
quarters
infected
with
Corynebacterium
bovis
and
staphylococcus
species.
J
Dairy
Sci
1988;
71:
2520-2525.
6.
TIMMS
LL,
SCHULTZ
LH.
Dynamics
and
significance
of
coagulase
negative
staphylo-
coccal
intramammary
infections.
J
Dairy
Sci
1987;
70:
2648-2657.
7.
HARMON
RJ,
LANGLOIS
BE.
Mastitis
due
to
coagulase-negative
staphylococcus
species.
Agri-Pract
1989;
10:
29-34.
8.
SMITH
RE,
HAGSTAD
HV.
Infection
of
the
bovine
udder
with
coagulase-negative
staphylococci.
In:
Progress
in
the
Control
of
Bovine
Mastitis.
F.R.
Germany:
Kieler
Milchwirtschaftliche,
Forschungsberichte
1985:
611-614.
9.
HOLMBERG
0.
Coagulase-negative
staphylococci
in
bovine
mastitis.
In:
Maidh
PA,
Schleifer
KH,
eds.
Coagulase
Negative
Staphylococci.
Stockholm:
Alinquist
and
Wiksell
Inlination,
1986:
203-211.
10.
LINDE
C.
The
effect
of
coagulase-negative
staphylococci
in
the
cow's
udder
on
experi-
mental
induction
of
mastitis
and
on
milk
production.
PhD
Thesis,
Swedish
University
of
Agricultural
Science,
Uppsala,
Sweden,
1982.
11.
LINDE
C.
HOLMBERG
0,
ASTROM
G.
The
interference
between
coagulase-negative
Staphylococci
and
Corynebacterium
bovis
and
the
common
udder
pathogens
in
the
lac-
tating
cow.
Nord
Vet
Med
1980;
32:
552-558.
12.
HOGAN
JS,
PANKEY
JW,
SMITH
KL.
Staphylococcus
species
other
than
Staphylo-
coccus
aureus.
Proceedings
of
the
National
Mastitis
Council,
26th
Annual
Meeting,
Orlando,
Florida,
1987:
21-32.
13.
RAINARD
P,
POUTREL
B.
Dynamics
of
nonclinical
bovine
intramammary
infections
with
major
and
minor
pathogens.
Am
J
Vet
Res
1982;
43:
2143-2146.
14.
BROWN
RW.
Intramammary
infections
produced
by
various
strains
of
Staphylococcus
epidermidis
and
micrococcus.
Cornell
Vet
1973;
63:
630-645.
15.
BRAMLEY
AJ.
The
effect
of
subclinical
Staphylococcus
epidermidis
infection
of
the
lactating
bovine
udder
on
its
susceptibility
to
infection
with
Streptococcus
agalactiae
or
Escherichia
coli.
Br
Vet
J
1978;
134:
146-151.
16.
POUTREL
B,
LERONDELLE
C.
Protective
effect in
the
lactating
bovine
mammary
gland
induced
by
coagulase-negative
staphylococci
against
experimental
Staphylococcus
aureus
infections.
Ann
Rech
Vet
1980;
11:
327-332.
17.
LINDE
C,
HOLMBERG
0,
ASTROM
G.
Interference
between
Staphylococcus
epider-
midis
(Se)
and
Staphylococcus
aureus
(Sa)
in
the
bovine
udder.
Acta
Vet
Scand
1975;
16:
146-148.
18.
LINDE
C,
HOLMBERG
0,
ASTROM
G.
Interference
between
Staphylococcus
epider-
midis
and
more
pathogenic
bacteria
in
the
bovine
udder.
Zentralbl
Bakteriol
1975;
(Suppl
5):
1035-1039.
19.
LINDE
C,
HOLMBERG
0,
ASTROM
G.
An
attempt
to
superimpose
Staphylococcus
aureus,
Streptococcus
agalactiae
and
Strep-
tococcus
dysgalactiae
upon
Staphylococcus
epidermidis
infections
in
the
cow's
udder.
Proceedings
of
Seminar
on
Mastitis
Control,
Doc
85,
International
Dairy
Federation,
Brussels,
Belgium,
1975:
391-394.
20.
TIMMS
LL,
SOMMER
DA,
SCHULTZ
LH.
Role of
minor
pathogens
in
mastitis.
J
Dairy
Sci
1985;
68
(Suppl
1):
202.
21.
GODKIN
A.
The
relationship
between
bulk
tank
culture
results,
management
of
factors
used
in
mastitis
control
and
the
herd
pre-
valence
of
mastitis.
DVSc
Thesis,
University
of
Guelph,
Canada,
1989.
22.
NATIONAL
MASTITIS
COUNCIL.
Lab-
oratory
and
Field
Handbook
on
Bovine
Mastitis.
Arlington,
Virginia:
The
National
Mastitis
Council
Inc.,
1987.
23.
ASHTON-TATE.
Learning
and
Using
dBASE
III
PLUS.
Torrance,
California:
Ashton
Tate,
1985,
1986.
24.
RYAN
BF,
JOINER
BL,
RYAN
AR.
Mini-
tab
Handbook
Second
Edition.
Boston,
Massachusetts:
Duxbury
Press,
1985.
25.
NATIONAL
MASITIS
COUNCIL.
Micro-
biological
Procedures
for
the
Diagnosis
of
Bovine
Udder
Infection.
The
National
Mastitis
Council
Inc.,
1990.
26.
BRAMLEY
AJ.
Infection
of
the
udder
with
coagulase-negative
micrococci
and
Coryne-
bacterium
bovis.
Proceedings
Seminar
on
Mastitis
Control,
Doc
85,
International
Dairy
Federation,
Brussels,
Belgium,
1975.
377-381.
27.
HOLMBERG
0.
Staphylococcus
epider-
midis
isolated
from
bovine
milk.
Acta
Vet
Scand
1973;
45(Suppl);
203-211.
28.
MURPHY
JM.
The
use
of
strict
foremilk
in
the
study
and
diagnosis
of
chronic
bovine
mastitis.
Cornell
Vet
1943;
33:
45-51.
29.
PANKEY
JW,
NICKERSON
SC,
BODDIE
RL,
HOGAN
JS.
Effect
of
Corynebacte-
rium
bovis
infection
on
susceptibility
to
major
mastitis
pathogens.
J
Dairy
Sci
1985;
68:
2684-2693.