Detection of Bovine Mastitis by Bromothymol Blue pH Indicator Test
R. J. M A R S C H K E and B. J. K I T C H E N l
ABSTRACT
A simple b r o m o t h y m o l blue indicator
test was evaluated for farm diagnosis of
mastitis. The test required highly ab-
sorbent blotting paper impregnated with
four spots of b r o m o t h y m o l blue. Indicator
color scores (1 to 4) for quarter foremilks
increased with somatic cell count and pH,
although variability within each color
score was large. Sensitivity of the bromo-
thymol blue test ranged from 51 to 56%
and specificity from 89 to 90% for most
reference criteria used to classify normal
and abnormal milk. Predictability of a
positive test ranged from 49 to 52% (false
positives 51 to 48%) and predictability of
a negative test from 90 to 97% (false
negatives 10 to 3%) for the same criteria.
Overall the b r o m o t h y m o l blue test
incorrectly diagnosed 11 to 20% of
3772 quarters. By classifying color score
2 as negative, predictability of a positive
result was 70 to 75% and sensitivity was
26 to 30%. The test can be used by dairy
producers to screen herds with a relatively
high incidence of mastitis or used in
combination with cow cell counts to
locate abnormal quarters. The bromo-
t h y m o l blue test was less sensitive than
the California Mastitis Test but offered
several practical advantages for use on
farm.
INTRODUCTION
Mastitis increases the alkalinity of milk (3,
16). The pH of normal cow's milk lies in the
range 6.5 to 6.8 (18) but may exceed 7.0 in
milks with high cell counts (7, 14). During
Received October 11, 1983.
~Gilbert Chandler Institute of Dairy Technology,
Department of Agriculture, Werribee, Victoria 3030,
Australia.
1985 J Dairy Sci 68:1263-1269
Otto Madsen Dairy Research Laboratory
Department of Primary Industries
Hamilton, Queensland 4007, Australia
inflammation of the mammary gland, per-
meability of the blood capillaries increases,
which allows alkaline blood constituents
(sodium and bicarbonate ions) to enter the milk
and consequently to increase milk pH (10, 18).
Indicator dyes such as bromocresol purple
(BCP) and b r o m o t h y m o l blue (BTB) have been
used for many years in mastitis diagnosis to
detect pH changes in milk (2, 4). Because of the
influence of fat content on the color produced
by BCP this dye largely has been replaced by
BTB. Diagnostic methods using BTB have
involved addition of the indicator solution to
milk (6), use of commercial indicator paper (7),
and absorbent paper impregnated with indicator
spots (20). However, it generally is accepted
that the BTB indicator test is not particularly
sensitive for detection of subclinical mastitis (4,
18), and with the introduction of rapid cowside
tests for cell count e.g., California Mastitis Test
(CMT) (19), the pH approach has fallen into
disfavor. The main disadvantage with indirect
tests such as CMT is that assessment of results is
based on the degree of gel formation in the
milk, and, consequently, untrained personnel
often have problems in distinguishing between
the various scoring categories. The BTB test,
however, is assessed on the basis of color
changes that are observed more readily.
The purpose of this study was to examine
the BTB test using laboratory prepared paper
strips impregnated with BTB indicator. In
addition, the use of these strips on commercial
dairy farms with varying mastitis problems was
investigated to determine the herd health
situations in which this rapid diagnostic test
would be most effective.
MATERIALS AND METHODS
Indicator Papers for pH
The pH test papers were prepared in the
laboratory from blotting paper (thickness 1.5
mm, weight .7 kg/m 2) (Rising Paper Co.,
MA) cut into strips 10 × 12 cm and pointed at
1263
1264 MARSCHKE AND KITCHEN

one end to aid quarter identification. Four
orange spots (diameter 3 cm) of BTB indicator
(.4 ml of .2% wt/vol in ethanol) were im-
pregnated in square formation on each test
paper (Figure 1).
Farms
Four commercial farms were sampled at
intervals of 1 to 2 mo. Three herds of ap-
proximately 50 cows were sampled on five oc-
casions, and a larger herd of approximately 100
cows was sampled twice. Animals were pre-
dominantly Friesians. Mastiffs incidence varied
on the four farms, as indicated by annual bulk
milk cell counts (annual means of monthly bulk
milk counts) ranging from 118 × 103 to 325 ×
103/ml. Prevalence of mastiffs in a particular
herd was gauged by the proportion of quarters
from that herd with somatic cell count (SCC)
greater than 500 × 103/ml.
Sampling
Sampling of a11 herds was at evening milking.
The BTB test was on quarter foremilks after
teats were washed thoroughly and dried with
paper towels. The pointed end of the test strip
was held toward the cow's head, and after the
first stream of foremilk was discarded, a small
amount of milk was squirted onto the cor-
responding indicator spots. Cross flow problems
because of excess milk were reduced by our
bending the test strip upward along the longi-
tudinal axis before sampling. Quarter samples
(25 m!) then were collected aseptically by the
scores 2, 3, and 4 were prepared by equal
quantities of the same bulk milk mixed with
1.0 M Tris-HC1 buffer solutions having pH 6.8,
7.1, and 7.4, respectively. Test colors were
scored by comparison with wet color standards.
Dried test colors again were scored against dried
reference colors the following day.
Determination of pH
Milk pH was determined on a TPS Digital pH
meter (Model 1852, "I'.P.S. Pty. Ltd., Old.). The
meter was calibrated with pH 7.0 and pH 4.0
standard buffers, and all measurements were at
20 + I°C.
Reference Tests
Mastiffs pathogens were identified by
procedures outlined by Griffin et al. (5) and
classified into major and minor groups (1).
Somatic cell counts were on a Fossomatic (Foss
Electric, Hillerod, Denmark). Milk conductivity
was determined at 25°C on a direct reading
conductivity meter fitted with a pipette type
conductivity cell (8). The CMT with Rapid
Mastitis Test reagent (P.D.S. Rural Products
j©
Cow No: Date:

O
procedure of Andrews et al. (1). The samples
were packed in ice, stored overnight at 1 to
2°C, and laboratory tested within 24 h.

Assessment of Bromothymol Blue Scores

Left Front Right Front
Within 1 to 2 min of sampling for the BTB
test, the color of each indicator spot was scored .~'.;~'..~.
on a scale 1 to 4 according to color standards.
Score 1 (pale green) was assessed as a normal
quarter, and scores 2, 3, and 4 (increasing from
moderate green to dark blue-green) were
assessed as abnormal quarters. Four color
standards were prepared by several drops of pH
adjusted milks placed onto different indicator
spots. The reference milk for score 1 was
Left Rear
ORight Rear

obtained by raw bulk milk (pH 6.6 to 6.7, fat Figure 1. Test strip impregnated with four spots
content 3.5 to 4.0%) diluted with an equal of bromothymol blue indicator. Longitudinal axis
volume of distilled water, whereas milks for is indicated by dotted line.

Journal of Dairy Science Vol. 68, No. 5, 1985

 BROMOTHYMOL BLUE TEST FOR MASTITIS
Ltd., Sydney, Aust.) was done in the laboratory
on 2 ml of the test milk mixed with 2 ml of rea-
gent. Results of the CMT test were scored on a
scale 1 to 4 corresponding to increasing viscosity
of the milk/reagent mixture. Score 1 was assessed
as a normal quarter and scores 2, 3, and 4 as
abnormal quarters.
Evaluation of the Bromothymol Blue Test
Quarter milks were classified as normal or
abnormal on four occasions according to
different reference criteria. For abnormal
milks the criteria were: a) a major pathogen
present, b) SCC /> 500 × 103/mi c) SCC/> 500
x 103/ml and a major pathogen present,
and d) conductivity /> 60 mM NaC1 or an
elevation of /> 5 mM NaC1 above that of the
lowest quarter for that cow. Quarter milks
containing minor pathogens were classified
normal. The BTB test was evaluated by its
sensitivity, specificity, and predictability posi-
tive and negative for each of the reference
criteria (11). Sensitivity was the proportion
of abnormal milks correctly identified and
specificity the proportion normal milks
correctly identified. Predictability positive
was the proportion of positive BTB results that
were true positives, and predictability negative
was the proportion of negative BTB results that
were true negatives.
RESULTS A N D DISCUSSION
The mean SCC and mean pH for each BTB
score are in Table 1 and represent data obtained
from all farms. An increase of BTB score was ac-
companied by increasing severity of mastitis as
indicated by both diagnostic measures. However,
the large variability about the means, particularly
for SCC, showed there was considerable over-
lapping of results and highlighted the possible
error in interpretation of indicator scores. The
wide variation of milk pH within each BTB
score showed that indicator results were not
closely related to pH. This may be explained by
variable changes of milk carbon dioxide content
that could affect pH between time of sampling
and laboratory pH test the following day.
McDowall (13) found that the pH of quarter
milks increased during standing for 2 to 5
h and showed that this was partly due to loss of
carbon dioxide. Table 1 also indicates that the
BTB test would not detect reliably milks with a
1265
SCC below about 1 x 106/ml. This was con-
firmed by only 39% of the 281 quarters in the
SCC range 500 × 103 to 1000 × 103/ml being
positive by the BTB test, this percentage
being reduced to 14% if score 2 results were
excluded.
The sensitivity and predictability positive of
the BTB test on all quarter milks are in Table 2.
For the classification criteria b through d,
slightly more than half of the abnormal quarters
were identified correctly. For milks containing
a major pathogen, sensitivity was considerably
less (39%). Poor relationships between presence
of pathogens and other mastitis measures such
as cell count have been reported (17). Pre-
dictability positive results for criteria b through
d showed that a positive BTB test had an
approximately equal chance of indicating either
normal or abnormal milk. With the exception
of results based on the presence and absence of
pathogens, sensitivity and predictability did not
vary greatly between the different criteria even
though these were measuring different responses
to mastitis. In Table 3 specificity and pre-
dictability negative are given for each classifi-
cation criteria. About 90% of normal quarters
were identified correctly by the test, and less
than 10% of BTB tests were false negatives.
Negative test results were, therefore, more
reliable than positive results, which produced
about 50% false readings. This is in direct
contrast to results of Fay et al. (4), possibly
because of the higher BTB color threshold and
different test procedures used by these workers.
Overall, the BTB test misclassified between 11
and 20% of all samples tested. A study by
Mijnen et al. (15) showed that pH test was of
little diagnostic value and correlated little with
TABLE 1. Relationship between milk composition
and bromothymol blue (BTB) scores.
BTB
score Quarters SCC1 X 10a/ml pH
(no.) X SE X SE
1 2128 252 13 6,63 .003
2 343 791 67 6.74 .006
3 91 1893 267 6.83 .018
4 63 3845 600 7.02 .035
aSomatic cell count.
Journal of Dairy Science Vol. 68, No. 5, 1985
1266 MARSCHKE AND KITCHEN

TABLE 2. Sensitivity and predictability of a positive bromothymol blue (BTB) test for various criteria of milk
abnorma/ity (4 herds).

Total
Classification Abnormal positive Predictability
criteria quarters by BTB test Sensitivity1 positive ~

(no.) (%)
Major pathogen present 422 649 39 25
SCC 3 /> 500 X lOa/ml 618 649 51 49
SCC/> 500 × lOS/ml and
major pathogen present 248 649 56 52
Conductivity 7> 60 mM
(or quarter difference i> 5 mM) 583 580 55 52

Proportion of abnormal quarters correctly identified.

2Proportion of positive BTB results that are true positives.
a Somatic cell count.

o t h e r variables such as SCC, c o n d u c t i v i t y , a n d r e d u c e s t h e n u m b e r o f false positive results,

lactose. In t h e i r i n v e s t i g a t i o n , actual pH a n d
n o t i n d i c a t o r scores were m e a s u r e d , a n d t h e pH
t e s t was assessed b y a q u a r t e r d i f f e r e n c e m e t h o d .
A n e x a m i n a t i o n o f t h e false positive BTB
results s h o w e d t h a t t h e s e were c o m p r i s e d
m a i n l y o f score 2 (score 2, 80%; score 3, 16%;
score 4, 4%). Increasing t h e BTB test t h r e s h o l d
b y classifying score 2 as n e g a t i v e r e d u c e d t h e
test sensitivity to 26 to 30% a n d increased
specificity t o 97 to 99% for criteria b t h r o u g h
d. P r e d i c t a b i l i t y o f a positive result increased
to 70 t o 75%, a n d p r e d i c t a b i l i t y negative
d e c r e a s e d to 87 to 95%. A l t h o u g h this s c h e m e
sensitivity o f t h e t e s t s i m u l t a n e o u s l y is r e d u c e d .
R e p o r t s b y o t h e r w o r k e r s w h o have used t h e
same scoring s y s t e m t o classify q u a r t e r s b u t
w i t h i n d i c a t o r s o l u t i o n m e t h o d s give variable
results o n test p e r f o r m a n c e . F o r a SCC t h r e s h o l d
of 5 0 0 × 1 0 3 / m l , F a y et al. (4) o b t a i n e d a
sensitivity of 21%, w h e r e a s P e n d s e a n d Nair
(16) r e p o r t e d 68% f o r t h e s a m e m e a s u r e .
Table 4 shows the effect of prevalence of
m a s t i t i s o n p r e d i c t a b i l i t i e s w h e n t h e BTB test
was used o n f o u r h e r d s . P r e d i c t a b i l i t y o f
a positive result increased m a r k e d l y w i t h
p r e v a l e n c e o f mastitis, b e i n g m o r e t h a n d o u b l e d

TABLE 3. Specificity and predictability of a negative bromothymol blue (BTB) test for various criteria of milk
normality (4 herds).

Total
Classification Normal negative Predictability
criteria quarters by BTB test Specificity 1 negative 2

- - ( n o . ) - (%)
Major pathogen absent 3347 3123 85 92
SCC 3 < 500 × 103/ml 3151 3123 89 90
SCC < 500 X 103/ml and major
pathogen absent 2977 3123 90 97
Conductivity < 60 mM
(or quarter difference < 5 raM) 3041 2824 90 92

1Proportion of normal quarters correctly identified.

2Proportion of negative BTB results that are true negatives.
s Somatic cell count.

Journal of Dairy Science Vol. 68, No. 5, 1985

 BROMOTH¥'MOL BLUE TEST FOR MASTITIS 1267

TABLE 4. Effect of prevalence of mastitis in different herds on predictabilities of bromothymol blue (BTB)
test.

BTB Predictability
Bulk No. of quarters
Herd milk SCC1 abnormal 2,3 Positive Negative Positive4 Negative
5

- - (no.) - - (%)
1 118 59 (5) 130 962 25 97
2 284 199(16) 201 1071 51 91
3 299 194(25) 178 612 57 85
4 325 166(27) 140 478 57 82

1Annual mean of monthly bulk milk SCC taken over the sampling period.
2Abnormal quarters are defined by SCC 1> 500 × 103/ml.
Numbers in parentheses represent the percentage of abnormal quarters.
4Proportion of positive BTB results that are true positives.
s Proportion of negative BTB results that are true negatives.

by a 2 to 3 times increase of bulk milk SCC or a
corresponding 3 to 5 times increase of per-
centage of abnormal quarters. In conjunction,
predictability negative decreased but less, and
changes of sensitivity and specificity were only
slight. Similar results to these were obtained by
McDermott et al. (12) in examining the re-
lationship between SCC and prevalence of
infection in different herds. These workers
concluded that application of SCC to different
herd situations may require appropriate altera-
tions to the test threshold. Our results showed
that increasing the threshold of the BTB test
(classifying score 2 as negative) markedly
reduced test sensitivity. Therefore, to optimize
all elements of the test (i.e., sensitivity, specifi-
city, and predictability), score 2 should be
classified as a positive result, and the test
should be used on herds with a relatively high
prevalence of mastiffs (bulk milk SCC greater
than 300 × 103/ml). The most effective ap-
plication of the test at this threshold would be
for initial screening purposes to obtain a
maximum percentage of infected animals.
Decisions to cull cows or to use antibiotic
therapy on cows showing a positive BTB result
should be based on other confirmatory tests. In
herds where cow cell counting schemes are in
operation and m o n t h l y results are available to
the dairy producer, initial screening is not
warranted, and the BTB test can be used to
locate abnormal quarters in cows with elevated
cell counts. Such quarters then can be monitored
regularly for recovery. If the dairy producer
wishes to treat subclinical infections, a higher
BTB test threshold should be used (score 2
negative) "and positive scores confirmed at two
consecutive milkings.
A comparative assessment of the rapid BTB
and CMT tests showed that the CMT was more
efficient than BTB in terms of sensitivity and
predictability positive and produced about half
the number of misclassifications (Table 5).
Specificity and predictability negative (not
shown in Table 5) were both 92% for CMT and
both 86% for BTB. However, a comparison of
the two tests on the basis of SCC may be of
doubtful significance, because CMT is an
indirect measure of SCC and could be expected
to give a more favorable result (9, 11).
In spite of the apparently better performance
of CMT, the BTB strip test has a number of
practical advantages for cowside use. It is
simpler to perform (no mixing of reagents) and
is more objective, being based on color reactions
and not viscosity changes. The paper strip
provides a permanent record of the test, and
information such as cow identification and date
of testing may be recorded on each strip.
Because the test colors and standards remain
unchanged for at least 2 wk if kept from direct
sunlight, the strips may be kept in the dairy for
reference in daily monitoring of problem cows.
The strips do not need to be read immediately
after the test but can be assessed later to save
milking time. In our trial, differences in reading

Journal of Dairy S c i e n c e VoL 68, No. 5, 1985

1268 MARSCHKE A N D KITCHEN
T A B L E 5. Comparison of efficiency of California Mastitis Test (CMT) and b r o m o t h y m o l blue (BTB) test for
a somatic cell c o u n t threshold of 500 X 103/ml (2 herds), t'2
Test Positive quarters Sensitivity 3
(no.)
CMT 234 76
BTB 228 55
1Total n u m b e r of quarters = 973.
2Number of abnormal quarters (SCC > 500 × 103/ml) = 246.
a Proportion of abnormal quarters correctly identified.
4 Proportion of positive results that are true positives.
the test papers wet (within 1 to 2 min of
sampling) or dry (the following day) were
infrequent, provided the test spots were com-
pared to correspondingly wet or dry standards.
F o r f i e l d a p p l i c a t i o n s , it w o u l d b e m o r e p r a c -
tical to use a set of dry color standards and
make comparisons accordingly. Unused test
strips could be stored indefinitely, being un-
affected by exposure to light.
ACKNOWLEDGMENTS
The authors are grateful for the assistance of
C. A n n a n d , B. H o l d i n g , N. H o l z a p f e l , R .
R o b e r t s , a n d W. W a r d in t h e c o l l e c t i o n a n d
testing of milk samples. This work was supported
in p a r t b y a g r a n t f r o m t h e A u s t r a l i a n D a i r y
Research Committee.
REFERENCES
1 Andrews, R. J., B. J. Kitchen, W. S. Kwee, and F.
Duncalfe. 1983. Relationship between individual
cow somatic cell counts and the mastitis infection
status of the udder. Aust. J. Dairy Technol. 38:71.
2 Baker, J. C., and L. L. Van Slyke. 1919. A m e t h o d
for the preliminary detection of abnormal milk
based on the hydrogen ion concentration. J. Biol.
Chem. 40:357.
3 Blood, D. C., and J. A. Henderson. 1968. Veterinary
medicine. 3rd ed. Bailliere Tindall and Cassell,
London.
4 Fay, A. C., H. W. Cave, and F. W. Atkeson. 1938.
Detection of mastitis by the b r o m t h y r n o l blue test,
leucocyte count, and the microscopic examination
of incubated samples of milk. Cornell Vet. 28:40.
5 Griffin, T. K., F. H. Dodd, F. K. Neave, D. R.
Westgarth, R. G. Kingwill, and C. D. Wilson. 1977.
A m e t h o d for diagnosing i n t r a m a m m a r y infection
in dairy cows for large experiments. J. Dairy Res.
44:25.
Journal o f Dairy Science Vol. 68, No. 5, 1985
Predictability
positive 4 Misclassifications
(%)
79 11
59 21
6 Hume, C. M. 1941. Studies on the detection of
mastitis in New Zealand dairy herds. I. A field
outfit for the h r o m t h y m o l blue test for mastitis.
N . Z . J . Sci. Technol. 22:322A.
7 Joshi, S. V., J. Prasad, and A. Rekib. 1976. Studies
on the field diagnosis of subclinical mastitis. Ind.
Vet. J. 53:752.
8 Kitchen, B. J., G. Middleton, I. G. Durward, R. J.
Andrews, and M. C. Salmon. 1980. Mastitis diag-
nostic tests to estimate m a m m a r y gland epithelial
cell damage. J. Dairy Sci. 63:978.
9 Linzell, J. L., and M. Peaker. 1975. Efficacy of the
m e a s u r e m e n t of the electrical conductivity of milk
for the detection of subclinical mastitis in cows:
Detection of infected cows at a single visit. Br. Vet.
J. 131:447.
10 Liick, H., and A. Smith. 1975. Relationship
between constituent concentrations and the pH
value of m a m m a r y gland secretions. S. Aft. J. Dairy
Technol. 7:27.
11 Martin, S. W. 1977. The evaluation of tests. Can. J.
Comp. Med. 41:19.
12 McDermott, M. P., H. N. Erb, and R. P. Natzke.
1982. Predictability by somatic cell counts related
to prevalence of i n t r a m a m m a r y infection within
herds. J. Dairy Sci. 65:1535.
13 McDowall, F. H. 1941. Studies on the detection of
mastitis in New Zealand dairy herds. II. Factors
influencing the b r o m t h y m o l blue test for mastitis.
N . Z . J . Sci. Technol. 22:328A.
14 McDowall, F. H. 1945. Studies on the detection of
mastitis in New Zealand dairy herds. V. Com-
position of milk from quarters reacting to the
b r o m t h y m o l blue test for mastitis. N . Z . J . Sci.
Technol. 27:258.
15 Mijnen, E., F.H.J. Jaartsveld, G.A.A. Albers,
M.W.A. Verstegen, and M.J.M. Tielen. 1982. The
value of cell count, lactose content, pH and con-
ductivity of milk for mastitis detection in individual
cows. Neth. Milk Dairy J. 36:65.
16 Pendse, M. D., and P. G. Nair. 1970. The develop-
m e n t and evaluation of a test for sub-clinical
mastitis. Ind. J. Dairy Sci. 23:7.
 B R O M O T H Y M O L BLUE TEST FOR MASTITIS 1269

17 Postle, D. S. 1967. Comparisons of mastitis screen-
ing test results from quarter, bucket and bulk milk
samples. J. Milk Food Technol. 30:7.
18 Schalm, O. W., E. J. Carroll, and N. C. Jain. 1971.
Bovine mastitis. Lea and Febiger, Philadelphia, PA.
19 Schalm, O. W., and D. O. Noorlander. 1957.
Experiments and observations leading to the
development o f the California Mastitis Test. J. A m .
Vet. Med. Assoc. 130:199.
20 SchiSnherr, W. 1958. Examination of the udder in
cows with the aid of an indicator paper. Monatsh.
Veterinaermed. 13 : 304.

Journal o f Dairy Science Vol. 68, No. 5, 1985