International Journal of Food Microbiology 64 (2001) 387393

www.elsevier.nl / locate / ijfoodmicro

Short communication

Evaluation of motility enrichment on modified semi-solid

RappaportVassiladis medium (MSRV) for the detection of
Salmonella in foods
a
a
b
a,
D. Worcman-Barninka , M.T. Destro , S.A. Fernandes , M. Landgraf *
a

Experimental, Faculdade de Ciencias

Paulo,

Departamento de Alimentos e Nutricao

Farmaceuticas
, Universidade de Sao
Paulo, Brazil
Av. Prof. Lineu Prestes 580, Bloco 14, 05508 -900 Sao
b
Paulo, Brazil
Instituto Adolfo Lutz, Sao
Received 16 November 1999; received in revised form 1 September 2000; accepted 4 November 2000

Abstract
The detection and identification of Salmonella spp. is still troublesome and time consuming to the food industry.
Employing the modified semi-solid RappaportVassiliadis medium (MSRV), presumptive results for Salmonella can be
obtained in 48 h, representing an interesting alternative to the standard methods. The specificity and sensitivity of the MSRV
method were evaluated in this research. The efficiency of this method was also compared with the methodology
recommended by the US Food and Drug Administration (FDA) using bismuth sulfite agar, XLT4 agar and Rambach agar. A
total of 146 food samples comprised of 41 chicken thighs, 35 Brazilian fresh pork sausages, 35 samples of cocoa powder
and / or granulated cocoa and 35 samples of grated fresh coconut, were examined. Overall, the rapid method (direct 1
indirect) and the standard culture detected 96.1% and 84.6% of the positive samples, respectively. No Salmonella was
detected in the coconut or cocoa samples by any of the methods. Eighteen (43.9%) chicken thigh samples were contaminated
with the microorganism. The rapid method (direct 1 indirect) and the standard culture detected 94.4% and 88.9% of these,
respectively. Salmonella was detected in eight (22.8%) fresh pork sausage samples. The MSRV method detected Salmonella
in all eight samples, while the standard gave positive results in six (75%). When compared with the standard method, the
indirect method showed 86.4% sensitivity and 96.8% specificity, while the direct MSRV showed a sensitivity of 71.4% and
specificity of 99.2%. Combined, both MSRV methods showed 95.5% sensitivity and 96.8% specificity. The MSRV medium
also reduces the time necessary for the isolation of Salmonella from foods. 2001 Elsevier Science B.V. All rights
reserved.
Keywords: Salmonella; MSRV medium; Chicken; Fresh sausage; Cocoa

1. Introduction
*Corresponding author. Tel.: 155-11-3818-7991; fax: 155-113815-4410.
E-mail address: landgraf@usp.br (M. Landgraf).

Despite the numerous enrichment and isolation

media for the detection of Salmonella, none of them

0168-1605 / 01 / $ see front matter 2001 Elsevier Science B.V. All rights reserved.
PII: S0168-1605( 00 )00484-0

388

D. Worcman-Barninka et al. / International Journal of Food Microbiology 64 (2001) 387 393

is considered absolute and there are several factors

involved that must be considered. According to the
standard methodology specified in the Bacteriological Analytical Manual (Andrews et al., 1995), isolation of Salmonella requires a minimum of 4 days and
several additional ones for the confirmation of
presumptive isolates. Since 90% of the food samples
analysed are negative for Salmonella, time and cost
effective measures should be considered for confirmation of potentially positive results (Oggel et al.,
1995).
The detection of Salmonella is still very dependent
on the use of appropriate culture media. Isolation is a
complex process involving several factors and, the
quality of the culture media is only one of these
factors (DAoust et al., 1992). According to the
standard method (Andrews et al., 1995), the recovery
of Salmonella requires three steps: pre-enrichment,
selective enrichment and plating, that takes a minimum of 4 days and several additional days for the
confirmation of presumptive results.
The plating media for isolation of Salmonella
employ a variety of selective and differential agents,
usually based on the inability of the majority of the
Salmonella strains to ferment lactose and, sometimes, other carbohydrates such as sucrose and
salicin. Although most of the isolates of Salmonella
do not ferment lactose, about 1% of S. enteritidis and
1585% of S. arizonae isolates do ferment this sugar
(Farmer et al., 1984). Therefore, the occurrence of
lactose-positive strains makes it difficult to distinguish Salmonella from non-Salmonella strains based
on lactose-dependent media (DAoust et al., 1992).
Associated with lactose fermentation, the ability of
Salmonella strains to produce H 2 S has also been
explored. However, the possibility of H 2 S-negative
strains must also be considered.
Due to the low incidence of non-motile Salmonella strains (Holbrook et al., 1989), the detection
methods based on the motility of this microorganism
are being used more often in routine laboratories.
This results in decreased detection times as well as
decreased storage and warehouse costs for the food
and feed industry (Oggel et al., 1995).
Goossens et al. (1984) introduced a procedure
involving motility enrichment on a semi-solid
medium in Petri dishes by adding 0.8% agar to the
RappaportVassiliadis (RV) formulation. When this
semi-solid medium was incorporated in the routine

analysis of faecal specimens, the isolation rate of

Salmonella rose by 22.5%. After reducing the agar
concentration to 0.27%, De Smedt et al. (1986)
introduced the modified semi-solid RV (MSRV) for
the isolation of Salmonella from foods. The selectivity of the medium is based on the ability of Salmonella to migrate through the highly selective semi-solid
gel. The relative low pH minimizes the migration of
the majority of the motile Enterobacteriaceae, except
Salmonella, Enterobacter cloacae and Citrobacter
freundii. The growth of the latter two genera is
inhibited by the addition of novobiocin and incubation at 428C. Other selective agents in MSRV are the
malachite green oxalate and high MgCl 2 concentration (De Smedt et al., 1994).
In the last few years, several studies have been
done to validate the MSRV methodology and to
compare its efficiency with that of standard and other
rapid methods (Dusch and Altwegg, 1995;
ODonoghue et al., 1992; Poppe and Duncan, 1996).
In most studies, the efficiency of the MSRV method
was found to be equivalent or superior to the
standard methods (De Smedt and Bolderdijk, 1990;
De Zutter et al., 1991; De Smedt et al., 1991). The
aim of the present study was to investigate the
efficiency of the MSRV motility enrichment method
with Salmonella spp. and other Enterobacteriaceae,
and to evaluate its specificity and sensitivity in the
detection of Salmonella from foods, as compared
with the US Food and Drug Administration (FDA)
methodology (Andrews et al., 1995).

2. Materials and methods

2.1. Cultures
Pure cultures of 24 different Salmonella serotypes
(S. enteritidis, S. brandenburg, S. dublin, S. give, S.
montevideo, S. inganda, S. panama, S. london, S.
ohio, S. cerro, S. muenchen, S. newport, S. bredeney,
S. belem, S. infantis, S. stanleyville, S. saintpaul, S.
madelia, S. derby, S. arizonae, S. anatum, S.
typhimurium, S. agona and S. oranienburg) and four
other microorganisms belonging to the family Enterobacteriaceae (Escherichia coli, Proteus mirabilis,
Enterobacter aerogenes and Citrobacter freundii)
were used. All the strains used belong to the Food

D. Worcman-Barninka et al. / International Journal of Food Microbiology 64 (2001) 387 393

Microbiology Laboratory, Faculty of Pharmaceutical

Sciences, University of Sao Paulo, Sao Paulo, Brazil.

2.2. Evaluation of the MSRV method with pure

cultures
Each Salmonella and non-salmonellae strain on
trypticase soya agar slants (TSA; Difco, Detroit, MI,
USA) were inoculated into 10 ml of buffered peptone water (BPW), followed by incubation at 358C
for 20 h. From each of these cultures, serial dilutions
in 0.1% peptone (Oxoid, Basingstoke, UK) and
0.85% (w / v) NaCl (Synth, Diadema, Brazil) were
prepared. Three equidistant drops, of 30 ml each,
from each dilution (10 0 , 10 22 , 10 24 and 10 26 ), were
inoculated onto the surface of the different plates of
the MSRV (Oxoid). The plates were incubated at
428C for 24 h and examined for growth and the
development of a migration zone. When no growth
or migration were observed, the MSRV plates were
reincubated and checked after an additional 24 h.
The cell concentrations of these cultures were
determined by plating 0.1 ml of the 10 24 and 10 25
dilutions onto TSA plates and incubating them at
358C for 24 h followed by enumeration.

2.3. Evaluation of the MSRV method with bacterial

mixtures
A 1 ml volume of each concentration (10 1 , 10 3 ,
10 5 and 10 7 CFU / ml) of S. enteritidis culture was
mixed with 1 ml of each concentration (10 2 , 10 4 , 10 6
and 10 8 CFU / ml) of each culture of E. coli, P.
mirabilis, Enterobacter aerogenes and C. freundii,
respectively. The inoculation procedure was the
same as described above for evaluation of the
method with pure cultures.

2.4. Evaluation of the MSRV and standard methods

with food samples
Chicken thighs (41), Brazilian fresh pork sausages
(35), samples of cocoa powder and / or granulated
cocoa (35) and samples of grated fresh coconut (35)
were analysed. All the samples were purchased on
different days, from different establishments, brands
and lots. Chicken thighs and Brazilian fresh pork

389

sausages were kept under refrigeration until analysed.

Chicken thighs were rinsed in buffered peptone
water (BPW) in a ratio of 1:1; and the other samples
were pre-enriched in BPW according to US FDA
methods (Andrews et al., 1995). The standard and
MSRV methods were based on a common pre-enrichment broth and isolation and identification steps were
carried out according to the following schemes:
(I) Standard method: The inoculated pre-enrichment broth was incubated at 358C for 24 h, and then
1 ml and 0.1 ml of broth were then transferred to 10
ml of MuellerKauffmann tetrathionate broth (TT;
Difco) and to 10 ml of RappaportVassiliadis broth
(RV; Difco), respectively, and incubated at 428C.
After 24 h of incubation, one loopful from each
broth was streaked onto xyloselysinetergitol 4
agar (XLT4; Difco), bismuth sulfite agar (BSA;
Oxoid) and Rambach agar (RAM; Merck, Darmstadt, Germany). The plates were incubated at 358C
for 24 h and examined for the presence of typical
Salmonella colonies. Three suspect colonies from
each plate were inoculated onto triple sugar iron
(TSI; Oxoid) and lysine iron agar (LIA; Difco).
After 24 h at 358C, isolates were inoculated onto
EPM ( L-tryptophan desaminase, H 2 S production,
urease production and glucose fermentation), MILi
(motility, indol and lysine) and Simmons citrate agar
Paulo, Brazil). After 24 h of
(Probac do Brasil, Sao
incubation at 378C, the isolates with typical Salmonella profile were confirmed using somatic O and
flagellar H antisera (Probac do Brasil). All the strains
were serotyped according to Popoff and LeMinor
(1992) at the Laboratory of Enteric Pathogens,
Paulo, Brazil.
Instituto Adolfo Lutz, Sao
(II) MSRV method: The pre-enrichment culture
was assayed by the direct and indirect MSRV
procedures. In direct MSRV methodology, three
drops of 30 ml each of the pre-enrichment culture
were transferred, in separate spots, onto an MSRV
plate and incubated for 24 h at 428C. In indirect
MSRV methodology, after incubation of TT and RV
broths at 37 and 428C, respectively, for 8 h, three
drops (30 ml each) were inoculated on separate
MSRV plates and incubated for 16 h at 428C. (Total
time 24 h; 8 h enrichment plus 16 h MSRV method).
In both cases, the plates were checked for migration
zones with a radius larger than 10 mm (De Smedt
and Bolderdijk, 1987). For migrated cultures, bac-

390

D. Worcman-Barninka et al. / International Journal of Food Microbiology 64 (2001) 387 393

terial suspensions from the edge of the zone of

migration were subjected to the same biochemical
tests as described for the standard method.

2.5. Statistical analysis

The presumptive and confirmed results obtained
with the standard and rapid methods were statistically analysed by using the McNemar test (Siegel,
1956). A x 2 value . 3.84 indicates a significant
difference at the 0.05 level.
The sensitivity, specificity and efficiency rates
between both methodologies were obtained according to Roitman and Boice (1982).

3.2. Evaluation of the MSRV method with bacterial

mixtures
Of the bacterial mixtures seeded onto the MSRV
surface, five combinations S. enteritidis (10 7 CFU /
ml) 1 E. coli (10 8 CFU / ml); S. enteritidis (10 7
CFU / ml) 1 Enterobacter aerogenes (10 6 CFU / ml);
S.
enteritidis
(10 3
CFU / ml) 1 Enterobacter
6
aerogenes (10 CFU / ml); S. enteritidis (10 7 CFU /
ml) 1 C. freundii (10 6 CFU / ml) and S. enteritidis
(10 7 CFU / ml) 1 C. freundii (10 2 CFU / ml) did not
show growth or characteristic migration zones within
24 h at 428C, but did so after 48 h.

3.3. Evaluation of the method with food samples

3. Results

3.1. Evaluation of the MSRV method with pure

strains
After 24 h of incubation, when seeded at levels
from 10 to 10 4 CFU / ml, the strains S. give, S. cerro,
S. belem, S. madelia, S. derby, S. arizonae, S. agona
and S. oranienburg did not develop characteristic
migration zones ( # 10 mm) on the MSRV surface.
The other strains, however, did ( $ 10 mm). After 24
additional hours of incubation, the strains with
initially negative results developed the characteristic
halo. The other Enterobacteriaceae strains (E. coli, P.
mirabilis, Enterobacter aerogenes and C. freundii)
did not show growth or showed minimum growth at
the point of inoculation, but without migration zones
within 24 or 48 h of incubation at 428C, independent
of the cell concentration (levels up to 10 8 CFU / ml).

Table 1 shows the positive results obtained with

the examination of food samples using the MSRV
(direct and indirect), standard culture and both
methods.
Salmonella spp. were not isolated from any samples of cocoa powder and / or granulated cocoa and
grated fresh coconut with the methodologies employed. No false-positive or false-negative reactions
were found with these products.
The results obtained with the standard culture and
the MSRV (direct 1 indirect) methods agreed for 141
samples out of 146. Using both methods, 21 samples
were positive and 120 were negative, showing 97.3%
agreement between them. When compared with the
standard method, the indirect method showed a
sensitivity of 86.4% and a specificity of 96.8%,
while the direct MSRV showed a sensitivity of
71.4% and a specificity of 99.2%. Combined, both

Table 1
Detection of foodborne Salmonella by a standard and MSRV methods
Sample type
(number)

Number of Salmonella positive samples

MSRV and
standard method
agreement (%)

MSRV 1 /
standard 2
(%)

MSRV 2 /
standard 1
(%)

Total positive
(%)

Chicken thighs (41)

Brazilian fresh pork sausages (35)
Cocoa powder and / or granulated cocoa (35)
Grated fresh coconut (35)

15 (83.3)
6 (75.0)
0 (0.0)
0 (0.0)

2 (11.1)
2 (25.0)
0 (0.0)
0 (0.0)

1
0
0
0

(5.5)
(0.0)
(0.0)
(0.0)

18 (43.9)
8 (22.8)
0 (0.0)
0 (0.0)

Total (146)

21 (14.4)

4 (2.7)

1 (0.7)

26 (17.8)

MSRV direct and indirect enrichment.

D. Worcman-Barninka et al. / International Journal of Food Microbiology 64 (2001) 387 393

MSRV methods (direct 1 indirect) showed a 95.5%

sensitivity and a 96.8% specificity. There were no
statistically significant differences between the
MSRV and standard methodologies.
Eighteen (43.9%) of the 41 chicken thighs sampled were positive for Salmonella. Seventeen
(94.4%) positive samples were obtained with the
MSRV method (direct and indirect). Of these, 11
were detected by direct enrichment (BPW/ MSRV)
and 16 by indirect MSRV methodology (Table 2).
Although more positive results were obtained with
MSRV medium than with standard culture, no
statistically significant differences were shown between test combinations.
Using the standard method, 16 (88.8%) samples
were positive for Salmonella spp. The selective
enrichment in RV broth allowed the detection of the
microorganisms in 15 samples and in TT broth, in 10
samples (Table 2). The best combination of selective
enrichment broth / selective agar was RV/ XLT4 (13
positive samples) (data not shown).
The most frequently isolated serotype from chicken samples was S. enteritidis. From 88 isolates, 87
(98.9%) belonged to this serotype. S. bradenburg
was recovered from one (1.1%) sample, that also
contained S. enteritidis.
Out of 35 samples of Brazilian fresh pork sausages
analysed, eight (23%) were found to be positive for
Salmonella (Table 1). The MSRV method detected
Salmonella in all of them (100%). The direct (BPW/
MSRV) enrichment detected five positive samples
and the indirect procedure in RV and TT broths
resulted in seven positive samples. Using the standard procedure, six (75%) samples were positive for
Salmonella spp. The RV broth in combination with
various plating media detected Salmonella in four
samples while the TT selective broth, detected five
positive samples. There were no statistically significant differences between the positive results using
rapid or standard methods.

391

Among the sausage samples, serotype diversity the

Salmonella strains was noted, probably due to the
variety of ingredients and / or product manipulation
during the manufacturing process. The isolates were
identified as S. enteritidis (30.5%), S. bredeney
(30.5%), S. infantis (19.4%), S. ohio (13.8%) and S.
agona (5.5%).

4. Discussion
The migration zone developed by motile Salmonella strains does not seem to be affected by
levels of the competitors, but is a consequence of the
levels of Salmonella. It was not possible to conclude
that the competition was sufficient to inhibit the
growth or migration of the Salmonella cells, since
we obtained satisfactory results in highly competitive
situations, like the bacterial mixtures composed of S.
enteritidis (10 1 CFU / ml) 1 E. coli (10 8 CFU / ml)
and S. enteritidis (10 1 CFU / ml) 1 Enterobacter
aerogenes (10 8 CFU / ml).
De Smedt and Bolderdijk (1987) obtained similar
results while studying method sensitivity with bacterial mixtures. In their study, Salmonella detection
using MSRV (direct 1 indirect) was successful in 49
out of 52 experiments. The three negative results
were those with very low concentrations of S.
typhimurium (less then 50 cells / ml), even after a
pre-enrichment step, which can be considered exceptional. However, these results do not agree with
those obtained by Silva and Eiroa (1993). Evaluating
12 different Salmonella serotypes, these authors
observed that only three strains (S. anatum, S.
havana and S. enteritidis) developed the characteristic migration zone with an inoculum level lower than
10 5 cells / ml, using the direct MSRV. The other
serotypes displayed migration only following inoculation of between 10 5 and 10 7 cells / ml. The strains
S. newport, S. typhimurium, S. infantis and S. derby

Table 2
Number of isolation with motility enrichment versus standard procedure
Food samples
(number)
Chicken thighs (41)
Sausages (35)
Total

Motility enrichment

Standard procedure

Direct

Indirect

Total

RV

TT

Total

11
5
16

16
7
23

17
8
25

15
4
19

10
5
15

16
6
22

392

D. Worcman-Barninka et al. / International Journal of Food Microbiology 64 (2001) 387 393

showed no migration after 24 h, even with an

inoculum of 10 7 cells / ml.
Although not all Salmonella strains migrated after
the first 24 h of incubation, most did after 48 h, and
the other Enterobacteriaceae did not show migration
even after an additional 24 h. We concluded that
reincubation for an additional 24 h, increases the
chance of recovery of Salmonella, without increasing
the number of false-positive results.
Previously, no statistically significant differences
were observed between an in house conventional
and the MSRV methodologies (ODonoghue and
Winn, 1993). On the other hand, Read et al. (1994)
obtained better results using MSRV medium as an
enrichment after pre-enrichment in buffered peptone
water, when compared to tetrathionate brilliant green
broth. However, Joosten et al. (1994) concluded that
the standard method showed better results than the
rapid one using the MSRV medium. Although
Wiberg and Norberg (1996) had found some disadvantages such as high cost and the inability to
detect non-motile salmonellae, they suggested the
use of the rapid method together with isolation onto
XLT4 agar, because of its rapidity and simplicity.
Dusch and Altwegg (1995) also found high sensitivity and specificity when using the MSRV
medium for the isolation of Salmonella from food.
When used in combination with a ferrioxamine Esupplemented pre-enrichment media, the motility
enrichment method (MSRV) can be more sensitive
and less time demanding (Pless and Reissbrodt,
1995).
Although no statistically significant differences
were found among the methodologies, the indirect
MSRV method performed slightly better (23 positive
samples) than the standard culture (22 positive
samples). The direct MSRV method ranked third,
with 16 positive samples. When combined, the
MSRV method (direct 1 indirect) gave 25 positive
samples. Therefore, the indirect MSRV or the combined methods (direct 1 indirect) can be used for
rapid detection of Salmonella in food samples.

Acknowledgements
de AmThe authors acknowledge the Fundacao
paro a Pesquisa do Estado de Sao Paulo grant No.
95 / 628-0 for financial support and the Coor-

de Aperfeicoamento de Pessoal de Ensino

denacao
Superior for D.W.-B.s fellowship.

References
Andrews, W.H., June, G.A., Sherrod, P., Hammack, T.S.,
Amaguana, R.M., 1995. Salmonella. In: Food and Drug
Administration Bacteriological Analytical Manual, 8th Edition.
AOAC International, Arlington, VA, pp. 5.015.19.
DAoust, J.Y., Sewell, A.M., Warburton, D.W., 1992. A comparison of standard cultural methods for the detection of
foodborne Salmonella. Int. J. Food Microbiol. 16, 4150.
De Smedt, J.M., Bolderdijk, R., 1987. Dynamics of Salmonella
isolation with modified semi solid RappaportVassiliadis
medium. J. Food Prot. 50, 658661.
De Smedt, J.M., Bolderdijk, R., Rappold, H., Lautenschlaeger, D.,
1986. Rapid Salmonella detection in foods by motility enrichment on a modified semi-solid RappaportVassiliadis medium.
J. Food Prot. 49, 510514.
De Smedt, J.M., Bolderdijk, R., 1990. Collaborative study of the
International Office of Cocoa, Chocolate, and Sugar Confectionery on the use of motility enrichment for Salmonella
detection in cocoa and chocolate. J. Food Prot. 53, 659664.
De Smedt, J.M., Bolderdijk, R., Milas, J., 1994. Salmonella
detection in cocoa and chocolate by motility enrichment on
modified semi solid RappaportVassiliadis medium: collaborative study. J. AOAC Int. 77, 365373.
De Smedt, J.M., Chartron, S., Cordier, J.L., Graff, E., Hoekstra,
H., Lecoupeau, J.P., Lindblom, M., Milas, J., Morgan, R.M.,
Nowacki, R., ODonoghue, D., Van Gestel, G., Varmedal, M.,
1991. Collaborative Study of the International Office of Cocoa,
Chocolate and Sugar Confectionery on Salmonella detection
from cocoa and chocolate processing environmental samples.
Int. J. Food Microbiol. 13, 301308.
De Zutter, L., De Smedt, J.M., Abrams, R., Beckers, H., Catteau,
M., de Borchgrave, J., Debevere, J., Hoekstra, J., Jonkers, F.,
Lengers, J., Notermans, S., Van Damme, L., Vandermeersch, R.,
Verbraeken, R., Waes, G., 1991. Collaborative study on the use
of motility enrichment on modified semisolid RappaportVassiliadis medium for the detection of Salmonella from foods.
Int. J. Food Microbiol. 13, 1120.
Dusch, H., Altwegg, M., 1995. Evaluation of five new plating
media for isolation of Salmonella species. J. Clin. Microbiol.
33, 802804.
Farmer, J.J., McWorther, A.C., Brenner, D.J., Morris, G.K., 1984.
The Salmonella arizona group of Enterobacteriaceae: nomenclature, classification and reporting. Clin. Microbiol. Newslett.
6, 6366.
Goossens, H., Wauters, G., De Boeck, M., Janssens, M., Butzler,
J.P., 1984. Semisolid selective motility enrichment medium
for isolation of Salmonellae from fecal specimens. J. Clin.
Microbiol. 19, 940941.
Holbrook, R., Anderson, J.M., Baird-Parker, A.C., Dodds, L.M.,
Sawhney, D., Stuchbury, S.H., Swaine, D., 1989. Rapid
detection of Salmonella in foods a convenient two-day
procedure. Lett. Appl. Microbiol. 8, 139142.

D. Worcman-Barninka et al. / International Journal of Food Microbiology 64 (2001) 387 393

Joosten, H.M.L.J., Van Dijck, W.G.F.M., Van der Velde, F., 1994.
Evaluation of motility enrichment on modified semisolid
RappaportVassiliadis medium (MSRV) and automated conductance in combination with Rambach agar for Salmonella
detection in environmental samples of a milk powder factory.
Int. J. Food Microbiol. 22, 201206.
ODonoghue, D., Winn, E., 1993. Comparison of the MSRV
method with an in-house standard method for the detection of
Salmonella in various high and low moisture foods. Lett. Appl.
Microbiol. 17, 174177.
ODonoghue, D., Morgan, R., Pugh, S., Davda, C., 1992. Comparison of the MSRV method with various rapid and standard
Salmonella detection methods for chocolate, confectionery and
biscuit ingredients. Lett. Appl. Microbiol. 15, 9295.
Oggel, J.J., Nundy, D.C., Ann Zebchuk, P., Shaw, S.J., 1995.
Reliability of the modified semi-solid RappaportVassiliadis
agar and the modified 12 test E system for detection of
Salmonella in poultry feeds. J. Food Prot. 58, 98101.
Pless, P., Reissbrodt, R., 1995. Improvement of Salmonella
detection on motility enrichment media by ferrioxamine Esupplementation of pre-enrichment culture. Int. J. Food Microbiol. 27, 147159.

Popoff, M.Y., LeMinor, L., 1992. In: Formules Antigeniques

des

393

Serovars
de Salmonella. Centre Collaborateur OMS de Reference et de Recherche pour les Salmonella, Paris, p. 145.
Poppe, C., Duncan, L., 1996. Comparison of the detection of
Salmonella by Tecra Unique E Salmonella test and the
modified RappaportVassiliadis medium. Food Microbiol. 13,
7581.
Read, S.C., Irwin, R.J., Poppe, C., Harris, J., 1994. A comparison
of two methods for isolation of Salmonella from poultry litter
samples. Poultry Sci. 73, 16171621.
Roitman, K.J.J., Boice, J.D., 1982. Epidemiologic Analysis With
Programmable Calculator. Epidemiology Resources, Boston,
MA.
Siegel, S., 1956. Nonparametric Statistics For the Behavioral
Sciences. McGraw Hill, New York.
do meio semi solido

Silva, N., Eiroa, M.U., 1993. Avaliacao

de
rapida

RappaportVassiliadis modificado para deteccao

de
Salmonella em alimentos. Colet. Inst. Tecnol. Aliment. 23,
6877.
Wiberg, C., Norberg, P., 1996. Comparison between a cultural
procedure using RappaportVassiladis broth and motility enrichments on modified semisolid RappaportVassiliadis
medium for Salmonella detection from food and feed. Int. J.
Food Microbiol. 29, 353360.