23

Original Article

Diagnosis of Typhoid Fever by Polymerase Chain

Reaction
S.R. Ambati, Gopal Nath1 and B.K. Das2

Miami Childrens Hospital, Miami, FL, USA, 1Department of Microbiology, Institute of Medical Sciences, Banaras
Hindu University, Varanasi, India, 2Department of Pediatrics, B.P. Koirala Institute of Health Sciences, Dharan,
Nepal

ABSTRACT
Objective. To determine the efficacy of nested polymerase chain reaction (PCR) in detecting Salmonella typhi gene sequences
in blood and urine specimens and to determine the cut-off titer of Widal test using PCR as gold standard test for diagnosis of
typhoid fever.

Methods. Study included 71 children between the ages of 8 months and 14 years; 52 of them were suspected cases of typhoid
fever, 11 were febrile non-typhoid controls, and 8 were apparently healthy children. Nested PCR in Blood and Urine, Blood
culture, Widal test and Urine culture were done and their results analyzed.

Results. Among suspected typhoid cases, PCR in blood and urine had positivity of 82.7% each. Blood culture, Widal test (at
cut off titer TO and / or TH ≥ 1:160) and urine culture had positivity of 26.9%, 50% and 3.8% respectively. In one case, urine
PCR was positive and blood PCR was negative. Similarly, in another case, PCR in blood was positive however urine tested
negative. Considering PCR as gold standard, the antibody cut off titer was evaluated. A cut-off titer of TO ≥ 1:80 and /or TH≥
1:160 had sensitivity and specificity of 72.7% and 84.2%, while the respective figures were 50% and 89.5% when the cut-off
titer was TO and/or TH ≥ 1:160.

Conclusion. The sensitivity, specificity, positive and negative predictive values, likelihood ratios were same for PCR based
detection of S. typhi in blood and urine samples. Nested PCR had higher efficacy in detecting typhoid fever than Widal test, blood
and urine cultures. A cut off titer of TO ≥1:80 and/or TH ≥ 1:160 was found to have better diagnostic value in this region. [Indian
J Pediatr 2007; 74 (10) : 909-913] E-mail : srikanthambati@hotmail.com

Key words : Blood culture; Polymerase chain reaction; Salmonella typhi; Typhoid fever; Widal test

Typhoid fever is the result of systemic infection caused
mainly by the bacteria Salmonella enterica subspecies
enterica serotype Typhi (S. typhi). According to the best
global estimates, there are at least 16 million new cases of
typhoid fever each yr, with 600,000 deaths.1 Typhoid
fever continues to be unabated in the developing
countries of Africa, Asia and Latin America where proper
sanitary facilities still remain a remote possibility. Until
now, the gold standard for the diagnosis was isolation of
S. typhi form blood, bone marrow, stool, urine or any
other body fluid. Blood culture alone may be positive in
45-70% of patients with typhoid if the volume of blood
collected is at least 20-30 ml.2 Isolation of the organism is
Correspondence and Reprint requests : Dr. Srikanth Reddy Ambati,
MD, 8225 Lake Drive, Building C # 206, Miami, FL-33166, USA.
Phone-001-7862185459.
[Received December 7, 2006; Accepted February 1, 2007]
often jeopardized in developing countries due to lack of
facilities or by prior antibiotic usage. So, the laboratory
diagnosis of typhoid fever relies heavily upon serological
tests such as Widal test. Newer and rapid diagnostic
modalities include antigen detection and identification of
specific nucleic acid sequence. Antigen detection yielded
unsatisfactory results because of its very low density in
the body fluids.3
The specific nucleic acid sequence of Salmonella Typhi
can be identified with or without amplification. DNA
probes were initially used by Rubin et al in 1989, with the
limitation of only few numbers of target DNA sequences
in the specimens collected from the patients.4 With the
advent of nested polymerase chain reaction (PCR), it is
possible to amplify and detect specific gene sequences of
S. typhi with in few hours.5,6,7 We used flagellin gene (H-
1d) for identification of S. typhi. This is a novel study
where PCR has been used to detect S. typhi in urine,
because bacteruria is well known phenomenon in

Indian Journal of Pediatrics, Volume 74—October, 2007 909


S.R. Ambati et al
typhoid fever and large amount of urine can be collected
form patients unlike blood. Further, the sensitivity and
specificity of different cutoff titers of Widal test have been
evaluated using PCR as gold standard test.
MATERIALS AND METHODS
This study was carried out at S.S. Hospital, Institute of
Medical Sciences, Banaras Hindu University, Varanasi,
Uttar Pradesh, India, from February 2003 to February
2004. A total of 71 children between the ages of 8 months
and 14 years were included in the study. Informed
consent was obtained from every patient and/or parent.
The subjects were divided into three groups. Group-A
included 52 clinically suspected cases of typhoid fever.
Cases were selected on the basis of high index of clinical
suspicion like continuous high grade fever, toxic
appearance, abdominal discomfort, relative bradycardia,
splenomegaly and hepatomegaly. Relevant investigations
were done to rule out other important causes of fever in
this region and those with a different etiology were
excluded from the study. Group-B included 11 febrile
non-typhoid controls in which an alternative diagnosis
has been established. The diagnoses in this control group
were Malaria (4), Urinary tract infection (3), Otitis media
(4). Group-C included 8 afebrile apparently healthy
controls with no history of fever in the preceding 6
months. Appropriate amount of blood was collected for
Widal test, Blood culture and PCR. Depending on the
age, 2-20 ml of blood was infused into brain heart
infusion broth with sodium polyanethol sulphonate for
culture. 3 ml of blood was collected in a tube containing
citrate phosphate buffer (pH 7) for nested PCR and 2-4 ml
of blood was taken in a sterile plain container for Widal
test. Approximately 200 ml of urine was collected in a
sterile container and transported to the laboratory
immediately.
After inoculation, the blood culture bottles (HiMedia,
Mumbai, India) were incubated for 7 days at 37°C. The
sub-culture plates were incubated overnight and
examined for the presence of bacterial colonies. Growth
tested positive for Salmonella typhi bio-chemically was
further confirmed by serological agglutination test by the
steps as per recommendation of Old. 8 Widal test was
performed to estimate the antibody titers against somatic
(TO) and flagellar (TH) antigens of Salmonella typhi using
commercial antigen kit (Span Diagnostics, Surat, India).
10 ml of urine was added to 10 ml of double strength
Selenite F broth for culture isolation. Subcultures were
made on appropriate solid plates at intervals.
Pure strain of S. typhi (ATCC 19430) provided by
Department of Microbiology, Institute of Medical
Sciences, Banaras Hindu University was used for the
standardization of PCR conditions and as a reference
strain. Strain was subcultured on nutrient agar and
910
24
verified by biochemical and serological methods. Phenol-
Chloroform method5 was used for DNA extraction from
whole blood samples, with some modifications.9 Nested
PCR was performed using oligonucleotide primers as
described by Song et al5 and modified by Frankel.6 ST1 (5’-
TAT GCC GCT ACA TAT GAT GAG-3’) and ST2 (5’-TTA
ACG CAG TAA AGA GAG-3’) were used for first round
PCR to amplify 495 bp sequence of flagellin (H-1d) gene
corresponding to nucleotides 1036-1056 and 1513-1530
respectively. For nested PCR, ST3 (5’-ACT GCT AAA
ACC ACT ACT-3’) and ST4 (5’-TGG AGA CTT CGG
TCG CGT AG-3’) were used to amplify a 364 bp
fragment(internal sequences) corresponding to
nucleotides 1072- 1089 and 1416- 1435 respectively. The
reaction mixture for the first round of PCR contained 2.5
µl 10x buffer (Genetix, USA), 1.1 µl of 1.5 mM MgCl 2
(Genetix, USA), 11 pmol of each primer (ST1 & ST2), 1 µl
of dNTP mix (MBI, Fermentas, USA), 1 U Taq DNA
polymerase (Genetix, USA) in a reaction mixture of 25 µl.
De-ionized water was used apart from the reactant
described above to achieve the reaction volume. In the
thermal cycler (Biometra, Goettingen, Germany), first
round amplification was done to 40 cycles and subjected
to denaturation for 1 min at 94oC, annealing for 75 sec at
57oC, and elongation for 1 min at 72oC followed by final
elongation step for 7 min.
For nested PCR, conditions remained same except it
had 21 pmol each of ST3 and ST4, annealing temperature
of 63oC and 4 µl of (1:6) diluted product of primary cycle
as template. 5 µl of amplified product was
electrophoresed on a 1.5% agarose gel containing 1.5 µg
ethidium bromide along with a tracking dye
bromophenol-blue initially at 100 volts for 5 min. and
then at 80 volts for 60 min with TBE buffer (10 mM Tris-
borate, 2 mM EDTA). The molecular marker (100 bp gene
Ladder; MBI, Fermentas, USA) was run concurrently to
see the amplicons of 495 bp and 364 bp (Fig. 1). Urine
sample was centrifuged at 1500g for 15 min. The pellet
thus obtained was washed twice with sterile phosphate
buffer saline (pH 7.2). The pellet was subjected for DNA
extraction. Amplification and detection of PCR product
was done using the method described above.
300bp-0 0-364 bp
100 bp-0
1-DNA ladder
2- Blood & 3-Urine from healthy control
4- Blood & 5-Urine from culture positive patient
6-Blood & 7-Urine from culture negative but high widal titers
8-Negative control
Fig 1. Detection of amplicons of Salmonella typhi specific flagellin
gene sequences in urine and blood
Indian Journal of Pediatrics, Volume 74—October, 2007
25

Diagnosis of Typhoid Fever by Polymerase Chain Reaction

Statistical analysis was done using Sensitivity,
Specificity, Positive predictive value (PV+), Negative
predictive value (PV-), Likelihood ratio for positive result
(LR+) and Likelihood ratio for negative result (LR -). 10
Significance level was determined using z test.
RESULTS
Out of 52 suspected cases of typhoid fever, PCR was able
to detect gene sequences specific for S. typhi in 43 cases in
blood and urine specimens each. However, S. typhi was
isolated from only 14 cases in blood culture and 2 cases in
urine culture. None of the febrile controls (group B) or
afebrile controls (group C) tested positive in blood
culture, urine culture or PCR. Widal test, at a cutoff titer
of TO ≥1:160 and/or TH ≥1:160, was positive in 50% (26/
52) cases in group A, while 18.2% (2/11) children in
group B were positive. However, none in group C were
positive by Widal test.
The comparative evaluation of PCR blood, Blood
culture, PCR urine, Urine culture and Widal test (cut off
titer TO and/or TH ≥ 1:160) is shown in table 1. PCR in
blood and urine samples had significantly (p<0.001)
higher sensitivities (82.7% each) compared to 26.9%, 3.8%,
50%, for blood culture, urine culture, Widal test
respectively. Specificity and Positive predictive value
were 100% for PCR and culture methods. Table 2 shows
TABLE 1. Evaluation of PCR, Cultures and Widal Test in the Diagnosis of Clinically Suspected Typhoid Fever Cases
the mean duration of fever in the cases positive for
various tests. In only one case PCR was positive in blood
and negative in urine. Similarly in another case PCR
could detect S. typhi in urine and not in blood. Detection
of S. typhi specific flagellin gene sequences in urine and
blood is shown in Fig. 1.
In group A, 19 (36.5%) cases had antibody titers of
≥1:160, 10 (19.2%) cases had titers of 1:80 and 23 (44.2%)
had titers ≤ 1:40 against TO and/or TH antigen. 19.2%
had titers of 1:80 and 44.2% had ≤1:40. The corresponding
figures in group-B were 18.2%, 9.1% and 72.7%,
respectively. While in group-C all had titres of ≤ 1:40.
Considering PCR as the gold standard, two cut-off titers
for widal test ‘TO ≥1:80 and/or TH ≥1:160’, ‘TO ≥1:160
and/or TH ≥1:160’ were evaluated as shown in table 3.
Antibody titers against TO and TH antigens of S. typhi in
PCR positive children is shown in table 4. About 40% of
PCR positive cases had TO and TH titers ≤1:40.
DISCUSSION
Till now, the gold standard test for the diagnosis of
typhoid fever is blood culture. The isolation rate of S.
typhi with standard culture techniques is between 40 to
70%. 11 In this study, blood culture was positive in 14
(26.9%) out of 52 clinically suspected typhoid patients.
The low isolation rate could be due to prior antibiotic

Tests Sensitivity Specificity PV + PV- LR+ LR -

Blood culture 26.9 100 100 33.3 8 0.731

PCR Blood 82.7 100 100 67 8 0.173
Urine Culture 3.8 100 100 27.5 8 0.962
PCR Urine 82.7 100 100 67 8 0.173
Widal test 50 89.5 92.8 39.5 4.8 0.566

PV+-Positive predictive value, PV--Negative predictive value, LR+-Likelihood ratio for positive result, LR-- Likelihood ratio for negative result.

TABLE 2. Status of Result of PCR Blood, Blood Culture, PCR Urine, and Widal Test and Mean Duration of Fever (n=52)

No. of cases PCR blood Blood Culture PCR Urine Widal test Mean duration of fever (days)

07 (13.5%) + + + + 12.5
07 (13.5%) + + + - 10.2
11 (21.5%) + - + + 08.0
17 (33.3%) + - + - 11.5
07 (13.5%) - - - + 10.1
01 (1.9%) + - - + 08.0
01 (1.9%) - - + - 05.0
01 (1.9%) - - - - 07.0

TABLE 3. Evaluation of Single Widal Test at Different Cut-off titer on Considering PCR as Gold Standard.

Test Sensitivity Specificity PV+ PV- LR+ LR-

Widal test (a) 72.7 84.2 91.4 57.1 4.6 0.32

Widal test (b) 50.0 89.5 91.7 43.6 4.76 0.56

a – Cutoff titre of TO ≥1:80 or TH ≥ 1:160 or both

b – Cutoff titre of TO ≥1:160 or TH ≥ 1:160 or both

Indian Journal of Pediatrics, Volume 74—October, 2007 911


S.R. Ambati et al
TABLE 4. Antibody titers against to and TH antigens of
Salmonella typhi in PCR positive children.
Titer Antibody
TO TH
No. % No. % was positive but PCR, blood and urine cultures were
≤1: 40 18 40.9 19 43.2
1: 80 9 20.5 10 22.7
≥ 1: 160 17 38.6 15 34.1
Total 44 100 44 100
intake, delay in presentation and non-application of
intracellular bacterial release technique. S. typhi can be
isolated from more than 90% of patients with typhoid
fever, if blood, bone marrow, duodenal aspirates was
cultured.11 Apart from blood culture, the feasibility of
other methods is limited due to technical and practical
reasons. Nucleic acid based detection of S. typhi was
developed in earlier studies using different target gene
sequences like ViaB region4,7 and flagellin gene.5,6 In this
study, nested PCR was used to yield 364 bp amplicon
specific for flagellin gene of S. typhi. Among the
suspected typhoid cases, PCR was able to detect S. typhi
in 82.7% (43/52) in both blood and urine samples. None
of the afebrile controls or febrile non-typhoid patients
was positive by PCR. The detection by PCR in blood and
urine was significantly higher (p<0.001 each) than blood
culture positivity. Fifty one out of 52 suspected cases
were positive by at least one of the 5 tests used. PCR in
blood and urine was positive in 44 cases collectively.
When PCR was considered as gold standard and the
other test were evaluated, a sensitivity of 31.8%, 50%,
4.5% and specificity of 100%, 50%, and 100% were
observed for blood culture, single tube Widal test and
urine culture, respectively.
Several earlier studies have reported sensitivities for
PCR ranging from 29.76% 12 to 71.9% 13,14 in clinically
suspected typhoid cases. The reasons for higher
sensitivity in our study are usage of nested PCR and
collection of 3 ml of blood for DNA extraction. Further,
the inhibitors of PCR like hemoglobin were eliminated by
taking only 4 ml of 6 times diluted product of 1st cycle as
template for the nested round. The sensitivity of PCR in
this study is comparable to the results obtained in another
work done at our institute.9 PCR was used in the past for
detection of Leptospires, Cytomegalovirus, Toxoplasma,
Hepatitis virus, Chlamydia and Neisseria in urine
samples. 15,16,17 Similarly, bacteruria is a well known
phenomenon with S. typhi. Hence, this study was
undertaken to assess the efficacy of PCR for the detection
of S. typhi in urine from patients suffering from typhoid
fever. Urine culture was positive in only 2 out of 52
suspected typhoid patients. So this cannot be used as
diagnostic test because of very less sensitivity though it
has 100% specificity. Out of 52 suspected typhoid cases,
only 7 (13.5%) cases were positive by all five tests. In
32.7% cases, only PCR was able to detect S. typhi. So,
912
26
without PCR, we would have missed one-third of
typhoid cases. In 7 cases (13.5%) Widal test was negative
but PCR and blood culture were positive which indicate
false negativity of Widal test. In 7 cases (13.5%) Widal
negative. These could be false positives due anamnestic
response. In 11 (21.1%) cases PCR and Widal test were
positive but blood culture was negative. These values
suggest that more than 50% cases would be missed if
blood culture is considered as gold standard test.
Despite several limitations, single tube widal test with
TO and/or TH titers ≥1:160 was taken as cut-off value in
endemic regions. 18 Rising antibody titers has been
traditionally regarded as a good diagnostic test, but it is
less practicable and the rise may not occur if there is early
use of antibiotics.19 Prior immunization or infection, cross
reaction with other Salmonellae, anamnestic response in
Malaria, Dengue are various reasons for false positivity of
this test.20 It may be stressed that single Widal test has got
little diagnostic significance until the sensitivity and
specificity of the test at different cut-off titers are known
for a defined population. Considering PCR as gold
standard diagnostic test, the significance of antibody
titers against TO and TH antigens has been evaluated in
this study. When two different cut-off titers were taken
i.e., (a) TO ≥1:80 and/or TH ≥1:160 (b) TO and/or TH
≥1:160, the sensitivity was significantly higher for the
former cut-off titer than the latter.
It can be inferred from the present study that
polymerase chain reaction has greatest diagnostic value
for the detection of S. typhi among all the diagnostic tests
used. PCR in urine has equal efficacy in detecting S.
Typhi to PCR in blood. Urine PCR is a better substitute
for establishing diagnosis of Typhoid fever as it is non-
invasive, rapid, sensitive and specific test. A cut off titer
of TO=1:80 and/or TH=1:160 for Widal test was found to
have better diagnostic value in this region.
REFERENCES
1. Ivanoff B, Levine MM, Lambert PH. Vaccination against
typhoid fever, Present status. Bull WHO 1994; 72: 957-971.
2. Hoffman SL, Edman DC, Punjabi NH et al. Bone marrow
aspirate culture superior to streptokinase clot culture and 8 ml
1:10 blood to broth ratio culture for diagnosis of typhoid fever.
Am J Trop Med Hyg 1986; 35: 836-839.
3. Forsyth JRL. Typhoid and Paratyphoid. In Forsyth JRL, ed.
Topley and Wilson’s Microbiology and Microbial Infections. Vol. 3.,
London; Arnold, 1998; 459-478.
4. Rubin FA, McWhirter PD, Punjabi NH et al. Use of a DNA
probe to detect Salmonella Typhi in the blood of patients with
typhoid fever. J Clin Microbiol 1989; 27: 1112-1114.
5. Song JH, Cho H, Pank MY, Na DS, Moon HB, Pai CH.
Detection of S. typhi in the blood of patients with typhoid
fever by PCR. J Clin Microbiol 1993; 31 : 1439-1443.
6. Frankel G. Detection of Salmonella Typhi by PCR. J Clin
Microbiol 1994; 32 : 1415.
Indian Journal of Pediatrics, Volume 74—October, 2007
27
Diagnosis of Typhoid Fever by Polymerase Chain Reaction
7. Hashimoto Y, Itho Y, Fujinaga Y et al. Development of Nested
PCR based on ViaB sequence to detect S. typhi. J Clin Microbiol
1995; 33: 775-777.
8. Old DC, Threlafall EJ. Salmonella. In Forsyth JRL, ed. Topley
and Wilson’s Microbiology and Microbial Infections. Vol. 2.,
London; Arnold, 1998; 970-997.
9. Prakash P, Mishra OP, Singh AK, Gulati AK, G Nath.
Evaluation of nested PCR in diagnosis of typhoid fever. J Clin
Microbiol 2005; 43: 431-432.
10. Greenberg RS, Daniels RS, Flanders WD, Eley JW, Boring JR.
Diagnostic testing. In Medical Epidemiology. New York;
McGraw-Hill; 1996; 77-88.
11. Gillman RH, Terminel M, Levine MM, Hernandez-Mendoza
P, Hornick RB. Relative efficacy of blood, urine, rectal swab,
bone marrow and rose spot cultures for recovery of Salmonella
Typhi in typhoid fever. Lancet 1975; 11: 1211-1213.
12. Chaudhary R, Laxmi BV, Nisar N, Ray K, Kumar D.
Standardization of PCR for the detection of S. Typhi in typhoid
fever. J Clin Pathol 1997; 50: 437-439.
13. Haque A, Ahmed J, Quereshi J. Early detection of Typhoid by
Polymerase Chain Reaction. Ann Saudi Med 1999; 19: 337-340.
14. Massi MN, Shirakawa T, Gotoh A, Bishnu A, Hatta M,
Indian Journal of Pediatrics, Volume 74—October, 2007
Kawabata M. Rapid diagnosis of typhoid fever by PCR assay
using one pair of primers from flagellin gene of Salmonella
Typhi. J Infect Chemother 2003; 9: 233-237.
15. Bal AE, Gravekamp C, Hartskeerl RA et al. Diagnosis of
leptospires in urine by PCR for early diagnosis of
leptospirosis. J Clin Microbiol 1994; 32 : 1894-1898.
16. Fuentes I, Rodriguez M, Domingo CJ, Castillo FD, Juncosa T,
Alwar J. Urine sample used for congenital toxoplasmosis
diagnosis by PCR. J Clin Microbiol 1996; 34: 2368-2371.
17. Whiley DM, Lecornec GM, Mackay IM, Siebert DJ, Sloots TP.
Real-time PCR assay for the detection of Neisseria
gonorrhoeae by Light cycler. Diag Microbiol Infect Dis 2002; 42:
85-89.
18. Rasaily R, Dutta P, Saha MR et al. Value of single Widal test
in the diagnosis of typhoid fever. J Clin Microbiol 1993; 97 : 104-
107.
19. Shukla S, Patel B, Chitnis DS. Hundred years Widal test and
its reappraisal in the endemic areas. Ind J Med Res 1997; 105 :
53-57.
20. Olapoenia LA, King AL. Widal agglutination test-100 years
later: still plagued by controversy. Prostgrad Med J 2000; 76: 80-
84.
913