Article

Tropical Doctor
0(0) 1–7
! The Author(s) 2018
Effectiveness of treatment regimens Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
for Typhoid fever in the nalidixic DOI: 10.1177/0049475518758884
journals.sagepub.com/home/tdo
acid-resistant S. typhi (NARST) era in
South India

Rini Bandyopadhyay1, Veeraghavan Balaji2, Bijesh Yadav3,

Sudha Jasmine4, Sowmya Sathyendra5 and Priscilla Rupali6

Abstract
The epidemiology of typhoid fever in South Asia has changed. Multi-drug resistant (MDR) Salmonella typhi (S. typhi) is now
frequently resistant to nalidixic acid and thus labelled NARST. Treatment failure with the use of fluoroquinolones has
been widely noted, forcing clinicians to adopt alternative treatment strategies. In this observational study, we looked at
various treatment regimens and correlated clinical and microbiological outcomes. In 146 hospitalised adults, the median
minimum inhibitory concentration (MIC) for ciprofloxacin was 0.38 mg/mL with a median fever clearance time (FCT) of
eight days (range ¼ 2–35 days). Of the regimens used, gatifloxacin and azithromycin had a shorter FCT of six days
compared to ceftriaxone (ten days; P < 0.001). Though mortality and relapse in our cohort was low, NARST seemed
to correlate with mortality (P ¼ 0.006). Gatifloxacin or azithromycin clearly emerge as the drugs of choice for treatment
of typhoid in South India.

Keywords
Typhoid fever, treatment failure, nalidixic acid-resistant S. typhi, gatifloxacin, azithromycin, India

which achieves good levels in macrophages but blood

Introduction
Quinolones have traditionally been considered the drug
of choice in the treatment of multidrug-resistant
(MDR) typhoid fever. However, clinical failure, as evi-
denced by late defervescence correlating with isolation
of Salmonella typhi (S. typhi) in blood cultures was
soon reported from all over the country. We have pre-
viously reported eight cases of treatment failure with
quinolones (both clinical and microbiological) from
our institution.1 Furthermore, different regimens have
been used worldwide where there was no response to
quinolones. Current data suggest that the best options
to treat nalidixic acid-resistant S. typhi (NARST) are
azithromycin or gatifloxacin.2 Concern, however, exists
that the presence of two or more mutations in the gyrA
gene is likely to render gatifloxacin ineffective in the
future. In addition, this drug has been banned in
recent years on the Indian subcontinent in spite of
being a good therapeutic choice for treatment of
typhoid, because of reports of dysglycaemia in elderly
Western populations.3 Azithromycin is a macrolide
levels are usually less than or just equal to the minimum
inhibitory concentration (MIC) for the organism;4 this
again suggests the possibility of future treatment fail-
ure. Ceftriaxone and cefotaxime, third-generation ceph-
alosporins, have traditionally been the drug of choice
for returning travellers from India with typhoid.5
1
Assistant Professor, Christian Medical College, Vellore, Tamil Nadu, India
2
Professor of Microbiology, Christian Medical College, Vellore, Tamil
Nadu, India
3
Consultant statistician, Department of Biostatistics, Christian Medical
College, Vellore, Tamil Nadu, India
4
Associate Professor, Department of Medicine, Christian Medical
College, Vellore, Tamil Nadu, India
5
Professor, Department of Medicine, Christian Medical College, Vellore,
Tamil Nadu, India
6
Professor and Head, Department of Infectious Diseases, Christian
Medical College, Vellore, Tamil Nadu, India
Corresponding author:
Priscilla Rupali, Christian Medical College, CMCH Vellore, Tamil Nadu,
India.
Email: prisci@cmcvellore.ac.in
2 Tropical Doctor 0(0)
However, present evidence, including meta-analyses,6
suggests that treatment outcomes are inferior to quin-
olones owing to higher rates of clinical failure and
relapse with the latter. The ongoing controversy regard-
ing treatment continues both in the endemic setting and
for travellers. Most parts of the world where typhoid is
endemic also have a pre-existing problem with Gram-
negative organisms having extended spectrum and
New Delhi metallo-beta-lactamase (NDM-1) which
are difficult to treat. Evaluation of new and recycled
old therapeutic options is necessary. In order to address
this question, we studied the clinical and microbio-
logical profile of adult patients admitted with NARST
and MDR typhoid fever in our hospital.
Participants and methods
Ours was a five-year retrospective cohort study con-
ducted at the Christian Medical College Hospital
(CMCH), Vellore, a tertiary-care medical centre
located in South India. All adult inpatients with cul-
ture-proven typhoid fever, admitted from January 2008
to December 2012, were included in the study. Cases
were searched from a list of patients with blood or bone
marrow cultures positive for S. enterica serotype typhi
or S. paratyphi obtained from the Department of
Microbiology. Corresponding inpatient charts were
obtained from Medical Records and the data were rec-
orded on pre-designed proformas. Adults (age >12
years) were recruited for the study. Patient data col-
lected included demographic details, presenting symp-
toms, physical and laboratory findings, and
complications. Response to treatment was assessed in
terms of time to fever clearance, clinical, microbio-
logical failure and mortality.
All isolates from blood culture had been identified
by standard biochemical tests and confirmed by slide
agglutination tests using specific Salmonella antiserum
(Murex Diagnostics Ltd., UK). Isolate antibiotic sus-
ceptibility was determined by the disc diffusion method
using a 5-mg disc of ciprofloxacin and a 30-mg disc of
nalidixic acid (HiMedia Laboratories Ltd., India),
according to the Clinical and Laboratory Standards
Institute (CLSI) guidelines and interpretive criteria.7
Isolates were also tested for susceptibility to chloram-
phenicol (30 mg), amoxycillin (10 mg), trimethoprim-sul-
famethoxazole (1.25/23.75 mg) and ceftriaxone (30 mg)
(HiMedia Laboratories Ltd., India).7 MICs of cipro-
floxacin were determined by the E-test method, using
commercially available strips (AB Biodisk, Sweden)
according to the manufacturer’s specifications.
During our study, we observed discordance between
the clinical response to therapy and the laboratory
interpretation of ciprofloxacin breakpoint susceptibil-
ities (MIC and disc diffusion) for S. typhi. However,
owing to a lack of clinical and microbiological correl-
ation and standardisation of the various
guidelines (CLSI), British Society of Antimicrobial
Chemotherapy (BSAC) and European Committee on
Antimicrobial Susceptibility Testing (EUCAST), we
decided to report the ciprofloxacin MIC for S. typhi
as resistant when 0.064 mg/mL according to
EUCAST guidelines.8 In 2012, CLSI, BSAC and
EUCAST guidelines were standardised and evidence-
based revision of the ciprofloxacin MIC breakpoints
and the disc diffusion interpretative criteria specifically
for S. typhi9 were amended as currently followed by our
microbiology department. This now reports ciprofloxa-
cin disk diffusion cut-offs > 31 mm as resistant or MIC
breakpoints < 0.06 mg/mL as ‘susceptible’.
Fever clearance time (FCT) was defined as from the
start of appropriate antibiotic therapy to the first
instance when oral temperature dropped below 37.5 C
and remained below this level continuously for 48 h.
Complete and sustained resolution of clinical symp-
toms and signs of fever defined cure, while the persist-
ence of fever > 6 days defined failure. Furthermore,
isolation of S. typhi or S. paratyphi from blood or
bone marrow cultures in patients who had been on
treatment for typhoid fever for > 7 days defined true
microbiological treatment failure. Relapse was defined
as recurrence of clinical symptoms and signs compat-
ible with typhoid fever < 6 weeks after completion of a
course of antibiotics for this disease, after having
had complete resolution of symptoms and signs
during the same treatment episode. Complications dir-
ectly attributed to typhoid were noted. An empiric anti-
biotic was started based solely on a clinical diagnosis of
typhoid, before microbiological confirmation. By con-
trast, the original/alternative antibiotic was the one
used after isolating S. typhi or S. paratyphi in culture
and interpretation of susceptibility data. The choice of
antibiotic therapy was at the discretion of the treating
physician.
All statistical analyses were done using a software
package (SPSS for Windows, Version 17.0.1, SPSS
Inc., Chicago, IL, USA). Continuous variables are pre-
sented as mean  standard deviation (SD) or as median
with interquartile range (IQR). Categorical variables
are expressed as proportions. The effect of infection
with NARST and MDR typhoid (S. typhi or S. para-
typhi) and the effect of different treatment regimens on
FCT were assessed by Kaplan–Meier survival analysis.
The log-rank test was applied to evaluate the equality
of fever clearance distributions across treatment regi-
mens. We performed a univariate analysis to determine
factors associated with adverse outcomes. Multivariate
analyses could not be carried out as the sample size was
too small. All tests were two-sided and a P value < 0.05
was considered statistically significant.
Bandyopadhyay et al.
Figure 1. Trend of typhoid isolates over the years. (a) Culture-confirmed S. typhi study isolates over the years with ciprofloxacin
MIC. (b) Number of NARST and MDR of typhoid isolates over the years.
Results
A total of 958 cases of typhoid were diagnosed and 146
(15.24%) adults admitted with culture-proven typhoid
were included in our study. The year-wise prevalence is
shown in Figure 1a. The mean age of the study group was
29  12.8 years and one-third of the patients were female
(52 [35.6%]).The study population was predominantly
urban (52.1%) in distribution. Only one-third (33.6%)
presented in the first week of illness with the mean dur-
ation of fever before admission being 17.8 days. Patients’
baseline characteristics are shown in Table 1.
Among the 146 isolates, 128 (87.7%) were S. typhi and
18 (12.3%) were S. paratyphi (Table 2). The organism
was isolated from blood in 118 (80.8%) and from bone
marrow in 16 (11%) patients, respectively. NARST iso-
lates were found in 116 (93.5%) patients and eight (6.5%)
patients had MDR S. typhi infection. All S. paratyphi
isolates were resistant to nalidixic acid and there were
no MDR S. paratyphi isolates. The MDR isolates were
resistant to nalidixic acid, ampicillin, chloramphenicol,
3
ciprofloxacin and trimethoprim-sulfamethoxazole.
There was no isolate reported as resistant to ceftriaxone
(MIC range ¼ 0.016–0.064 mg/mL; susceptible if
MIC  8 mg/mL). The antimicrobial susceptibility pat-
terns of typhoid isolates and the trend of NARST and
MDR typhoid isolates over the years are shown in Figure
1a and 1b, respectively.
Clinical outcomes
The mean temperature at admission in the study popu-
lation was 38.8 C with NARST and MDR typhoid iso-
lates having maximum temperatures (Tmax) of 40.5 C
and 40 C, respectively. Clinical failure was seen in 87
(59.6%) with a median FCT of eight days (IQR ¼ 6–10
days). In four patients, the fever failed to defervesce
before death. The median FCT in patients with nali-
dixic acid-susceptible S. typhi (NASST) was six days
(IQR ¼ 2–7 days), NARST eight days (IQR ¼ 6–10
days) and MDR S. typhi 6.5 days (IQR ¼ 6–13 days);
this was not statistically significant.
4 Tropical Doctor 0(0)

Table 1. Baseline characteristics of patients with typhoid fever. Table 2. Patterns of typhoid isolates.

Variable Value Variable Value

Age (years), mean  SD 29.32  12.82 S. typhi, n (%) 128 (87.7)

Sex S. paratyphi, n (%) 18 (12.3)
Male, n (%) 94 (64.4) Sample in which the organism is isolated, n (%)
Female, n (%) 52 (35.6) Blood 118 (80.8)
Demography Bone marrow 16 (11.0)
Rural, n (%) 70 (47.9) Blood and bone marrow 6 (4.1)
Urban, n (%) 76 (52.1) Pus 5 (3.4)
Duration of fever before hospitalisation 17.77  18.7 Stool 0
(days), mean  SD Bile 1 (0.7)
Length of hospital stay (days), mean  SD 9.64  5.24 NASST S. typhi, n (%) 4 (3.1)
Temperature at admission ( F), mean  SD 102.9  1.38 NARST S. typhi, n (%) 116 (93.5)
Associated symptoms, n (%) NARST S. paratyphi, n (%) 18 (12.3)
Diarrhoea 69 (47.3) MDR S. typhi, n (%) 8 (6.5)
Abdominal pain 43 (29.5) MDR S. paratyphi, n (%) 0
Headache 36 (24.7) MIC (mg/mL)
Vomiting 71 (48.6)  0.25 51 (34.9)
Altered sensorium 5 (3.4) 0.25–0.75 62 (42.5)
None 23 (15.8) 0.75–1 5 (3.4)
Associated signs, n (%) 1 10 (6.8)
Relative bradycardia* 19 (13.0) Microbiological failure, n (%) 3 (2.1)
Coated tongue 14 (9.6)
NASST, nalidixic acid-sensitive S. typhi; NARST, nalidixic acid-resistant S.
Hepatomegaly 27 (18.5) paratyphi; MDR, multidrug-resistant; MIC, minimum inhibitory
Hepatosplenomegaly 24 (16.4) concentration.
Splenomegaly 15 (10.3)
None 40 (27.4)
Empiric antibiotic used at admission, n (%) Empiric antibiotics were started on 109 (74.7%) with
None 37 (25.3) the initial choice being ciprofloxacin in 36 (33%), cef-
Ciprofloxacin 36 (24.7) triaxone in 43 (39.4%), piperacillin-tazobactam in 12
Ceftriaxone 43 (29.5) (11%) and azithromycin in four (3.7%) patients. In
Piperacillin-Tazobactam 12 (8.2) 30/109 (27.5%) patients, the initial empiric antibiotic
Others 18 (12.4)
regimen was continued as the final antibiotic of
choice. In the remaining 79 (72.5%), if the initial anti-
Time to fever defervescence 8.54  4.4
(days), mean  SD
biotic choice was a fluoroquinolone or a third-genera-
tion cephalosporin, it was changed to an alternative
Overall mortality, n (%) 4 (2.7)
single antibiotic in 51 (64.6%), dual antibiotic combin-
Clinical failure, n (%) 87 (59.6)
ation in 26 (32.9%) and triple antibiotic combination in
*Defined as pulse rate < 80/min when temperature > 101 F. two (2.5%). Among the 37 patients (25.3%) in whom
empiric antibiotics were deferred pending blood culture
Complete clinical cure was seen in 142 (97.3%) results, the final antibiotic was a single antibiotic in 29
patients but 19 (13%) developed complications, which (78.4%), whereas eight (22%) were given a combin-
included encephalopathy (4), haemophagocytic syn- ation of antibiotics. Gatifloxacin was used as a single
drome (3), visceral abscess (3), gastrointestinal bleeding antibiotic in 36 (45%), followed by azithromycin in 20
(2) and sepsis with multiorgan dysfunction (2). Among (25%), ceftriaxone in 16 (20%) and cotrimoxazole in
those with complications, 84.2% had NARST and five (6.3%). The antibiotic combinations in the dual
10.5% MDR isolates. The overall mortality in the antibiotic combinations used in the study cohort
study cohort was four (2.7%). Clinical outcomes in were: azithromycin-based in 17 (51.5%), gatifloxacin-
patients with NARST, NASST and MDR S. typhi iso- based in ten (30.3%) and ceftriaxone-based in
lates could not be compared owing to small numbers. six (18.2%).
Only one out of the 146 patients had a relapse of The median FCT was significantly prolonged with
typhoid fever. dual and triple antibiotic regimens compared to those
Bandyopadhyay et al. 5

Figure 2. Median FCT among different antibiotics.

Figure 3. Response of typhoid fever to various antimicrobials. Kaplan–Meier survival plots of time to fever clearance in patients with
blood culture-confirmed Typhoid fever. The y-axis represents the cumulative probability of event-free survival, i.e. remaining febrile.

treated with a single appropriate antibiotic regimen gatifloxacin or azithromycin (median FCT ¼ 6 days)
(P ¼ 0.004). A comparison of median FCT among dif- alone as the final antibiotic compared to those treated
ferent antibiotic combinations is shown in Figure 2. with ceftriaxone alone (median FCT ¼ 10 days;
The median FCT was shorter in patients treated with P < 0.001) (Figure 3).
6 Tropical Doctor 0(0)
Microbiological outcomes
Out of the 146 isolates, MIC of ciprofloxacin was avail-
able for 128 isolates and the median MIC in the cohort
was 0.38 mg/mL (Table 2). The increase in MIC of
ciprofloxacin had no adverse impact on the clinical
response to antimicrobial therapy, both in terms of
FCT (P ¼ 0.85) and mortality (P ¼ 0.4). In the study
cohort, 46 patients (31.5%) had a second blood culture
at a median time period of four days (IQR ¼ 2–6 days)
and S. typhi was isolated from eight (17.4%) patients
and S. paratyphi from one patient. True microbio-
logical failure was found in only three patients with
NARST isolates. There was no statistically significant
correlation between rising MIC and microbiological
failure (P ¼ 0.78). All patients with microbiological fail-
ure were initially treated with ciprofloxacin empirically
and were subsequently given a triple drug regimen con-
sisting of gatifloxacin, azithromycin and co-trimoxa-
zole. There was no increase in mortality or
complications in patients with microbiological failure.
Discussion
Our study demonstrates a very high prevalence (93.5%)
of NARST and highlights a high rate of clinical failure
with commonly used antibiotic regimens in India. The
prevalence of nalidixic acid resistance in our study is
higher compared to earlier Indian studies.10
Community-based prospective studies have shown
high rates of clinical failure with fluoroquinolone ther-
apy (14%),11 an increased prevalence of NARST
strains with a consistent increase in the MIC levels of
ciprofloxacin for S. typhi isolates.12
NARST and MDR S. typhi infections have been
associated with an increased severity of illness and
higher rates of complications13 contributing to a signifi-
cantly increased morbidity, loss of productivity and
increased cost of healthcare. In our study, mortality
was significantly higher in patients with NARST iso-
lates but the FCTs between NASST, NARST and
MDR groups were similar. This may be explained by
late presentations to hospital and easily accessible over-
the-counter antibiotics, thus delaying appropriate ther-
apy. There was no significant association between age,
sex, duration of illness or MIC of ciprofloxacin and
severity of typhoid fever as shown in previous studies.14
Furthermore, there was no mortality in patients with
MDR isolates and the MDR phenotype was not asso-
ciated with severe disease compared to previous reports
in children,15 perhaps due to the low prevalence of
MDR S. typhi infections in the study cohort (6.5%).
The notable decline in MDR strains of S. typhi has
been corroborated by studies from North India,16
implying a trend towards re-emergence of susceptibility
to first-line antibiotics which may be recyclable for
effective treatment. Studies have suggested that re-
emergence of susceptibility to these drugs may be a
result of the emergence of de novo susceptible strains,
loss of high molecular weight self-transmissible plas-
mids17 or a lack of antibiotic selection pressure on
these drugs.18 Further, reports indicate that MDR S.
typhi strains are genetically homogenous and may coex-
ist as distinct, independent clones.19
Among the antibiotics used, gatifloxacin had the
lowest FCT. Where typhoid is endemic and there are
limited therapeutic options, judicious use of this drug
may prove to be cost-effective through earlier deferves-
cence, shorter duration of hospitalisation and fewer
complications.
Despite being the current World Health
Organization recommendation for the treatment of
typhoid fever,20 azithromycin has an inherent risk of
developing resistance. The efficacy of azithromycin is
related to the tissue rather than the serum concentra-
tion4 as it concentrates in macrophages. Consequently,
it achieves low blood levels, usually less than or just
equal to the organism’s MIC, although it has a long
half-life (2–3 days). There is a theoretical consideration
that this lower concentration in the blood as compared
to tissue could easily lead to resistance. Moreover,
azithromycin is extensively used in the treatment of
other diseases such as rickettsial infections, sinusitis
and atypical pneumonia. Its indiscriminate use may
contribute to resistance in the future.
Systematic studies to establish the efficacy of differ-
ent treatment regimens for typhoid fever in the
NARST era are scarce. The numerous antibiotic regi-
mens used in our hospital are reflective of current
chaotic clinical practice. The relatively small size of
our study, the retrospective analyses of the data lead-
ing to bias (sicker patients could have received com-
bination therapy) and constantly changing laboratory
reporting of S. typhi MIC breakpoint susceptibilities
are the main limitations of our study. Robust molecu-
lar assays for early diagnosis, a better understanding
of the disease epidemiology in the community and
widespread awareness of appropriate antibiotic use
would be useful in controlling the spread of resistant
isolates and establishing an optimal treatment
regimen.
Conclusion
Nonetheless, azithromycin and gatifloxacin clearly
emerge as drugs of choice for the treatment of typhoid.
Widespread preventive strategies, sensitive and specific
rapid diagnostic tests, and a randomised controlled trial
comparing various therapeutic regimens leading to the
development of an appropriate treatment guideline are
urgently required.
Bandyopadhyay et al.
Acknowledgements
We thank all staff from the The Medical Records and
Department of Microbiology, Christian Medical College,
Vellore, for their kind help and cooperation.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.
Funding
The author(s) received no financial support for the research,
authorship, and/or publication of this article.
References
1. Rupali P, Abraham OC, Jesudason MV, et al. Treatment
failure in typhoid fever with ciprofloxacin susceptible
Salmonella enterica serotype Typhi. Diagn Microbiol
Infect Dis 2004; 49: 1–3.
2. Dolecek C, Tran TP, Nguyen NR, et al. A multi-center
randomised controlled trial of gatifloxacin versus azithro-
mycin for the treatment of uncomplicated typhoid fever in
children and adults in Vietnam. PLoS One 2008; 3: e2188.
3. Park-Wyllie LY, Juurlink DN, Kopp A, et al. Outpatient
gatifloxacin therapy and dysglycemia in older adults. N
Engl J Med 2006; 354: 1352–1361.
4. Capoor MR, Rawat D, Nair D, et al. In vitro activity of
azithromycin, newer quinolones and cephalosporins in
ciprofloxacin-resistant Salmonella causing enteric fever. J
Med Microbiol 2007; 56: 1490–1494 .
5. Basnyat B, Maskey AP, Zimmerman MD, et al. Enteric
(typhoid) fever in travelers. Clin Infect Dis 2005; 41:
1467–1472.
6. Thaver D, Zaidi AK, Critchley J, et al. A comparison of
fluoroquinolones versus other antibiotics for treating
enteric fever: meta-analysis. BMJ 2009; 338: b1865.
7. Clinical and Laboratory Standards Institute (CLSI).
Performance Standards for Antimicrobial Disk
Susceptibility Tests; Approved Standard—Ninth Edition.
Wayne, PA: CLSI, 2006.
8. European Committee on Antimicrobial Susceptibility
Testing. Breakpoint tables for interpretation of MICs and
zone diameters. Version 2.0. Basel: EUCAST, 2012.
7
9. Clinical and Laboratory Standards Institute (CLSI).
Performance Standards for Antimicrobial Susceptibility
Testing; Twenty-Third Informational Supplement. M100-
S23. Wayne, PA: CLSI, 2013.
10. Kadhiravan T, Wig N, Kapil A, et al. Clinical outcomes
in typhoid fever: adverse impact of infection with nali-
dixic acid-resistant Salmonella typhi. BMC Infect Dis
2005; 5: 37.
11. Sinha A, Sazawal S, Kumar R, et al. Typhoid fever in
children aged less than 5 years. Lancet 1999; 354:
734–737.
12. Jesudason MV, Malathy B and John TJ. Trend of
increasing levels of minimum inhibitory concentration
of ciprofloxacin to Salmonella typhi. Indian J Med Res
1996; 103: 247–249.
13. Kabra SK, Madhulika, Talati A, et al. Multidrug-resis-
tant typhoid fever. Trop Doct 2000; 30: 195–197.
14. Parry CM, Thompson C, Vinh H, et al. Risk factors for
the development of severe typhoid fever in Vietnam.
BMC Infect Dis 2014; 14: 73.
15. Bhutta ZA, Naqvi SH, Razzaq RA, et al. Multidrug-
resistant typhoid in children: presentation and clinical
features. Rev Infect Dis 1991; 13: 832–836.
16. Singhal L, Gupta PK, Kale P, et al. Trends in antimicro-
bial susceptibility of Salmonella Typhi from North India
(2001-2012). Indian J Med Microbiol 2014; 32: 149–152.
17. Shanahan PM, Jesudason MV, Thomson CJ, et al.
Molecular analysis of and identification of antibiotic
resistance genes in clinical isolates of Salmonella typhi
from India. J Clin Microbiol 1998; 36: 1595–1600.
18. Dutta S, Sur D, Manna B, et al. Rollback of Salmonella
enterica serotype Typhi resistance to chloramphenicol
and other antimicrobials in Kolkata, India. Antimicrob
Agents Chemother 2005; 49: 1662–1663.
19. Thong KL, Bhutta ZA and Pang T. Multidrug-resistant
strains of Salmonella enterica serotype typhi are genetic-
ally homogenous and coexist with antibiotic-sensitive
strains as distinct, independent clones. Int J Infect Dis
2000; 4: 194–197.
20. World Health Organization. Background document: the
diagnosis, treatment and prevention of typhoid fever.
Geneva: World Health Organization, 2003.