Emergence of Salmonella paratyphi A as a Major Cause of Enteric Fever: Need for Early Detection, Preventive Measures, and Effective Vaccines

Monthida Fangtham MD, Henry Wilde MD, FACP
DOI: http://dx.doi.org/10.1111/j.1708-8305.2008.00237.x 344-350 First published online: 1 September 2008

Enteric fever (EF) is a systemic infection caused by Salmonella enterica serotype typhi (typhoid fever) or serotype paratyphi A, B, or C (paratyphoid fever). However, other organisms may also cause the EF syndrome. 1 EF used to be an important cause of morbidity and death in war‐torn and developing countries including Thailand. However, improved sanitation has reduced the prevalence of EF worldwide. It nevertheless remains a serious public health problem in parts of Asia where sanitary conditions remain poor.

Salmonella enterica serotype paratyphi A (S paratyphi A) was thought to cause a smaller portion of EF cases in the past with Salmonella enterica serotype typhi (S typhi) as the major problem worldwide. This is no longer the case. Several recent reports showed an increasing incidence of S paratyphi A causing the EF syndrome. This is especially true in developing countries that have a high disease burden (incidence >100/100,000 cases/y) such as South Central Asia and Southeast Asia. 2 Reports from Western countries also found an increasing incidence of S paratyphi A among returned travelers from endemic regions. 3 Furthermore, S paratyphi A, B, and C have been shown to cause localized infections such as meningitis, 4 brain abscess, 5 liver abscess, 6,7 splenic abscess, 8,9 thyroid abscess, 10 osteomyelitis, and psoas abscess. 11 It has been suggested that a disproportionately higher number of cases of S paratyphi A in Thailand may have been due to mass vaccination of schoolchildren with the crude heat‐inactivated typhoid vaccine in 1977. 12 This resulted in a sharp decline of the EF incidence, but isolation rates of S paratyphi A did not significantly decrease. 12

In the past, paratyphoid fever was believed to cause a milder disease than typhoid fever, but recent reports showed indistinguishable clinical features in paratyphoid and typhoid fever. One report from travelers in Israel showed a greater complication rate in paratyphoid fever patients. 13 Because of indistinguishable clinical features between S typhi and S paratyphi disease and public health concerns, diagnostic tests were developed to identify the etiologic agent. However, rapid diagnostic tools are still not generally available. This contributes to lack of accurate epidemiological data and much empirical and often inappropriate treatment. Currently available typhoid vaccines do not protect against S paratyphi A, and it appears that S paratyphi A is an emerging problem.


Data used in this review were identified by searching the PubMed database for the terms “Enteric fever,”“typhoid fever,”“paratyphoid fever,”“Salmonella paratyphi A,”“Widal test,”“typhoid vaccine,” and “Salmonella paratyphi A vaccine.” Only articles published in English in the past 20 years were considered. Reference lists in recent book chapters and review articles were also used. Furthermore, any new work of relevance on Web sites was included.


Over the past decade, S paratyphi A was shown to be an increasingly reported cause of EF worldwide. An outbreak of EF caused by S paratyphi A has been reported from India and Nepal with a restricted number of clones. 14,15 Several reports also documented an increased incidence of S paratyphi A among returning travelers from endemic areas. 13,16 Reports from India showed an increasing incidence of EF due to S paratyphi A in many regions such as the urban slums at New Delhi, 17 Mumbai, 18 Nagpur, 19 Kolkata, 20 south India (Karnataka), 21 and north India. 22 One of the highest increases of S paratyphi A isolation came from New Delhi, India. It has shown a rise from 6.5% in 1994 to 44.9% in 1998, excluding the strain from the urban residential area of the New Delhi outbreak in 1996. 17 Consistent with a current report from Pondicherry, India, it showed that 47% of isolation during 2004 to 2005 was nalidixic acid–resistant S paratyphi A. 23 A recent report from New Delhi, India, demonstrated a significant increase in S paratyphi A isolation from 1.7% in 2001 to 18% in 2005 to 2006. There was an increase of 3.8% in patients requiring hospitalization. One of the largest reports from Nepal also indicated an increased incidence of S paratyphi A from 23% during 1993 to 1998 to 34% during 1999 to 2003 (2,677 of 9,124 EF cases during 1993–2003). 24 This is consistent with other reports from Pakistan, China, Indonesia, and the Philippines, which showed a high proportion of S paratyphi A especially in Southeast China where S paratyphi A is more frequently isolated than S typhi. Other reports from China demonstrated a high incidence of S paratyphi A with infection rates up to 64% among all EF cases. 25–28

The incidence of EF appears to vary with age groups and seasons. Paratyphoid fever has the highest incidence in teenagers and young adults in contrast to typhoid fever, which is more common in children. 16,29 This is possibly due to a different mode of transmission between S typhi and S paratyphi A. Salmonella typhi is mainly transmitted by household contact. In contrast, S paratyphi A is mainly transmitted by contaminated food from a street vendor. 30 Several studies report seasonal variation of the EF incidence but with conflicting data. Studies from Indonesia and Vietnam showed a peak incidence in the dry season, 30,31 whereas in India and Nepal, the monsoon season has a higher incidence that is explained by flooding causing a greater risk of water contamination. 15,29

The incidence of EF declined greatly with provision of good sanitation and clean water. With the advent of increasing international travel, EF became associated with travelers returning from endemic areas, particularly the Indian subcontinent. An increasing isolation rate of S paratyphi A in travelers has been noted. Studies from Israel and Sweden have shown that S paratyphi A can be responsible for up to 50% of EF cases in returning travelers. 13,16 This high incidence rate among tourists may be due to the fact that most travelers from developed countries to Asia are teenagers or young adults. They are more likely to acquire a food‐borne disease from street vendors, the main route of transmission for S paratyphi A. 28 The true prevalence of EF caused by S paratyphi A may still be underestimated due to lack of reliable diagnostic tools. It has been suggested that most cases present with nonspecific symptoms, often in outpatient settings, and are treated empirically without full investigation and isolation of the responsible organism. 25

Transmission and risk factors

Humans are the only natural host and reservoir for S paratyphi A and S typhi. 32 The routes of transmission of EF are ingestion of contaminated food and water and close contact with acutely infected patients or with a chronic typhoid carrier. The infectious dose for S typhi is estimated as >1,000 organisms orally, 33,34 and a greater infective dose is required for S paratyphi A. 35 There may be differences in transmission and risk factors between S paratyphi A and S typhi. A recent study from Indonesia found that infection caused by S typhi is transmitted within households by poor hand washing, sharing of food, and recent typhoid fever in a household member. This contrasted to S paratyphi A that was predominantly transmitted outside the household by consumption of contaminated foods from street vendors and flooding. 30 It has also been suggested that S paratyphi A can multiply in food to reach an infective dose. 35

Clinical features

Salmonella paratyphi A is believed to cause a milder disease than S typhi, but several recent studies showed that S paratyphi A and S typhi produce indistinguishable clinical features. 15,16,27,36,37 There is one report showing a greater complication rate with S paratyphi A infections. 16 EF patients typically present with fever. It is initially low grade and rises by the second week of illness. It is accompanied by dull frontal headache, malaise, myalgia, dry cough, and occasionally abdominal symptoms such as diarrhea or constipation as well as encephalopathy. Physical signs and laboratory parameters are similar and include abdominal tenderness, relative bradycardia, hepatomegaly, splenomegaly, rose spots (rarely seen in Africans and Asians), and a coated tongue. Laboratory test may show mild anemia, leukopenia, thrombocytopenia, and elevated liver enzymes. 38,39 Complications of EF occur in approximately 10% to 15% of the cases. These usually appear where there is delay in treatment. 39 The most common complications of EF are gastrointestinal bleeding, intestinal perforation, and typhoid encephalopathy. Several studies showed no difference between complication rates between S paratyphi A and S typhi. 27,36,37 However, a study in Israeli travelers found that EF caused by S paratyphi A had a higher rate of complications (18.7% vs 3.2%) when compared to S typhi. 13


An evidence‐based clinical diagnosis of EF cannot be made as the presentations are nonspecific. It requires isolation of S paratyphi or S typhi from blood, stool, urine, or other body fluids. Blood culture is the mainstay diagnosis of EF. Bone marrow culture is more sensitive and gives positive results in up to 80% to 95% of the case even when the patient has already been on antibiotics. It has the disadvantage of being an invasive procedure and is rarely performed in the field. 38 Facilities for culture and isolation are often not readily available in endemic regions, and even if they are, it will take 3 to 5 days for results to become available. Most EF is diagnosed on clinical grounds and treated empirically without identification of the causative organism. This may result in unnecessary and inappropriate antibiotic use that can induce resistant strains.

Efforts to develop a serodiagnostic test for EF have been attempted for 100 years. The main focus was on the detection of S typhi. The classical Widal test was developed in 1896, and it is often the only available method to try to diagnose EF in highly endemic countries. The main concern is that the classical Widal test cannot differentiate between S typhi and S paratyphi A infections. This is not the only problem because the Widal test also reacts with other infectious agents that are present in acute febrile illnesses including malaria and dengue fever. This results in misleading diagnoses. 40 The classical Widal test requires acute and convalescent serum samples taken approximately 10 to 14 days apart. This is of little help in instituting urgently needed evidence‐based early therapy. A recent study evaluated a single Widal test. It was found to have no acceptable sensitivity or specificity in endemic areas or in vaccinated patients due to difficulty in establishing a baseline titer. 40 It cannot reliably identify the etiologic organism.

Several laboratories in endemic areas have developed newer immunological methods for diagnosis of EF including rapid antibody‐ and antigen‐based tests. However, most focused on detection of S typhi. Multiple commercial antibody‐based test kits such as Typhi‐Dot (Malaysian Biodiagnostic Research SDN BHD, Kuala Lumper, Malaysia), TUBEX (IDL Biotech, Sollentuna, Sweden), and Multi‐Test Dip‐S‐Ticks (PANBIO INDX Inc., Baltimore, MD, USA) focus on the detection of immunoglobulin (Ig) M and/or G of S typhi. None target S paratyphi A. The sensitivity and specificity of these tests showed promising results that are superior to the Widal test but do not diagnose S paratyphi A. 41 Antibody‐based tests have the highest sensitivity after the second week of illness. Waiting for such results may cause delay in diagnosis and thus limits the value of antibody‐based tests. Most of these tests are not readily available in the field. A rapid test, which becomes positive soon after onset of EF, would allow early evidence‐based therapy. Asian scientists have made antigen‐based tests for early detection of S typhi. One Thai laboratory developed a monoclonal antibody in 1998 for the detection of S typhi antigen 9 in clinical specimens using an enzyme‐linked immunosorbent assay method. The test showed promising results in urine specimens with a specificity and sensitivity of 100 and 65%, respectively. 42 Serially collected urine specimens improved sensitivity to 95%. 43 Despite the availability of antigen‐based diagnostic technology, these test kits are not widely used, and most were not developed for the detection of S paratyphi A, which is an emerging problem in Asia. It is ironic that a locally made S paratyphi A antigen–based test kit is available only for the food industry, and even this valuable test is not well known and should be improved into a human format. 44 Better methods for rapid, sensitive, and specific detection of S paratyphi A in clinical specimens are needs not only by clinicians but also by epidemiologists.

Antibiotic resistance

Antibiotic resistance of S paratyphi A appears to be an emerging problem. The 1996 outbreak of paratyphoid fever in India demonstrated that isolates were sensitive to all antibiotics including chloramphenicol, amoxicillin, cotrimoxazole, gentamicin, ciprofloxacin, and ceftriaxone. 14,17 Two years later, there was a report from New Delhi, India, concerning drug‐resistant S paratyphi A. The incidence of resistance to ciprofloxacin increased to 24%, and 32% of isolates had decreased susceptibility to ciprofloxacin [minimum inhibitory concentration (MIC) >2 mg/mL], the drug of choice for EF in India. 45 There were also reports of an increasing incidence of drug‐resistant S paratyphi A from EF‐endemic regions. Reports from a north Indian tertiary care hospital showed increasing multidrug‐resistant S paratyphi A strains. 22 In an outbreak of paratyphoid fever in 2001 in Nepal, 84% of the isolated strains were reported as resistant to nalidixic acid, which is considered the best predictor of clinical response to fluoroquinolones. 15 One of the largest prospective studies of EF reported a worrisome result: S paratyphi A was significantly more likely to be resistant to nalidixic acid (75.25% vs 50.5%) and ofloxacin (3.6% vs 0.5%) than S typhi. 36 The currently largest and longest retrospective study in Nepal showed that the trend of fluoroquinolone resistance in S paratyphi A is more rapidly increasing compared with that of S typhi. 24 Moreover, MICs of other antibacterials were also higher in S paratyphi A. 36 A high‐level ciprofloxacin‐resistant strain has been reported in India (MIC 8 mg/mL) 46 and Japan (MIC 128 mg/mL). 47 The development of ciprofloxacin resistance can be explained by exposure to this drug at concentrations near the MIC. It is due to a chromosomally mediated trait and different from other drugs where resistance is mainly plasmid mediated. Furthermore, rising antibiotic‐resistant strains of S paratyphi A have also been reported in travelers. A study from 10 European countries showed an increasing incidence of multidrug‐resistant S paratyphi A. It rose from 9% in 1999 to 25% in 2001, and the incidence of decreased susceptibility to ciprofloxacin also increased from 6% to 18%. Most of these resistant strains have been associated with travelers returning from the Indian subcontinent where resistant strains are endemic. 48 The increasing isolation rates of antibiotic‐resistant S paratyphi A herald serious clinical and public health consequences such as delayed response to treatment and prolonged bacterial shedding time.


Currently available typhoid vaccines are the live attenuated oral vaccine (Ty21a) (Vivotif Berna, Bern, Switzerland) and the parenteral Vi‐based vaccine (Typhim Vi) (Sanofi Pasteur , Lyon, France) (Table 1). A major disadvantage of both vaccines is lack of significant efficacy against S paratyphi A. However, several studies suggested some protection by Ty21a against S paratyphi A infection. Salmonella paratyphi A and B share the somatic O12‐antigen with S typhi, and this might be responsible for the Ty21a vaccine providing some cross‐protection against S paratyphi by a S typhi vaccine. Immunological studies have shown that volunteers vaccinated with Ty21a displayed a significant increase in cross‐reactive activity against S paratyphi A and B. 49 More recently, a nationwide study on the incidence of EF in Israeli travelers from 1995 to 2003 showed that the incidence of S paratyphi A during Vi vaccine use was three times higher than that during the period when Ty21a vaccine was used. 13 However, data from Indonesia showed conflicting results regarding Ty21a where a 30‐month surveillance showed no protection against S paratyphi A infection. 50 A recent review of the pooled data from two randomized, placebo‐controlled field trials of efficacy of Ty21a in Area Norte 51 and Area Occidente, 52 Santiago, Chile, suggested that Ty21a confers moderate protection against S paratyphi B (efficacy 49%). However, it was not possible to ascertain whether Ty21a also protects against S paratyphi A as there were few cases of S paratyphi A during the surveillance period. 53 Vi‐based vaccines contain only purified Vi antigen of S typhi. Salmonella paratyphi A and B lack Vi antigen. This renders Vi‐based S typhi vaccines ineffective for preventing paratyphoid infection. This concept is supported by southwestern China field trials, which showed that Vi‐based vaccines do not provide protection against S paratyphi A. 54,55

View this table:
Table 1

Current typhoid vaccines and their major manufacturers 32,38,39,56,60,61

Typhoid vaccineAgeRegimenRouteEfficacyAdverse effect and comments
Parenteral whole‐cell vaccine2‐dose regimenParenteral, subcutaneousVary 50%–90%Discontinued due to side effects
Live attenuated oral vaccine (Ty21a) (Vivotif Berna)Children >6 y and adult1 capsule on days 1, 3, 5, and 7 or 3 doses given 2 days apartOralVary 50%–80%Well tolerated; cannot be given to immunocompromised host, pregnancy; delay >24 h before administering antibiotics or mefloquine; must be refrigerated; multidose administration
Parenteral Vi‐based vaccine (Sanofi Pasteur)Children >2 y and adultSingle dose, booster every  2 yIntramuscularVary 50%–80%Well tolerated; can be given simultaneously with other vaccines and antimalarials
Vi vaccine conjugated to nontoxic recombinant Pseudomonas aeruginosa exotoxin A (Vi‐rEPA)Children 2–5 y and adult2‐dose regimenParenteralUp to 90%Well tolerated; vaccine is not commercially available at present

Efforts to develop a S paratyphi A vaccine started already in 1915 with attempts to improve the old parenteral whole‐cell typhoid vaccine. This trivalent vaccine (T.A.B) consisted of 1,000 million S typhi, 750 million S paratyphi A, and 750 S paratyphi B. 56 One study among expatriates in Nepal in 1987 to 1988 showed that whole‐cell typhoid vaccine offered 70% to 75% protection against S paratyphi A and 100% protection against S typhi. 57 This contrasted with a 1987 mass immunization study using whole‐cell typhoid vaccine among Thai schoolchildren that did not demonstrate protection against S paratyphi A. 12 These whole‐cell vaccines have been largely discontinued because of adverse side effects. A S paratyphi A vaccine has been developed by the National Institutes of Health, Bethesda, MD, and is being studied in field trials in Vietnam. It consists of the detoxified O‐specific polysaccharides of S paratyphi A that have been conjugated with tetanus toxoid. It has been shown to be safe and able to elicit IgG antibodies with bactericidal activity in the serum of adults, teenagers, and 2‐ to 4‐year‐old children. 58,59 A phase III clinical trial is planned in China (http://eclipse.nichd.nih.gov/nichd/annualreport/2004/ldmi/biot.htm). The University of Maryland is developing a live attenuated S paratyphi A oral vaccine against paratyphoid fever, which is in preclinical evaluation (http://www.ord.umaryland.edu/tech_com/Portfolio/vaccines/CV2006-034.html).

The ultimate solution to EF is provision of good public health, safe water, and food sanitation. Vaccination is an interim solution, but the problem of EF will remain in many parts of the world and for travelers to such regions. Current typhoid vaccines seem to allow paratyphoid fever to have become an emerging problem in many developing parts of the world. This appears to be the right time to encourage development of a bivalent vaccine that protects against Salmonella typhosa and Salmonella paratyphosa A (and perhaps also B).

Declaration of interests

H. W. has received research funding from the Swiss Serum and Vaccine Institute and from Aventis Pasteur. The authors state that they have no conflicts of interest.


The authors’ work was supported by the Division of Research Affairs, Faculty of Medicine, Chulalongkorn University and the National Center for Genetic Engineering and Biotechnology of Thailand. We thank Khun Domnern Garden for his careful review of this manuscript and valuable style suggestions.


View Abstract