One (0.8%) of the 121 Salmonella Paratyphi A isolates was MDR; the rest were susceptible to the first-line antimicrobial drugs ampicillin, chloramphenicol, and trimethoprim/sulfamethoxazole. In contrast, 97/480 (20.2%) Salmonella Typhi isolates were MDR (p<0.001 for the difference in occurrence between Salmonella Paratyphi A and Salmonella Typhi). Of note, azithromycin resistance was significantly more frequent (p<0.001) among Salmonella Paratyphi A isolates (14/121 [11.6%]) than among Salmonella Typhi isolates (9/483 [1.9%]). However, we did not observe resistance to any third-generation cephalosporin tested for either pathogen. Almost all Salmonella Paratyphi A (119/121 [98.3%]) and Salmonella Typhi (456/483 [94.4%]) isolates were resistant to nalidixic acid and had intermediate resistance to ciprofloxacin (119/121 [98.3%] Salmonella Paratyphi A and 411/483 [85.1%] Salmonella Typhi isolates) (Table 6).
In an urban slum population in Dhaka, Bangladesh, we identified a moderate incidence of paratyphoid fever (defined as 10-100/100,000 PY) and a high incidence of typhoid fever (defined as >100/100,000 PY) (26). The lack of private toilets and safe drinking water was one of the key risk factors for both diseases. Our findings highlight a high prevalence of both diseases in the population predominantly affecting the same age group and sharing similar risk factors; integrated control and preventive measures might include the use of bivalent vaccines for both typhoid and paratyphoid fevers and improved WASH. We observed a higher burden of MDR in Salmonella Typhi, whereas Salmonella Paratyphi A remained largely susceptible to former first-line drugs but exhibited higher azithromycin resistance. Both pathogens exhibited intermediate resistance to ciprofloxacin.
This study is one of the few detailed comparative epidemiologic investigations on paratyphoid and typhoid fever in Bangladesh undertaken concurrently for both diseases. We found a higher incidence of typhoid fever than paratyphoid fever, consistent with previous studies (1,27,28). The Surveillance for Enteric Fever in Asia Project in 2016-2019 (27) estimated a crude incidence of typhoid fever as 103/100,000 PY and of paratyphoid fever as 16/100,000 PY in Bangladesh. A population-based study under the STRATAA (Strategic Typhoid Alliance across Africa and Asia) consortium also conducted in Dhaka in 2016-2018 (28) showed crude incidences of typhoid as 161/100,000 PY and paratyphoid as 42/100,000 PY. However, during the TCV efficacy study, the incidence of typhoid fever in Nepal (342/100,000 PY) (29) and Bangladesh (213/100,000 PY) (20) was higher than in Malawi (182.7/100,000 PY) (30) in the control group.
Previous studies from Dhaka noted that some environmental factors, such as the preceding rainy season, higher temperature, higher rainfall, or water level of nearby water sources, were associated with a higher incidence of typhoid fever (31,32). Estimating the typhoid prevalence is challenging, given that the factors influencing blood culture positivity can vary widely. Existing studies have shown a strong correlation between enteric fever incidence and poor sanitation and limited access to clean drinking water (33,34) which aligns with our WASH risk factor analysis. Although establishing optimal WASH infrastructure is resource heavy, our findings suggest that public health interventions to improve private toilets and safe drinking water sources should be considered as a near-term investment for tackling both diseases.
The STRATAA study showed the highest reported typhoid incidence in Kathmandu and Dhaka among the 5-9-year age group (28), whereas in this study, we found that children <16 years of age had the highest risk for enteric fever. This finding suggests a bivalent vaccine against both pathogens could be targeted for children <16 years of age. Among children <16 years of age, the highest incidence for both diseases was among children 2-4 years of age, emphasizing that immunization programs with a bivalent vaccine should prioritize infants <2 years of age, as has been done with the implementation of TCV in typhoid-endemic countries (20,29,30).
AMR constitutes a compelling driver for the development of new-generation bivalent vaccines as well as improved WASH interventions. Although first-line antimicrobial drugs remain effective for treating paratyphoid fever in our setting, MDR Salmonella Typhi is highly prevalent. Because the 2 diseases are difficult to distinguish based on clinical features, the selection of appropriate first-line antimicrobial drugs is difficult in many settings (35). Moreover, nalidixic acid resistance and ciprofloxacin intermediate resistance in both pathogens, as well as the substantial level of resistance to azithromycin in Salmonella Paratyphi A, indicate that empiric use of these drugs for treating enteric fever might not be effective in most patients (2,9-11,36). Furthermore, third-generation cephalosporin-resistant Salmonella Typhi has been reported in Pakistan (37). Our study found both organisms to be sensitive to cephalosporin, but Pakistan's rapid cephalosporin resistance spread signals a need for vigilant monitoring and prudent antimicrobial use.
TCVs have been approved for implementation in the Expanded Program of Immunization of Bangladesh. However, TCVs do not provide cross-protection against Salmonella Paratyphi A. After a period of TCV implementation, the prevalence of paratyphoid fever may increase, perhaps because of strain replacement, as was observed in China with the Vi-polysaccharide typhoid vaccine (38). In aggregate, our findings thus strongly support preventing both diseases by devising effective bivalent vaccines and WASH interventions (19,39,40).
The first limitation of our study was its 2-year duration; seasonality analysis would be strengthened with a longer surveillance period. Second, some cases might have been missed because we enrolled a fraction of eligible participants from the 8 healthcare facilities. Third, the available clinical data were categorized according to systems, not on the basis of individual symptoms or signs, which limited our ability to identify distinguishing clinical features. Fourth, the WASH data was based on a simple questionnaire and could have missed important information. Last, our study was limited to an urban setting in Dhaka, so the findings might not be generalizable to rural areas or other geographic locations. Despite all those limitations, the strength of this study was the prospective, comprehensive, and concurrent surveillance of a large study population for both paratyphoid and typhoid fever, together with repeated censuses, which enabled us to analyze the dynamic cohort in a highly mobile population along with closed cohort from the baseline study population.
In conclusion, we found that, although the incidence of Salmonella Typhi in the study area in Dhaka is greater than that of Salmonella Paratyphi A, the effect of Salmonella Paratyphi A is not negligible, especially in children. Vaccination with a bivalent vaccine should be programmatically feasible given the similar age-specific patterns of incidence, and the similarities of WASH factors associated with the risk for each pathogen suggest that simple WASH interventions might be effective against both pathogens.