The global disease burden of Salmonella infections in 2017 included 135 900 deaths caused by Salmonella Typhi and Paratyphi and 77 500 deaths caused by invasive non-typhoidal Salmonella, with increasing antimicrobial resistance (AMR) exacerbating morbidity, mortality and costs. The aim of our systematic review and meta-analysis is to estimate the length of hospital stay and associated treatment costs for patients with susceptible and antibiotic-resistant Salmonella Typhi, Paratyphi and non-typhoidal Salmonella infections.

Systematic review and meta-analysis.

We searched EMBASE, Medline/PubMed, Scopus, Hinari and LILACS databases for studies published between 1 January 2005 and 15 May 2024, with no language restrictions.

We included 30 studies that reported the length of hospital stay or treatment costs for patients with susceptible or antibiotic-resistant Salmonella Typhi, Paratyphi and non-typhoidal Salmonella infections. We excluded studies with sample sizes of less than 30 patients, those focused on non-human subjects and those not reporting our outcomes of interest.

Two reviewers independently screened studies and extracted data on the length of hospital stay and associated costs, with monetary values converted to 2019 USD. We aggregated data according to GDP per capita quantiles using a random-effects meta-analysis. We conducted a quality assessment using an adapted Joanna Briggs Institute tool.

Patients with drug-resistant Salmonella infections had longer hospital stays, with an additional 0.5–2.2 days compared with drug-susceptible Salmonella infections. Based on our meta-analysis, the mean hospital stay for typhoidal Salmonella infections was 6.4 days (95% CI 4.9 to 7.8) for drug-susceptible cases and 8.4 days (95% CI 5.1 to 11.7) for resistant cases in the lowest income quartiles. While there were insufficient data to perform a pooled analysis, individual studies inferred that treatment costs for resistant typhoidal Salmonella infections were higher than for susceptible infections, and resistant non-typhoidal Salmonella infections had longer hospital stays and higher costs compared with susceptible infections. Data were scarce from high-Salmonella-burden countries, particularly in sub-Saharan Africa and parts of Asia.

Patients with antibiotic-resistant Salmonella infections experience a greater healthcare burden in terms of hospitalisation length and direct costs compared with those with susceptible infections. We highlight the economic burden of AMR in Salmonella infections and emphasise the need for preventive measures.

Data are available in a public, open access repository. We used R software (V.4.2.3) for all analyses and uploaded the data and code to GitHub (https://github.com/ckim0509/amr_cost).

https://creativecommons.org/licenses/by/3.0/igo/

This is an open access article distributed under the terms of the Creative Commons Attribution IGO License (CC BY 3.0 IGO), which permits use, distribution,and reproduction in any medium, provided the original work is properly cited. In any reproduction of this article there should not be any suggestion that WHO or this article endorse any specific organization or products. The use of the WHO logo is not permitted. This notice should be preserved along with the article’s original URL.

Salmonella spp are gram-negative bacteria that cause foodborne illnesses and are classified into two species: Salmonella enterica and Salmonella bongori. Salmonella enterica subspecies enterica includes both typhoidal and non-typhoidal Salmonella (NTS). Typhoidal Salmonella refers to serovars Typhi and Paratyphi (A, B and C) that cause enteric fever. NTS refers to the other serovars (e.g., Typhimurium and Enteritidis,) that primarily cause gastroenteritis and invasive diseases.1 2

The primary mode of Salmonella transmission is through the faecal-oral route via contaminated food and water. While infections caused by various Salmonella serovars can manifest as mild and self-limiting conditions, they can progress to severe symptoms that may require hospitalisation.3 NTS, in particular, can cause febrile invasive disease (iNTS), which has a high case fatality rate of 20%–28% among children in Africa.4 Enteric fever caused by Salmonella Typhi and Salmonella Paratyphi accounted for 135 900 deaths globally in 2017.5 There were also an estimated 5 35 000 cases of iNTS worldwide in the same year, leading to around 77 500 deaths.6

The increasing antibiotic resistance in Salmonella enterica serovars threatens the effective treatment of Salmonella infections, with high resistance rates reported in various regions.7 In South Asia, overall antimicrobial resistance (AMR) in Salmonella increased from 53% to 77% over 10 years.8 In South America, AMR in Salmonella increased significantly between 2012 and 2021, with resistance rates to key antibiotics such as ciprofloxacin, ampicillin, ampicillin/sulbactam and ceftriaxone. In particular, ciprofloxacin resistance in Salmonella isolates increased from negligible levels in 2012–2013 to up to 60% in 2020–2021.9

Salmonella Typhi isolates are classified as multidrug-resistant (MDR) when they are resistant to chloramphenicol, ampicillin and trimethoprim-sulfamethoxazole and as extensively drug-resistant (XDR) when they meet the MDR criteria and are also resistant to fluoroquinolones and third-generation cephalosporins.10–12 The emerging resistance has rendered first-line antibiotics ineffective for treating Salmonella infections, while resistance to these critical second-line treatments, such as fluoroquinolones and cephalosporins, is also on the rise. In Salmonella Typhi, resistance markers were found in 100% of sequenced isolates from Pakistan and 75%–97% from Nepal, Cambodia and India. For NTS, fluoroquinolone resistance rates varied by serovar, country and source, with up to 13% of Typhimurium isolates from Scotland and 7% from the USA showing resistance.13

With rising AMR in Salmonella, assessing the health and economic burden and planning strategic responses are essential. Patients with drug-resistant Salmonella infections have excess mortality compared with those with susceptible infections.14 15 Additionally, they are associated with severe disease that results in an excess risk of hospitalisation, longer length of hospital stays and increased hospital costs compared with patients with susceptible Salmonella infections.16 17 Although individual studies have explored these outcomes in particular contexts, there has yet to be a comprehensive systematic review that synthesises the global economic burden of drug-resistant Salmonella infections.

Therefore, we conducted a systematic review and meta-analysis to estimate the length of hospital stay and associated treatment costs for patients with susceptible or antibiotic-resistant Salmonella Typhi, Paratyphi and NTS infections. Our specific study questions examine two key areas: (1) morbidity, utility and productivity impacts (Q1–Q2) and (2) hospital and healthcare system costs (Q3–Q6). Specifically, we sought to answer the following questions: (Q1) What is the length of symptomatic infection with a drug-susceptible, mixed or unknown resistant Salmonella infection? (Q2) What is the length of symptomatic infection with a drug-resistant Salmonella infection? (Q3) What is the length of hospitalisation with a drug-susceptible, mixed or unknown resistant Salmonella infection? (Q4) What is the length of hospitalisation with a drug-resistant Salmonella infection? (Q5) What is the per-patient cost of infection with a drug-susceptible, mixed or unknown resistant Salmonella infection? (Q6) What is the per-patient cost of an infection with a drug-resistant Salmonella infection?

The PRISMA flow diagram illustrates the identification, screening, eligibility and inclusion stages of our systematic review (see figure 1). We streamlined to 30 studies in our systematic review and included studies with populations affected by Salmonella Typhi (n=9), Salmonella Typhi and Salmonella Paratyphi (n=11) and NTS (n=10). The studies were conducted across diverse geographical locations, including Australia (n=1), Bangladesh (n=3), China (n=3), India (n=6), Indonesia (n=1), Italy (n=1), Nepal (n=1), New Zealand (n=1), Pakistan (n=8), Spain (n=1), Sweden (n=1), Taiwan (n=2), Turkey (n=2), the UK (n=1), the USA (n=2) and Vietnam (n=2). In some instances, multiple countries were involved in the same study.22 23

The 30 studies included 15 case series studies, 4 cohort studies, 4 cost of illness studies, 2 cross-sectional studies, 4 prevalence studies and 1 randomised controlled trial. The average quality scores were 0.93 for case series studies, 0.83 for cohort studies, 0.91 for cost of illness studies, 0.83 for cross-sectional studies, 1 for prevalence studies and 0.55 for the randomised controlled trial (see table 1).

Our systematic review and meta-analysis of studies estimated the length of hospital stay and associated treatment costs for patients with susceptible or antibiotic-resistant Salmonella Typhi, Paratyphi and NTS infections. Our findings show that patients with drug-resistant Salmonella infections had longer hospital stays, with an additional 0.5–2.2 days compared with drug-susceptible Salmonella infections. Our meta-analysis results show that the mean hospital stay for typhoidal Salmonella infections was 6.4 days (95% CI 4.9 to 7.8) for drug-susceptible cases and 8.4 days (95% CI 5.1 to 11.7) for resistant cases in the lowest income quartiles. Individual studies have suggested that treatment costs for resistant typhoidal Salmonella infections were higher than for susceptible infections and resistant NTS infections had longer hospital stays and higher costs compared with susceptible infections. Duration of fever varied across studies without clear patterns related to resistance.

Increased costs and hospital stays for resistant Salmonella infections are expected, given that resistant infections require more time to clear than susceptible infections, as has been demonstrated for other pathogens.44–46 This could be explained by increased antibiotic treatment failure for resistant Salmonella infections. Duong et al showed that changes to secondary or tertiary antimicrobials from the initial primary treatment were significantly correlated with prolonged hospitalisation.35

Our systematic review found few primary studies from regions with high typhoid fever burdens, including sub-Saharan Africa, parts of Southeast Asia (e.g., Cambodia, Laos, Myanmar), and the remaining Oceanic countries. Typhoid fever remains highly endemic in these settings. Furthermore, iNTS infections are especially prevalent in sub-Saharan Africa. Despite the high disease burden in these regions, our review found limited research quantifying healthcare costs, hospitalisation duration and economic consequences associated with resistant versus susceptible Salmonella infections. Better utilisation of routine healthcare records, including electronic health records where available, will help fill the gaps in data and evidence for Salmonella.

Our study has limitations. First, the timing of antibiotic administration, hospital presentation and prehospitalisation antibiotic use have not been controlled in our primary studies and have varied. This would likely not differ between resistant and non-resistant infections, but this might have contributed to the heterogeneity in our meta-analysis outcomes. Similarly, variability in study design and quality might have reduced the strength of the pooled estimates derived from our meta-analysis. The quality of evidence on length of hospital stay and costs varied across the included studies. Many studies were observational without control groups, limiting the ability to attribute outcomes to antibiotic resistance status. In addition, while we have pooled estimates across countries, healthcare costs can vary substantially across different healthcare systems and settings. We combined estimates to gain a broader overview of the economic burden and used random effects models along with gross domestic product (GDP) per capita quantiles to account for between-study heterogeneity and minimise differences in costs across countries. However, there remains uncertainty around applying cost estimates from one context to another.

Quantifying the impact on healthcare utilisation and costs through this systematic approach is critical for informing economic evaluations and policy decisions regarding interventions to curb Salmonella infections and AMR. The estimates from our study make a valuable contribution to understanding the overall disease and economic burden of AMR Salmonella. Our findings can be incorporated into disease models used for estimating the health and economic impact of interventions (e.g., vaccines, improved diagnostics, and antimicrobial stewardship programmes) on reducing the burden of AMR Salmonella.47 48

Preventive measures such as improved access to safe water, sanitation and hygiene (WASH) are essential to block transmission pathways.49 Further, vaccines mitigate AMR by preventing diseases and reducing antibiotic use associated with infections.50 51 WHO recommends introducing and scaling up typhoid conjugate vaccines (TCV) alongside WASH interventions in typhoid-endemic areas.52 Pakistan, Liberia and Zimbabwe have included TCV in immunisation programmes and campaigns.53 However, TCV introduction to other endemic regions has faced barriers including regulatory processes, development of implementation strategies and competing health priorities.54 Furthermore, vaccines are still lacking for Salmonella Paratyphi and NTS, although progress is being made to develop new vaccines.55 56

Robust data on the health and economic burden of resistant Salmonella strains are essential for evaluating the value of these interventions. Our systematic review addresses this critical evidence gap by providing estimates to quantify the healthcare and economic costs of AMR Salmonella. Furthermore, our methodological approach establishes a framework that can be applied to other high-priority pathogens where the evidence on the health and economic burden of AMR is limited.

In conclusion, based on our systematic review and meta-analysis, we confirm that resistant strains of Salmonella are associated with an increased economic burden in terms of increased hospitalisation costs and length of stay. However, there are remaining gaps in understanding the specific healthcare costs and extended durations of hospitalisation linked to antibiotic resistance in Salmonella infections, especially in high-Salmonella-burden countries, particularly in sub-Saharan Africa and parts of Asia, and this warrants future studies to address these evidence gaps.

Data are available in a public, open access repository. We used R software (V.4.2.3) for all analyses and uploaded the data and code to GitHub (https://github.com/ckim0509/amr_cost).

We thank Ana Belen Ibarz Pavon for her review and valuable feedback. KA is supported by the International Vaccine Institute, Save the Children, Vaccine Impact Modelling Consortium (INV-034281) and the Japan Agency for Medical Research and Development (JP223fa627004).