Prevalence, serotype diversity, and antibiotic resistance of non-typhoidal Salmonella in East African poultry: A systematic review and meta-analysis

Abstract

Non-typhoidal salmonellosis, caused by Salmonella enterica, threatens poultry production and food safety in East Africa, where poultry is vital for livelihoods. This systematic review and meta-analysis assessed the pooled prevalence, serotype distribution, and antimicrobial resistance patterns of non-typhoidal Salmonella in East African poultry. Following PRISMA guidelines, we searched on PubMed, Web of Science, Scopus, African Journals Online, and Google Scholar for studies published up to December, 2025. A random-effects model was used to calculate pooled prevalence with 95% confidence intervals in R (Version 4.6.0). Subgroup analyses and I² statistics were used to explore the source of heterogeneity, and a heatmap visualized the antibiotic resistance patterns. Thirteen studies comprising 422 positive isolates were included. The pooled prevalence of non-typhoidal Salmonella in East Africa was 6% (95% CI: 4-8%), with high heterogeneity (I² = 95.3%). The subgroup analysis showed prevalence was highest in Uganda (11%, 95% CI: 7-16%) and lowest in Tanzania (1%, 95% CI: 0-6%). By diagnostic method, slide agglutination was higher (7%, 95% CI: 5-10%) than PCR (4%, 95% CI: 1-8%). By sample type, internal organs had the highest prevalence (17%, 95% CI: 12-21%) and environmental samples had the lowest (1%, 95% CI: 0-3%). By source, slaughterhouses recorded the highest rate (12%, 95% CI: 7-18%) and extensive farms the lowest rate (3%, 95% CI: 0-8%). The pooled antibiotic resistance was 65% (95% CI: 55-79%). Oxytetracycline is the most resistant antibiotic (84%, 95% CI: 70-98%). S. Typhimurium exhibited the highest pooled prevalence (18%, 95% CI: 15-22%), and S. Kentucky was the most multidrug-resistant serotype and had the highest number of resistance genes. In conclusion, non-typhoidal Salmonella is prevalent in East African poultry, with multidrug-resistant serotypes distributed across the region. Therefore, collaborative interventions are urgently needed to address this zoonotic threat.

Background

Non-typhoidal Salmonellosis is a zoonotic disease caused by Gram-negative, broad-host-range serotypes of the genus Salmonella, commonly referred to as non-typhoidal Salmonella (NTS) (Bangera et al., 2019). NTS is a major food safety concern in both developed and developing countries, most commonly associated with acute gastroenteritis, bacteremia, and focal infections in humans and animals (Chousalkar and Willson, 2022). It is the major cause of invasive bacteremia in the world, especially in sub-Saharan Africa (in immunocompromised and malnourished children and the elderly). It can also lead to chronic disorders, such as neoplasms and autoimmune disorders (Feasey et al., 2012; Lamichhane et al., 2024; Sánchez-Vargas et al., 2011; Se Eun et al., 2021; Sima et al., 2024). The 2017 report showed that the invasive form of NTS has caused approximately half a million illnesses and 77,500 deaths worldwide, particularly in sub-Saharan Africa, reaching 78.9% of the total cases and 85.9% of deaths (Crump et al., 2023; Stanaway JD et al., 2019). Zoonotically, NTS is transmitted to humans from poultry, beef, pork, fish, dairy, and dairy products throughout the farm-to-fork chain (Lamichhane et al., 2024). Poultry is the most common reservoir, and its products (eggs and meat) are the most common sources of foodborne non-typhoidal salmonellosis outbreaks in humans (Shaji et al., 2023).

Even though Poultry is the most significant source of protein in the world, East African countries are still not achieving this. Likely due to infectious diseases, lack of genetic improvement, traditional management systems, and a shortage of formulated feed (De Bruyn et al., 2015; Pius et al., 2021; Vernooij et al., 2018). Diseases, especially salmonellosis, are the main constraint for the poultry production industry and an important public health concern (Grace et al., 2024). According to the 2023 European Union One Health report, Salmonellosis is the second most reported human zoonosis, after campylobacteriosis (EFSA, 2024). Salmonellosis is transmitted both horizontally and vertically, causing high mortality among young chickens and leading to significant production losses. It is characterised by asymptomatic infection in adult chickens, non-host-specific colonization, and high zoonotic potential (Shaji et al., 2023; Wang et al., 2018).

In poultry, several Salmonella serotypes are of major concern. They can be divided into two groups based on their host range and public health impact. The first group, Salmonella Gallinarum and Salmonella Pullorum, are avian-specific pathogenic Salmonella serotypes that cause severe systemic infections in poultry, resulting in economic losses due to mortality and production reductions (Hu et al., 2024; Zhou et al., 2022). The second and more critical group is the NTS serotypes; unlike host-specific serotypes, they are zoonotic and significantly impact both poultry production and public health, contributing to food spoilage and food poisoning (Peter et al., 2023). Worldwide, Salmonella Typhimurium and Salmonella Enteritidis are the two common NTS responsible for this dual burden (Arya et al., 2017; Billah and Rahman, 2024). Across the African continent, the prevalence of poultry salmonellosis is approximately 14.4%. Specifically, several zoonotic NTS serotypes are widely reported in Africa, including Salmonella Enteritidis, Salmonella Typhimurium, Salmonella Kentucky, and Salmonella Virchow (Kabeta et al., 2024).

Antibiotic therapy is the first-line treatment for salmonellosis; however, improper antibiotic use is common on poultry farms. As a result, antibiotic resistance is becoming a critical challenge. Antibiotic residues also contaminate chicken products, which are a source of emerging public health diseases (Castro-Vargas et al., 2020; Punchihewage-Don et al., 2022). Research in sub-Saharan Africa indicates that 70% of NTS isolates have developed resistance to ampicillin, trimethoprim-sulfamethoxazole, chloramphenicol, cephalosporin, and fluoroquinolone (Tack et al., 2020). Despite numerous studies in East African countries, no consolidated evidence has been summarized on the overall burden, serotypes, and antibiotic resistance patterns of NTS in poultry across the region. Therefore, this systematic review and meta-analysis study addresses this gap by providing pooled data on prevalence, serotype distribution, and antibiotic resistance of NTS in this region.

Methods

Data sources and search strategy

We conducted the systematic review and meta-analysis following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (supplementary Table 1) as previously described (Shamseer et al., 2015). The outcomes of interest were the pooled prevalence, serotype diversity, and antibiotic resistance in NTS in East African poultry. This was conducted by systematically searching peer-reviewed articles in PubMed, Web of Science, Scopus, Google Scholar, and African Journals Online (AJOL) databases. A search strategy employed the following keywords and Boolean operators: (“Poultry” OR “Chicken” OR “Broiler” OR “Layer”) AND (“Salmonella” OR “Salmonellosis” OR “Salmonella enterica” OR “Nontyphoidal”) AND (“Prevalence” OR “Seroprevalence” OR “Epidemiology” OR “Sero-epidemiology”) AND (“Antibiotic resistance” OR “Antimicrobial resistance” OR “Drug resistance”) AND (“East Africa” OR “Ethiopia” OR “Kenya” OR “Uganda” OR “Tanzania” OR “Rwanda” OR “Sudan” OR “South Sudan” OR “Burundi” OR “Somalia” OR “Eritrea” OR “Djibouti”) covering the period from January 2000 to December 2025.

Inclusion and exclusion criteria

The study included Primary research articles (cross-sectional, cohort, case-control, and surveillance reports) from 11 East African countries (Ethiopia, Kenya, Uganda, Tanzania, Rwanda, Burundi, South Sudan, Somalia, Sudan, Djibouti, Eritrea) reporting prevalence, serotype identification, and antibiotic resistance of NTS in poultry, published between 2000 and 2025. Studies were excluded if they were outside East Africa, lacked serological or molecular diagnosis, serotype distribution, or antibiotic resistance profile, involved non-poultry sources, or were experimental in design. Additionally, we excluded non-English-language publications, as well as reviews, editorials, preprints, theses, government reports, and opinion pieces.

Study selection and data extraction

Two reviewers independently conducted a manual screening of the search results based on article titles and abstracts. Whenever disagreements arose regarding study relevance, a third reviewer intervened to reach a final decision. Studies that passed this initial screening proceeded to a full-text review. During that review, the complete articles were obtained to confirm they fully met the inclusion criteria. Once confirmed, we assessed for risk of bias, extracted key information, and organized it into a Microsoft Excel sheet. The collected data included authors, title, publication year, study area, study design, sample size, management systems involved, sample sources, diagnostic methods, prevalence rates, identified serotypes, antibiotic resistance patterns, any related resistance genes, and the study's main conclusions.

Quality assessment of the selected studies

The quality, reporting, and risk of bias of the selected papers were assessed by two independent authors using the appraisal tool for cross-sectional studies (AXIS tool) (Downes et al., 2016), shown in supplementary Table 2. This appraisal checklist comprises 20 items that evaluate the title, abstract, introduction, methods, results, and discussion sections. It covers study objectives, methodological aspects, results, limitations, and funding declarations.

Data management and statistical analysis

All data extracted from the included studies were managed and organized using Microsoft Excel. Statistical analyses, including pooled prevalence, serotype distribution, antibiotic resistance patterns, and subgroup analyses, were conducted in R software (version 4.6.0) using the meta and metafor packages. The pooled prevalence was calculated using a random effects model with the Restricted Maximum Likelihood (REML) method. Forest plots were generated using the metaprop function in R, with raw proportions (Balduzzi et al., 2019). Heterogeneity among studies was assessed using Cochran’s Q test, with a p-value < 0.05 indicating significant heterogeneity. To identify potential sources of heterogeneity, subgroup meta-analyses were conducted based on study area, sample source, sample type, and diagnostic method. The inverse -variance index (I²) and true -variance index (τ²) were estimated to quantify heterogeneity, with significance set at p < 0.05. The heatmap analysis was used to illustrate the relationship between the serotypes and their antibiotic resistance rates. Publication bias was assessed qualitatively using funnel plots (effect size versus standard error) and statistically using a regression test and a rank correlation test. All results were reported with 95% confidence intervals, and a p-value < 0.05 was considered statistically significant.

Results

Literature search and PRISMA flow diagram

The selection process was conducted in accordance with the PRISMA 2020 guidelines (supplementary Table 1). From the initial identification of 562 articles, 58 duplicates were excluded. After screening titles and abstracts, 47 reports were selected for full-text evaluation. Following rigorous review, 34 articles were excluded for failing to meet the inclusion criteria. As a result, 13 articles were included in the final analysis (Figure 1).

Fig. 1.

PRISMA flowchart for the selection of studies. The prevalence, serotype diversity, and antibiotic resistance of poultry NTS in East Africa.

Study characteristics and quality assessment

A total of 13 studies published between 2003 and 2024 met the inclusion criteria, including those from Ethiopia (n = 6), Uganda (n = 2), Tanzania (n = 2), Kenya (n = 1), Sudan (n = 1), and South Sudan (n = 1). No studies were identified from Rwanda, Burundi, Somalia, Djibouti, or Eritrea. All included studies were cross-sectional in design and were conducted across various countries in East Africa, with sample sizes ranging from 208 to 1515. The characteristics of the studies included in this systematic review and meta-analysis are summarized in Table 1.

Table 1.

Descriptive statistics of the studies included in the systematic review and meta-analysis.

Author, year (Ref)	Study period	Study Area	Study design	Sample Source	Sample type	Diagnostic technique	Sample size
(Eguale, 2018)	July 2013–January 2014	Ethiopia	cross-sectional	Intensive farm	Fecal droppings	Slide agglutination	549
(Waktole et al., 2024)	January 2021 - June 2022	Ethiopia	cross-sectional	Intensive farm	Cloacal swabs, faecal droppings, litter samples, and feed and drinking water	PCR	1515
(Dagnew et al., 2020)	November 2017 -May 2018	Ethiopia	cross-sectional	Intensive farm	Cloacal swabs, faecal droppings, feed samples, and floor swabs	Slide agglutination	836
(Zewdu and Cornelius, 2009)	September 2003- February 2004	Ethiopia	cross-sectional	Supermarkets	Chicken meat	Slide agglutination	208
(Molla et al., 2003)	October 2001 - April 2002	Ethiopia	cross-sectional	Slaughter house	Liver, gizzard, heart, muscle, and skin	Slide agglutination	378
(Abayneh et al., 2023)	December 2020 - May 2021	Ethiopia	cross-sectional	Intensive farm	Cloacal swab and faecal dropping	Slide agglutination	390
(Rukambile et al., 2021)	September 2017 - February 2018	Tanzania	cross-sectional	Extensive farm	Fecal dropping	PCR	810
(Thye et al., 2025)	April 2019 - June 2020	Tanzania	cross-sectional	Intensive farm	Fecal dropping	Slide agglutination	777
(Odoch et al., 2018)	June 2015 - August 2016	Uganda	cross-sectional	Intensive farm	Fecal droppings	PCR	796
(Ball et al., 2020)	March 2017 -August 2018	Uganda	cross-sectional	Intensive farm	Cloacal swab	Slide agglutination	379
(Saad et al., 2022)	September 2019- December 2019.	South Sudan	cross-sectional	Extensive farm	Cloacal swab	Slide agglutination	270
(Hag Elsafi et al., 2022)	October 2004 - May 2005.	Sudan	cross-sectional	Slaughter house	Livers, intestinal contents, skin swabs and egg specimens	Slide agglutination	679
(Nyabundi et al., 2017)	December 2013 - October 2014	Kenya	cross-sectional	Local market	Cloacal swabs and eggs	PCR	408

The evaluation of the selected studies used the AXIS tool, which comprises 20 questions to assess quality, reporting, and risk of bias. Among 13 studies, total scores ranged from 12 to 17. One study scored 17, five scored 16, three scored 14, three scored 13, and one scored 12. Consequently, no studies fulfilled all 20 criteria (supplementary Table 2).

Pooled prevalence of NTS

The pooled prevalence of NTS was calculated from 13 studies across 6 of the 11 East African countries, comprising a total sample size of 7995. The estimated pooled effect size was 6% (95% CI: 4-8%). The analysis revealed significant heterogeneity across the studies (Q (12) = 253.29; p < 0.001; I² = 95.3%; τ² = 0.0009). Despite this heterogeneity, weights under the random-effects model were evenly distributed (5.4-8.7%), as shown in Fig. 2.

Fig. 2.

Forest plot of the pooled prevalence of NTS in East African poultry.

Subgroup meta-analysis and sources of heterogeneity

The results of the subgroup meta-analysis across different moderators are summarized in Table 2, and the corresponding forest plots are provided in supplementary Figs. 1, 2, 3, and 4. Subgroup analysis revealed substantial variability across categories. Based on study region, Uganda had the highest pooled prevalence of 11% (95% CI: 7-16%) with substantial heterogeneity (I² =74.5%; τ² = 0.0009; p = 0.0475), while Tanzania reported the lowest pooled prevalence at 1% (95% CI: 0–6%) with high heterogeneity (I² = 82.6%; τ² = 0.0009; p = 0.0167) (supplementary Figure 1). For diagnostic methods, the pooled estimate for slide agglutination was higher at 7% (95% CI: 5-10%) with the greatest heterogeneity (I² = 95.9%; τ² = 0.0012; p < 0.0001) compared to PCR, which was 4% (95% CI: 1-8%) with heterogeneity (I² =94.7%; τ² =0.0012; p < 0.0001) (supplementary Figure 2). Sample types were categorized into five main groups: intestinal content (faecal drops and cloacal swabs), environmental (feed, water, litter, and floor swabs), egg, meat (muscle and skin), and internal organs (liver, heart, and gizzard). Internal organs exhibited the highest estimated pooled prevalence at 17% (95% CI: 12-21%) with heterogeneity (I² = 93.6%; τ² = 0.0006; p < 0.0001), while environmental types recorded the lowest pooled prevalence at 1% (95% CI: 0-3%) with the lowest heterogeneity (I²=28.6%; τ²=0.0006; p = 0.2309) (supplementary Figure 3). The analysis based on the sample source revealed that slaughterhouses recorded the highest pooled prevalence of 12% (95% CI: 7-18%) with heterogeneity (I² = 97.9; τ² = 0.0009; p < 0.0001), while sample from extensive farms were recorded the lowest pooled estimate at 3% (95% CI: 0-8%) without heterogeneity (I² = 0: τ ² = 0.0009; p = 0.3581) (supplementary Figure 4).

Table 2.

Pooled prevalence of NTS, stratified by subgroup analysis.

Moderators	K	Category	N	Case	Pooled Prevalence (95% CI)	Heterogeneity			Test for subgroup difference (CE)
						I2 (%)	τ2	P-value	Q	P-value
Pooled prevalence	13	Overall	7995	422	6% (95% CI: 4-8%)	95.3	0.0009	<0.001	253.29	<0.0001
Study area	6	Ethiopia	5901	205	7% (95% CI: 5-10%)	95.8	0.0009	<0.001	123.68	<0.0001
	2	Uganda	1175	126	11% (95% CI: 7-16%)	74.5	0.0009	0.0475
	2	Tanzania	1587	24	1% (95% CI: 0-8%)	82.6	0.0009	0.0167
	1	Kenya	408	20	5% (95% CI: 3-7%)	-	-	-
	1	South Sudan	270	8	3% (95% CI: 2-6%)	-	-	-
	1	Sudan	679	38	6% (95% CI: 2-8%)	-	-	-
Diagnostic Method	4	PCR	3529	138	4% (95% CI: 1-8%)	94.7	0.0012	<0.0001	1.6	=0.2058
	9	Slide agglutination	4466	284	7% (95% CI: 5-19%)	95.9	0.0012	<0.0001
Sample type	11	Intestinal content	5991	269	5 (95% CI: 3-6%)	93.1	0.0006	<0.0001	100.18	<0.0001
	4	Internal Organ	269	61	17 (95% CI: 12-21%)	93.6	0.0006	<0.0001
	2	Egg	316	20	6 (95% CI: 2-11%)	0	0.0006	0.7223
	2	Environmental	682	5	1 (95% CI: 0-3%)	28.6	0.0006	0.2309
	3	Meat	616	61	9 (95% CI: 6-12%)	80.3	0.0006	0.0016
Sample source	7	Intensive farm	5242	228	5 (95% CI: 3-7%)	95.4	0.0009	<0.0001	75.57	<0.0001
	2	Extensive Farm	1080	27	3 (95% CI: 0-7%)	0	0.009	0.3581
	2	Slaughterhouse	1057	118	12 (95% CI: 8-17%)	97.9	0.0009	<0.0001
	1	Local Market	408	20	5 (95% CI: 3-7%)	-	-	-
	1	Supermarket	208	29	14 (95% CI: 10-19%)	-	-	-

K=number of included studies, N=total number of poultry population, Case=positive for non-typhoidal Salmonella, CE=common effect.

Fig. 3.

Forest plot of serotype-specific proportions from meta-analysis.

Fig. 4.

Heat map presentation of the geographical distribution of NTS in East Africa. The rows represent a list of isolated serotypes (n), and columns represent countries where serotypes are distributed. The burden of distribution is represented by a white-to-deep-red sequential colour intensity (0% to 100%).

Prevalence of individual serotypes

Across the 13 studies, a total of 422 NTS were identified, representing 34 distinct serotypes. Among these, S. Typhimurium exhibited the highest pooled prevalence at 18% (95% CI: 15-22%), followed by S. Braenderup at 11% (95% CI: 9-15%), S. Enteritidis at 8% (95% CI: 6-11%), S. Kentucky at 8% (95% CI: 5-11%), S. Newport and S. Saintpaul each at 6% (95% CI: 4-9%). Together, these top 6 serotypes accounted for the majority of the NTS burden in East Africa. In contrast, S. Havana, S. Barranquilla, S. Haardt, S. Tarshyne, S. Millesi, S. Montevideo, and S. Kiambu showed significantly lower pooled prevalence, each below 1%. The remaining serotypes were detected sporadically, with wide confidence intervals indicating high variability across studies. Despite this heterogeneity, weights under the random-effects model were evenly distributed (ranging from 2.3% to 3.1%), as shown in Fig. 3.

Geographical distribution of serotypes

Of the 422 NTS serotype isolates, 48.6% were from Ethiopia, followed by 29.9% from Uganda, with the remaining 21.5% from Kenya, Tanzania, Sudan, and South Sudan. S. Typhimurium, S. Enteritidis, and S. Saintpaul were the most common serotypes in Ethiopia, while S. Enteritidis and S. Newport were highly prevalent in Uganda. S. Typhimurium was found across all six countries, whereas S. Enteritidis was detected in every country except Sudan. In third place, S. Kentucky was identified in Ethiopia, Uganda, Tanzania, and Sudan. S. Hadar appeared in three countries (Ethiopia, Uganda, Tanzania). Four serotypes, such as S. Newport, S. Virchow, S. Aberdeen, and S. Uganda, were present in two countries, while the other 25 serotypes were country-specific, as shown in Fig. 4.

Pooled antibiotic resistance prevalence

A total of 13 articles were included in the meta-analysis to estimate the antibiotic resistance of poultry-derived NTS in East Africa. Fig. 5 indicates that the pooled antibiotic resistance was 67% (95% CI: 55-79%). The analysis showed substantial heterogeneity among the studies (I² = 94.4%; Q test = 4050.42; df = 12; p < 0.001). Despite significant heterogeneity between studies, they were weighted roughly equally, with individual study weights ranging from 4.8% to 8.9%. Three studies from Ethiopia, including Eguale, 2018, Waktole et al., 2024, and Dagnew et al., 2020, reported 100% resistance to a given antibiotic.

Fig. 5.

A forest plot of pooled antibiotic resistance prevalence across selected studies.

Proportion of antibiotic resistance

A random-effect meta-analysis was conducted to estimate resistance proportion across 35 antibiotic agents. While 13 studies were included in the overall pooled prevalence of antibiotic resistance, one study was excluded from the drug-specific analysis because it lacked drug-specific reporting. Substantial heterogeneity was observed across all antibiotic agents (I² =97.5; τ² =0.00053; p < 0.0001). Resistance proportions varied widely, ranging from 0% to 84%. oxytetracycline is the most resistant antibiotic (84%, 95% CI: 70-98%), followed by sulfisoxazole (54%, 95% CI: 47-60%), spectinomycin (46%, 95% CI: 37-56%), streptomycin (37%, 95% CI: 31-42%), and tetracycline (37%, 95% CI: 32-42%). All the isolated serotypes were 100% susceptible to ceftiofur, cefoperazone, cefotaxime, apramycin, meropenem, enrofloxacin, ceftriaxone, and ceftazidime, as shown in Fig, 6.

Fig. 6.

A forest plot of antibiotic resistance proportions.

Antibiotic-serotype resistance patterns

The integration of 34 NTS serotypes with the commonly used 11-antibiotic panel in a heatmap (Fig. 7) identifies distinct phenotypic clusters. Of the 34 serotypes, 38% were resistant to at least one of the selected antibiotics, with a resistance rate greater than 50%. S. Kentucky is >50% resistant to five different antibiotics, such as nalidixic acid, streptomycin, sulfisoxazole, tetracycline, and ciprofloxacin, whereas S. Saintpaul is highly resistant to streptomycin, sulfisoxazole, and chloramphenicol. S. Typhimurium has high resistance to ampicillin only, but moderate resistance to amoxicillin and clavulanic acid, chloramphenicol, sulfisoxazole, and tetracycline. Despite their low prevalence, S. Mbandaka and S. Havana exhibited an extreme resistance profile (100%) to 3 different antibiotics individually. Resistance to at least one antibiotic was observed in 24 of the 34 isolated serotypes. All highly prevalent and multidrug-resistant serotypes were still susceptible to ceftriaxone.

Fig. 7.

Heat map visualization of serotype-specific resistance to commonly used antibiotics in NTS serovars. Rows represent individual serotypes, columns represent antibiotics. A white- to-deep red sequential colour scale indicates the percentage of resistance from 0% to 100%. The grey colour indicates serotypes that are not detected by the respective antibiotics.

Antibiotic resistance genes

Among the 13 studies reviewed, only 4 reported specific antibiotic resistance genes (ARGs). Overall, 41.1% (60 out of 145) of the tested isolates carried at least one ARG. S. Kentucky exhibited the highest number of resistance genes, with tetA being the most frequently detected, as shown in Fig. 8.

Fig. 8.

Heat map presentation of serotype-specific abundance of antibiotic resistance genes. Rows represent the number of detected isolates and the percentage of ARGs, and columns represent a list of ARGs. White colour indicates absence of ARGs, red indicates the presence of ARGs, and grey indicates non-detected serotypes.

Publication bias assessment

Publication bias and small-study effects were assessed using a logit-transformed funnel plot, a regression test and a rank correlation test (Fig. 9). Although the funnel plot suggested asymmetry in the slope distribution, neither test indicated significant publication bias. The mixed effects meta-regressions test was not significant (z = 1.1290; p = 0.2589; b = 0.0228; 95% CI: 0.0514, 0.0970), and the rank correlation test similarly showed no evidence of bias (τ² =−0/0769, P = 0.7650).

Fig. 9.

Funnel plot showing publication bias and small study effects.

Discussion

This systematic review and meta-analysis provide the first consolidated report on the prevalence, serotype distribution and antibiotic resistance patterns of NTS in East African poultry. Based on 13 studies with a total population of 7995 and 422 positive samples, the estimated pooled prevalence was 6% (95% CI: 4-8%). This finding is lower than the continental African estimate of 14.4% (Kabeta et al., 2024) and reports from Vietnam (45.6%) (Trung et al., 2017), Egypt (33.3%) (Barac et al., 2024), the USA (17.9%) (Diaz et al., 2022), and Iran (14.64%) (Bangera et al., 2019). However, these studies report higher rates than those in Brazil (5,3%) (Voss-Rech et al., 2015), Chile (<4%) (Alegria-Moran et al., 2017), and Ghana (4.69%) (Sarkodie-Addo et al., 2025). The variation in prevalence across regions may reflect differences in production systems, sampling technique, diagnostic method, or true epidemiological differences in NTS distribution.

A subgroup meta-analysis based on the region indicated that the highest estimated pooled NTS prevalence in East Africa was recorded in Uganda at 11%. In contrast, this figure is much lower than the prevalence reported in Nigeria (31.6%) and Cameroon (71.9%) (Kabeta et al., 2024). The higher prevalence of NTS in Nigeria may be associated with the larger poultry population and inadequate biosecurity protocols (Sanni et al., 2022). The pooled prevalence of NTS in Ethiopia identified in this study is lower than the previously reported figure of 12.46% (Basazinew et al., 2025). This discrepancy may show the earlier report did not differentiate between serotypes, potentially including avian-specific serotypes in its overall prevalence estimate. Based on the diagnostic technique, our analysis revealed that the slide agglutination technique yielded a higher pooled prevalence of NTS (7%) than the PCR-detection method (4%), reflecting its high sensitivity for serotype- specific antigen detection in the included studies.

Subgroup analysis by sample type showed that internal organs, including the heart, gizzard, and liver had a higher pooled prevalence of NTS (17%). This study reflects a similar pattern observed in Egypt. Although a cross-sectional study in Egypt reported the highest contamination rate of 30% (Barac et al., 2024), both studies highlight the gizzard and liver as high-risk tissues within the poultry carcass. The isolated number of NTS from eggs was 6%, which is notably higher than the rates of 2.3% reported in Korea (Jung and Lee, 2024), 2.2% reported in Southwest Benin (Hinson et al., 2026), and 5.5% reported in Ghana (Sarkodie-Addo et al., 2025). The high prevalence of NTS in eggs from East Africa might be associated with improper vaccination of layers, high stocking density, faecal contamination, and poor egg storage conditions.

The pooled prevalence from intestinal contents, such as cloacal swabs and faecal droppings, was estimated at 5%, which was lower than the prevalence from internal organs. Similarly, studies from Egypt showed that the prevalence in internal organs was almost double that of cloacal swabs (Salem et al., 2025). A 22% prevalence was also detected in studies from India (Sudhanthirakodi et al., 2016). In the present subgroup analysis, environmental samples (feed, water, and floor swabs) recorded the lowest estimated pooled prevalence at 1%. This finding significantly deviates from recorded data in the USA, where the environmental prevalence was reported at 29,5%, higher than that detected in live birds and products (Diaz et al., 2022), and in Nepal, where pooled prevalence from environmental samples was 9% (Sharma et al., 2021). This difference is likely due to poultry in East Africa being predominantly raised under extensive production systems, in which the environment may not be conducive for the survival and multiplication of Salmonella.

The estimated prevalence of NTS from poultry meat in this study ranked second (9%), just after internal organ records, but remains far lower than the 35.03 % reported in Israel (Malkiely et al., 2025). This underscores why poultry meat is a common source of NTS and poses a serious public health risk. Subgroup analysis by sample source revealed that samples from supermarkets had a 14% positivity rate, higher than those detected in slaughterhouses (9%) and local markets (5%). These higher-prevalence reports from supermarket samples revealed that chickens might be sourced from outside formal slaughterhouses, with contamination occurring during transport and storage in East Africa. In contrast, it is much lower than the report of Mexico’s slaughterhouses at 27.2% (Regalado-Pineda et al., 2020) and Vietnam at 42.8% (Xin et al., 2025). In the USA, 7.67% of retail chicken samples were reported as positive for NTS (Sodagari et al., 2024). The prevalence of intensive farms was 5%, higher than that of extensive farms (3%). These results showed that intensive farms were favorable environments for the horizontal transmission of NTS.

In the current study, the heat map visualization showed that 34 NTS serotypes were distributed across East Africa. Among these, S. Typhimurium, S. Braenderup, S. Enteritidis, S. Kentucky, S. Newport, and S. Saintpaul were the most dominant serotypes, accounting for around 57% of the total number of isolated serotypes. Specifically, S. Typhimurium shared 19% of the pooled prevalence. The current data align with a meta-analysis conducted in African poultry, which reported S. Typhimurium and S. Enteritidis as the dominant serotypes (Kabeta et al., 2024). Likewise, S. Typhimurium and S. Chester (19.05% for each serotype) in Morocco (Zahli et al., 2022). Similarly, S. Typhimurium was the dominant serotype in India at 58% and in Uganda at 36.7% (Sarkodie-Addo et al., 2025), but at higher rates than in the current study (Grakh et al., 2026). Additionally, a study from a poultry breeding farm in China indicated that S. Typhimurium is the second most abundant serotype (29%), next to S. Thompson at 38% (Ju et al., 2023). In contrast, studies from the USA reported a different serotype distribution, with S. Kentucky (35%), S. Infantis (19%), S. Enteritidis (15%), and S. Typhimurium (13.8%) (Sodagari et al., 2024), and from Iran, S. Infantis (43.1%) (Bangera et al., 2019). In Vietnam, S. Typhimurium was the predominant serotype, but at a relatively lower rate (10.1%) (Nhung et al., 2024). Across specific regions in East Africa, among 34 identified strains, the majority (73.5%) showed high geographic specificity; meanwhile, only 26.5% of serotypes were regionally non-specific, detected from two or more study areas. This 26.5% might have arisen from the broad host range and adaptability, regional trade pathways, or poor biosecurity measures across borders. The other scenario for regional specificity may be due to limited research being done in most regions of East Africa, environmental factors, and the narrow host range of the strains.

The meta-analysis of pooled antibiotic resistance prevalence of NTS in the current study revealed 65%, with heterogeneity (94.4%) across the included studies. This data aligns with the findings of a systematic review and meta-analysis conducted in the USA, which reported that the prevalence of antibiotic resistance to at least one antibiotic was 64.6% (Yin et al., 2021). These similarities may reflect comparable patterns of antibiotic usage, particularly the routine use of antibiotics as growth promoters. In contrast, the current results are significantly higher than the pooled resistance reported from the Middle East and North Africa (14.4%) (Bellil et al., 2023). This discrepancy could be attributed to regional variations in antibiotic stewardship, including stricter regulations on antibiotic use on poultry farms and better access to alternative interventions, such as vaccines and probiotics. Furthermore, the pooled prevalence of specific antibiotic agents in the current study showed the highest pooled resistance was recorded for oxytetracycline (84%), followed by sulfisoxazole (54%), spectinomycin (46%), streptomycin (37%), and tetracycline (37%). Similarly, a meta-analysis report from Ethiopia showed that the highest pooled resistance rates were observed for tetracycline (75%) and oxytetracycline (64%) (Basazinew et al., 2025). A study in China showed that resistance to nalidixic acid is 72.3%, followed by ampicillin (55.3%) and streptomycin (48.7%) (Yang et al., 2020). This finding showed that the wide availability of tetracycline, aminoglycosides, and sulfonamides in East African markets often leads to frequent use for both therapeutic and prophylactic purposes. Additionally, an inadequate number of veterinarians in the region leads to the use of antibiotics without prescription, a factor that likely drives the significant resistance patterns identified in this study. By contrast, resistance to third-generation cephalosporins (ceftriaxone, ceftiofur) and carbapenems (meropenem) was notably absent, likely because the high cost of these drugs limits their use in poultry production. This finding offers a window of opportunity: preserving susceptibility to these critically important antibiotics should be a priority.

The Heatmap analysis of serotype-antibiotic resistance patterns revealed distinct clustering among the isolated serotypes based on susceptibility to 11 commonly used antibiotics. Of the 34 serotypes, 38% were resistant to at least one of the selected antibiotics, with a resistance rate greater than 50%. Among these, S. Kentucky was recorded as resistant to five different antibiotics, all at rates exceeding 50%. This finding is in line with the surveillance in France, England and Wales, Denmark, and the United States, which identified S. Kentucky as MDR and alarmingly increased resistance to ciprofloxacin (Le Hello et al., 2011). Additionally, a highly multidrug-resistant strain of S. Kentucky sequence type (ST) 198 emerged in North Africa (Regalado-Pineda et al., 2020). In China, this strain carried even higher antibiotic resistance genes (Ju et al., 2023). Although all MDR strains in the present study remained susceptible to ceftriaxone, the emergence of S. Kentucky as a multidrug-resistant serotype in East African poultry warrants close monitoring. Additionally, S. Kentucky exhibited the highest number of resistance genes, with tetA being the most frequently detected.

This study has some limitations, including high heterogeneity (95.3%) across studies, which reflects substantial variability in sampling strategies, diagnostic methods, and study populations. It included only 6 of the 11 East African countries; no eligible studies were found for Rwanda, Burundi, Eritrea, Somalia, and Djibouti, and the review protocol was not prepared and registered in PROSPERO before commencement.

Conclusion and recommendations

In conclusion, the pooled prevalence, serotype distribution, and antibiotic resistance patterns indicate that NTS from poultry sources pose a major threat to public and poultry production. High resistance to commonly used antibiotics represents a significant risk to humans through the direct and indirect consumption of contaminated poultry and poultry products. Therefore, policymakers and governments must apply One Health strategies for prevention and control. Implementation of a robust antibiotic stewardship program, adoption of One Health approaches, enforcing stricter policies on agricultural antibiotic use, and launching public awareness campaigns on safe handling and cooking of poultry and poultry products.

List of abbreviations

ARG: antibiotic resistance gene, CI: confidence interval, df: degree of freedom, MDR: multidrug-resistant, NTS: non-typhoidal Salmonella, PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

All authors have read and approved the final manuscript.

Availability of data and materials

All data generated or analyzed in this study are available in the main text and supplementary files.

Competing interests

The authors declared no competing interests.

Author contributions

Dinku Yigezaw Mebratie conceived and designed the study, developed the search strategy, performed literature searches, screened records, extracted data, assessed study quality, conducted the meta-analysis, interpreted the data, and drafted the manuscript. Xinyu Wang, Weiqi Guo, Beibei Zhang, and Chang Liu participated in literature searching, screening, and data extraction, and contributed to data interpretation. Lei Deng, Jiangang Hu, Jingjing Qi, Yanqing Bao, Mingxing Tian, and Haileyesus Dejene performed quality assessment, statistical analysis, and meta-analysis. Shaohui Wang conceptualized the study, designed the methodology, interpreted the data, supervised the project, and acquired the funding. All authors critically revised the manuscript for important intellectual content and approved the final version.

Disclosures

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.