A systemic review of literatures on human Salmonella enterica serovars in Nigeria (1999-2018)

Introduction: Salmonella infections are endemic in Nigeria. There is lack of reliable data on culture-positive Salmonella with national coverage. This systemic review of literatures was undertaken to aggregate data on culture proven cases of human Salmonellae and to determine the prevailing serotypes for disease burden estimations. Methodology: This involved comprehensive search engines of Pubmed, Google Scholar, Google and Embase for the literatures on culture positive human Salmonellae from 1999-2018. This review documented the prevalence, common Salmonella serotypes. antibiotic resistance and risk factors associated with human Salmonella infections. Results: This study revealed that 21out of 36 States in Nigeria reported Salmonella-associated diseases, spanning the six geopolitical zones. Our study revealed prevalence of 1.9% (2,732/143,756) Salmonella-bacteraemia and 16.3% (1,967/12,081) Salmonella-associated gastroenteritis. Fifty-three 53 Salmonella serotypes were identified. 39 serotypes were associated with Salmonella-bacteraemia and 31 serotypes with Salmonella-gastroenteritis. Salmonella typhi remains the commonest serotype accounting for 85.2% for Salmonella-bacteraemia and 73.1% Salmonella-gastroenteritis. S. typhimurium (3.8%) was mostly implicated invasive non-typhoidal serotype followed S. enteritidis (2.8%) among others. Human Immunodeficiency Virus-infected individuals, malnutrition was among factors predisposing Salmonella infections. Over 60% of the reported Salmonella isolates developed resistance to two or more of 23 antibiotics recorded, mostly ampicillin, cotrimoxazole, tetracycline and amoxicillin. Conclusions: This study revealed 39 Invasive and 31 non-invasive Salmonella serotypes. Ampicillin, cotrimoxazole, amoxicillin-clavulanate and tetracycline are the most frequently reported antibiotics resisted by Salmonella isolates. This antimicrobial resistance exhibited poses a threat to public health. Data generated from this review would serve as a baseline information for future surveillance studies.


Introduction
Salmonella infections are grouped into typhoidal salmonellosis (enteric fever) caused by Salmonella enterica serovars typhi (typhoid fever) and S. paratyphi A, B and C (paratyphoid fever) which are restricted to human host. The second is non-typhoidal Salmonellaassociated infections caused by other Salmonella enterica serovars such as S. typhimurium, S. enteritidis and S. choleraesuis otherwise called non-typhoidal Salmonellae (NTS). Salmonella infections generally result in considerable morbidity and mortality with a significant socioeconomic impact worldwide [1]. The Global Burden of Disease (GBD) through the Institute for Health Metrics and Evaluation estimated enteric fever morbidity at approximately 15.5 million with 154,000 deaths in 2016 [2]. These estimates rely on a modeling approach and heterogeneity of diagnostic tools and resource limitations. Typhoid fever is an acute, life-threatening, febrile illness with an estimated global disease burden ranges between 11 and 21 million cases and approximately 145,000 to 161,000 deaths annually, majority of cases occur in South/South-East Asia, and sub-Saharan Africa [3,4]. Invasive nontyphoidal Salmonella (iNTS) disease continues to be a public health problem with a global estimate between 2.1 and 6.5 million cases annually. The sub-Saharan Africa and other underdeveloped countries, accounting for 25% of the mortality rate [5]. Invasive nontyphoidal Salmonella disease remains a major neglected tropical disease and problem in Africa. The possible reasons had been attributed to host immunity, pathogen virulence and routes of transmission. Invasive nontyphoidal Salmonella disease is caused mainly by Salmonella enterica serovars S. typhimurium and S. enteritidis, but other serovars have been implicated [6]. Non-typhoidal Salmonella-gastroenteritis is understood to acquire from animal reservoirs. The transmission to humans can occur via many routes such as animal wastes, contaminated water, animal food products, poultry, undercooked meat and environment among others [6][7][8] however, little is known about the reservoirs of infection and routes of transmission of iNTS [6]. Definitive diagnosis of Salmonella infections by automated blood culture apparatus such as BacTec or BacT/ALERT usually enhances the sensitivity of the culture that is mostly unaffordable by many developing countries. The conventional culture method adopted by developing nations often comes with low sensitivity [9,10], a similar situation observed over the years in Nigeria. The rate of prolonged illness in patients with Salmonella blood stream infections in high-burden countries is a continuing public health challenge due to treatment failure. Antimicrobial resistance is a growing concern in typhoidal and non-typhoidal Salmonellae in Nigeria. Resistance to ampicillin, chloramphenicol, and trimethoprim-sulfamethoxazole also known as multidrug resistance (MDR) had since been documented [11]. More recently, fluoroquinolone resistance has been reported in parts of Nigeria [12]. The paucity of epidemiological data on number of culture positive Salmonella isolates and prevailing Salmonella serotypes across geographical zones in Nigeria over the years has made it difficult to estimate the true burden of human salmonellosis. In 2017, World Health Organization (WHO) called for country and continent-wide approach in generating more accurate disease data regarding Salmonella bacteraemia and non-invasive Salmonella disease in Asia and sub-Saharan Africa for effective vaccine administration [3]. Unfortunately, the last estimated data on global burden of Salmonella blood stream infections excluded Nigeria [1] due to absence of reliable data with national coverage. Cultured proven cases of Salmonellae remain pivotal determinants to estimate the real burdens of Salmonella-associated diseases. Although, there have been several published reports on Salmonella infections in Nigeria, most of these reports are either serologically dependent or sparse, and have not been aggregated [13,14]. The aim of the current study is to address the knowledge gap of reliable data with national coverage by conducting a systemic review of available reports on culture positive human Salmonella bacteremia and gastroenteritis, and to determine the prevailing Salmonella serotypes and their geographical spread in Nigeria. The information data will help in health decision-making process and program, and serve as a policy guide for prompt interventions.

Methodology
The systematic review protocols (PRISMA-P) 2015 [15] checklist was followed for the review of metaanalysis of original research studies that reported culture positive human Salmonella serotypes from blood and stool samples of patients across 36 states of six geopolitical zones of Nigeria including Federal Capital Territory (FCT) Abuja. The PubMed (Mediline) (http://www.ncbi.nlm.nih.gov/pubmed), the Google search and Google scholar (https://scholar.google.com/) electronic databases were searched. The following search strings were also used which are: "fever" OR "febrile" OR "bacteremia" OR "septicaemia" OR "septicaemias" OR "septicaemic" OR "Salmonella" OR "Salmonellas" OR "Salmonellae" OR "bloodstream infections" OR Gastroenteritis`' OR "bloodstream pathogen" OR "febrile AND (infant)" OR "child" OR "adolescent" OR "adult" OR "patients" OR "human" OR "travel" OR :community" OR "village" OR" incidence" OR" prevalence" OR "hospital" OR "man". The national list of the 36 states of Nigeria was used as the basis for searching. English key words (i.e. MESH terms in PubMed) and English full-length research articles were considered. The search results were limited to publications from 1st Jan 1999 up to 31st December, 2018, because initial searches failed to find any papers prior to 1999. To ensure a comprehensive search of the literature, an independent search on and Embase-Ovid (https://library.maastrichtuniversity.nl/collections/data bases/embase/) database were considered using a similar search string and strategy. Additional publications (original full-length article) were obtained by the manual scanning of the reference list from the retrieved publications on Salmonella infections.

Selection criteria
The full text of the search results of online articles/abstracts were reviewed independently with the aim of including articles that performed culture on human clinical samples (blood and stool) to obtain Salmonella isolates in Nigeria. The full text version either obtained on line or ordered when only the abstract was available online. Relevant articles were retrieved and thoroughly reviewed using inclusion and exclusion criteria set for this study. The reference sections of retrieved publications were reviewed thoroughly in search of further potential articles for inclusion.

Inclusion criteria
Research articles on cross sectional and cohort studies were included if they reported Salmonella isolated from blood and stool culture of human samples, if samples were collected from a named hospital or within a named community, if the symptomatic study subjects were recruited and research was conducted in Nigeria.

Exclusion criteria
Studies were excluded if the Salmonellaassociated diseases reported were not based on blood and stool culture reports, if Salmonella species were not from human samples, while reports based on serological surveys, clinical diagnosis (such as typhoid perforations without culture), ecological correlation, systematic reviews, books and book chapters were also excluded from the studies and research works that were not conducted within Nigeria. Duplicated reported works were verified and excluded using two deduplication options which include Mendeley citation manager and Ovid multifile search.

Validity assessment
Study validity was established by accurately assessing all original research articles for inclusion and exclusion criteria described above, ensuring that only reliable data were used.

Review of the selected literatures
In data extraction and collection process, relevant descriptive and quantitative variables were extracted from each of the selected articles. A standardized template was used for the data extraction in form of a Microsoft Excel 2013 workbook with each column of the database corresponding to one of the fields in the template. Double data extraction and entry was performed to ensure accuracy. The following parameters were used to extract information from the articles used in this study which included name of journal, title of article, publication date, study location (including city and region in Nigeria), study setting, (hospital-based, community based, hospital and community-based), study period, patient age, patient gender and the inclusion criteria. Quantitative data collected included number of potential study participants, subjects enrolled, specimen type (blood and stool), source of sample (human), number positive culture Salmonella isolates, Salmonella at both genus and serotypes level (if stated), risk factors such as infection with malaria or human immunodeficiency virus (HIV) (if known), the antibiotics tested and phenotypic resistance recorded.

Data analysis
Data were cleaned and a descriptive analysis performed. All reported Salmonella isolates and serotypes were directly drawn from the literatures. Duplicate reports were sorted and removed from the final analysis. The sorting was done by grouping all reports according to state and then comparing variables, which included study dates, name and location of the study site, age group of subjects reported, number of blood cultures taken and number of Salmonella isolated across the studies from each state for potential overlaps. Prevalence was calculated in this study as percentage of positive samples for Salmonella in the total samples recorded. Proportion was determined as number of Salmonella bacteraemia or Salmonella gastroenteritis positive culture in each zone divided by total number of Salmonella bacteraemia or Salmonella gastroenteritis in six zones of Nigeria. Relevant studies included prospective hospital-based or community-based culture series for subjects presenting with fever or with no known focus of infection or groups of subjects not selected for possible associated risk factors as anaemia, HIV, malaria and malnutrition. Summary data were extracted from the articles where available for risk factors associated with Salmonella diseases. Resistance phenotype was defined when at least 50% of identified Salmonella isolates developed resistance to test antibiotics. Summary data for risk factors associated with culture-proven human Salmonella infections were extracted directly from the studies where available. Statistically significant measures of association, such as odds ratio (OR), risk ratio (RR) and prevalence were extracted from the studies. The meta-analysis MedCalc version 19.4.1 software was used to pool the individual studies of interest from each geopolitical zone. Heterogenicity was assessed visually using Forrest plot and quantitatively using I² statistic and associated pvalue. Results were considered statistically significant (p-value <0.05) at 95% confidence intervals. The analysis was presented as odd ratio based on the likelihood of the null hypothesis of possible publication bias was ruled out.

Results
The online database search performed on Google search, Google Scholar and PubMed from January 1, 1999 to December 31, 2018, yielded 18,603 articles. These were reviewed manually for relevance, based on the inclusion criteria and 102 articles were found to be eligible. A similar search strategy was used on the Embase database with 40 eligible articles identified. Critical screening revealed 86 articles used for descriptive analysis in this study.           The articles were further reviewed quantitatively with duplicate studies removed leaving only 56 articles. A total 53 articles were finally considered for this study, after exclusion of authors whose articles were reflected in multiple zones (Figure 1). Thirty seven out of the 53 published reports [11, 12, 13, 14, 16, 17-36, 39-45, 47-53] were hospital-based, nine [38,46,[57][58][59][60] were community based and seven [37,[61][62][63][64][65][66] were both community and hospital-based studies.  Table 2). S. typhi is the most common Salmonellabacteraemia (85.2%) and Salmonella gastroenteritis (73.1%) while S. typhimurium remains the most common serotype reported (3.8%) for iNTS disease ( Table 2). Figure 2 shows the map of Nigeria indicating the six geopolitical zones and prevalence of Salmonella infections. The prevalence of Salmonella-bacteraemia was highest in South-South 12.3% followed by South-East 10%, South-West 3.4% and the least was recorded in North-Central 1.4%. South-West recorded 40.3% prevalence for Salmonella-associated gastroenteritis followed by South-South 27.8% while the least 4.2% was found in North-East ( Figure 2).

States in six geopolitical zones reporting Salmonellaassociated diseases in Nigeria
Salmonella-associated diseases were reported in 21 of the 36 states in Nigeria plus Federal Capital Territory (FCT) Abuja. The majority of the reports 23/53 (43.4%) were from South-west (SW) geopolitical zone (including 10 reports from Lagos, 3 from Ekiti, 2 from Ogun and 4 each from Osun and Oyo States). Specifically, 22.6% (12/53) from North-west (NW) (5 reports from Kano, 4 from Kaduna and 3 from Kastina State), 11.3% (6/53) from South-east (SE) (3 reports from Imo, 2 from Ebonyi State and one from Anambra State) and 9.4% (5/53) from North-central (NC) (2 reports from Nasarawa, one each from Niger, Plateau and FCT Abuja).    The South-south (SS) zone recorded 9.4% (5/53) (one report each from Akwa-Ibom, Bayelsa, Delta, Edo and Cross-river State) and 3.8% (2/53) from North-east (NE) (2 reports were from Bauchi State) ( Table 1). The four most frequently reported human associated serotypes are S. typhi, S. paratyphi, S typhimurium and S. enteritidis. However, S. typhi-associated bacteraemia was mostly reported from NW (1,077) and SW (780). The highest number of serotypes, typhi was reported in SW (841) and NC (175) from cases of gastroenteritis. Serotype arizona was confined to SW and dublin to NC and SW zones. Other reported serotypes are listed in ( Table 2).

Trends of reported culture positive human Salmonella infections in Nigeria (1999-2018)
The earliest report on salmonellosis for positive human cultured samples was 1999 in SW [46]. There  (Table 3). There was general fluctuation in the trend of Salmonella gastroenteritis in each of the zones studied, with highest between 2008 Table 3. Geographical distribution of culture positive human Salmonella infections Nigeria (1999-2018).

Region
Year of publication     (Table  3 and Figure 4).

Reported risks associated with Salmonella-diseases in Nigeria
In this study, 48 out of 53 reviewed articles with age-associated Salmonella infections reports were observed. Forty of these articles reported on all age groups and eight articles were mainly on young children of less than 12 years. Over 46% (1,264/2,732) Salmonella bloodstream infections were recorded in children <12 years as reported by five studies [20,28,41,44,64]. About 50% of Salmonella bacteraemia were observed from all age groups as reported by 20 authors [11-12, 16, 18, 19, 22, 31, 32, 34, 36, 40, 41, 43, 45, 50, 51, 61, 63, 65]. Only two authors [53,58] reported mainly on adult population. Similar observations were noted in Salmonella-associated gastroenteritis ( Table  1). One study [28] reported sickle cell anaemia as a risk factor in Salmonella bloodstream infection in children, although this risk factor is not statistically significant (p >0.05) (Tables 1 and 4). Two studies observed an association between HIV and Salmonella-associated diseases [21]. It was reported that HIV infected patients with bacteraemia had significantly lower CD4+ count, while fever complication arising from multiple drug resistant Salmonella infection among HIV patients with bacteraemia was more than those HIV patients without Non-typhoidal Salmonellae were more predominant in HIV patients than non-HIV and were more resistant to common antibiotics. Malnutrition (30) Odds ratio = 2.56, CI = 1. 45-5.22 Bacteremia was highly prevalent among malnourished children. Sickle cell anaemia (SCD) (28) p > 0.05 Bacteremia in SCD confirmed children in three hospitals reportedly caused by a combination of 11 bacteria genera with lesser bacteremia found in non-SCD children, possessing increased resistant to most commonly used antibiotics. Age (25) p > 0.05 Patients under age bracket of 0-9 years reportedly found to be more susceptible to typhoidal Salmonella bacteraemia with 6 (43%) positive cases.
p > 0.05 Highest prevalence was observed among the age group of 21-25 years with 86.96% (60/69) students tested positive. Gender (21,29,31,33,34,54) p > 0.05 All the six authors reported high prevalence of Salmonellae in male patients than female patients with a statistical significant association. bacteraemia (median 28 vs. 88 cells/ml, p=0.01) [21]. Malnutrition was identified as a risk factor by a study [30] in Salmonella bacteraemia and high proportion of this infection occurred more among malnourished children (Table 4). Furthermore, Plasmodium falciparum co-infection with Salmonella bacteraemia was reported by two studies [46,64] with significant association put at p > 0.05.

Reports on Antimicrobial Resistance
More than 80% of the authors reported antimicrobial resistance on the Salmonella isolates. In all, Salmonella isolates were found to develop resistance to 23 different antibiotics (Table 1 and Figure  5). The periods of antibiotic resistance development as reported by different authors is shown in Table 1. For example, in 2012, one author [22] reported six antibiotics resistance (amoxicillin-clavulanate, amoxicillin, ciprofloxacin, ceftriaxone, cefipime and ampicillin) in Salmonella-bacteraemia (Table 1). Generally, at least 60% of the reported isolates developed resistance to one antibiotic or the other. Ampicillin, cotrimoxazole, amoxicillin-clavulanate and tetracycline are the most frequently reported antibiotics resisted by Salmonella isolates. Two authors reported fluoroquinolones antibiotic resistance (ciprofloxacin and perfloxacin) in Salmonella bacteraemia [18,22]. An author reported each colistin resistant Salmonella gastroenteritis [65] and colistin resistant-Salmonella bacteraemia [37]. This study revealed that Salmonellaassociated bacteraemia and Salmonella-associated gastroenteritis mostly developed resistant to three cephalosporin antibiotics: ceftriaxone [22,24,35], cefotaxime [12,33] and cefuroxime [18,45,47,56,60].

Discussion
To the best of knowledge, this is the first systemic review of human Salmonellae and diversity of its serotypes in Nigeria. It was found that 47% of eligible published articles reported culture-proven Salmonellaassociated infections in human from 21 of the 36 states (including FCT, Abuja). Lack of data from the remaining fifteen other states is an indication of gaps in the published articles partly due to a lack of culturebased studies, poor infrastructure to undertake blood culture for definitive diagnosis of Salmonellaassociated diseases and lack of researcher interest to conduct research and publish reports on Salmonellae.
In this review, 53 human Salmonella serotypes were recorded. Four of these serotypes; S. typhi, S. paratyphi, S typhimurium and S. enteritidis were frequently reported. The prevalence of 1.9% and 16.3% cultureproven Salmonella blood-stream and Salmonellaassociated gastroenteritis was recorded respectively. This result clearly showed that typhoidal and nontyphoidal salmonellosis is a public health issue. It is also interesting to note that prevalence of S. typhi, the most common serotype in blood stream infections is 1.6% (2,286/143,756). This serotype accounted for 83.7% Salmonellae bacteraemia and over 70% Salmonella gastroenteritis in proportion. This high prevalence and proportion of S. typhi reported may be due to population explosion, poor health care facilities, poor portable water distribution network, inadequate sewage treatment/ disposal system and low per capital income as identified in previous study [16]. Study from the systemic reviews elsewhere indicated wide variation in S. typhi positivity rates between countries over time, from 0.05% among ambulatory patients in Thailand in 2007 to 24.1% in Cambodia in 2009 and to less than 4% in Kuwait [67] was reported. In this study, nearly half of the included literatures reported data on typhoid fever outcomes, with paratyphoid fever being relatively under-represented. Specifically, 0.06% prevalence and 5.1% proportion of S. paratyphi (A, B and C) in blood stream infections were obtained. However, the sub-national studies from Burkina Faso, Ghana, Guinea-Bissau, Kenya, Madagascar, Mozambique, Tanzania and Senegal, reported the prevalence of S. paratyphi A, B, and C to be less than 0.7%, whereas in Kuwait it was between 0.11% and 0.82% [67]. Specifically, 36 different iNTS serotypes were reported in nine published articles with S. typhimurium being the most common reported serotype (39.8%) associated with iNTS disease mainly from NC, NW and SW zones. Close to this, is S. enteritidis from NC, NW, SW and SS. This study revealed that Salmonella serovars enteritidis, S. choleraesuis, S. dublin, S. arizonae, S. bagney, S. saint paul and S, poona are also associated with invasive diseases. NTS serotypes other than S. typhimurium such as Salmonella serovars S. heidelberg, S. dublin, and S, choleraesuis had been implicated in hospitalized invasive disease in patients more than S. typhimurium [68].
There is variability in occurrence of Salmonellaassociated diseases and proportion of Salmonella pathogens across the states with published articles, based on the summation of reported published data in those states. For instance, SS recorded highest occurrence (12.3%) of Salmonella blood stream infection from four scientific publications [12,[51][52][53] and the SW geo-political zone with highest published articles [11, 16-18, 19, 38, 41, 42-44, 58, 63-66] recorded 3.4% prevalence of Salmonella bacteraemia. Interestingly, highest proportion of Salmonella bacteraemia 46.9% (1,281/2,732) was found in NW zone from eight studies (Table 1). Furthermore, Salmonella gastroenteritis has its peak proportion in SW (48%). There was statistically significant difference (p < 0.05) in the number of published articles and the number of Salmonella isolates recorded between zones. The high heterogeneity in the estimated Salmonella occurrence and proportion in each zone may be attributed to many factors but not limited to; number of published articles, detection method used, and amount of sample processed, available `facilities, accessibility to health facilities and conflict or stability of the zone. In NE for instance, a general death of published articles and very low proportion of Salmonella isolates were noticed. These are likely connected to the security issues and crisis confronting the zone in the last one decade with attendance humanitarian assistance from international communities.
There was general fluctuation in the trend of Salmonella infections over time. It was observed that zone with occurrence of typhoidal Salmonellae and NTS are historically with an increasing trend over time ( Table 2 and Figure 4). These observations could in part, be attributed to decreased access to improved sanitation facilities and quality water, as well as malnutrition and poverty. Malnutrition was identified by a study [30] to have a significant association with increased Salmonella infection.
In this review, 75% of scientific articles evaluated, reported Salmonella infections, to cut across all age groups. However, a high proportion (46%) of Salmonella bacteraemia in children under 12 years of age was found. Statistical significance association of typhoidal salmonellosis with age had been documented and with children at the high risk of salmonellosis [21,24,25,33,54]. HIV infection, malaria parasitaemia coinfection with typhoid [17,21] sickle cell anaemia [28] and gender among others are certain predisposing factors reported to influence occurrence of Salmonellaassociated diseases in this study.
Antimicrobial resistance in Salmonella serotypes was reported by 80% of the published articles in this review. Of which more than 60% of Salmonella isolates developed resistant to one or more of 23 antibiotics reportedly tested and mostly to first line empirical antibiotics. In addition, fluoroquinolone antibiotic (ofloxacin and ciprofloxacin) resistance in Salmonella serotypes were recorded. It has been documented that infections caused by multiple drug resistant strains are more severe than those caused by susceptible strains [69]. In Nigeria, efficacies of the first line empirical antibiotics such as ampicillin, tetracycline, cotrimoxazole, chloramphenicol against Salmonella blood stream infections had been documented to be doubtful and are such no longer effective [16,20]. The observed reported fluoroquinolone resistance is likely to be the consequence of widespread replacement of these traditional first-line antibiotics with fluoroquinolones, which has now become a threat to public health. Multiple drug resistant and/or reduced fluoroquinolone susceptibility had been an issue in Nigeria [16] and other West African countries such as Malawi and Tanzania [6]. Salmonella isolates were found to develop resistant to ceftazidime, cefotaxime, and cefuroxime in this review, an indication that alternative antibiotics to Salmonella infections is becoming doubtful. The complete suspension of ampicillin, tetracycline, chloramphenicol and cotrimoxazole in the treatment of Salmonella-blood stream infections is suggested for now in Nigeria and other developing countries where high burden of selective pressure on these antibiotics is rampant. It is obvious that findings from this systemic review would serve as baselines information on Salmonella blood stream infections. It would resolve the problem of paucity of aggregated national data on culture proven Salmonella cases and might help in the estimation of the true burden of Salmonella blood stream infections in Nigeria. For now, information data on culture proven salmonellosis is unavailable and/ or inaccessible at both State and Federal Ministries. In addition, data on culture positive salmonellosis in most of the referral hospitals are mirage, while data from private hospitals remain a conjecture. This observation is critical for policy makers and stakeholders in the health sector. The need to insist on culture-proven bloodstream infection diagnosis, even when symptomatology and serological diagnosis are inevitable during emergency, should form parts of Health Ministry policy directions. The action taken by some departments in the Ministry of Health in 2011 to establish Nigeria Centre for Disease Control (NCDC) recently became a reality, when the Federal Government of Nigeria signed a Bill for an Act to establish NCDC [70] in December 2018. The decision of the government is an intervention geared towards developing and sustaining a network of reference specialized laboratories, effective surveillance coordination systems and strategic management of diseases of public health importance.
Limitations: There are limitations and gaps in knowledge in this systemic review on the availability and quality of data in the published literatures. The possibility of estimating mortality rate and case fatality ratio due to the dearth of data on Salmonella-associated diseases was not determined. There was also a notable number of variations in testing methodologies in the included published literatures. Presently, blood and stool cultures are the gold standard for diagnosing typhoid and non-typhoidal diseases in most developing countries, thus, the standard of identification techniques and interpretation of results used for Salmonella isolates is another challenge. The source of isolation and variability of strains among the same serovars may influence antibiotic susceptibility test results as observed in this study. Proper coordinated hospital and laboratory-based studies would have added to the quality of the results. New prospective hospital-culture based studies in 36 States in Nigeria would have been more useful.

Conclusions
The study revealed between 1999 to 2018 prevalence of 1.9% and 16.3% culture-proven Salmonella blood-stream infections and gastroenteritis in Nigeria respectively. Typhoidal and non-typhoidal salmonellosis remain a public health challenge. S. typhi is the most common serotype associated with Salmonella infections out of 53 identified human Salmonella serotypes currently circulating in Nigeria, while S. typhimurium remains the most common serotype reported for iNTS disease. Over 46% Salmonella bloodstream infections were recorded in children <12 years. Human Immunodeficiency Virus (HIV)-infected individuals, sickle cell-anaemic individuals, children, malaria-coinfection and malnutrition are identified factors predisposing Salmonella infections. The study revealed that at least 60% of the reported isolates are multiple drug resistance. Fluoroquinolone and cephalosporin resistant isolates were recorded. A comprehensive epidemiological data to estimate the true burdens of Salmonella blood stream infections is paramount and achievable. Lack of researcher interest to conduct research and publish reports on Salmonellae among other gaps identified in this study, could be addressed by examining the health facilities, services, records and administration, in those states with little or no tools for definitive diagnosis of Salmonella-associated diseases. The National Surveillance program should be strengthened to ameliorate imperil caused by invasive non-typhoidal Salmonella disease. The data generated from this review would serve as a baseline information for future surveillance studies. A combined approach strategy that would include improvement in sanitation, water safety and use of vaccine is necessary.