Bacterial bloodstream infections (BSIs) represent a major clinical challenge for both adult and pediatric hospital admissions. This underscores the critical need to consider BSI in the differential diagnosis, especially for febrile patients. In Ethiopia, the pooled prevalence of culture-confirmed pediatric BSIs is 30.66% (95% CI: 27.18–34.15%), reflecting a substantial public health concern. The analysis of 23 studies revealed considerable heterogeneity (I² = 91.3%, p < 0.001). In contrast, a separate meta-analysis focused exclusively on neonates hospitalized with clinical sepsis, which included nine studies, reported a higher pooled prevalence of bacteremia at 40. % (95% CI: 34.0–46.0%). This study also exhibited substantial heterogeneity (I² = 92.19%, p < 0.01) [37].
Compared to previous global and regional studies, Ethiopia’s current pooled prevalence of pediatric BSIs is much higher. The review of Community-Acquired BSIs in Africa found only 8.2% in children [38], whereas Community-Onset BSIs in Africa and Asia had a median frequency of 12.5% [39]. Similarly, the Pediatric BSIs in LMICs analysis found a positive rate of 19.1%, which is still significantly lower than Ethiopia’s result [40]. In comparison, in Africa, the overall frequency of healthcare-associated infections was 12.76%, with BSIs accounting for only 17.07% [40]. The higher prevalence of pediatric BSIs in Ethiopia compared to global and regional studies may be due to limited healthcare access, poor infection prevention, and diagnostic challenges. Additionally, high antimicrobial resistance especially among Klebsiella and E. coli reflects gaps in antibiotic stewardship and infection control, exacerbating the burden in the Ethiopian setting.
Reviews on antimicrobial resistance, such as those on carbapenem resistant Gram-Negative Bacteria in African neonates and carbapenem resistant Enterobacteriaceae in West Africa, consistently identified K. pneumoniae and E. coli as leading resistant pathogens, mirroring findings in Ethiopia where Klebsiella spp. was the most frequently isolated organism (30.6%) with a high MDR rate (72.2%). Lastly, the impact of antibiotic resistant BSIs in LMICs meta-analysis revealed increased mortality, prolonged hospital stays, and elevated healthcare costs associated with resistant infections, reinforcing the urgency of the current study’s call for enhanced infection prevention, diagnostics, and antimicrobial stewardship in Ethiopia.
In this review, when examining the types of bacterial pathogens causing bloodstream infections in Ethiopia, Gram-negative bacteria were found to be more prevalent, comprising 56.65% (95% CI: 43.23–70.07%) of the isolates. This aligns with global trends where Gram-negative pathogens are often more difficult to treat due to their higher rates of multidrug resistance. The most common Gram-negative pathogen identified was Klebsiella spp., which accounted for 30.6% (95% CI: 20.76–40.45%) of the infections, posing significant treatment challenges due to its resistance to multiple classes of antibiotics. In comparison to studies conducted in Africa [38, 41], where Klebsiella spp. has been reported with a pooled prevalence of 2.2–28% in Asia [42], this prevalence in Ethiopia reflects a more considerable burden of infection, which might be associated with regional epidemiological factors such as higher HIV prevalence and hospital-acquired infections.
On the other hand, Gram-positive bacteria accounted for 44.51% (95% CI: 32.68–56.34%) of the isolates, with Staphylococcus aureus being the second most common pathogen, responsible for 20.51% (95% CI: 13.63–27.39%) of infections. This is comparable to other studies in Africa [38, 41], where S. aureus is also a predominant cause of bloodstream infections. Additionally, coagulase-negative Staphylococci (CONS) made up 18.24% (95% CI: 10.77–25.71%) of the isolates. Both S. aureus and CONS are of concern due to their potential for multidrug resistance, especially in the case of methicillin-resistant S. aureus (MRSA), which complicates treatment options. This finding aligns with reports from other regions, where MRSA remains a significant challenge in managing bloodstream infections.
Overall, the high prevalence of both Gram-negative and Gram-positive bacteria, particularly multidrug-resistant strains like Klebsiella spp. and S. aureus, underscores the urgent need for targeted antimicrobial stewardship, improved diagnostic strategies, and strengthened infection control measures to mitigate the impact of bloodstream infections in Ethiopia. The escalating rates of multidrug resistance are further complicating treatment protocols and intensifying the burden of infection in the region. Moreover, the inconsistency in the measurement and reporting of antimicrobial resistance data hampers meaningful comparisons across different countries and even within regions of the same country. This inconsistency underscores the critical need for reliable and standardized data on pathogen resistance. Access to both routine and research data on resistance patterns is essential to developing focused and effective strategies to address the growing global AMR crisis. This review highlights the alarmingly high prevalence of resistant sepsis-causing pathogens in Ethiopia, calling for urgent action to confront this escalating threat to public health.
The pooled prevalence of MDR in bacterial blood stream infections in Ethiopia is alarmingly high, with the overall pooled MDR estimate reaching 80.54% (95% CI: 77.24–83.85%), indicating significant resistance to commonly used antibiotics. The variation in MDR rates across studies is substantial, ranging from 65.0% (95% CI: 58.82–71.18%) to 91.3% (95% CI: 87.84–94.76%), with high heterogeneity (I² = 90.4%, p < 0.001), reflecting the complexity of the issue and suggesting regional or methodological factors at play. When comparing Gram-positive and Gram-negative bacteria, it is clear that Gram-negative pathogens exhibit much higher levels of resistance. Klebsiella, Acinetobacter, and Serratia species showed a 100% MDR rate, signaling complete resistance to multiple antibiotic classes, which poses a severe treatment challenge. Other Gram-negative bacteria, such as Salmonella (98.43%), Proteus (90.72%), and Enterobacter (90.12%), also demonstrate alarmingly high resistance. In contrast, Gram-positive bacteria like Coagulase-negative Staphylococci (CONS) and Enterococcus species exhibit lower but still concerning MDR rates of 66.46%, while S. aureus and S. pyogenes both show 58.77% MDR. S. pneumoniae has a somewhat lower MDR rate of 51.08%, though it remains a significant concern. Overall, the Gram-negative bacteria exhibit higher and more worrisome resistance compared to Gram-positive bacteria, particularly with the 100% resistance seen in certain species. Implementing antimicrobial stewardship programs in hospital and pediatric care settings is critical to address the burden of MDR pathogens.
Among Gram-positive bacteria, Coagulase-negative Staphylococci (CoNS) exhibited notable resistance to trimethoprim-sulfamethoxazole (72.05%) and penicillin (71.5%), but displayed lower resistance to amoxicillin (12.3%) and vancomycin (13%). S. aureus was highly resistant to ampicillin (85.2%) and penicillin (81.92%), but had relatively lower resistance to vancomycin (21.27%) and ciprofloxacin (21.34%). S. pyogenes showed high resistance to SXT (91.7%) and penicillin (75%), but lower resistance to ciprofloxacin, vancomycin, and ceftriaxone (16.7% for each). Enterococcus species demonstrated resistance to amoxicillin (35.13%) and ceftriaxone (44.36%), with vancomycin resistance at 26.27%. S. pneumoniae exhibited resistance to penicillin (80%), amoxicillin (38.3%), and ceftriaxone (21.7%), indicating considerable but lower resistance compared to other Gram-positive pathogens. These patterns are consistent with findings from other regions, including Africa and Sub-Saharan Africa, where similar resistance to penicillin and ampicillin has been reported in S. aureus and CoNS.
In Gram-negative bacteria, the resistance profiles are more alarming, particularly among the most common pathogens associated with sepsis. E. coli exhibited 100% resistance to ampicillin, along with 93.75% resistance to tetracycline and 93.05% resistance to SXT. Klebsiella species showed 100% resistance to ampicillin, erythromycin, doxycycline, and cefotaxime, with complete resistance to ceftriaxone. Salmonella species exhibited 98.43% resistance to ampicillin, 93.73% resistance to tetracycline, and high resistance to chloramphenicol (93.73%), but lower resistance to ciprofloxacin (40.65%). Acinetobacter species demonstrated significant resistance to amoxicillin (71.65%), gentamicin (74.66%), and ceftriaxone, with complete resistance to doxycycline. Similarly, Serratia species showed resistance to amoxicillin, gentamicin, and tetracycline. These high levels of resistance across Gram-negative pathogens underscore the challenges in treating sepsis effectively with first-line antibiotics like ampicillin, tetracycline, and SXT.
The resistance rates found in Ethiopia are consistent with reports from Africa [41] and Sub-Saharan Africa [38], where high resistance to penicillin, ampicillin, tetracycline, and SXT has been observed in both Gram-positive and Gram-negative bacteria. However, discrepancies are noted in certain antibiotic-bacterium combinations. For example, S. pyogenes isolates in our analysis showed a higher resistance to penicillin than previously reported in Africa [41], Asia [42] and Salmonella spp. showed higher resistance to ceftriaxone compared to previous studies in Africa [41] and Sub-Saharan Africa [38]. These differences may arise from variations in antimicrobial resistance testing methodologies and highlight the need for standardized and harmonized testing protocols across regions.
This study employed sensitivity analysis, subgroup analysis, and meta-regression to identify potential sources of heterogeneity in the data. Subgroup analyses in Ethiopia demonstrated significant variation in the prevalence of pediatric BSIs and MDR pathogens based on demographic, clinical, and geographic characteristics. Neonatal and NICU patients had the greatest BSI rates (34.07% and 35.83%, respectively), with regional peaks in Sidama and Tigray. Gram-negative bacteria were more prevalent, and studies with bigger sample sizes revealed increased BSI prevalence, presumably due to improved identification. MDR prevalence was disturbingly high at 80.54%, with no significant differences observed across age groups, hospital wards, research periods, or sample sizes, indicating widespread and durable resistance (76.75%). resistant pediatric bloodstream infections in Ethiopia. These findings highlight the urgent need for targeted, region-specific infection control and antimicrobial stewardship efforts to curb the burden of resistant pediatric BSIs in Ethiopia.
The sensitivity analysis demonstrated that excluding any single study had minimal impact on the pooled estimate, confirming the robustness and reliability of the overall result. The studies included in the analysis had prevalence estimates ranging from 26.54 to 34.76%, with most estimates falling between 27% and 31%. The combined estimate for these studies was 30.66% (95% CI: 27.18–34.15%), showing consistency across the included studies. Notably, the pooled estimate remained within the 95% confidence interval of the overall estimate, confirming that no single study significantly influenced the pooled prevalence of bacterial infections in Ethiopia. This reinforces the stability of the pooled estimate and indicates a consistent pattern across the studies.
The potential impact of small-study effects and publication bias on the pooled prevalence estimate of bloodstream infections was evaluated. While the visual inspection of funnel plots suggested some degree of asymmetry, the Egger’s test results indicated that there was no significant publication bias, with a p-value greater than or equal to 0.05. This suggests that small-study effects did not notably influence the overall prevalence estimate of bloodstream infections in Ethiopia. To further assess the effect of potential publication bias, a trim-and-fill analysis was conducted. Initially, the pooled estimate from the 23 studies included in the analysis was 30.66% (95% CI: 27.18–34.15%). After performing the trim-and-fill procedure to adjust for publication bias, the adjusted pooled prevalence of bloodstream infections in Ethiopia was found to be 29.52% (95% CI: 28.52–30.53%). These results indicate that publication bias had a minor influence on the pooled estimate, and the adjustments made were relatively small.
This study is limited by potential publication bias due to the inclusion of only published articles, excluding gray literature. All included data were phenotypic, lacking genotypic analysis of resistance mechanisms. Additionally, variability in local AMR patterns and non-standardized diagnostic methods across studies may have influenced the pooled estimates.