1 Background

The rise of antimicrobial resistance and the resulting scarcity of effective antibiotics are two global concerns [1]. Nosocomial infections pose a serious risk to people everywhere, and the improper treatment with broad-spectrum antibiotics encourages the spread of hospital-associated multi-drug resistant pathogens [2]. Enterococci are important nosocomial pathogens. They are considered the primary cause of nosocomial infections in the USA and the second-highest cause of such infections globally [3]. They are facultative anaerobic Gram-positive cocci organisms belonging to the lactic acid bacteria. In 1984, group D streptococci were separated from the streptococci and were recognized as a distinct genus, which was named Enterococcus [4]. As per the LPSN List of Prokaryotic Names with Standing in Nomenclature, there are currently 80 verified species within the Enterococcus genus [5]. There are numerous diverse environments where enterococci may be found, including soil, water, on plants, and in the gastrointestinal tracts of both humans and animals [6]. It is a also frequently found in animal-derived foods such as meat, fermented and cooked meats, and cheese [7, 8].

Vancomycin is one of the therapeutic options for enterococci infections, acting by inhibiting peptidoglycan cross-linking by binding to the terminal D-Ala-D-Ala pentapeptide that compromises the integrity of the peptidoglycan layers, eventually leading to cell death [9]. As a tool to facilitate antibiotic stewardship and optimal use, the WHO Model List of Essential Medicines classified antibiotics into Access, Watch, and Reserve (AWaRe) categories for the treatment of priority bacterial infections. Based on the WHO AWaRe classification, vancomycin is classified within the Watch group, which includes most of the “highest-priority critically important antimicrobials” for human medicine and veterinary use. Antibiotics within the watch category are recommended only for specific limited indications. It is also included in the WHO Model List of Essential Medicines that contains the medications considered to be the most effective, safe, and meet the most important needs in a health system [10].

Unfortunately, enterococci evolved a resistance to vancomycin that was first detected in the late 1980s in the United Kingdom and Europe [11, 12].The vancomycin resistance (Van) operon confers resistance to glycopeptides in The Enterococcus species and this operon can be carried on mobile genetic elements and/or chromosomally [13]. One of the components of the Van operon is a variable ligase and, so far, 9 variant genes have been identified [14]. They are classified into two categories based on the ligases they encode: the operons vanA, vanB, vanD, and vanM, which encode for D-Ala-D-Lac ligase; and the operons vanC, vanE, vanG, vanL, and vanN which encodes for the D-Ala-D-Ser ligase [15].

Infections caused by VRE have a great impact on the healthcare system including longer hospital stays, higher death rates, and higher healthcare costs when compared to vancomycin-susceptible enterococci [16, 17]. Furthermore, vancomycin-resistant E. faecium bacteremia is associated with a bad prognosis and a higher mortality rate than vancomycin-resistant E. faecalis bacteremia [18, 19].

In the USA, 54,500 estimated cases are hospitalized; 5400 are estimated fatalities. Healthcare costs were estimated to be $539 million in 2017. In 2019, the CDC classified VRE as a serious threat in the United States’ Antibiotic Resistance Threats [20].

Although there are several reports from various Egyptian regions, the pooled incidence of VRE among enterococci clinical isolates in Egypt is unclear. Given the significant impact of VRE on patient mortality and healthcare costs, we conducted a systematic review with meta-analysis to determine the pooled prevalence of VRE among total enterococci clinical isolates, Enterococcus faecium and Enterococcus faecalis among total enterococci, Vancomycin-resistant Enterococcus faecalis, Vancomycin-resistant Enterococcus faecium, VanA and VanB genes among VRE, and linezolid, gentamicin, and ampicillin among total enterococci isolates and VRE.

2 Methods

2.1 Search strategy

Using the following keywords: enterococci, enterococcus, vancomycin-resistant enterococci, vancomycin-resistant enterococcus, VRE, and Egypt, a thorough literature search was carried out in the following databases: Medline (through PubMed), Scopus, Google Scholar, and Web of Science. Using ZOTERO version 6, search findings were combined, and duplicates were eliminated. Only original publications published in English were included in the search, which was limited to articles published between 2010 and 2022. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement during the preparation of this meta-analysis [21]. The PRISMA Checklist is presented in Fig. S1.

2.2 Inclusion and exclusion criteria

Two independent reviewers (A.Az and M.Y) selected the included studies based on the inclusion and exclusion criteria.

Any study that fulfilled all of the following criteria was included: studies conducted in Egypt; clinical isolates were only included (isolates from patients); studies that reported the prevalence of VRE; and studies using standard methods for detecting VRE. Studies were excluded if they had any of the following: studies published in languages other than English, full text not available, case report studies, review articles, and conference abstracts. Furthermore, articles with fewer than 30 subjects were excluded to reduce any potential bias brought on by a small sample size.

2.3 Data extraction

From each included study, the following data were extracted by two separate reviewers (A.A and H.E) and reviewed by a third (A.A.E): the first author's last name, publication date, government or city, total isolates of enterococci, total count of VRE, E. faecalis, and E. faecium among total enterococci, Vancomycin-resistant E. faecalis, Vancomycin-resistant E. faecium, VanA, and VanB genes among VRE, method of detection, specimen (urine/blood/wound, other sources), and resistance to linezolid, high content gentamicin, and ampicillin among total isolates of enterococci and VRE isolates.

2.4 Quality assessment

Two independent reviewers (A.Az and H.K) evaluated the quality of the included studies using a checklist derived from Ding et al. (2017) [22]. Disagreements were settled by consensus.

2.5 Meta-analysis

Statistical analysis has been performed using MedCalc Version 20.113. I2 and Q test were used to measure the heterogeneity between the studies and based on the random effects model, results were reported as proportions with a 95% confidence interval (CI). By visually examining the funnel plot, the risk of bias within studies was assessed. It was then tested using the non-parametric rank test, Begg’s test, and the parametric regression test (also known as “Egger’s test”). Low p values (P < 0.05) are considered a sign of publication bias in both situations since they show asymmetry. Publication bias testing was not performed when the number of studies was less than 10 [23]. Analyses of the subgroups were conducted based on region and the method used. Sensitivity analysis was conducted by Open Meta-Analyst Software using leave-one-out approach.

3 Results

3.1 Characteristics of the included studies

As depicted (Fig. 1), 4 separate databases were searched, yielding a total of 1093 results. Eight hundred fifty-one articles with irrelevant titles, 57 reviews, and 95 duplicates were removed. The remaining 90 publications were then evaluated by reading the abstracts, and 18 were eliminated. By reading the full text, 72 articles were reviewed for eligibility, and by the end, a total of 25 studies fulfilled our inclusion and exclusion criteria and were included in our review [24–48] (Table S1 and S2 summarize the characteristics and the quality of the included studies respectively; see supplementary material).

Fig. 1
figure 1

Flow chart depicting the selection of publications

Full size image

3.2 The prevalence of VRE among total enterococci clinical isolates

The pooled prevalence of VRE among enterococci clinical isolates was estimated to be 26% (95% CI 16.9 to 36.3). The studies had a significant degree of heterogeneity, as evident by (I2 = 95.45%) and Cochrane Q test = 528. The funnel plot showed a slight asymmetry by visual inspection; as evidenced by the Egger’s test and Begg’s test these were statistically insignificant (P = .0468, and P = .0258), respectively (Fig. 2).

Fig. 2
figure 2

The prevalence of VRE among clinical isolates in Egypt. a Forest plot of VRE among total enterococci. b Funnel plot of VRE among total enterococci

Full size image

3.3 Subgroup analysis

The VRE prevalence was also determined depending on location and antibiotic susceptibility technique. Of 25 studies, 13 studies used the disc diffusion method and 12 studies used the MIC-based methods to determine the susceptibility of enterococci to vancomycin. The results of the meta-analysis based on subgroup were summarized in Table 1.

Table 1 Pooled prevalence of VRE in Egypt by subgroups
Full size table

The pooled VRE prevalence among total enterococci by the disc diffusion method was 24.02% (95% CI 11.36 to 39.6; I2 = 96.05%; P < 0.0001).Visual observation of the funnel plot revealed asymmetry, and there was a statistically significant funnel plot asymmetry as evidenced by Egger’s test and Begg’s test (P = 0.0288, P = 0.0082) (Figs. S2 and S3; see supplementary material).

On the other hand, the prevalence of VRE based on MIC-based methods was 28.25% (95% CI 15.83 to 42.64; I2 = 94.85% P < 0.0001) and the funnel plot showed symmetry that was evidenced by both Egger’s test and Begg’s test (P = 0.9875, P = 0.7297, for broth microdilution 26.76% (95% CI 5.9 to 55.6; I2 = 95.81%; P < 0.0001), E test 38.24% (95% CI 24.7 to 52.78; I2 = 87.60%; P 0.0001) and for vitek 2 automated system 13% (95% CI 5.17 to 23.64; I2 = 66.53%; P = 0.0504) (Figs. S4–S8; see supplementary material).

In our review, the majority of research was reported from Mansoura (6 studies), Cairo (6 studies), Minia (3), Tanta (2), Menoufia (2), Zagazig (2), Sohag (2), Ismailia (1), and Alexandria and EL Beheira (1).

Table 1 summarizes the prevalence, which ranged from 12.5% (95% CI 3.9 to 24.8; I2 = 90.61%; P 0.0001) and I2 (95% CI 3.2 to 25.5; I2 = 69.93%; P = 0.0682) for Mansoura and Tanta, respectively, to 62.5% (95% CI 51 to 73.2; I2 = 0.00%; P = 0.7605) for Menoufia (Figs. S9–S15, see supplementary material).

3.4 Prevalence of E. faecalis and E. faecium among total enterococci

The frequency of E. faecalis and E. faecium has been co-reported in 20 studies. E. faecalis had a greater pooled prevalence than E. faecium, with 61.22% (95% CI 53.65 to 68.53) and 32.47% (95% CI 27 to 38.2), respectively. There was no evidence of funnel plot asymmetry by visual inspection of the funnel plot and by both Egger’s test and Begg’s test as shown in (Table 2) and (Figs. 3 and 4).

Table 2 Pooled prevalence of E. faecium and E. faecium among enterococci, their Vancomycin resistance and VanA and VanB genes among VRE
Full size table
Fig. 3
figure 3

The prevalence of E. faecalis and E. faecium among total enterococci isolates. a Forest plot of E. faecalis among total enterococci. b Forest plot E. faecium among total enterococci

Full size image
Fig. 4
figure 4

Funnel plot of publication bias for E. faecalis and E. faecium among total enterococci isolates. a Funnel plot of E. faecalis among total enterococci. b Funnel plot of E. faecium among total enterococci

Full size image

3.5 The prevalence of vancomycin-resistant E. faecalis among total E. faecalis

In 13 publications, vancomycin-resistant E. faecalis was shown to be common among all E. faecalis. The pooled prevalence was 31.7% (95% CI 18.6 to 46.4). By visual inspection, the funnel plot displayed asymmetry, which was further confirmed by both Egger’s test and Begg’s test, which showed statistically significant funnel plot asymmetry (as shown in Table 2 and Fig. 5).

Fig. 5
figure 5

The prevalence of Vancomycin-resistant E. faecalis among total E. faecalis. a Forest plot Vancomycin-resistant E. faecalis. b Funnel plot of Vancomycin-resistant E. faecalis

Full size image

3.6 Prevalence of Vancomycin-resistant E. faecium among total E. faecium

Twelve studies gave an account of the prevalence of vancomycin-resistant E. faecium among total E. faecium. It had a pooled prevalence of 46.1% (95% CI 25.7 to 67.1). Again, there was no evidence of funnel plot asymmetry by visual inspection of the funnel plot and by both Egger’s test and Begg’s test (as presented in Table 2 and Fig. 6).

Fig. 6
figure 6

The prevalence of Vancomycin-resistant E. faecium among total E. faecium. a Forest plot of Vancomycin-resistant E. faecium. b Funnel plot of Vancomycin-resistant E. faecium

Full size image

3.7 VanA and VanB gene among VRE

VanA was more frequent than VanB among VRE, with pooled prevalence of 63.3% (95% CI 52.1 to 73.7) and 17.95 (95% CI 7.8 to 31), respectively, as reported by 7 studies that indicate the prevalence of both VanA and VanB genes among VRE (as shown in Fig. 7).

Fig. 7
figure 7

The dissemination of VanA and VanB among VRE. a Forest plot of VanA among VRE. b Forest plot of VanB gene among VRE

Full size image

3.8 The resistance profile of enterococci to linezolid, ampicillin, and high content gentamicin

As depicted in Table 3, the pooled resistance rate of linezolid was substantially lower than that of ampicillin and high-level gentamicin (HLG), 5.54% (95% CI 2.33 to 10%), 65.7% (95% CI 50.8 to 79.2%), and 61.1% (95% CI; 47.4 to 73.9%), respectively (Figs. 8 and 9). There was no funnel plot asymmetry (Fig. 10) that was evident by Egger’s test and Begg’s test. (The characteristics of the antibiotic resistance profile among total Enterococci and VRE are summarized in Table S3 and S4; see Supplementary material.)

Table 3 Pooled resistance profile of enterococci isolates in Egypt
Full size table
Fig. 8
figure 8

The resistance profile of enterococci isolates to linezolid and ampicillin. a Forest plot of linezolid resistance among total enterococci isolates. b Forest plot of ampicillin resistance among total enterococci isolates

Full size image
Fig. 9
figure 9

Forest plot of enterococci isolates resistant to high-level gentamicin

Full size image
Fig. 10
figure 10

Funnel plots of the resistance profile of the enterococcal isolates to linezolid and ampicillin. a Funnel plot of linezolid resistance. b Funnel plot of resistance ampicillin

Full size image

3.9 The resistance profile of VRE to linezolid, and ampicillin

There were only 4 studies that reported linezolid and ampicillin resistance rates among VRE. The pooled resistance rate of linezolid was much lower than that of ampicillin, 5.2% (95% CI 1.3 to 11.5%) and 85% (95% CI 49 to 100%) respectively (Figs. S16 and S17).

3.10 Sensitivity analysis

Sensitivity analysis, using the leave-one-out approach, indicated that the combined estimate of VRE among total enterococci clinical isolates is reliable and is not dependent on any single study (Fig. S18).

4 Discussion

To our knowledge, this is the first systematic review and meta-analysis study to evaluate the pooled prevalence of VRE and the antimicrobial resistance profile of enterococci in Egypt. This study is based on data analysis from published literature on the prevalence of VRE in patients in Egypt published between 2010 and 2022. The pooled prevalence of VRE among enterococci clinical isolates in Egypt was estimated to be 26%. E. faecalis had a greater pooled prevalence than E. faecium. The VanA gene was more frequent than the VanB gene among VRE. The pooled resistance rate of linezolid was substantially lower than that of ampicillin and high-level gentamicin (HLG).

In our review, the pooled prevalence of VRE among enterococci clinical isolates was 26% (95% CI 16.9 to 36.3), which was higher than the pooled prevalence of VRE among clinical specimens in Iran and Asia, which was 9.4% (95% CI 7.3–12) and 8.10% (95% CI 7–9), respectively [49, 50].

Subgroup analysis based on the antimicrobial susceptibility methods yielded heterogeneous results. The pooled VRE prevalence among total enterococci by the disc diffusion method was 24.02% (11.36 to 39.6) and by MIC-based methods it was 28.25% (95% CI 15.83 to 42.64). The results obtained by vitek 2 were lower than those of the E-test and broth microdilution (9.22% (95% CI 2.69 to 19.1), 26.76% (95% CI 5.9 to 55.6), and 38.24% (95% CI 24.697 to 52.78), but the 95% confidence interval overlapped. These heterogeneous results could be explained by different resistance patterns based on region, specimen source, and technique variability [51].

We also found that the frequency of E. faecalis and E. faecium has been reported in 20 studies. E. faecalis had a greater pooled prevalence than E. faecium, with 61.22% (95% CI 53.65 to 68.53) and 32.47% (95% CI 27 to 38.2), respectively. These results were consistent with a meta-analysis by Moghimbeig et al. [52]. However, several reports from the UK, Denmark, the Netherlands, Poland, and Iran have shown a trend towards the replacement of E. faecalis by E. faecium [53–57].

The prevalence of vancomycin resistance among E. faecium and E. faecalis was co-reported in 12 studies and showed a higher rate of vancomycin resistance among E. faecium than that of E. faecalis 46.1% (95% CI 25.7 to 67.1), and 31.7% (95% CI 18.6 to 46.4), respectively, but the 95% CI overlapped. Other studies have also reported similar findings [49, 58, 59]. This point is particularly important as vancomycin-resistant E. faecium bacteremia is associated with a bad prognosis and a higher mortality rate than vancomycin-resistant E. faecalis bacteremia [18, 19].

According to 7 studies that co-reported the dissemination of VanA and VanB variants, VRE harboring VanA variants were more prevalent than those harboring VanB variants, with a pooled prevalence rate of 63.3% (95%CI 52.1 to 73.7) and 17.95% (95% CI 7.8 to 31), respectively. The dissemination of VanA and VanB variants among enterococci varies worldwide. For instance, VanA is predominant in North America and Europe. On the other hand, the VanB variant is dominant in Australia and New Zealand and is increasingly being reported in Europe [60].

Linezolid is the first member of the oxazolidinone family of antibiotics and is considered one of the last-resort antibiotics for management of VRE infections [61]. the ZAAPS and LEADER surveillance programs, which were set up to monitor linezolid resistance in non-USA and USA countries, respectively, revealed that enterococci were susceptible to linezolid in more than 99% of cases [62, 63]. According to current meta-analysis outcomes, resistance rates of 927 enterococci clinical isolates to linezolid were documented in 13 studies, with a pooled resistance rate of 5.54% (95% CI 2.33 to 10%).The pooled resistance rate of linezolid remained almost consistent against enterococci showing a VRE phenotype similar to the ZAAPS surveillance program [62]. Given this high resistance rate, linezolid should be reserved for treatment of confirmed or suspected infections due to multi-drug-resistant organisms and be de-escalated wherever possible.

Assessing the occurrence of ampicillin and gentamicin resistance in enterococci is clinically important, as a combination of both is recommended in the treatment of ampicillin-sensitive VRE when bactericidal activity is needed [64, 65].

Our study demonstrated a high level of ampicillin resistance among enterococci clinical isolates with a pooled resistance rate of 65.7% (95% CI 50.8 to 79.2%), similar to a study conducted in India that revealed a 75.5% ampicillin resistance rate [66], and much higher than a surveillance study in Europe conducted between 2011 and 2019 that revealed a 10.6% ampicillin resistance rate [67].

High-level gentamicin resistance (HLGR) was documented in seven studies with a pooled resistance rate of 61.1% (95% CI 47.4 to 73.9%). This is somewhat comparable to a meta-analysis done in Iran that revealed a 49.4% (95% CI 42.2 to 56.6%) HLGR to enterococci [68].

Several factors may explain this high prevalence of vancomycin, ampicillin and high-level gentamicin resistance among enterococci isolates in Egypt. First, infection control programs are not very adequate in Egypt. Workload, inadequate resources, limited opportunities for infection control training and insufficient staff were the most common obstacles complained about by healthcare workers against the practice of standard precautions [69–73]. Second, the inappropriate use of antibiotics and antibiotic self-medication are prevalent in Egypt [74–76].

We think the following measures may be needed to limit further increases in antibiotic resistance among enterococci or other pathogens. First, a national Antimicrobial Resistance Policy in Egypt to understand the emergence, spread, and factors influencing antimicrobial resistance. Second, a prohibition on antibiotic self-medication. Third, efforts to educate healthcare workers and patients about the proper use of antimicrobials. Fourth, rapid molecular diagnostics to support appropriate antimicrobial use. Fifth, if not previously established, infection control strategies and antimicrobial stewardship practices should be followed. Sixth, research is needed to define “inappropriate” antimicrobial prescribing and to better understand the primary drivers of such use.

There are some limitations to our study. First, our results do not fully reflect the prevalence of VRE in Egypt, as not all regions in Egypt reported the prevalence of VRE. Second, there was a high heterogeneity in VRE prevalence between studies that could stem from the difference in antibiotic resistance pattern from region to region or from AST methods themselves. Third, the small number of included studies is of concern.

5 Conclusion

Given the high incidence of resistance to vancomycin, linezolid, high-level gentamicin, and ampicillin in clinical specimens from enterococci in Egypt, we strongly advise that healthcare settings develop and follow their own antibiogram to guide choosing an appropriate empirical therapy as well as implementing infection control programs to prevent further escalation of the problem.