Neonatal sepsis in Iran: A systematic review and meta-analysis on national prevalence and causative pathogens

Background Neonatal sepsis is accounted for 30–50% of annual neonatal deaths in developing countries. We performed a systematic review and meta-analysis study to evaluate the national prevalence and identification of the etiological pathogens of neonatal sepsis in Iran. Methods A comprehensive literature search was done on the national and international databases for studies published between 2000 and 2019. The DerSimonian and Laird random-effects model was used to calculate pooled prevalence estimates, with 95% confidence intervals (CIs). Subgroup analyses and meta-regressions regarding the gender, type of sepsis and time during were also performed. Data were extracted, analyzed, and presented according to PRISMA guideline. Results Of 944 publications identified, 22 studies containing 14,683 neonates met the eligibility criteria. The pooled national prevalence of sepsis in Iran was 15.98% (95%CI, 11.96–20.46%; 1,367/14,683). Prevalence rate in boys (20.42%; 95%CI, 9.03–34.8%) was slightly higher than girls (18.5%; 95%CI, 7.4–32.8). A decreasing trend in prevalence of neonatal sepsis was found in recent years, although not statistically significant (c = -0.005; P value = 0.4). The most prevalent causative bacterial pathogens were Enterobacter spp. (23.04%), followed by Klebsiella pneumoniae (17.54%), coagulase-negative Staphylococci (14.06%), Escherichia coli (13.92%), Pseudomonas aeruginosa (12.67%), and Staphylococcus aureus (11.48%). Conclusion Our findings showed a high prevalence of neonatal sepsis in suspected neonates, suggesting the need to implement preventive measures, routine assessment, and close monitoring of neonates. Also, Enterobacter spp. and Klebsiella pneumoniae were identified as the principal bacterial pathogens responsible for neonatal septicemia in Iran.


Introduction
Sepsis is a serious cause of morbidity and mortality among neonates, killing approximately 3 million newborns each year [1]. The global burden of neonatal sepsis was measured as 2,202 per 100,000 live births [2]. Sepsis is a potentially life-threatening condition resulting from an extreme systemic immune response of the body to fight against an infection. This invasive infection, frequently bacterial, characterized by systemic signs of infection and isolation of bacteria from the bloodstream. Sepsis is one of the most important reasons for hospitalization of newborns in neonatal intensive care units (NICUs). Studies showed that neonatal sepsis is fatal and may quickly lead to septic shock and death if it's left untreated [3]. The therapy is often associated with an incontinent increase in antibiotic administration leading to development of antibiotic resistance.
Neonatal sepsis is categorized into either an early-onset neonatal sepsis (EOS) or late-onset sepsis (LOS) [4]. EOS is defined by bacteremia or meningitis occurring in newborns less than 3-days old [5]. EOS could be vertically transmitted infection that usually occurs as an ascending infection from the mother's cervix. Group B Streptococcus (GBS) is the leading cause of EOS [6] followed by Escherichia coli (E. coli) and Listeria monocytogenes [7]. Prematurity or low birth weight, birth asphyxia, prolonged rupture of membranes, and traumatic delivery are the major risk factors that contribute to the incidence of EOS [8]. LOS is defined as sepsis in infants during 4-90 days of life and could be caused either by vertically or horizontally transmitted infections [9]. The most important microorganisms involved in LOS include coagulase-negative Staphylococci (CoNS), Enterobacter spp., Escherichia coli, Pseudomonas aeruginosa, Klebsiella pneumonia, Staphylococcus aureus and Candida albicans [9,10]. Some main risk factors associated with LOS are prematurity, low birth weight, poor hand hygiene, central venous catheters and prolonged mechanical ventilation [11].
In the past few years several studies were performed in Iran regarding the prevalence of neonatal sepsis in, although most of them were focused to limited geographical areas. Here, we performed a comprehensive systematic review and meta-analysis study to assess the prevalence of neonatal sepsis and identify the causative pathogens of the infection in Iranian neonates.

Materials and methods
We used stepwise approach specified in the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines [12] to perform of this systematic review and meta-analysis (S1 Checklist).

Search strategy and selection criteria
Four major Iranian biomedical literature database servers (Iranmedex, Magiran, Irandoc and SID) and five international electronic databases (PubMed/MEDLINE, EMBASE, Scopus, Web of Science and Google Scholar) were searched for articles reporting the prevalence of sepsis in neonates and published between 2000 and 2019. The search terms utilized were "Iran", "sepsis", "septicaemia", "bacteraemia", "neonatal", "neonate" and "neonates" (See S1 Text: Supplementary file for the details of the databases searches). We restricted the literature search to reports in English and Persian languages and human subjects. Bibliographies of the obtained studies and the relevant review articles were carefully evaluated to identify the studies that primarily cover the sepsis and its causative pathogens, but were not found in databases search. We only included observational original studies that had information on the prevalence of sepsis, reported data on the number of children with sepsis and the causative organisms, and neonatal sepsis cases diagnosed according to the standard guidelines (blood culture). Studies were excluded if they were done in adults, investigated only a single pathogen of sepsis, had sample sizes less than 50 cases, duplicate studies, reviews, letters, case-reports or case series without original data.

Data extraction and quality assessment
After duplicates removal and screening of titles and abstracts, the full text of eligible studies were reviewed in depth by two independent reviewers (Z.A.R, and S.E). Data were extracted from all the eligible studies and compiled in Microsoft Excel spreadsheets. Inconsistencies were resolved by discussion and consensus with a third reviewer (A.R.). The following information were extracted from each study: first author last name, year of publication, study period, sample size, number of infected children, and number of bacterial pathogens. The quality of the studies was determined using Newcastle-Ottawa scale. Each item was scored for a maximum of six scores. Publications which scored 0-2, 3, 4 and 5 were classified as "unsatisfactory", "satisfactory", "good" and "very good", respectively.

Meta-analyses
The pooled prevalence of sepsis in Iranian neonates at a 95% confidence interval (CI) was estimated using the DerSimonian and Laird random-effects model (REM). The I 2 statistic was used to assess the heterogeneity of the estimates. Subgroup analyses were performed according to the gender, type of sepsis (EOS and LOS) and time during. The pooled prevalence of each bacterial pathogen was also calculated using REM. Forest plots were used for showing of pooled prevalence rates by REM. We did not undertake an assessment of publication bias, as it is not relevant for the prevalence studies [13]. All analyses were performed using STATA v.13 (STATA Corp., College Station, Texas, USA). Statistical tests were significant if P value was <0.05.

Results
In the preliminary search, a total of 944 potentially relevant studies were identified through the search of the national and international databases, and 909 studies were excluded after duplicates removal and further review of titles and abstracts. Finally, 35 studies were reviewed in depth and 22 study papers  involving 14,683 neonates were found to be eligible for meta-analysis (Fig 1). The studies included in meta-analysis represented 12 different provinces covering all parts of the country. The province with the highest number of studies was Tehran (7/22). Five and seven studies had extractable data regarding the gender (including 847 boys and 664 girls) and type of sepsis (EOS and LOS). All studies used blood culture methods to detect sepsis, while information regarding the bacterial pathogens was reported in 19 studies. All these studies cover only the hospitalized neonates. Main characteristics of studies included are summarized in S1 File and Table 1.
Among the bacterial pathogens causing the neonatal sepsis, our results indicated that the most prevalent causative pathogens belong to Enterobacter spp. (

Discussion
Neonatal sepsis continues to be a major health issue worldwide. A recent recommendation by World Health Organization (WHO) listed it as a life-threatening infection and a key healthcare priority for the coming decades [36,37]. Therefore, studies that estimate the national and global prevalence data could be very useful to prioritize the control measures for this preventable disease. In this study, we report a country-level estimate of neonatal sepsis in Iran. Our estimates showed that the prevalence of sepsis in Iranian neonates is about 16%, with high variability in different provinces. The prevalence estimates were limited to hospitalized neonates and do not cover the entire population. Prevalence rate estimated in Iran is lower than those reported from Egypt (44%) [38], Tanzania (39%) [39], Cameroon (37%) [40], and higher than those reported from Oman (3.5%) [41], USA (5.16 per 1,000) [42], and Australia (0.5 per 1,1000) [43]. Also, the prevalence of sepsis in other studies from Nepal and India was reported 36% and 46.8%, respectively [44,45]. These geographical variation could be attributed to preventive strategies adopted in each country, clinical criteria for sepsis diagnosis, differential sensitivity and specificity of the culture methods in different laboratories, health status of mothers during pregnancy, sanitary status in delivery section, and social-economic status in different countries [46,47]. In consistent with previous studies in some parts of the world [38,41,42,48], we found that prevalence of neonatal sepsis in boys is higher than girls. It is assumed that variation in genetic factors especially genes located in chromosome X may be responsible for these gender based differences [49]. In meta-regression analysis we have found a statistically non-significant but a decreasing trend in the prevalence of neonatal sepsis in recent years. This could be due to sustainable improvement of the sanitary conditions and efforts undertaken to improve mother and child health in Iran during last years [50,51].
With respect to the causative pathogens of neonatal sepsis, our study indicated that Enterobacter spp. (23.0%) and Klebsiella pneumoniae (17.4%) were the most common organisms responsible for neonatal sepsis. Another large systematic review on causative pathogens of   The current study had some limitations that need to be considered when interpreting the results of this work. First, the majority of studies included were published in local journals and may have resulted in a publication bias. Second, although we have performed a comprehensive literature search, data from several provinces were limited, therefore the same estimates cannot be guaranteed from further studies that include these regions in the future. Third, we could not perform a detailed sub-group analyses other than what was presented in the manuscript, because the majority of studies included in the meta-analysis did not classify the neonatal sepsis cases based on type of delivery, birth weight and community-onset/hospital-associated infections. Forth, our meta-analysis demonstrated a substantial heterogeneity of neonatal sepsis prevalence across studies, thus making it difficult for a direct comparison. Nevertheless, our study performed on a large sample size and a rigorous methodology provides useful information for national authorities to prioritize prevention efforts and intervention programs to reduce the burden of neonatal sepsis in Iran. Our finding showed that about 16% of newborn infants hospitalized in Iran had sepsis, a statistic that is higher than those reported from developed countries. This high rate of neonatal sepsis can lead to several neonatal death annually, and call for further measures and interventions to control and prevent sepsis in Iranian neonates.