Antibiotic therapy in neonates and impact on gut microbiota and antibiotic resistance development: a systematic review

Antibiotic therapy in neonates and impact on gut microbiota and antibiotic resistance... Abstract Objectives To systematically review the impact of antibiotic therapy in the neonatal period on changes in the gut microbiota and/or antibiotic resistance development. Methods Data sources were PubMed, Embase, Medline and the Cochrane Database, supplemented by manual searches of reference lists. Randomized controlled trials (RCTs) and observational studies were included if they provided data on different categories of antibiotic treatment (yes versus no, long versus short duration and/or broad- versus narrow-spectrum regimens) and subsequent changes in the gut microbiota and/or antibiotic resistance development. We evaluated risk of bias using the Cochrane Handbook, adapted to include observational studies. When appropriate, we used the vote-counting method to perform semi-quantitative meta-analyses. We applied the Grades of Recommendation, Assessment, Development and Evaluation approach to rate the quality of evidence (QoE). Study protocol registration: PROSPERO CRD42015026743. Results We included 48 studies, comprising 3 RCTs and 45 observational studies. Prolonged antibiotic treatment was associated with reduced gut microbial diversity in all three studies investigating this outcome (very low QoE). Antibiotic treatment was associated with reduced colonization rates of protective commensal anaerobic bacteria in four of five studies (very low QoE). However, all three categories of antibiotic treatment were associated with an increased risk of antibiotic resistance development, in particular MDR in Gram-negative bacteria, and we graded the QoE for these outcomes as moderate. Conclusions We are moderately confident that antibiotic treatment leads to antibiotic resistance development in neonates and it may also induce potentially disease-promoting gut microbiota alterations. Our findings emphasize the need to reduce unnecessary antibiotic treatment in neonates. Introduction Upon birth, infants are suddenly exposed to a wide range of bacteria colonizing mucoepithelial surfaces, including the gut.1 The subsequent development of the infant gut microbiota is dynamic, non-resilient and shaped by factors such as mode of delivery, feeding, diet and environment.2–4 A healthy gut microbiota plays a crucial role in the development of the immune system, digestive functions and protection against infections.4–6 The commensal aerobic and anaerobic bacteria are also essential for colonization resistance—the ability to prevent invasion and persistent carriage of pathogenic and antibiotic-resistant bacteria.7 Antibiotics are the most commonly prescribed medications in the neonatal unit.8 However, antibiotic overuse in early life disrupts the actively developing gut microbiota, causing ‘bacterial dysbiosis’, which is associated with an increased risk of early adverse outcomes such as necrotizing enterocolitis and fungal infections.9 Early antibiotic exposure has also been associated with allergic diseases, obesity, diabetes and inflammatory bowel disease later in life.10–14 Overuse of antibiotics, particularly broad-spectrum antibiotics, applies a selection pressure that favours antibiotic-resistant bacteria and decreases colonization resistance.7,15 The currently observed increase in resistance to aminoglycosides and ampicillin among Gram-negative bacteria has begun to threaten this traditional combination as empirical treatment for neonatal sepsis.16,17 Moreover, the emergence worldwide of ESBL-producing Enterobacteriaceae presents major challenges in managing neonatal sepsis.18 Globally, an estimated 200 000 neonatal deaths are attributed to resistant organisms each year.19 However, the relative impact of different types of antibiotic exposure on the actively developing gut microbiota composition and antibiotic resistance development is not fully understood. The purpose of this systematic review is to identify, critically appraise and synthesize evidence from studies reporting different categories of antibiotic therapy in neonates and their impact on the gut microbiota and/or antibiotic resistance development. We included both observational studies and randomized controlled trials (RCTs) in line with suggestions from the Cochrane handbook stating that systematic reviews of adverse effects will usually need to include non-randomized studies in addition to RCTs. Methods This review was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses following a registered protocol and according to the recommendations given by the Cochrane Handbook for Systematic Reviews and Interventions.20–22 We recently published a systematic review on early clinical adverse effects of neonatal antibiotic treatment based on the same research protocol.9 For this review, our primary research question was: ‘Are different categories of antibiotic treatment in neonates associated with different changes in gut microbiota composition and/or differences in antibiotic resistance development?’ Study protocol registration: PROSPERO CRD42015026743. Search strategy We developed our search strategy in consultation with an epidemiologist, a librarian, a paediatric pharmacologist and a neonatologist. We searched PubMed, Embase, Medline and the Cochrane Database using MeSH terms and free-text searches with no time restrictions (last search 22 December 2016). The first search was conducted with MeSH terms in PubMed, Medline and the Cochrane Database by combining ‘Infant, Newborn’ and ‘Anti-Bacterial Agents’ with one of two outcome terms: ‘Drug Resistance, Bacterial’ or ‘Microbiota’. The Embase database uses its own key words, and we combined ‘Newborn’ and ‘Antibiotic Agent’ with either ‘Antibiotic Resistance’ or ‘Microbiome’. The second search was conducted using free text in PubMed, Medline and Embase combining the keywords: ‘Infant, Low Birth Weight’ or ‘Infant, Postmature’ or ‘Infant, Premature’ or ‘Infant, Newborn’ with ‘Anti-Bacterial Agents’ or ‘Antibiotics’ and one of the following: ‘Antibiotic Resistance’ or ‘Antibacterial Drug Resistance’ or ‘Microbiota’ or ‘Microbiome’ or ‘Microbiomes’ or ‘Gut Flora’. We examined reference lists of included studies and relevant reviews to identify additional eligible studies. We then combined all citations and excluded duplicates or triplicates. We did not contact authors for supplementary information and we did not perform searches in the grey literature. Study selection and eligibility criteria A study was eligible for review if it reported different categories of intravenous antibiotic treatment in the neonatal period and evaluated their impact on changes in the gut microbiota and/or antibiotic resistance development. If infants were born prematurely we defined the neonatal period as up to 44 weeks postmenstrual age. We compared three different categories of antibiotic therapy: (i) antibiotic treatment yes versus no; (ii) antibiotic treatment long versus short; and (iii) antibiotic treatment broad versus narrow spectrum. For category (ii), we suggested in advance that ‘prolonged’ antibiotic exposure was always ≥3 days or the longest regimen among two antibiotic regimens compared. For category (iii), we always defined regimens including third-generation cephalosporins or carbapenems as broad-spectrum regimens when compared with regimens containing aminoglycosides for coverage against Gram-negative bacteria. This definition was based on previous reports indicating that empirical therapy containing a third-generation cephalosporin for Gram-negative coverage induces significantly more resistance than a regimen containing an aminoglycoside.15 If two similar regimens were compared, the regimen with the broadest spectrum was labelled broad spectrum. Both RCTs and observational studies such as cohort, case–control and cross-sectional studies were eligible for inclusion. We included case–control studies reporting on the prespecified outcomes if data on antibiotic therapy prior to the outcomes were presented as extractable data on cases and controls. We excluded case reports and case series, studies with a non-neonatal or non-human population, studies that were written in languages other than English, and studies that investigated antenatal antibiotics, oral antibiotics and/or low-dose intravenous vancomycin prophylaxis. Screening, data extraction and management Two reviewers (J. W. F. and E. E.) independently screened search results and assessed each potentially eligible study per our predetermined inclusion and exclusion criteria. We only excluded studies that we agreed were irrelevant according to our predefined criteria. A third researcher (C. K.) had the deciding vote in cases of disagreement. We extracted the following information from included studies: author, year, country, study design, study population, including gestational age (GA) and birth weight (BW), comparison of outcomes between groups with different categories of antibiotic treatment, and, if available, risk estimates with 95% CI for the specific outcome. Gut microbiota analyses were based on faecal samples using both standard culture-based methods and culture-independent methods relying on DNA amplification and sequencing.23 After reviewing the articles presenting data on gut microbiota we decided to present data from these studies in three main categories: microbial load, microbial diversity and microbial composition, clearly acknowledging some overlap between these categories. We defined microbial load as the total number of bacteria in a sample, microbial diversity as the number of different bacterial genera or species in a sample, and microbial composition as the taxonomical composition in a sample. Antibiotic resistance development was based on detection of antibiotic susceptibility patterns in bacteria isolated from blood, urine, cerebrospinal fluid, faeces, tracheal aspirates and/or the skin surface. We defined MDR bacteria as bacteria resistant to ≥2 unrelated classes of antibiotics or broad-spectrum antibiotics.24–28 Included in this category were ESBL-producing Gram-negative bacteria, carbapenem-resistant Acinetobacter baumannii and Gram-negative bacteria resistant to third-generation cephalosporins. Antibiotic-resistant bacteria that did not meet any of these criteria were defined as ‘other antibiotic-resistant bacteria’. We applied a simple vote-counting method to investigate whether the different categories of antibiotic therapy had any effect on the outcomes of interest.22 Studies were classified based on whether they showed a reduction in the outcome measure, no effect or an increase in the outcome measure following antibiotic treatment. When appropriate, outcomes were presented in vote-count figures. Assessment of methodological quality Methodological quality was assessed using the Cochrane Handbook of Systematic Reviews of Interventions and recently published suggestions on how to assess risk of bias and confounding in observational studies.22,29 Five domains related to risk of bias were assessed for each study and included: Selection, Performance, Detection, Reporting and Confounding. Risks of bias were judged as low, high or unclear for each domain. The risk of reporting bias was considered unclear in studies that did not have a previously published protocol. The risk of detection bias was considered high in studies that examined gut microbiota with culture-based methods, unclear in studies that applied 16S rRNA sequencing techniques and low in studies that applied shotgun metagenome sequencing techniques. Two reviewers (J. W. F. and E. E.) assessed the risks of bias for each study. Disagreements in the categorization process were resolved after discussion between J. W. F., E. E. and C. K. We applied the Grades of Recommendation, Assessment, Development and Evaluation Working Group (GRADE) approach to rate the quality of evidence (QoE) for each relevant outcome category.30 This approach specifies four levels of quality from high to very low, which define the degree to which its estimates of effects or associations can be trusted.22,30 RCTs started as high QoE and observational studies started as low QoE.30 Several factors could either downgrade or upgrade the quality rating. Results Overview of included studies From 3380 identified studies, we reviewed 137 potentially eligible full-text articles. Forty-eight studies met our inclusion criteria: 3 RCTs published between 2000 and 2013,15,31,32 and 45 observational studies published between 1974 and 2016 (Figure 1).24–28,33–73 Two articles presented data from the same study population and were defined as one study.34,35 Antibiotic treatment was the randomized intervention in two out of the three included RCTs.15,31,32 Among the 45 observational studies, there were 22 prospective cohort studies, 12 case–control studies, 7 before–after studies and 4 retrospective cohort studies. There were large variations regarding the categories of antibiotic therapy studied. Tables S1 and S2 (available as Supplementary data at JAC Online) display the main characteristics and primary outcomes of interest from the 48 included studies. Figure 1. View largeDownload slide Preferred reporting items for systematic reviews and meta-analysis (PRISMA) flow diagram. Figure 1. View largeDownload slide Preferred reporting items for systematic reviews and meta-analysis (PRISMA) flow diagram. Risk of bias and QoE Figure S1(a and b) displays risk of bias assessments for each included study. Outcomes adjusted for differences in populations were reported in 16/45 (36%) observational studies.25,26,28,39,44,46,50,52,55,57,63,64,67,69–71 Five of these studies used stratification or multivariate analysis to adjust for antenatal antibiotic treatment as a potentially confounding variable. None of the RCTs were included in public registries. We graded the QoE as very low for the outcomes microbial load and microbial diversity in relation to the three different categories of antibiotic treatment owing to inclusion of observational studies with serious risk of bias and inconsistent results. We graded the QoE as very low for the outcomes relating to microbial composition after antibiotic treatment (Figure 2a–d). We graded the QoE as moderate for the outcomes relating to antibiotic resistance development owing to inclusion of observational studies that either had large effect sizes (yes versus no and broad versus narrow) or a dose–response effect (long versus short) after antibiotic treatment (Figure 3a–c). Figure 2. View large Download slide View large Download slide Vote count on gut microbial composition after antibiotic exposure compared with no antibiotic exposure. The sizes of the squares are proportional to study populations. An asterisk symbolizes a lack of testing for statistical significance. Figure 2. View large Download slide View large Download slide Vote count on gut microbial composition after antibiotic exposure compared with no antibiotic exposure. The sizes of the squares are proportional to study populations. An asterisk symbolizes a lack of testing for statistical significance. Figure 3. View large Download slide View large Download slide Vote count on infection and/or colonization with MDR Gram-negative bacteria following antibiotic exposure. The sizes of squares are proportional to study populations. A dagger symbolizes multivariate regression analysis. Figure 3. View large Download slide View large Download slide Vote count on infection and/or colonization with MDR Gram-negative bacteria following antibiotic exposure. The sizes of squares are proportional to study populations. A dagger symbolizes multivariate regression analysis. Gut microbiota composition Nineteen studies reported on antibiotic exposure and impact on the gut microbiota composition (Table S1); these comprised 2 RCTs31,32 and 17 observational studies.33–49,73 Three studies reported outcome data from both antibiotic treatment yes versus no and broad versus narrow spectrum,34,37,47 and one study reported outcome after antibiotic treatment yes versus no and long versus short.42 The remaining 15 studies reported outcome data from one category of antibiotic treatment. To examine gut microbiota composition, nine studies used 16S rRNA gene-sequence analysis,33,38–40,42,44–46,49 one used fluorescence in situ hybridization techniques,32 one used deep shotgun metagenome sequence analysis48 and eight used standard culture-based methods.31,34–37,41,43,47,73 The included studies reported primarily taxonomic data with different hierarchical details on (i) Enterobacteriaceae, (ii) obligate commensal anaerobic bacteria (bacteroides, lactobacilli and bifidobacteria, etc.), (iii) clostridia and/or (iv) Gram-positive cocci. Microbial load Three studies (296 neonates) compared the impact of antibiotic treatment (yes versus no) on microbial loads.32,34,40 One study (165 neonates) found increased microbial loads,34 one RCT (113 preterm neonates) found decreased microbial loads32 and one study (18 term neonates) found no significant differences in microbial loads following antibiotic treatment.40 A small study of extremely low-birth-weight neonates found an inverse correlation between the duration of antibiotic therapy and the microbial load on day 30 of life.41 Microbial diversity Four studies (159 neonates) compared microbial diversity after antibiotic treatment (yes versus no).40,42,44,49 Two studies (112 preterm neonates) reported decreased diversity among antibiotic-treated neonates42,49 and two studies (47 neonates) reported no significant differences.40,44 Three studies (224 preterm neonates) examined the impact of antibiotic therapy duration (long versus short) on microbial diversity, and all three found decreased diversity following prolonged therapy.41,44,48 Microbial composition Figure 2 displays the results of studies reporting the impact of antibiotic treatment (yes versus no) on microbial composition. Nine studies focused on Enterobacteriaceae; four reported an increase and five studies reported unchanged composition after antibiotic treatment, mainly ampicillin plus an aminoglycoside (Figure 2a).33,34,36,37,40,42,43,46,47 Five studies focused on different commensal obligate anaerobes, showing a clear trend towards reduced colonization rates following treatment (Figure 2b).35,36,38,40,43 In the five studies focusing on clostridia, there were equivocal results (Figure 2c).36,39,40,45,46 Finally, four studies focused on Gram-positive cocci, and these studies showed either unchanged or higher colonization rates after antibiotic treatment (Figure 2d).33,36,37,40 Two studies (n = 983) reported Enterobacteriaceae colonization rates after treatment with broad- versus narrow-spectrum antibiotics.37,47 Both studies found lower colonization rates following third-generation cephalosporin treatment. One study of preterm infants (n = 76) reported lower colonization rates of clostridia in those who received ≥10 days of antibiotic therapy compared with shorter duration.39 Another study of preterm infants (n = 74) reported higher colonization rates of staphylococci in those who received ≥5 days of antibiotic treatment compared with shorter-duration therapy.42 Finally, two studies (n = 104) compared the impact of antibiotic therapy (broad versus narrow) on abundance and/or colonization rates with staphylococci, but neither found any significant differences.37,42 Antibiotic resistance development Thirty-one studies, 2 RCTs15,31 and 29 observational studies,24–28,37,50–72 evaluated the risk of antibiotic resistance development after antibiotic exposure (Table S2). Five studies reported outcome after antibiotic treatment yes versus no and broad versus narrow spectrum.27,37,53,55,67 Two studies reported outcome after antibiotic treatment long versus short duration and broad versus narrow spectrum.26,64 Two studies reported outcome after antibiotic treatment yes versus no and long versus short duration.25,57 The remaining 23 studies assessed only one category of antibiotic therapy. Nine studies reported on both infections and colonization with antibiotic-resistant bacteria,24,51,57,58,60,62,65,67,68 whereas 15 studies only reported on colonization,15,25–27,31,37,53–56,59,61,66,69,72 and 7 studies only reported on infections.28,50,52,63,64,70,71 MDR bacteria were varyingly defined as bacteria that were resistant to both third-generation cephalosporins and aminoglycosides55,58 or bacteria resistant to ≥2 or ≥3 unrelated classes of antibiotics.24–28 Thirty of 31 studies focused solely on antibiotic resistance development in Gram-negative bacteria. Among these, 20 studies focused on MDR Gram-negative bacteria. MDR Gram-negative bacteria Figure 3 displays the results of the 20 studies reporting the impact of antibiotic exposure on rates of infection and/or colonization with MDR Gram-negative bacteria. Nine studies reported data after antibiotic treatment yes versus no, and the majority reported increased rates of MDR Gram-negative bacteria following treatment (Figure 3a).25,27,55,57,59,63,67,69,70 Five studies reported data after long versus shorter duration of treatment, and the majority found significantly more MDR Gram-negative bacteria after prolonged treatment (Figure 3b).25,26,56,57,64 Thirteen studies reported data after treatment with broad-spectrum versus narrow-spectrum antibiotics, and the overwhelming majority reported higher rates of MDR Gram-negative bacteria following treatment with broad-spectrum antibiotics (Figure 3c).15,24,26–28,50,51,55,58,64,65,67,71 Other antibiotic-resistant bacteria Four studies (n = 1825) compared the impact of antibiotic treatment (yes versus no) on antibiotic-resistant bacteria that were not MDR according to our definition.37,52,53,66 One study (n = 584) found a higher rate of prior antibiotic treatment in neonates colonized with antibiotic-resistant Escherichia coli and/or Klebsiella pneumoniae.66 One study (n = 953) found an increased incidence of TEM-1 genes in E. coli strains in neonates following antibiotic therapy.53 Two studies (n = 288) found no statistically significant associations between antibiotic treatment (yes versus no) and subsequent antibiotic resistance development.37,52 Two studies compared the impact of antibiotic therapy duration;61,72 one of them (n = 1180) found significantly longer prior antibiotic treatment among neonates colonized with antibiotic-resistant Gram-negative bacteria,72 whereas the other (unknown number of neonates) found no correlation between the duration of treatment and gentamicin-resistant Gram-negative bacteria.61 Eight studies (n = 3029) compared the impact of broad- versus narrow-spectrum antibiotic treatment.31,37,53,54,60,62,68,72 One RCT (n = 276) found higher colonization rates with ampicillin-resistant A. baumannii following treatment with penicillin and gentamicin compared with ampicillin and gentamicin.31 One study (n = 440) found a higher rate of both ampicillin and cefuroxime resistance in Gram-negative bacteria following treatment with ampicillin compared with cefuroxime.62 One study (n = 118) found a higher rate of gentamicin resistance among Gram-negative bacteria following treatment with gentamicin compared with amikacin.68 The remaining five studies (n = 2195) did not formally test for statistically significant differences when comparing broad- versus narrow-spectrum regimens,37,53,54,60,72 although three of these studies (n = 1258) reported increased rates of antibiotic resistance following broad-spectrum therapy.54,60,72 Discussion Key findings To our knowledge, this is the first systematic review to examine antibiotic therapy in neonates and its impact on gut microbiota and/or antibiotic resistance development. The primary findings were the lack of RCTs and large high-quality observational studies and the heterogeneity regarding methodology and outcomes among the included studies. Despite this, there were several salient features in this review. First, prolonged antibiotic therapy was associated with reduced gut microbial diversity.41,44,48 Decreased gut microbial diversity has been associated with early adverse outcomes such as necrotizing enterocolitis, and may have potential long-lasting consequences such as increased likelihood of obesity and inflammatory diseases.10,49,74–77 Combined, these findings imply that prolonged exposure to antibiotic treatment in the neonatal period may increase the likelihood of disease, either in the neonatal period or later in life. However, the QoE for this outcome was graded as very low. It is possible that neonatal antibiotic therapy, regardless of treatment length, leads to decreased microbial diversity, but the studies included in this category were small, and two out of four studies did not detect a significant difference.40,42,44,49 Second, four out of nine studies reported increased abundance and/or colonization rates of Enterobacteriaceae following neonatal antibiotic treatment, whereas none of the studies reported reduced abundance.33,34,36,37,40,42,43,46,47 In the majority of these studies, the empirical regimens consisted of ampicillin and gentamicin. We speculate that intravenous ampicillin also has an impact on Gram-positive gut bacteria despite being mainly secreted through the kidneys,78 whereas intravenous gentamicin, mainly covering Gram-negative bacteria in the bloodstream,79 has a very low penetration into the gut. Combined, this may give undue benefits to Gram-negative Enterobacteriaceae. In contrast, third-generation cephalosporin therapy may lead to a relatively lower abundance of Enterobacteriaceae as both Gram-negative and Gram-positive bacteria are within their spectrum of activity.79 However, the QoE for this outcome was again graded as very low, and even though overgrowth of Enterobacteriaceae in the human gut has previously been associated with necrotizing enterocolitis, inflammatory bowel disease and chronic fatigue syndrome, there is no strong evidence of any causal relationship.74,76,80–82 Third, antibiotic treatment in the neonatal period was strongly associated with reduced abundance of protective commensal anaerobic bacteria such as bifidobacteria, lactobacilli and/or bacteriodes.35,40,43 These bacteria provide colonization resistance against antibiotic-resistant bacteria and potentially pathogenic bacteria such as Enterobacteriaceae and Clostridium difficile.7 Moreover, it is well known that bifidobacteria may reduce expression of inflammatory response genes and stimulate genes promoting the integrity of the mucosal barrier. The QoE for this outcome was graded as very low, but our results are in line with findings in adult populations showing decreased diversity, reduced colonization rates of obligate anaerobes and increased colonization rates of Proteobacteria following antibiotic exposure.83–85 Furthermore, our findings are biologically plausible as reduced numbers of bifidobacteria and lactobacilli seem to increase the risk of necrotizing enterocolitis in preterm infants with an exaggerated inflammatory response.82,86–90 In adults some studies have found larger changes in the gut microbiota than oral microbiota following antibiotic treatment, with greater resilience in the oral communities.84,85 However, we believe that the gut microbiota is of highest clinical relevance, both as the largest reservoir for antibiotic-resistant bacteria and because the gut is characterized as the motor of multiple organ dysfunction syndrome. Fourth, all three categories of neonatal antibiotic treatment investigated in this review were clearly associated with an increased risk of antibiotic resistance development, in particular ESBL-producing Gram-negative bacteria and other MDR bacteria. These findings were based on moderate QoE. Antibiotic resistance genes exist even in the absence of antimicrobial drugs.91,92 Moreover, overuse of antibiotics may lead to increased antibiotic resistance through several mechanisms.91,93 Antibiotics apply a direct selection pressure that gives significant advantages to bacteria expressing resistance genes.94 Antibiotic treatment also contributes to changes in the human gut-associated resistome, which comprises numerous functional antibiotic resistance genes in the gut microbiota.95 Gibson et al.96 recently found that only a fraction of antibiotic resistance genes that are enriched after a specific antibiotic therapy are unique to the particular antibiotic given. Finally, antibiotic treatment appears to reduce colonization resistance against antibiotic-resistant bacteria through the collateral destruction of obligate anaerobic bacteria.7,97 An increase in the gut resistome and a decrease in colonization resistance could theoretically increase horizontal transfer of antibiotic resistance genes from commensals to potential pathogens.98 Although in vivo horizontal transfer between commensals and pathogens in the gut microbiota remains to be shown, there is evidence of exchange of antibiotic resistance genes between environmental bacteria and human pathogens.99 Strengths and limitations The primary strength of this study is our rigorous and sensitive search strategy based on a previously registered search protocol. Additionally, the adverse impact of the developing infant gut microbiota is of great clinical and scientific interest. The main limitations were the lack of RCTs and the diverse study outcomes, which made meta-analysis impossible to perform. Instead, we applied a semi-quantitative vote-counting method to assess the effect of neonatal antibiotic treatment on relevant outcomes. This method has limitations as it usually fails to account for the population size and methodological quality of pooled studies. Nevertheless, vote counting may be an effective method to assess the ranking of outcomes.100 Moreover, we attempted to improve the method by presenting the differential weight of each study with squares corresponding to sample size. The majority of studies included were small and there was considerable heterogeneity in study designs, outcomes, categories of antibiotic treatment and methodological quality. Observational studies are prone to biases and confounding, and only a third of the included studies attempted to adjust for confounding through multivariable regression analysis. Evidence from observational studies is usually considered to be of low quality. However, well-designed observational studies have been shown to provide similar results to RCTs and they can therefore be useful for detecting rare adverse outcomes by allowing larger sample sizes and longer lengths of follow-up than RCTs for lower costs.101 We used the GRADE approach to assess the QoE. Overall, we graded the QoE as very low for all outcomes presented in the gut microbiota category. In contrast, we considered the QoE to be moderate in the antibiotic resistance category owing to large effect sizes and a dose–response effect. Based on current evidence we are therefore moderately confident that all types of antibiotic treatment lead to increased rates of antibiotic resistance. All included studies published prior to 2007 used culture-based techniques to examine the gut microbiota composition. It has been estimated that <20% of environmental bacteria can be grown in defined growth media. This increases the risk of detection bias in older studies.102 Sequencing-based techniques also have limitations. Studies relying on 16S rRNA analysis allow only a coarse sorting of bacteria mainly at phylum level. Deep shotgun metagenome sequencing allows for finer distinction at the genus or species level, but it is of crucial importance to standardize sampling and temperature control during the pipeline up to DNA extraction in order to obtain valid results.103 Moreover bioinformatic presentations are often challenging to understand and interpret. We also acknowledge that our definition of broad-spectrum and narrow-spectrum antibiotics is somewhat arbitrary as most of the narrow-spectrum regimens covered both Gram-negative and Gram-positive bacteria. However, our study confirms previous findings, clearly suggesting that antibiotic regimens containing third-generation cephalosporins or carbapenems are more frequently associated with antibiotic resistance development than regimens with aminoglycosides for Gram-negative coverage.15,24,26,28,50,55,64,65,67,71 Finally, we decided to exclude studies that only examined antenatal antibiotic treatment, despite the frequent use of intrapartum antibiotic prophylaxis for prevention of neonatal infections and its reported effects on the infant gut microbiota and carriage of antibiotic resistance genes.104 The focus of this review was neonatal antibiotic treatment given for suspected neonatal infection, and the isolated effects of antenatal antibiotics given to infants who did not receive antibiotics after birth were beyond the scope of this study. Implications and conclusions This systematic review highlights the profound impact on the gut microbiota and antibiotic resistance development exerted by antibiotic treatment in neonates. Antibiotic exposure in the neonatal period appears to induce various potentially disease-promoting alterations in the gut microbiota, but the QoE was very low for the outcomes investigated in this review. However, we are moderately confident, based on data from this review, that antibiotic treatment leads to antibiotic resistance development, in particular in Gram-negative bacteria. This clearly threatens current empirical antibiotic regimens and is a finding of great concern. In conclusion, the findings from this systematic review, along with the findings from our recent systematic review on early adverse outcome of neonatal antibiotic therapy,9 strongly emphasize the need to reduce unnecessary antibiotic treatment in neonates. Important steps to reduce the burden of neonatal antibiotic therapy include improving preventive measures such as hand hygiene, stopping antibiotic therapy after 36–48 h if infection is only vaguely suspected and there is no growth in the blood culture, and restricting the empirical use of broad-spectrum antibiotic treatment.105,106 Funding J. W. F. and E. E. are recipients of funding for their PhD studies from the Northern Norway Regional Health Authority. This study was supported by internal funding. The funding source had no role in the: design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review or approval of the manuscript; and decision to submit the manuscript for publication. Transparency declarations None to declare. Author contributions J. W. F. reviewed all relevant titles, abstracts and full-text articles, assessed quality, extracted data and drafted the initial manuscript. E. E. reviewed all relevant titles, abstracts and full-text articles, extracted data, assessed quality and revised the manuscript. L. K. J. contributed to study design, methodological assessment and revised the manuscript. J. N. v. d. A. contributed to study design and revised the manuscript. C. K. conceptualized and designed the study, reviewed relevant abstracts and articles, assessed quality and revised the manuscript. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work. J. W. F., E. E. and C. K. have full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Supplementary data Tables S1 and S2 and Figure S1 are available as Supplementary data at JAC Online. References 1 Neu J. The microbiome during pregnancy and early postnatal life. Semin Fetal Neonatal Med  2016; 21: 373– 9. Google Scholar CrossRef Search ADS PubMed  2 Azad MB, Konya T, Persaud RR et al.   Impact of maternal intrapartum antibiotics, method of birth and breastfeeding on gut microbiota during the first year of life: a prospective cohort study. BJOG  2016; 123: 983– 93. Google Scholar CrossRef Search ADS PubMed  3 Hartz LE, Bradshaw W, Brandon DH. Potential NICU environmental influences on the neonate’s microbiome: a systematic review. Adv Neonatal Care  2015; 15: 324– 35. Google Scholar CrossRef Search ADS PubMed  4 Koenig JE, Spor A, Scalfone N et al.   Succession of microbial consortia in the developing infant gut microbiome. Proc Natl Acad Sci U S A  2011; 108 Suppl 1: 4578– 85. Google Scholar CrossRef Search ADS PubMed  5 Tarr PI, Warner BB. Gut bacteria and late-onset neonatal bloodstream infections in preterm infants. Semin Fetal Neonatal Med  2016; 21: 388– 93. Google Scholar CrossRef Search ADS PubMed  6 Gensollen T, Iyer SS, Kasper DL et al.   How colonization by microbiota in early life shapes the immune system. Science  2016; 352: 539– 44. Google Scholar CrossRef Search ADS PubMed  7 Pamer EG. Resurrecting the intestinal microbiota to combat antibiotic-resistant pathogens. Science  2016; 352: 535– 8. Google Scholar CrossRef Search ADS PubMed  8 Hsieh EM, Hornik CP, Clark RH et al.   Medication use in the neonatal intensive care unit. Am J Perinatol  2014; 31: 811– 21. Google Scholar CrossRef Search ADS PubMed  9 Esaiassen E, Fjalstad JW, Juvet LK et al.   Antibiotic exposure in neonates and early adverse outcomes: a systematic review and meta-analysis. J Antimicrob Chemother  2017; 72: 1858– 70. Google Scholar CrossRef Search ADS PubMed  10 Turnbaugh PJ, Hamady M, Yatsunenko T et al.   A core gut microbiome in obese and lean twins. Nature  2009; 457: 480– 4. Google Scholar CrossRef Search ADS PubMed  11 Saari A, Virta LJ, Sankilampi U et al.   Antibiotic exposure in infancy and risk of being overweight in the first 24 months of life. Pediatrics  2015; 135: 617– 26. Google Scholar CrossRef Search ADS PubMed  12 Sim K, Shaw AG, Randell P et al.   Dysbiosis anticipating necrotizing enterocolitis in very premature infants. Clin Infect Dis  2015; 60: 389– 97. Google Scholar CrossRef Search ADS PubMed  13 Hviid A, Svanstrom H, Frisch M. Antibiotic use and inflammatory bowel diseases in childhood. Gut  2011; 60: 49– 54. Google Scholar CrossRef Search ADS PubMed  14 van Nimwegen FA, Penders J, Stobberingh EE et al.   Mode and place of delivery, gastrointestinal microbiota, and their influence on asthma and atopy. J Allergy Clin Immunol  2011; 128: 948– 55.e1–3. Google Scholar CrossRef Search ADS PubMed  15 de Man P, Verhoeven BA, Verbrugh HA et al.   An antibiotic policy to prevent emergence of resistant bacilli. Lancet  2000; 355: 973– 8. Google Scholar CrossRef Search ADS PubMed  16 Downie L, Armiento R, Subhi R et al.   Community-acquired neonatal and infant sepsis in developing countries: efficacy of WHO’s currently recommended antibiotics–systematic review and meta-analysis. Arch Dis Child  2013; 98: 146– 54. Google Scholar CrossRef Search ADS PubMed  17 Blackburn RM, Verlander NQ, Heath PT et al.   The changing antibiotic susceptibility of bloodstream infections in the first month of life: informing antibiotic policies for early- and late-onset neonatal sepsis. Epidemiol Infect  2014; 142: 803– 11. Google Scholar CrossRef Search ADS PubMed  18 Stapleton PJ, Murphy M, McCallion N et al.   Outbreaks of extended spectrum β-lactamase-producing Enterobacteriaceae in neonatal intensive care units: a systematic review. Arch Dis Child Fetal Neonatal Ed  2016; 101: F72– 8. Google Scholar CrossRef Search ADS PubMed  19 Laxminarayan R, Matsoso P, Pant S et al.   Access to effective antimicrobials: a worldwide challenge. Lancet  2016; 387: 168– 75. Google Scholar CrossRef Search ADS PubMed  20 Liberati A, Altman DG, Tetzlaff J et al.   The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol  2009; 62: e1– 34. Google Scholar CrossRef Search ADS PubMed  21 Klingenberg C, Fjalstad J, Esaiassen E et al.   A Systematic Review of Early Adverse Effects Associated with Antibiotic Exposure in the Neonatal Period. PROSPERO. http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42015026743. 22 Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0. 2011. http://www.handbook.cochrane.org/. 23 Hiergeist A, Glasner J, Reischl U et al.   Analyses of intestinal microbiota: culture versus sequencing. ILAR J  2015; 56: 228– 40. Google Scholar CrossRef Search ADS PubMed  24 de Araujo OR, da Silva DC, Diegues AR et al.   Cefepime restriction improves gram-negative overall resistance patterns in neonatal intensive care unit. Braz J Infect Dis  2007; 11: 277– 80. Google Scholar CrossRef Search ADS PubMed  25 Giuffre M, Geraci DM, Bonura C et al.   The increasing challenge of multidrug-resistant gram-negative bacilli: results of a 5-year active surveillance program in a neonatal intensive care unit. Medicine (Baltimore)  2016; 95: e3016. Google Scholar CrossRef Search ADS PubMed  26 Mammina C, Di Carlo P, Cipolla D et al.   Surveillance of multidrug-resistant gram-negative bacilli in a neonatal intensive care unit: prominent role of cross transmission. Am J Infect Control  2007; 35: 222– 30. Google Scholar CrossRef Search ADS PubMed  27 Millar M, Philpott A, Wilks M et al.   Colonization and persistence of antibiotic-resistant Enterobacteriaceae strains in infants nursed in two neonatal intensive care units in East London, United Kingdom. J Clin Microbiol  2008; 46: 560– 7. Google Scholar CrossRef Search ADS PubMed  28 Thatrimontrichai A, Apisarnthanarak A, Chanvitan P et al.   Risk factors and outcomes of carbapenem-resistant Acinetobacter baumannii bacteremia in neonatal intensive care unit: a case-case-control study. Pediatr Infect Dis J  2013; 32: 140– 5. Google Scholar CrossRef Search ADS PubMed  29 Viswanathan M, Berkman ND, Dryden DM et al.   AHRQ Methods for Effective Health Care. Assessing Risk of Bias and Confounding in Observational Studies of Interventions or Exposures: Further Development of the RTI Item Bank . Rockville, MD: Agency for Healthcare Research and Quality (US), 2013. 30 GRADE Working Group. Handbook for Grading the Quality of Evidence and the Strength of Recommendations Using the GRADE Approach. http://gdt.guidelinedevelopment.org/app/handbook/handbook.html#h.g2dqzi9je57e. 31 Parm U, Metsvaht T, Sepp E et al.   Impact of empiric antibiotic regimen on bowel colonization in neonates with suspected early onset sepsis. Eur J Clin Microbiol Infect Dis  2010; 29: 807– 16. Google Scholar CrossRef Search ADS PubMed  32 Westerbeek EA, Slump RA, Lafeber HN et al.   The effect of enteral supplementation of specific neutral and acidic oligosaccharides on the faecal microbiota and intestinal microenvironment in preterm infants. Eur J Clin Microbiol Infect Dis  2013; 32: 269– 76. Google Scholar CrossRef Search ADS PubMed  33 Arboleya S, Sanchez B, Milani C et al.   Intestinal microbiota development in preterm neonates and effect of perinatal antibiotics. J Pediatr  2015; 166: 538– 44. Google Scholar CrossRef Search ADS PubMed  34 Bennet R, Eriksson M, Nord CE et al.   Fecal bacterial microflora of newborn infants during intensive care management and treatment with five antibiotic regimens. Pediatr Infect Dis  1986; 5: 533– 9. Google Scholar CrossRef Search ADS PubMed  35 Bennet R, Nord CE. Development of the faecal anaerobic microflora after Caesarean section and treatment with antibiotics in newborn infants. Infection  1987; 15: 332– 6. Google Scholar CrossRef Search ADS PubMed  36 Blakey JL, Lubitz L, Barnes GL. Development of gut colonisation in pre-term neonates. J Med Microbiol  1982; 15: 519– 29. Google Scholar CrossRef Search ADS PubMed  37 Bonnemaison E, Lanotte P, Cantagrel S et al.   Comparison of fecal flora following administration of two antibiotic protocols for suspected maternofetal infection. Biol Neonate  2003; 84: 304– 10. Google Scholar CrossRef Search ADS PubMed  38 Butel MJ, Suau A, Campeotto F et al.   Conditions of bifidobacterial colonization in preterm infants: a prospective analysis. J Pediatr Gastroenterol Nutr  2007; 44: 577– 82. Google Scholar CrossRef Search ADS PubMed  39 Ferraris L, Butel MJ, Campeotto F et al.   Clostridia in premature neonates’ gut: incidence, antibiotic susceptibility, and perinatal determinants influencing colonization. PLoS One  2012; 7: e30594. Google Scholar CrossRef Search ADS PubMed  40 Fouhy F, Guinane CM, Hussey S et al.   High-throughput sequencing reveals the incomplete, short-term recovery of infant gut microbiota following parenteral antibiotic treatment with ampicillin and gentamicin. Antimicrob Agents Chemother  2012; 56: 5811– 20. Google Scholar CrossRef Search ADS PubMed  41 Gewolb IH, Schwalbe RS, Taciak VL et al.   Stool microflora in extremely low birthweight infants. Arch Dis Child Fetal Neonatal Ed  1999; 80: F167– 73. Google Scholar CrossRef Search ADS PubMed  42 Greenwood C, Morrow AL, Lagomarcino AJ et al.   Early empiric antibiotic use in preterm infants is associated with lower bacterial diversity and higher relative abundance of Enterobacter. J Pediatr  2014; 165: 23– 9. Google Scholar CrossRef Search ADS PubMed  43 Hall MA, Cole CB, Smith SL et al.   Factors influencing the presence of faecal lactobacilli in early infancy. Arch Dis Child  1990; 65: 185– 8. Google Scholar CrossRef Search ADS PubMed  44 Jacquot A, Neveu D, Aujoulat F et al.   Dynamics and clinical evolution of bacterial gut microflora in extremely premature patients. J Pediatr  2011; 158: 390– 6. Google Scholar CrossRef Search ADS PubMed  45 Jenke AC, Postberg J, Mariel B et al.   S100A12 and hBD2 correlate with the composition of the fecal microflora in ELBW infants and expansion of E. coli is associated with NEC. Biomed Res Int ; 2013: 150372. PubMed  46 La Rosa PS, Warner BB, Zhou Y et al.   Patterned progression of bacterial populations in the premature infant gut. Proc Natl Acad Sci U S A  2014; 111: 12522– 7. Google Scholar CrossRef Search ADS PubMed  47 Tullus K, Berglund B, Fryklund B et al.   Influence of antibiotic therapy of faecal carriage of P-fimbriated Escherichia coli and other Gram-negative bacteria in neonates. J Antimicrob Chemother  1988; 22: 563– 8. Google Scholar CrossRef Search ADS PubMed  48 Ward DV, Scholz M, Zolfo M et al.   Metagenomic sequencing with strain-level resolution implicates uropathogenic E. coli in necrotizing enterocolitis and mortality in preterm infants. Cell Rep  2016; 14: 2912– 24. Google Scholar CrossRef Search ADS PubMed  49 Zhou Y, Shan G, Sodergren E et al.   Longitudinal analysis of the premature infant intestinal microbiome prior to necrotizing enterocolitis: a case-control study. PLoS One  2015; 10: e0118632. Google Scholar CrossRef Search ADS PubMed  50 Abdel-Hady H, Hawas S, El-Daker M et al.   Extended-spectrum β-lactamase producing Klebsiella pneumoniae in neonatal intensive care unit. J Perinatol  2008; 28: 685– 90. Google Scholar CrossRef Search ADS PubMed  51 Acolet D, Ahmet Z, Houang E et al.   Enterobacter cloacae in a neonatal intensive care unit: account of an outbreak and its relationship to use of third generation cephalosporins. J Hosp Infect  1994; 28: 273– 86. Google Scholar CrossRef Search ADS PubMed  52 Bergin SP, Thaden JT, Ericson JE et al.   Neonatal Escherichia coli bloodstream infections: clinical outcomes and impact of initial antibiotic therapy. Pediatr Infect Dis J  2015; 34: 933– 6. Google Scholar CrossRef Search ADS PubMed  53 Burman LG, Haeggman S, Kuistila M et al.   Epidemiology of plasmid-mediated β-lactamases in enterobacteria Swedish neonatal wards and relation to antimicrobial therapy. Antimicrob Agents Chemother  1992; 36: 989– 92. Google Scholar CrossRef Search ADS PubMed  54 Burman LG, Berglund B, Huovinen P et al.   Effect of ampicillin versus cefuroxime on the emergence of β-lactam resistance in faecal Enterobacter cloacae isolates from neonates. J Antimicrob Chemother  1993; 31: 111– 6. Google Scholar CrossRef Search ADS PubMed  55 Calil R, Marba ST, von Nowakonski A et al.   Reduction in colonization and nosocomial infection by multiresistant bacteria in a neonatal unit after institution of educational measures and restriction in the use of cephalosporins. Am J Infect Control  2001; 29: 133– 8. Google Scholar CrossRef Search ADS PubMed  56 Cantey JB, Wozniak PS, Pruszynski JE et al.   Reducing unnecessary antibiotic use in the neonatal intensive care unit (SCOUT): a prospective interrupted time-series study. Lancet Infect Dis  2016; 16: 1178– 84. Google Scholar CrossRef Search ADS PubMed  57 Crivaro V, Bagattini M, Salza MF et al.   Risk factors for extended-spectrum β-lactamase-producing Serratia marcescens and Klebsiella pneumoniae acquisition in a neonatal intensive care unit. J Hosp Infect  2007; 67: 135– 41. Google Scholar CrossRef Search ADS PubMed  58 De Champs C. Clinical and bacteriological survey after change in aminoglycoside treatment to control an epidemic of Enterobacter cloacae. J Hosp Infect  1994; 28: 219– 29. Google Scholar CrossRef Search ADS PubMed  59 Duman M, Abacioglu H, Karaman M et al.   β-lactam antibiotic resistance in aerobic commensal fecal flora of newborns. Pediatr Int  2005; 47: 267– 73. Google Scholar CrossRef Search ADS PubMed  60 Gaynes RP, Simpson D, Reeves SA et al.   A nursery outbreak of multiple-aminoglycoside-resistant Escherichia coli. Infect Control  1984; 5: 519– 24. Google Scholar CrossRef Search ADS PubMed  61 Isaacs D, Catterson J, Hope PL et al.   Factors influencing colonisation with gentamicin resistant gram negative organisms in the neonatal unit. Arch Dis Child  1988; 63: 533– 5. Google Scholar CrossRef Search ADS PubMed  62 Kalenic S, Francetic I, Polak J et al.   Impact of ampicillin and cefuroxime on bacterial colonization and infection in patients on a neonatal intensive care unit. J Hosp Infect  1993; 23: 35– 41. Google Scholar CrossRef Search ADS PubMed  63 Kumar A, Randhawa VS, Nirupam N et al.   Risk factors for carbapenem-resistant Acinetobacter baumanii blood stream infections in a neonatal intensive care unit, Delhi, India. J Infect Dev Ctries  2014; 8: 1049– 54. Google Scholar PubMed  64 Le J, Nguyen T, Okamoto M et al.   Impact of empiric antibiotic use on development of infections caused by extended-spectrum β-lactamase bacteria in a neonatal intensive care unit. Pediatr Infect Dis J  2008; 27: 314– 8. Google Scholar CrossRef Search ADS PubMed  65 Linkin DR, Fishman NO, Patel JB et al.   Risk factors for extended-spectrum β-lactamase-producing Enterobacteriaceae in a neonatal intensive care unit. Infect Control Hosp Epidemiol  2004; 25: 781– 3. Google Scholar CrossRef Search ADS PubMed  66 Noy JH, Ayliffe GA, Linton KB. Antibiotic-resistant Gram-negative bacilli in the faeces of neonates. J Med Microbiol  1974; 7: 509– 20. Google Scholar CrossRef Search ADS PubMed  67 Pessoa-Silva CL, Meurer Moreira B, Camara Almeida V et al.   Extended-spectrum β-lactamase-producing Klebsiella pneumoniae in a neonatal intensive care unit: risk factors for infection and colonization. J Hosp Infect  2003; 53: 198– 206. Google Scholar CrossRef Search ADS PubMed  68 Raz R, Sharir R, Shmilowitz L et al.   The elimination of gentamicin-resistant gram-negative bacteria in a newborn intensive care unit. Infection  1987; 15: 32– 4. Google Scholar CrossRef Search ADS PubMed  69 Rettedal S, Hoyland LI, Natas O et al.   Risk factors for acquisition of CTX-M-15 extended-spectrum β-lactamase-producing Klebsiella pneumoniae during an outbreak in a neonatal intensive care unit in Norway. Scand J Infect Dis  2013; 45: 54– 8. Google Scholar CrossRef Search ADS PubMed  70 Sehgal R, Gaind R, Chellani H et al.   Extended-spectrum β lactamase-producing Gram-negative bacteria: clinical profile and outcome in a neonatal intensive care unit. Ann Trop Paediatr  2007; 27: 45– 54. Google Scholar CrossRef Search ADS PubMed  71 Thatrimontrichai A, Techato C, Dissaneevate S et al.   Risk factors and outcomes of carbapenem-resistant Acinetobacter baumannii ventilator-associated pneumonia in the neonate: a case-case-control study. J Infect Chemother  2016; 22: 444– 9. Google Scholar CrossRef Search ADS PubMed  72 Toltzis P, Dul MJ, Hoyen C et al.   Molecular epidemiology of antibiotic-resistant Gram-negative bacilli in a neonatal intensive care unit during a nonoutbreak period. Pediatrics  2001; 108: 1143– 8. Google Scholar CrossRef Search ADS PubMed  73 Goldmann DA, Leclair J, Macone A. Bacterial colonization of neonates admitted to an intensive care environment. J Pediatr  1978; 93: 288– 93. Google Scholar CrossRef Search ADS PubMed  74 Warner BB, Deych E, Zhou Y et al.   Gut bacteria dysbiosis and necrotising enterocolitis in very low birthweight infants: a prospective case-control study. Lancet  2016; 387: 1928– 36. Google Scholar CrossRef Search ADS PubMed  75 Sundin J, Rangel I, Fuentes S et al.   Altered faecal and mucosal microbial composition in post-infectious irritable bowel syndrome patients correlates with mucosal lymphocyte phenotypes and psychological distress. Aliment Pharmacol Ther  2015; 41: 342– 51. Google Scholar CrossRef Search ADS PubMed  76 Giloteaux L, Goodrich JK, Walters WA et al.   Reduced diversity and altered composition of the gut microbiome in individuals with myalgic encephalomyelitis/chronic fatigue syndrome. Microbiome  2016; 4: 30. Google Scholar CrossRef Search ADS PubMed  77 Bisgaard H, Li N, Bonnelykke K et al.   Reduced diversity of the intestinal microbiota during infancy is associated with increased risk of allergic disease at school age. J Allergy Clin Immunol  2011; 128: 646– 52.e1–5. Google Scholar CrossRef Search ADS PubMed  78 Zhang L, Huang Y, Zhou Y et al.   Antibiotic administration routes significantly influence the levels of antibiotic resistance in gut microbiota. Antimicrob Agents Chemother  2013; 57: 3659– 66. Google Scholar CrossRef Search ADS PubMed  79 Roberts JK, Stockmann C, Constance JE et al.   Pharmacokinetics and pharmacodynamics of antibacterials, antifungals, and antivirals used most frequently in neonates and infants. Clin Pharmacokinet  2014; 53: 581– 610. Google Scholar CrossRef Search ADS PubMed  80 Torrazza RM, Ukhanova M, Wang X et al.   Intestinal microbial ecology and environmental factors affecting necrotizing enterocolitis. PLoS One  2013; 8: e83304. Google Scholar CrossRef Search ADS PubMed  81 Belkaid Y, Hand TW. Role of the microbiota in immunity and inflammation. Cell  2014; 157: 121– 41. Google Scholar CrossRef Search ADS PubMed  82 Jiang H, Ling Z, Zhang Y et al.   Altered fecal microbiota composition in patients with major depressive disorder. Brain Behav Immun  2015; 48: 186– 94. Google Scholar CrossRef Search ADS PubMed  83 Yap TW, Gan HM, Lee YP et al.   Helicobacter pylori eradication causes perturbation of the human gut microbiome in young adults. PLoS One  2016; 11: e0151893. Google Scholar CrossRef Search ADS PubMed  84 Zaura E, Brandt BW, Teixeira de Mattos MJ et al.   Same exposure but two radically different responses to antibiotics: resilience of the salivary microbiome versus long-term microbial shifts in feces. MBio  2015; 6: e01693-15. Google Scholar CrossRef Search ADS PubMed  85 Jakobsson HE, Jernberg C, Andersson AF et al.   Short-term antibiotic treatment has differing long-term impacts on the human throat and gut microbiome. PLoS One  2010; 5: e9836. Google Scholar CrossRef Search ADS PubMed  86 Ahmad OF, Akbar A. Microbiome, antibiotics and irritable bowel syndrome. Br Med Bull  2016; 1: 91– 9. Google Scholar CrossRef Search ADS   87 Dermyshi E, Wang Y, Yan C et al.   The “golden age” of probiotics: a systematic review and meta-analysis of randomized and observational studies in preterm infants. Neonatology  2017; 112: 9– 23. Google Scholar CrossRef Search ADS PubMed  88 Chow J, Tang H, Mazmanian SK. Pathobionts of the gastrointestinal microbiota and inflammatory disease. Curr Opin Immunol  2011; 23: 473– 80. Google Scholar CrossRef Search ADS PubMed  89 Kamada N, Chen GY, Inohara N et al.   Control of pathogens and pathobionts by the gut microbiota. Nat Immunol  2013; 14: 685– 90. Google Scholar CrossRef Search ADS PubMed  90 Wang M, Monaco MH, Donovan SM. Impact of early gut microbiota on immune and metabolic development and function. Semin Fetal Neonatal Med  2016; 21: 380– 7. Google Scholar CrossRef Search ADS PubMed  91 Pallecchi L, Bartoloni A, Paradisi F et al.   Antibiotic resistance in the absence of antimicrobial use: mechanisms and implications. Expert Rev Anti Infect Ther  2008; 6: 725– 32. Google Scholar CrossRef Search ADS PubMed  92 Zhang L, Kinkelaar D, Huang Y et al.   Acquired antibiotic resistance: are we born with it? Appl Environ Microbiol  2011; 77: 7134– 41. Google Scholar CrossRef Search ADS PubMed  93 Patel SJ, Saiman L. Antibiotic resistance in neonatal intensive care unit pathogens: mechanisms, clinical impact, and prevention including antibiotic stewardship. Clin Perinatol  2010; 37: 547– 63. Google Scholar CrossRef Search ADS PubMed  94 Davies J, Davies D. Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev  2010; 74: 417– 33. Google Scholar CrossRef Search ADS PubMed  95 Sommer MO, Church GM, Dantas G. The human microbiome harbors a diverse reservoir of antibiotic resistance genes. Virulence  2010; 1: 299– 303. Google Scholar CrossRef Search ADS PubMed  96 Gibson MK, Wang B, Ahmadi S et al.   Developmental dynamics of the preterm infant gut microbiota and antibiotic resistome. Nat Microbiol  2016; 1: 16024. Google Scholar CrossRef Search ADS PubMed  97 Donskey CJ, Chowdhry TK, Hecker MT et al.   Effect of antibiotic therapy on the density of vancomycin-resistant enterococci in the stool of colonized patients. N Engl J Med  2000; 343: 1925– 32. Google Scholar CrossRef Search ADS PubMed  98 Martinez JL. General principles of antibiotic resistance in bacteria. Drug Discov Today Technol  2014; 11: 33– 9. Google Scholar CrossRef Search ADS PubMed  99 Forsberg KJ, Reyes A, Wang B et al.   The shared antibiotic resistome of soil bacteria and human pathogens. Science  2012; 337: 1107– 11. Google Scholar CrossRef Search ADS PubMed  100 Rikke BA, Wynes MW, Rozeboom LM et al.   Independent validation test of the vote-counting strategy used to rank biomarkers from published studies. Biomark Med  2015; 9: 751– 61. Google Scholar CrossRef Search ADS PubMed  101 Cameron C, Fireman B, Hutton B et al.   Network meta-analysis incorporating randomized controlled trials and non-randomized comparative cohort studies for assessing the safety and effectiveness of medical treatments: challenges and opportunities. Syst Rev  2015; 4: 147. Google Scholar CrossRef Search ADS PubMed  102 Ward DM, Weller R, Bateson MM. 16S rRNA sequences reveal numerous uncultured microorganisms in a natural community. Nature  1990; 345: 63– 5. Google Scholar CrossRef Search ADS PubMed  103 Hanage WP. Microbiology: microbiome science needs a healthy dose of scepticism. Nature  2014; 512: 247– 8. Google Scholar CrossRef Search ADS PubMed  104 Nogacka A, Salazar N, Suarez M et al.   Impact of intrapartum antimicrobial prophylaxis upon the intestinal microbiota and the prevalence of antibiotic resistance genes in vaginally delivered full-term neonates. Microbiome  2017; 5: 93. Google Scholar CrossRef Search ADS PubMed  105 Polin RA; Committee on Fetus and Newborn. Management of neonates with suspected or proven early-onset bacterial sepsis. Pediatrics  2012; 129: 1006– 15. Google Scholar CrossRef Search ADS PubMed  106 National Institute for Health and Care Excellence (NICE). Antibiotics for Early-Onset Neonatal Infection: Antibiotics for the Prevention and Treatment of Early-Onset Neonatal Infection. https://www.nice.org.uk/guidance/cg149. © The Author 2017. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please email: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Antimicrobial Chemotherapy Oxford University Press

Antibiotic therapy in neonates and impact on gut microbiota and antibiotic resistance development: a systematic review

Loading next page...
 
/lp/ou_press/antibiotic-therapy-in-neonates-and-impact-on-gut-microbiota-and-CU0JUBDkIc
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please email: journals.permissions@oup.com.
ISSN
0305-7453
eISSN
1460-2091
D.O.I.
10.1093/jac/dkx426
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives To systematically review the impact of antibiotic therapy in the neonatal period on changes in the gut microbiota and/or antibiotic resistance development. Methods Data sources were PubMed, Embase, Medline and the Cochrane Database, supplemented by manual searches of reference lists. Randomized controlled trials (RCTs) and observational studies were included if they provided data on different categories of antibiotic treatment (yes versus no, long versus short duration and/or broad- versus narrow-spectrum regimens) and subsequent changes in the gut microbiota and/or antibiotic resistance development. We evaluated risk of bias using the Cochrane Handbook, adapted to include observational studies. When appropriate, we used the vote-counting method to perform semi-quantitative meta-analyses. We applied the Grades of Recommendation, Assessment, Development and Evaluation approach to rate the quality of evidence (QoE). Study protocol registration: PROSPERO CRD42015026743. Results We included 48 studies, comprising 3 RCTs and 45 observational studies. Prolonged antibiotic treatment was associated with reduced gut microbial diversity in all three studies investigating this outcome (very low QoE). Antibiotic treatment was associated with reduced colonization rates of protective commensal anaerobic bacteria in four of five studies (very low QoE). However, all three categories of antibiotic treatment were associated with an increased risk of antibiotic resistance development, in particular MDR in Gram-negative bacteria, and we graded the QoE for these outcomes as moderate. Conclusions We are moderately confident that antibiotic treatment leads to antibiotic resistance development in neonates and it may also induce potentially disease-promoting gut microbiota alterations. Our findings emphasize the need to reduce unnecessary antibiotic treatment in neonates. Introduction Upon birth, infants are suddenly exposed to a wide range of bacteria colonizing mucoepithelial surfaces, including the gut.1 The subsequent development of the infant gut microbiota is dynamic, non-resilient and shaped by factors such as mode of delivery, feeding, diet and environment.2–4 A healthy gut microbiota plays a crucial role in the development of the immune system, digestive functions and protection against infections.4–6 The commensal aerobic and anaerobic bacteria are also essential for colonization resistance—the ability to prevent invasion and persistent carriage of pathogenic and antibiotic-resistant bacteria.7 Antibiotics are the most commonly prescribed medications in the neonatal unit.8 However, antibiotic overuse in early life disrupts the actively developing gut microbiota, causing ‘bacterial dysbiosis’, which is associated with an increased risk of early adverse outcomes such as necrotizing enterocolitis and fungal infections.9 Early antibiotic exposure has also been associated with allergic diseases, obesity, diabetes and inflammatory bowel disease later in life.10–14 Overuse of antibiotics, particularly broad-spectrum antibiotics, applies a selection pressure that favours antibiotic-resistant bacteria and decreases colonization resistance.7,15 The currently observed increase in resistance to aminoglycosides and ampicillin among Gram-negative bacteria has begun to threaten this traditional combination as empirical treatment for neonatal sepsis.16,17 Moreover, the emergence worldwide of ESBL-producing Enterobacteriaceae presents major challenges in managing neonatal sepsis.18 Globally, an estimated 200 000 neonatal deaths are attributed to resistant organisms each year.19 However, the relative impact of different types of antibiotic exposure on the actively developing gut microbiota composition and antibiotic resistance development is not fully understood. The purpose of this systematic review is to identify, critically appraise and synthesize evidence from studies reporting different categories of antibiotic therapy in neonates and their impact on the gut microbiota and/or antibiotic resistance development. We included both observational studies and randomized controlled trials (RCTs) in line with suggestions from the Cochrane handbook stating that systematic reviews of adverse effects will usually need to include non-randomized studies in addition to RCTs. Methods This review was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses following a registered protocol and according to the recommendations given by the Cochrane Handbook for Systematic Reviews and Interventions.20–22 We recently published a systematic review on early clinical adverse effects of neonatal antibiotic treatment based on the same research protocol.9 For this review, our primary research question was: ‘Are different categories of antibiotic treatment in neonates associated with different changes in gut microbiota composition and/or differences in antibiotic resistance development?’ Study protocol registration: PROSPERO CRD42015026743. Search strategy We developed our search strategy in consultation with an epidemiologist, a librarian, a paediatric pharmacologist and a neonatologist. We searched PubMed, Embase, Medline and the Cochrane Database using MeSH terms and free-text searches with no time restrictions (last search 22 December 2016). The first search was conducted with MeSH terms in PubMed, Medline and the Cochrane Database by combining ‘Infant, Newborn’ and ‘Anti-Bacterial Agents’ with one of two outcome terms: ‘Drug Resistance, Bacterial’ or ‘Microbiota’. The Embase database uses its own key words, and we combined ‘Newborn’ and ‘Antibiotic Agent’ with either ‘Antibiotic Resistance’ or ‘Microbiome’. The second search was conducted using free text in PubMed, Medline and Embase combining the keywords: ‘Infant, Low Birth Weight’ or ‘Infant, Postmature’ or ‘Infant, Premature’ or ‘Infant, Newborn’ with ‘Anti-Bacterial Agents’ or ‘Antibiotics’ and one of the following: ‘Antibiotic Resistance’ or ‘Antibacterial Drug Resistance’ or ‘Microbiota’ or ‘Microbiome’ or ‘Microbiomes’ or ‘Gut Flora’. We examined reference lists of included studies and relevant reviews to identify additional eligible studies. We then combined all citations and excluded duplicates or triplicates. We did not contact authors for supplementary information and we did not perform searches in the grey literature. Study selection and eligibility criteria A study was eligible for review if it reported different categories of intravenous antibiotic treatment in the neonatal period and evaluated their impact on changes in the gut microbiota and/or antibiotic resistance development. If infants were born prematurely we defined the neonatal period as up to 44 weeks postmenstrual age. We compared three different categories of antibiotic therapy: (i) antibiotic treatment yes versus no; (ii) antibiotic treatment long versus short; and (iii) antibiotic treatment broad versus narrow spectrum. For category (ii), we suggested in advance that ‘prolonged’ antibiotic exposure was always ≥3 days or the longest regimen among two antibiotic regimens compared. For category (iii), we always defined regimens including third-generation cephalosporins or carbapenems as broad-spectrum regimens when compared with regimens containing aminoglycosides for coverage against Gram-negative bacteria. This definition was based on previous reports indicating that empirical therapy containing a third-generation cephalosporin for Gram-negative coverage induces significantly more resistance than a regimen containing an aminoglycoside.15 If two similar regimens were compared, the regimen with the broadest spectrum was labelled broad spectrum. Both RCTs and observational studies such as cohort, case–control and cross-sectional studies were eligible for inclusion. We included case–control studies reporting on the prespecified outcomes if data on antibiotic therapy prior to the outcomes were presented as extractable data on cases and controls. We excluded case reports and case series, studies with a non-neonatal or non-human population, studies that were written in languages other than English, and studies that investigated antenatal antibiotics, oral antibiotics and/or low-dose intravenous vancomycin prophylaxis. Screening, data extraction and management Two reviewers (J. W. F. and E. E.) independently screened search results and assessed each potentially eligible study per our predetermined inclusion and exclusion criteria. We only excluded studies that we agreed were irrelevant according to our predefined criteria. A third researcher (C. K.) had the deciding vote in cases of disagreement. We extracted the following information from included studies: author, year, country, study design, study population, including gestational age (GA) and birth weight (BW), comparison of outcomes between groups with different categories of antibiotic treatment, and, if available, risk estimates with 95% CI for the specific outcome. Gut microbiota analyses were based on faecal samples using both standard culture-based methods and culture-independent methods relying on DNA amplification and sequencing.23 After reviewing the articles presenting data on gut microbiota we decided to present data from these studies in three main categories: microbial load, microbial diversity and microbial composition, clearly acknowledging some overlap between these categories. We defined microbial load as the total number of bacteria in a sample, microbial diversity as the number of different bacterial genera or species in a sample, and microbial composition as the taxonomical composition in a sample. Antibiotic resistance development was based on detection of antibiotic susceptibility patterns in bacteria isolated from blood, urine, cerebrospinal fluid, faeces, tracheal aspirates and/or the skin surface. We defined MDR bacteria as bacteria resistant to ≥2 unrelated classes of antibiotics or broad-spectrum antibiotics.24–28 Included in this category were ESBL-producing Gram-negative bacteria, carbapenem-resistant Acinetobacter baumannii and Gram-negative bacteria resistant to third-generation cephalosporins. Antibiotic-resistant bacteria that did not meet any of these criteria were defined as ‘other antibiotic-resistant bacteria’. We applied a simple vote-counting method to investigate whether the different categories of antibiotic therapy had any effect on the outcomes of interest.22 Studies were classified based on whether they showed a reduction in the outcome measure, no effect or an increase in the outcome measure following antibiotic treatment. When appropriate, outcomes were presented in vote-count figures. Assessment of methodological quality Methodological quality was assessed using the Cochrane Handbook of Systematic Reviews of Interventions and recently published suggestions on how to assess risk of bias and confounding in observational studies.22,29 Five domains related to risk of bias were assessed for each study and included: Selection, Performance, Detection, Reporting and Confounding. Risks of bias were judged as low, high or unclear for each domain. The risk of reporting bias was considered unclear in studies that did not have a previously published protocol. The risk of detection bias was considered high in studies that examined gut microbiota with culture-based methods, unclear in studies that applied 16S rRNA sequencing techniques and low in studies that applied shotgun metagenome sequencing techniques. Two reviewers (J. W. F. and E. E.) assessed the risks of bias for each study. Disagreements in the categorization process were resolved after discussion between J. W. F., E. E. and C. K. We applied the Grades of Recommendation, Assessment, Development and Evaluation Working Group (GRADE) approach to rate the quality of evidence (QoE) for each relevant outcome category.30 This approach specifies four levels of quality from high to very low, which define the degree to which its estimates of effects or associations can be trusted.22,30 RCTs started as high QoE and observational studies started as low QoE.30 Several factors could either downgrade or upgrade the quality rating. Results Overview of included studies From 3380 identified studies, we reviewed 137 potentially eligible full-text articles. Forty-eight studies met our inclusion criteria: 3 RCTs published between 2000 and 2013,15,31,32 and 45 observational studies published between 1974 and 2016 (Figure 1).24–28,33–73 Two articles presented data from the same study population and were defined as one study.34,35 Antibiotic treatment was the randomized intervention in two out of the three included RCTs.15,31,32 Among the 45 observational studies, there were 22 prospective cohort studies, 12 case–control studies, 7 before–after studies and 4 retrospective cohort studies. There were large variations regarding the categories of antibiotic therapy studied. Tables S1 and S2 (available as Supplementary data at JAC Online) display the main characteristics and primary outcomes of interest from the 48 included studies. Figure 1. View largeDownload slide Preferred reporting items for systematic reviews and meta-analysis (PRISMA) flow diagram. Figure 1. View largeDownload slide Preferred reporting items for systematic reviews and meta-analysis (PRISMA) flow diagram. Risk of bias and QoE Figure S1(a and b) displays risk of bias assessments for each included study. Outcomes adjusted for differences in populations were reported in 16/45 (36%) observational studies.25,26,28,39,44,46,50,52,55,57,63,64,67,69–71 Five of these studies used stratification or multivariate analysis to adjust for antenatal antibiotic treatment as a potentially confounding variable. None of the RCTs were included in public registries. We graded the QoE as very low for the outcomes microbial load and microbial diversity in relation to the three different categories of antibiotic treatment owing to inclusion of observational studies with serious risk of bias and inconsistent results. We graded the QoE as very low for the outcomes relating to microbial composition after antibiotic treatment (Figure 2a–d). We graded the QoE as moderate for the outcomes relating to antibiotic resistance development owing to inclusion of observational studies that either had large effect sizes (yes versus no and broad versus narrow) or a dose–response effect (long versus short) after antibiotic treatment (Figure 3a–c). Figure 2. View large Download slide View large Download slide Vote count on gut microbial composition after antibiotic exposure compared with no antibiotic exposure. The sizes of the squares are proportional to study populations. An asterisk symbolizes a lack of testing for statistical significance. Figure 2. View large Download slide View large Download slide Vote count on gut microbial composition after antibiotic exposure compared with no antibiotic exposure. The sizes of the squares are proportional to study populations. An asterisk symbolizes a lack of testing for statistical significance. Figure 3. View large Download slide View large Download slide Vote count on infection and/or colonization with MDR Gram-negative bacteria following antibiotic exposure. The sizes of squares are proportional to study populations. A dagger symbolizes multivariate regression analysis. Figure 3. View large Download slide View large Download slide Vote count on infection and/or colonization with MDR Gram-negative bacteria following antibiotic exposure. The sizes of squares are proportional to study populations. A dagger symbolizes multivariate regression analysis. Gut microbiota composition Nineteen studies reported on antibiotic exposure and impact on the gut microbiota composition (Table S1); these comprised 2 RCTs31,32 and 17 observational studies.33–49,73 Three studies reported outcome data from both antibiotic treatment yes versus no and broad versus narrow spectrum,34,37,47 and one study reported outcome after antibiotic treatment yes versus no and long versus short.42 The remaining 15 studies reported outcome data from one category of antibiotic treatment. To examine gut microbiota composition, nine studies used 16S rRNA gene-sequence analysis,33,38–40,42,44–46,49 one used fluorescence in situ hybridization techniques,32 one used deep shotgun metagenome sequence analysis48 and eight used standard culture-based methods.31,34–37,41,43,47,73 The included studies reported primarily taxonomic data with different hierarchical details on (i) Enterobacteriaceae, (ii) obligate commensal anaerobic bacteria (bacteroides, lactobacilli and bifidobacteria, etc.), (iii) clostridia and/or (iv) Gram-positive cocci. Microbial load Three studies (296 neonates) compared the impact of antibiotic treatment (yes versus no) on microbial loads.32,34,40 One study (165 neonates) found increased microbial loads,34 one RCT (113 preterm neonates) found decreased microbial loads32 and one study (18 term neonates) found no significant differences in microbial loads following antibiotic treatment.40 A small study of extremely low-birth-weight neonates found an inverse correlation between the duration of antibiotic therapy and the microbial load on day 30 of life.41 Microbial diversity Four studies (159 neonates) compared microbial diversity after antibiotic treatment (yes versus no).40,42,44,49 Two studies (112 preterm neonates) reported decreased diversity among antibiotic-treated neonates42,49 and two studies (47 neonates) reported no significant differences.40,44 Three studies (224 preterm neonates) examined the impact of antibiotic therapy duration (long versus short) on microbial diversity, and all three found decreased diversity following prolonged therapy.41,44,48 Microbial composition Figure 2 displays the results of studies reporting the impact of antibiotic treatment (yes versus no) on microbial composition. Nine studies focused on Enterobacteriaceae; four reported an increase and five studies reported unchanged composition after antibiotic treatment, mainly ampicillin plus an aminoglycoside (Figure 2a).33,34,36,37,40,42,43,46,47 Five studies focused on different commensal obligate anaerobes, showing a clear trend towards reduced colonization rates following treatment (Figure 2b).35,36,38,40,43 In the five studies focusing on clostridia, there were equivocal results (Figure 2c).36,39,40,45,46 Finally, four studies focused on Gram-positive cocci, and these studies showed either unchanged or higher colonization rates after antibiotic treatment (Figure 2d).33,36,37,40 Two studies (n = 983) reported Enterobacteriaceae colonization rates after treatment with broad- versus narrow-spectrum antibiotics.37,47 Both studies found lower colonization rates following third-generation cephalosporin treatment. One study of preterm infants (n = 76) reported lower colonization rates of clostridia in those who received ≥10 days of antibiotic therapy compared with shorter duration.39 Another study of preterm infants (n = 74) reported higher colonization rates of staphylococci in those who received ≥5 days of antibiotic treatment compared with shorter-duration therapy.42 Finally, two studies (n = 104) compared the impact of antibiotic therapy (broad versus narrow) on abundance and/or colonization rates with staphylococci, but neither found any significant differences.37,42 Antibiotic resistance development Thirty-one studies, 2 RCTs15,31 and 29 observational studies,24–28,37,50–72 evaluated the risk of antibiotic resistance development after antibiotic exposure (Table S2). Five studies reported outcome after antibiotic treatment yes versus no and broad versus narrow spectrum.27,37,53,55,67 Two studies reported outcome after antibiotic treatment long versus short duration and broad versus narrow spectrum.26,64 Two studies reported outcome after antibiotic treatment yes versus no and long versus short duration.25,57 The remaining 23 studies assessed only one category of antibiotic therapy. Nine studies reported on both infections and colonization with antibiotic-resistant bacteria,24,51,57,58,60,62,65,67,68 whereas 15 studies only reported on colonization,15,25–27,31,37,53–56,59,61,66,69,72 and 7 studies only reported on infections.28,50,52,63,64,70,71 MDR bacteria were varyingly defined as bacteria that were resistant to both third-generation cephalosporins and aminoglycosides55,58 or bacteria resistant to ≥2 or ≥3 unrelated classes of antibiotics.24–28 Thirty of 31 studies focused solely on antibiotic resistance development in Gram-negative bacteria. Among these, 20 studies focused on MDR Gram-negative bacteria. MDR Gram-negative bacteria Figure 3 displays the results of the 20 studies reporting the impact of antibiotic exposure on rates of infection and/or colonization with MDR Gram-negative bacteria. Nine studies reported data after antibiotic treatment yes versus no, and the majority reported increased rates of MDR Gram-negative bacteria following treatment (Figure 3a).25,27,55,57,59,63,67,69,70 Five studies reported data after long versus shorter duration of treatment, and the majority found significantly more MDR Gram-negative bacteria after prolonged treatment (Figure 3b).25,26,56,57,64 Thirteen studies reported data after treatment with broad-spectrum versus narrow-spectrum antibiotics, and the overwhelming majority reported higher rates of MDR Gram-negative bacteria following treatment with broad-spectrum antibiotics (Figure 3c).15,24,26–28,50,51,55,58,64,65,67,71 Other antibiotic-resistant bacteria Four studies (n = 1825) compared the impact of antibiotic treatment (yes versus no) on antibiotic-resistant bacteria that were not MDR according to our definition.37,52,53,66 One study (n = 584) found a higher rate of prior antibiotic treatment in neonates colonized with antibiotic-resistant Escherichia coli and/or Klebsiella pneumoniae.66 One study (n = 953) found an increased incidence of TEM-1 genes in E. coli strains in neonates following antibiotic therapy.53 Two studies (n = 288) found no statistically significant associations between antibiotic treatment (yes versus no) and subsequent antibiotic resistance development.37,52 Two studies compared the impact of antibiotic therapy duration;61,72 one of them (n = 1180) found significantly longer prior antibiotic treatment among neonates colonized with antibiotic-resistant Gram-negative bacteria,72 whereas the other (unknown number of neonates) found no correlation between the duration of treatment and gentamicin-resistant Gram-negative bacteria.61 Eight studies (n = 3029) compared the impact of broad- versus narrow-spectrum antibiotic treatment.31,37,53,54,60,62,68,72 One RCT (n = 276) found higher colonization rates with ampicillin-resistant A. baumannii following treatment with penicillin and gentamicin compared with ampicillin and gentamicin.31 One study (n = 440) found a higher rate of both ampicillin and cefuroxime resistance in Gram-negative bacteria following treatment with ampicillin compared with cefuroxime.62 One study (n = 118) found a higher rate of gentamicin resistance among Gram-negative bacteria following treatment with gentamicin compared with amikacin.68 The remaining five studies (n = 2195) did not formally test for statistically significant differences when comparing broad- versus narrow-spectrum regimens,37,53,54,60,72 although three of these studies (n = 1258) reported increased rates of antibiotic resistance following broad-spectrum therapy.54,60,72 Discussion Key findings To our knowledge, this is the first systematic review to examine antibiotic therapy in neonates and its impact on gut microbiota and/or antibiotic resistance development. The primary findings were the lack of RCTs and large high-quality observational studies and the heterogeneity regarding methodology and outcomes among the included studies. Despite this, there were several salient features in this review. First, prolonged antibiotic therapy was associated with reduced gut microbial diversity.41,44,48 Decreased gut microbial diversity has been associated with early adverse outcomes such as necrotizing enterocolitis, and may have potential long-lasting consequences such as increased likelihood of obesity and inflammatory diseases.10,49,74–77 Combined, these findings imply that prolonged exposure to antibiotic treatment in the neonatal period may increase the likelihood of disease, either in the neonatal period or later in life. However, the QoE for this outcome was graded as very low. It is possible that neonatal antibiotic therapy, regardless of treatment length, leads to decreased microbial diversity, but the studies included in this category were small, and two out of four studies did not detect a significant difference.40,42,44,49 Second, four out of nine studies reported increased abundance and/or colonization rates of Enterobacteriaceae following neonatal antibiotic treatment, whereas none of the studies reported reduced abundance.33,34,36,37,40,42,43,46,47 In the majority of these studies, the empirical regimens consisted of ampicillin and gentamicin. We speculate that intravenous ampicillin also has an impact on Gram-positive gut bacteria despite being mainly secreted through the kidneys,78 whereas intravenous gentamicin, mainly covering Gram-negative bacteria in the bloodstream,79 has a very low penetration into the gut. Combined, this may give undue benefits to Gram-negative Enterobacteriaceae. In contrast, third-generation cephalosporin therapy may lead to a relatively lower abundance of Enterobacteriaceae as both Gram-negative and Gram-positive bacteria are within their spectrum of activity.79 However, the QoE for this outcome was again graded as very low, and even though overgrowth of Enterobacteriaceae in the human gut has previously been associated with necrotizing enterocolitis, inflammatory bowel disease and chronic fatigue syndrome, there is no strong evidence of any causal relationship.74,76,80–82 Third, antibiotic treatment in the neonatal period was strongly associated with reduced abundance of protective commensal anaerobic bacteria such as bifidobacteria, lactobacilli and/or bacteriodes.35,40,43 These bacteria provide colonization resistance against antibiotic-resistant bacteria and potentially pathogenic bacteria such as Enterobacteriaceae and Clostridium difficile.7 Moreover, it is well known that bifidobacteria may reduce expression of inflammatory response genes and stimulate genes promoting the integrity of the mucosal barrier. The QoE for this outcome was graded as very low, but our results are in line with findings in adult populations showing decreased diversity, reduced colonization rates of obligate anaerobes and increased colonization rates of Proteobacteria following antibiotic exposure.83–85 Furthermore, our findings are biologically plausible as reduced numbers of bifidobacteria and lactobacilli seem to increase the risk of necrotizing enterocolitis in preterm infants with an exaggerated inflammatory response.82,86–90 In adults some studies have found larger changes in the gut microbiota than oral microbiota following antibiotic treatment, with greater resilience in the oral communities.84,85 However, we believe that the gut microbiota is of highest clinical relevance, both as the largest reservoir for antibiotic-resistant bacteria and because the gut is characterized as the motor of multiple organ dysfunction syndrome. Fourth, all three categories of neonatal antibiotic treatment investigated in this review were clearly associated with an increased risk of antibiotic resistance development, in particular ESBL-producing Gram-negative bacteria and other MDR bacteria. These findings were based on moderate QoE. Antibiotic resistance genes exist even in the absence of antimicrobial drugs.91,92 Moreover, overuse of antibiotics may lead to increased antibiotic resistance through several mechanisms.91,93 Antibiotics apply a direct selection pressure that gives significant advantages to bacteria expressing resistance genes.94 Antibiotic treatment also contributes to changes in the human gut-associated resistome, which comprises numerous functional antibiotic resistance genes in the gut microbiota.95 Gibson et al.96 recently found that only a fraction of antibiotic resistance genes that are enriched after a specific antibiotic therapy are unique to the particular antibiotic given. Finally, antibiotic treatment appears to reduce colonization resistance against antibiotic-resistant bacteria through the collateral destruction of obligate anaerobic bacteria.7,97 An increase in the gut resistome and a decrease in colonization resistance could theoretically increase horizontal transfer of antibiotic resistance genes from commensals to potential pathogens.98 Although in vivo horizontal transfer between commensals and pathogens in the gut microbiota remains to be shown, there is evidence of exchange of antibiotic resistance genes between environmental bacteria and human pathogens.99 Strengths and limitations The primary strength of this study is our rigorous and sensitive search strategy based on a previously registered search protocol. Additionally, the adverse impact of the developing infant gut microbiota is of great clinical and scientific interest. The main limitations were the lack of RCTs and the diverse study outcomes, which made meta-analysis impossible to perform. Instead, we applied a semi-quantitative vote-counting method to assess the effect of neonatal antibiotic treatment on relevant outcomes. This method has limitations as it usually fails to account for the population size and methodological quality of pooled studies. Nevertheless, vote counting may be an effective method to assess the ranking of outcomes.100 Moreover, we attempted to improve the method by presenting the differential weight of each study with squares corresponding to sample size. The majority of studies included were small and there was considerable heterogeneity in study designs, outcomes, categories of antibiotic treatment and methodological quality. Observational studies are prone to biases and confounding, and only a third of the included studies attempted to adjust for confounding through multivariable regression analysis. Evidence from observational studies is usually considered to be of low quality. However, well-designed observational studies have been shown to provide similar results to RCTs and they can therefore be useful for detecting rare adverse outcomes by allowing larger sample sizes and longer lengths of follow-up than RCTs for lower costs.101 We used the GRADE approach to assess the QoE. Overall, we graded the QoE as very low for all outcomes presented in the gut microbiota category. In contrast, we considered the QoE to be moderate in the antibiotic resistance category owing to large effect sizes and a dose–response effect. Based on current evidence we are therefore moderately confident that all types of antibiotic treatment lead to increased rates of antibiotic resistance. All included studies published prior to 2007 used culture-based techniques to examine the gut microbiota composition. It has been estimated that <20% of environmental bacteria can be grown in defined growth media. This increases the risk of detection bias in older studies.102 Sequencing-based techniques also have limitations. Studies relying on 16S rRNA analysis allow only a coarse sorting of bacteria mainly at phylum level. Deep shotgun metagenome sequencing allows for finer distinction at the genus or species level, but it is of crucial importance to standardize sampling and temperature control during the pipeline up to DNA extraction in order to obtain valid results.103 Moreover bioinformatic presentations are often challenging to understand and interpret. We also acknowledge that our definition of broad-spectrum and narrow-spectrum antibiotics is somewhat arbitrary as most of the narrow-spectrum regimens covered both Gram-negative and Gram-positive bacteria. However, our study confirms previous findings, clearly suggesting that antibiotic regimens containing third-generation cephalosporins or carbapenems are more frequently associated with antibiotic resistance development than regimens with aminoglycosides for Gram-negative coverage.15,24,26,28,50,55,64,65,67,71 Finally, we decided to exclude studies that only examined antenatal antibiotic treatment, despite the frequent use of intrapartum antibiotic prophylaxis for prevention of neonatal infections and its reported effects on the infant gut microbiota and carriage of antibiotic resistance genes.104 The focus of this review was neonatal antibiotic treatment given for suspected neonatal infection, and the isolated effects of antenatal antibiotics given to infants who did not receive antibiotics after birth were beyond the scope of this study. Implications and conclusions This systematic review highlights the profound impact on the gut microbiota and antibiotic resistance development exerted by antibiotic treatment in neonates. Antibiotic exposure in the neonatal period appears to induce various potentially disease-promoting alterations in the gut microbiota, but the QoE was very low for the outcomes investigated in this review. However, we are moderately confident, based on data from this review, that antibiotic treatment leads to antibiotic resistance development, in particular in Gram-negative bacteria. This clearly threatens current empirical antibiotic regimens and is a finding of great concern. In conclusion, the findings from this systematic review, along with the findings from our recent systematic review on early adverse outcome of neonatal antibiotic therapy,9 strongly emphasize the need to reduce unnecessary antibiotic treatment in neonates. Important steps to reduce the burden of neonatal antibiotic therapy include improving preventive measures such as hand hygiene, stopping antibiotic therapy after 36–48 h if infection is only vaguely suspected and there is no growth in the blood culture, and restricting the empirical use of broad-spectrum antibiotic treatment.105,106 Funding J. W. F. and E. E. are recipients of funding for their PhD studies from the Northern Norway Regional Health Authority. This study was supported by internal funding. The funding source had no role in the: design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review or approval of the manuscript; and decision to submit the manuscript for publication. Transparency declarations None to declare. Author contributions J. W. F. reviewed all relevant titles, abstracts and full-text articles, assessed quality, extracted data and drafted the initial manuscript. E. E. reviewed all relevant titles, abstracts and full-text articles, extracted data, assessed quality and revised the manuscript. L. K. J. contributed to study design, methodological assessment and revised the manuscript. J. N. v. d. A. contributed to study design and revised the manuscript. C. K. conceptualized and designed the study, reviewed relevant abstracts and articles, assessed quality and revised the manuscript. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work. J. W. F., E. E. and C. K. have full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Supplementary data Tables S1 and S2 and Figure S1 are available as Supplementary data at JAC Online. References 1 Neu J. The microbiome during pregnancy and early postnatal life. Semin Fetal Neonatal Med  2016; 21: 373– 9. Google Scholar CrossRef Search ADS PubMed  2 Azad MB, Konya T, Persaud RR et al.   Impact of maternal intrapartum antibiotics, method of birth and breastfeeding on gut microbiota during the first year of life: a prospective cohort study. BJOG  2016; 123: 983– 93. Google Scholar CrossRef Search ADS PubMed  3 Hartz LE, Bradshaw W, Brandon DH. Potential NICU environmental influences on the neonate’s microbiome: a systematic review. Adv Neonatal Care  2015; 15: 324– 35. Google Scholar CrossRef Search ADS PubMed  4 Koenig JE, Spor A, Scalfone N et al.   Succession of microbial consortia in the developing infant gut microbiome. Proc Natl Acad Sci U S A  2011; 108 Suppl 1: 4578– 85. Google Scholar CrossRef Search ADS PubMed  5 Tarr PI, Warner BB. Gut bacteria and late-onset neonatal bloodstream infections in preterm infants. Semin Fetal Neonatal Med  2016; 21: 388– 93. Google Scholar CrossRef Search ADS PubMed  6 Gensollen T, Iyer SS, Kasper DL et al.   How colonization by microbiota in early life shapes the immune system. Science  2016; 352: 539– 44. Google Scholar CrossRef Search ADS PubMed  7 Pamer EG. Resurrecting the intestinal microbiota to combat antibiotic-resistant pathogens. Science  2016; 352: 535– 8. Google Scholar CrossRef Search ADS PubMed  8 Hsieh EM, Hornik CP, Clark RH et al.   Medication use in the neonatal intensive care unit. Am J Perinatol  2014; 31: 811– 21. Google Scholar CrossRef Search ADS PubMed  9 Esaiassen E, Fjalstad JW, Juvet LK et al.   Antibiotic exposure in neonates and early adverse outcomes: a systematic review and meta-analysis. J Antimicrob Chemother  2017; 72: 1858– 70. Google Scholar CrossRef Search ADS PubMed  10 Turnbaugh PJ, Hamady M, Yatsunenko T et al.   A core gut microbiome in obese and lean twins. Nature  2009; 457: 480– 4. Google Scholar CrossRef Search ADS PubMed  11 Saari A, Virta LJ, Sankilampi U et al.   Antibiotic exposure in infancy and risk of being overweight in the first 24 months of life. Pediatrics  2015; 135: 617– 26. Google Scholar CrossRef Search ADS PubMed  12 Sim K, Shaw AG, Randell P et al.   Dysbiosis anticipating necrotizing enterocolitis in very premature infants. Clin Infect Dis  2015; 60: 389– 97. Google Scholar CrossRef Search ADS PubMed  13 Hviid A, Svanstrom H, Frisch M. Antibiotic use and inflammatory bowel diseases in childhood. Gut  2011; 60: 49– 54. Google Scholar CrossRef Search ADS PubMed  14 van Nimwegen FA, Penders J, Stobberingh EE et al.   Mode and place of delivery, gastrointestinal microbiota, and their influence on asthma and atopy. J Allergy Clin Immunol  2011; 128: 948– 55.e1–3. Google Scholar CrossRef Search ADS PubMed  15 de Man P, Verhoeven BA, Verbrugh HA et al.   An antibiotic policy to prevent emergence of resistant bacilli. Lancet  2000; 355: 973– 8. Google Scholar CrossRef Search ADS PubMed  16 Downie L, Armiento R, Subhi R et al.   Community-acquired neonatal and infant sepsis in developing countries: efficacy of WHO’s currently recommended antibiotics–systematic review and meta-analysis. Arch Dis Child  2013; 98: 146– 54. Google Scholar CrossRef Search ADS PubMed  17 Blackburn RM, Verlander NQ, Heath PT et al.   The changing antibiotic susceptibility of bloodstream infections in the first month of life: informing antibiotic policies for early- and late-onset neonatal sepsis. Epidemiol Infect  2014; 142: 803– 11. Google Scholar CrossRef Search ADS PubMed  18 Stapleton PJ, Murphy M, McCallion N et al.   Outbreaks of extended spectrum β-lactamase-producing Enterobacteriaceae in neonatal intensive care units: a systematic review. Arch Dis Child Fetal Neonatal Ed  2016; 101: F72– 8. Google Scholar CrossRef Search ADS PubMed  19 Laxminarayan R, Matsoso P, Pant S et al.   Access to effective antimicrobials: a worldwide challenge. Lancet  2016; 387: 168– 75. Google Scholar CrossRef Search ADS PubMed  20 Liberati A, Altman DG, Tetzlaff J et al.   The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol  2009; 62: e1– 34. Google Scholar CrossRef Search ADS PubMed  21 Klingenberg C, Fjalstad J, Esaiassen E et al.   A Systematic Review of Early Adverse Effects Associated with Antibiotic Exposure in the Neonatal Period. PROSPERO. http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42015026743. 22 Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0. 2011. http://www.handbook.cochrane.org/. 23 Hiergeist A, Glasner J, Reischl U et al.   Analyses of intestinal microbiota: culture versus sequencing. ILAR J  2015; 56: 228– 40. Google Scholar CrossRef Search ADS PubMed  24 de Araujo OR, da Silva DC, Diegues AR et al.   Cefepime restriction improves gram-negative overall resistance patterns in neonatal intensive care unit. Braz J Infect Dis  2007; 11: 277– 80. Google Scholar CrossRef Search ADS PubMed  25 Giuffre M, Geraci DM, Bonura C et al.   The increasing challenge of multidrug-resistant gram-negative bacilli: results of a 5-year active surveillance program in a neonatal intensive care unit. Medicine (Baltimore)  2016; 95: e3016. Google Scholar CrossRef Search ADS PubMed  26 Mammina C, Di Carlo P, Cipolla D et al.   Surveillance of multidrug-resistant gram-negative bacilli in a neonatal intensive care unit: prominent role of cross transmission. Am J Infect Control  2007; 35: 222– 30. Google Scholar CrossRef Search ADS PubMed  27 Millar M, Philpott A, Wilks M et al.   Colonization and persistence of antibiotic-resistant Enterobacteriaceae strains in infants nursed in two neonatal intensive care units in East London, United Kingdom. J Clin Microbiol  2008; 46: 560– 7. Google Scholar CrossRef Search ADS PubMed  28 Thatrimontrichai A, Apisarnthanarak A, Chanvitan P et al.   Risk factors and outcomes of carbapenem-resistant Acinetobacter baumannii bacteremia in neonatal intensive care unit: a case-case-control study. Pediatr Infect Dis J  2013; 32: 140– 5. Google Scholar CrossRef Search ADS PubMed  29 Viswanathan M, Berkman ND, Dryden DM et al.   AHRQ Methods for Effective Health Care. Assessing Risk of Bias and Confounding in Observational Studies of Interventions or Exposures: Further Development of the RTI Item Bank . Rockville, MD: Agency for Healthcare Research and Quality (US), 2013. 30 GRADE Working Group. Handbook for Grading the Quality of Evidence and the Strength of Recommendations Using the GRADE Approach. http://gdt.guidelinedevelopment.org/app/handbook/handbook.html#h.g2dqzi9je57e. 31 Parm U, Metsvaht T, Sepp E et al.   Impact of empiric antibiotic regimen on bowel colonization in neonates with suspected early onset sepsis. Eur J Clin Microbiol Infect Dis  2010; 29: 807– 16. Google Scholar CrossRef Search ADS PubMed  32 Westerbeek EA, Slump RA, Lafeber HN et al.   The effect of enteral supplementation of specific neutral and acidic oligosaccharides on the faecal microbiota and intestinal microenvironment in preterm infants. Eur J Clin Microbiol Infect Dis  2013; 32: 269– 76. Google Scholar CrossRef Search ADS PubMed  33 Arboleya S, Sanchez B, Milani C et al.   Intestinal microbiota development in preterm neonates and effect of perinatal antibiotics. J Pediatr  2015; 166: 538– 44. Google Scholar CrossRef Search ADS PubMed  34 Bennet R, Eriksson M, Nord CE et al.   Fecal bacterial microflora of newborn infants during intensive care management and treatment with five antibiotic regimens. Pediatr Infect Dis  1986; 5: 533– 9. Google Scholar CrossRef Search ADS PubMed  35 Bennet R, Nord CE. Development of the faecal anaerobic microflora after Caesarean section and treatment with antibiotics in newborn infants. Infection  1987; 15: 332– 6. Google Scholar CrossRef Search ADS PubMed  36 Blakey JL, Lubitz L, Barnes GL. Development of gut colonisation in pre-term neonates. J Med Microbiol  1982; 15: 519– 29. Google Scholar CrossRef Search ADS PubMed  37 Bonnemaison E, Lanotte P, Cantagrel S et al.   Comparison of fecal flora following administration of two antibiotic protocols for suspected maternofetal infection. Biol Neonate  2003; 84: 304– 10. Google Scholar CrossRef Search ADS PubMed  38 Butel MJ, Suau A, Campeotto F et al.   Conditions of bifidobacterial colonization in preterm infants: a prospective analysis. J Pediatr Gastroenterol Nutr  2007; 44: 577– 82. Google Scholar CrossRef Search ADS PubMed  39 Ferraris L, Butel MJ, Campeotto F et al.   Clostridia in premature neonates’ gut: incidence, antibiotic susceptibility, and perinatal determinants influencing colonization. PLoS One  2012; 7: e30594. Google Scholar CrossRef Search ADS PubMed  40 Fouhy F, Guinane CM, Hussey S et al.   High-throughput sequencing reveals the incomplete, short-term recovery of infant gut microbiota following parenteral antibiotic treatment with ampicillin and gentamicin. Antimicrob Agents Chemother  2012; 56: 5811– 20. Google Scholar CrossRef Search ADS PubMed  41 Gewolb IH, Schwalbe RS, Taciak VL et al.   Stool microflora in extremely low birthweight infants. Arch Dis Child Fetal Neonatal Ed  1999; 80: F167– 73. Google Scholar CrossRef Search ADS PubMed  42 Greenwood C, Morrow AL, Lagomarcino AJ et al.   Early empiric antibiotic use in preterm infants is associated with lower bacterial diversity and higher relative abundance of Enterobacter. J Pediatr  2014; 165: 23– 9. Google Scholar CrossRef Search ADS PubMed  43 Hall MA, Cole CB, Smith SL et al.   Factors influencing the presence of faecal lactobacilli in early infancy. Arch Dis Child  1990; 65: 185– 8. Google Scholar CrossRef Search ADS PubMed  44 Jacquot A, Neveu D, Aujoulat F et al.   Dynamics and clinical evolution of bacterial gut microflora in extremely premature patients. J Pediatr  2011; 158: 390– 6. Google Scholar CrossRef Search ADS PubMed  45 Jenke AC, Postberg J, Mariel B et al.   S100A12 and hBD2 correlate with the composition of the fecal microflora in ELBW infants and expansion of E. coli is associated with NEC. Biomed Res Int ; 2013: 150372. PubMed  46 La Rosa PS, Warner BB, Zhou Y et al.   Patterned progression of bacterial populations in the premature infant gut. Proc Natl Acad Sci U S A  2014; 111: 12522– 7. Google Scholar CrossRef Search ADS PubMed  47 Tullus K, Berglund B, Fryklund B et al.   Influence of antibiotic therapy of faecal carriage of P-fimbriated Escherichia coli and other Gram-negative bacteria in neonates. J Antimicrob Chemother  1988; 22: 563– 8. Google Scholar CrossRef Search ADS PubMed  48 Ward DV, Scholz M, Zolfo M et al.   Metagenomic sequencing with strain-level resolution implicates uropathogenic E. coli in necrotizing enterocolitis and mortality in preterm infants. Cell Rep  2016; 14: 2912– 24. Google Scholar CrossRef Search ADS PubMed  49 Zhou Y, Shan G, Sodergren E et al.   Longitudinal analysis of the premature infant intestinal microbiome prior to necrotizing enterocolitis: a case-control study. PLoS One  2015; 10: e0118632. Google Scholar CrossRef Search ADS PubMed  50 Abdel-Hady H, Hawas S, El-Daker M et al.   Extended-spectrum β-lactamase producing Klebsiella pneumoniae in neonatal intensive care unit. J Perinatol  2008; 28: 685– 90. Google Scholar CrossRef Search ADS PubMed  51 Acolet D, Ahmet Z, Houang E et al.   Enterobacter cloacae in a neonatal intensive care unit: account of an outbreak and its relationship to use of third generation cephalosporins. J Hosp Infect  1994; 28: 273– 86. Google Scholar CrossRef Search ADS PubMed  52 Bergin SP, Thaden JT, Ericson JE et al.   Neonatal Escherichia coli bloodstream infections: clinical outcomes and impact of initial antibiotic therapy. Pediatr Infect Dis J  2015; 34: 933– 6. Google Scholar CrossRef Search ADS PubMed  53 Burman LG, Haeggman S, Kuistila M et al.   Epidemiology of plasmid-mediated β-lactamases in enterobacteria Swedish neonatal wards and relation to antimicrobial therapy. Antimicrob Agents Chemother  1992; 36: 989– 92. Google Scholar CrossRef Search ADS PubMed  54 Burman LG, Berglund B, Huovinen P et al.   Effect of ampicillin versus cefuroxime on the emergence of β-lactam resistance in faecal Enterobacter cloacae isolates from neonates. J Antimicrob Chemother  1993; 31: 111– 6. Google Scholar CrossRef Search ADS PubMed  55 Calil R, Marba ST, von Nowakonski A et al.   Reduction in colonization and nosocomial infection by multiresistant bacteria in a neonatal unit after institution of educational measures and restriction in the use of cephalosporins. Am J Infect Control  2001; 29: 133– 8. Google Scholar CrossRef Search ADS PubMed  56 Cantey JB, Wozniak PS, Pruszynski JE et al.   Reducing unnecessary antibiotic use in the neonatal intensive care unit (SCOUT): a prospective interrupted time-series study. Lancet Infect Dis  2016; 16: 1178– 84. Google Scholar CrossRef Search ADS PubMed  57 Crivaro V, Bagattini M, Salza MF et al.   Risk factors for extended-spectrum β-lactamase-producing Serratia marcescens and Klebsiella pneumoniae acquisition in a neonatal intensive care unit. J Hosp Infect  2007; 67: 135– 41. Google Scholar CrossRef Search ADS PubMed  58 De Champs C. Clinical and bacteriological survey after change in aminoglycoside treatment to control an epidemic of Enterobacter cloacae. J Hosp Infect  1994; 28: 219– 29. Google Scholar CrossRef Search ADS PubMed  59 Duman M, Abacioglu H, Karaman M et al.   β-lactam antibiotic resistance in aerobic commensal fecal flora of newborns. Pediatr Int  2005; 47: 267– 73. Google Scholar CrossRef Search ADS PubMed  60 Gaynes RP, Simpson D, Reeves SA et al.   A nursery outbreak of multiple-aminoglycoside-resistant Escherichia coli. Infect Control  1984; 5: 519– 24. Google Scholar CrossRef Search ADS PubMed  61 Isaacs D, Catterson J, Hope PL et al.   Factors influencing colonisation with gentamicin resistant gram negative organisms in the neonatal unit. Arch Dis Child  1988; 63: 533– 5. Google Scholar CrossRef Search ADS PubMed  62 Kalenic S, Francetic I, Polak J et al.   Impact of ampicillin and cefuroxime on bacterial colonization and infection in patients on a neonatal intensive care unit. J Hosp Infect  1993; 23: 35– 41. Google Scholar CrossRef Search ADS PubMed  63 Kumar A, Randhawa VS, Nirupam N et al.   Risk factors for carbapenem-resistant Acinetobacter baumanii blood stream infections in a neonatal intensive care unit, Delhi, India. J Infect Dev Ctries  2014; 8: 1049– 54. Google Scholar PubMed  64 Le J, Nguyen T, Okamoto M et al.   Impact of empiric antibiotic use on development of infections caused by extended-spectrum β-lactamase bacteria in a neonatal intensive care unit. Pediatr Infect Dis J  2008; 27: 314– 8. Google Scholar CrossRef Search ADS PubMed  65 Linkin DR, Fishman NO, Patel JB et al.   Risk factors for extended-spectrum β-lactamase-producing Enterobacteriaceae in a neonatal intensive care unit. Infect Control Hosp Epidemiol  2004; 25: 781– 3. Google Scholar CrossRef Search ADS PubMed  66 Noy JH, Ayliffe GA, Linton KB. Antibiotic-resistant Gram-negative bacilli in the faeces of neonates. J Med Microbiol  1974; 7: 509– 20. Google Scholar CrossRef Search ADS PubMed  67 Pessoa-Silva CL, Meurer Moreira B, Camara Almeida V et al.   Extended-spectrum β-lactamase-producing Klebsiella pneumoniae in a neonatal intensive care unit: risk factors for infection and colonization. J Hosp Infect  2003; 53: 198– 206. Google Scholar CrossRef Search ADS PubMed  68 Raz R, Sharir R, Shmilowitz L et al.   The elimination of gentamicin-resistant gram-negative bacteria in a newborn intensive care unit. Infection  1987; 15: 32– 4. Google Scholar CrossRef Search ADS PubMed  69 Rettedal S, Hoyland LI, Natas O et al.   Risk factors for acquisition of CTX-M-15 extended-spectrum β-lactamase-producing Klebsiella pneumoniae during an outbreak in a neonatal intensive care unit in Norway. Scand J Infect Dis  2013; 45: 54– 8. Google Scholar CrossRef Search ADS PubMed  70 Sehgal R, Gaind R, Chellani H et al.   Extended-spectrum β lactamase-producing Gram-negative bacteria: clinical profile and outcome in a neonatal intensive care unit. Ann Trop Paediatr  2007; 27: 45– 54. Google Scholar CrossRef Search ADS PubMed  71 Thatrimontrichai A, Techato C, Dissaneevate S et al.   Risk factors and outcomes of carbapenem-resistant Acinetobacter baumannii ventilator-associated pneumonia in the neonate: a case-case-control study. J Infect Chemother  2016; 22: 444– 9. Google Scholar CrossRef Search ADS PubMed  72 Toltzis P, Dul MJ, Hoyen C et al.   Molecular epidemiology of antibiotic-resistant Gram-negative bacilli in a neonatal intensive care unit during a nonoutbreak period. Pediatrics  2001; 108: 1143– 8. Google Scholar CrossRef Search ADS PubMed  73 Goldmann DA, Leclair J, Macone A. Bacterial colonization of neonates admitted to an intensive care environment. J Pediatr  1978; 93: 288– 93. Google Scholar CrossRef Search ADS PubMed  74 Warner BB, Deych E, Zhou Y et al.   Gut bacteria dysbiosis and necrotising enterocolitis in very low birthweight infants: a prospective case-control study. Lancet  2016; 387: 1928– 36. Google Scholar CrossRef Search ADS PubMed  75 Sundin J, Rangel I, Fuentes S et al.   Altered faecal and mucosal microbial composition in post-infectious irritable bowel syndrome patients correlates with mucosal lymphocyte phenotypes and psychological distress. Aliment Pharmacol Ther  2015; 41: 342– 51. Google Scholar CrossRef Search ADS PubMed  76 Giloteaux L, Goodrich JK, Walters WA et al.   Reduced diversity and altered composition of the gut microbiome in individuals with myalgic encephalomyelitis/chronic fatigue syndrome. Microbiome  2016; 4: 30. Google Scholar CrossRef Search ADS PubMed  77 Bisgaard H, Li N, Bonnelykke K et al.   Reduced diversity of the intestinal microbiota during infancy is associated with increased risk of allergic disease at school age. J Allergy Clin Immunol  2011; 128: 646– 52.e1–5. Google Scholar CrossRef Search ADS PubMed  78 Zhang L, Huang Y, Zhou Y et al.   Antibiotic administration routes significantly influence the levels of antibiotic resistance in gut microbiota. Antimicrob Agents Chemother  2013; 57: 3659– 66. Google Scholar CrossRef Search ADS PubMed  79 Roberts JK, Stockmann C, Constance JE et al.   Pharmacokinetics and pharmacodynamics of antibacterials, antifungals, and antivirals used most frequently in neonates and infants. Clin Pharmacokinet  2014; 53: 581– 610. Google Scholar CrossRef Search ADS PubMed  80 Torrazza RM, Ukhanova M, Wang X et al.   Intestinal microbial ecology and environmental factors affecting necrotizing enterocolitis. PLoS One  2013; 8: e83304. Google Scholar CrossRef Search ADS PubMed  81 Belkaid Y, Hand TW. Role of the microbiota in immunity and inflammation. Cell  2014; 157: 121– 41. Google Scholar CrossRef Search ADS PubMed  82 Jiang H, Ling Z, Zhang Y et al.   Altered fecal microbiota composition in patients with major depressive disorder. Brain Behav Immun  2015; 48: 186– 94. Google Scholar CrossRef Search ADS PubMed  83 Yap TW, Gan HM, Lee YP et al.   Helicobacter pylori eradication causes perturbation of the human gut microbiome in young adults. PLoS One  2016; 11: e0151893. Google Scholar CrossRef Search ADS PubMed  84 Zaura E, Brandt BW, Teixeira de Mattos MJ et al.   Same exposure but two radically different responses to antibiotics: resilience of the salivary microbiome versus long-term microbial shifts in feces. MBio  2015; 6: e01693-15. Google Scholar CrossRef Search ADS PubMed  85 Jakobsson HE, Jernberg C, Andersson AF et al.   Short-term antibiotic treatment has differing long-term impacts on the human throat and gut microbiome. PLoS One  2010; 5: e9836. Google Scholar CrossRef Search ADS PubMed  86 Ahmad OF, Akbar A. Microbiome, antibiotics and irritable bowel syndrome. Br Med Bull  2016; 1: 91– 9. Google Scholar CrossRef Search ADS   87 Dermyshi E, Wang Y, Yan C et al.   The “golden age” of probiotics: a systematic review and meta-analysis of randomized and observational studies in preterm infants. Neonatology  2017; 112: 9– 23. Google Scholar CrossRef Search ADS PubMed  88 Chow J, Tang H, Mazmanian SK. Pathobionts of the gastrointestinal microbiota and inflammatory disease. Curr Opin Immunol  2011; 23: 473– 80. Google Scholar CrossRef Search ADS PubMed  89 Kamada N, Chen GY, Inohara N et al.   Control of pathogens and pathobionts by the gut microbiota. Nat Immunol  2013; 14: 685– 90. Google Scholar CrossRef Search ADS PubMed  90 Wang M, Monaco MH, Donovan SM. Impact of early gut microbiota on immune and metabolic development and function. Semin Fetal Neonatal Med  2016; 21: 380– 7. Google Scholar CrossRef Search ADS PubMed  91 Pallecchi L, Bartoloni A, Paradisi F et al.   Antibiotic resistance in the absence of antimicrobial use: mechanisms and implications. Expert Rev Anti Infect Ther  2008; 6: 725– 32. Google Scholar CrossRef Search ADS PubMed  92 Zhang L, Kinkelaar D, Huang Y et al.   Acquired antibiotic resistance: are we born with it? Appl Environ Microbiol  2011; 77: 7134– 41. Google Scholar CrossRef Search ADS PubMed  93 Patel SJ, Saiman L. Antibiotic resistance in neonatal intensive care unit pathogens: mechanisms, clinical impact, and prevention including antibiotic stewardship. Clin Perinatol  2010; 37: 547– 63. Google Scholar CrossRef Search ADS PubMed  94 Davies J, Davies D. Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev  2010; 74: 417– 33. Google Scholar CrossRef Search ADS PubMed  95 Sommer MO, Church GM, Dantas G. The human microbiome harbors a diverse reservoir of antibiotic resistance genes. Virulence  2010; 1: 299– 303. Google Scholar CrossRef Search ADS PubMed  96 Gibson MK, Wang B, Ahmadi S et al.   Developmental dynamics of the preterm infant gut microbiota and antibiotic resistome. Nat Microbiol  2016; 1: 16024. Google Scholar CrossRef Search ADS PubMed  97 Donskey CJ, Chowdhry TK, Hecker MT et al.   Effect of antibiotic therapy on the density of vancomycin-resistant enterococci in the stool of colonized patients. N Engl J Med  2000; 343: 1925– 32. Google Scholar CrossRef Search ADS PubMed  98 Martinez JL. General principles of antibiotic resistance in bacteria. Drug Discov Today Technol  2014; 11: 33– 9. Google Scholar CrossRef Search ADS PubMed  99 Forsberg KJ, Reyes A, Wang B et al.   The shared antibiotic resistome of soil bacteria and human pathogens. Science  2012; 337: 1107– 11. Google Scholar CrossRef Search ADS PubMed  100 Rikke BA, Wynes MW, Rozeboom LM et al.   Independent validation test of the vote-counting strategy used to rank biomarkers from published studies. Biomark Med  2015; 9: 751– 61. Google Scholar CrossRef Search ADS PubMed  101 Cameron C, Fireman B, Hutton B et al.   Network meta-analysis incorporating randomized controlled trials and non-randomized comparative cohort studies for assessing the safety and effectiveness of medical treatments: challenges and opportunities. Syst Rev  2015; 4: 147. Google Scholar CrossRef Search ADS PubMed  102 Ward DM, Weller R, Bateson MM. 16S rRNA sequences reveal numerous uncultured microorganisms in a natural community. Nature  1990; 345: 63– 5. Google Scholar CrossRef Search ADS PubMed  103 Hanage WP. Microbiology: microbiome science needs a healthy dose of scepticism. Nature  2014; 512: 247– 8. Google Scholar CrossRef Search ADS PubMed  104 Nogacka A, Salazar N, Suarez M et al.   Impact of intrapartum antimicrobial prophylaxis upon the intestinal microbiota and the prevalence of antibiotic resistance genes in vaginally delivered full-term neonates. Microbiome  2017; 5: 93. Google Scholar CrossRef Search ADS PubMed  105 Polin RA; Committee on Fetus and Newborn. Management of neonates with suspected or proven early-onset bacterial sepsis. Pediatrics  2012; 129: 1006– 15. Google Scholar CrossRef Search ADS PubMed  106 National Institute for Health and Care Excellence (NICE). Antibiotics for Early-Onset Neonatal Infection: Antibiotics for the Prevention and Treatment of Early-Onset Neonatal Infection. https://www.nice.org.uk/guidance/cg149. © The Author 2017. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

Journal

Journal of Antimicrobial ChemotherapyOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial