Skip to main content
  • Research Articles
  • Open access
  • Published:

Lower gut dysbiosis and mortality in acute critical illness: a systematic review and meta-analysis



The human gastrointestinal tract harbours a complex multi-kingdom community known as the microbiome. Dysbiosis refers to its disruption and is reportedly extreme in acute critical illness yet its clinical implications are unresolved. The review systematically evaluates the association between gut dysbiosis and clinical outcomes of patients early in critical illness.


Following PRISMA guidelines, a prospectively registered search was undertaken of MEDLINE and Cochrane databases for observational studies undertaking metagenomic sequencing of the lower gastrointestinal tract of critically ill adults and children within 72 h of admission. Eligible studies reported an alpha diversity metric and one or more of the primary outcome, in-hospital mortality, or secondary clinical outcomes. After aggregate data were requested, meta-analysis was performed for four studies with in-hospital mortality stratified to high or low Shannon index.


The search identified 26 studies for systematic review and 4 had suitable data for meta-analysis. No effect of alpha diversity was seen on in-hospital mortality after binary transformation of Shannon index (odds ratio 0.52, CI 0.12–4.98, I2 = 0.64) however certainty of evidence is low. Pathogen dominance and commensal depletion were each more frequently associated with in-hospital mortality, adverse clinical and ecological sequelae, particularly overabundance of Enterococcus.


There is a paucity of large, rigorous observational studies in this population. Globally, alpha diversity was dynamically reduced in early ICU admission in adults and children and was not associated with in-hospital mortality. The abundance of taxa such as Enterococcus spp. appears to offer greater predictive capacity for important clinical and ecological outcomes.


The lower gastrointestinal tract microbiome is a complex multi-kingdom community of bacteria, archaea, fungi and viruses. In humans, this community is shaped in early life, but modified by age, environment, diet, disease states and antibiotic administration [1]. Due to the taxonomic complexity of the system, microbiomes are typically characterised using measures of relative abundance of important families or genera, or diversity. Alpha diversity represents the richness (number) and evenness (spread) of species as a summary statistic of a single population, often expressed by Shannon or Simpson's indices (see Additional file 2: Glossary) [2]. Beta diversity describes the absolute or relative taxonomic overlap between clinical samples over time, between individuals or between body sites (biogeography) [3]. A typical healthy microbiome sees high alpha diversity with a predominance of Firmicutes and Bifidobacteria. In this state, the microbiome’s impact as a modulator of metabolism and immune function is thought to be optimised, though the mechanistic links for these processes are only now being elucidated [4]. Disruption of microbiome community structure and metabolic function is termed dysbiosis. Whilst there is no global consensus definition, the prototypical disease state associated with dysbiosis is Clostridioides difficile associated diarrhoea which is accompanied by profound loss of alpha diversity with increased relative abundance of Enterobacteriaceae and Enterococcaceae families [5]. However, dysbiosis is also associated with many other chronic inflammatory, autoimmune and neuropsychiatric conditions [6, 7].

Patients presenting with acute critical illness are at especially high risk of dysbiosis [8, 9]. Co-morbidities, severe metabolic disturbance and the frequent use of antibiotics in intensive care units are thought to drive the disproportionate depletion of commensal organisms and emergence of opportunistic pathogens [10]. Whilst severe dysbiosis in these settings is thought to confer a poorer prognosis, the association between conventionally reported microbiome characteristics and adverse clinical outcomes has not been systematically reported.

In preparation for a large observational study of lower gut microbiome in patients with acute critical illness admitted to a tertiary intensive care unit (ICU) and a trial of salvage faecal microbiota transplantation (FMT) for patients with presumed dysbiosis, we aimed to systematically review publications reporting associations between lower gut dysbiosis by metagenomic techniques and adverse clinical outcomes in critically ill adults or children. We hypothesised that amongst patients who undergo sampling of lower gut microbiota within 72 h of admission with critical illness, lower alpha-diversity will be associated with in-hospital mortality and certain pathogenic genera (Enterococcus, Streptococcus, Pseudomonas) will exhibit greater relative abundance.


Search strategy and selection criteria

This systematic review was prospectively registered (PROSPERO CRD 42,022,281,530) and conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. The inclusion criteria included prospective or retrospective observational studies examining critically ill adults or children admitted to an intensive care unit who underwent microbiome analyses by molecular methods, including 16S rRNA and shotgun metagenomic sequencing, where Shannon index or other alpha diversity index was reported as exposure variable. Neonates were excluded given known dynamic differences in microbiome composition. To capture acute critical illness, we restricted the analysis to studies reporting a majority of rectal or stool sampling within 72 h of ICU admission and only samples within this timeframe were included in meta-analysis. Strong concordance of stool and swab samples has been demonstrated [11,12,13] and results were therefore pooled.

Representing the weighted geometric mean of proportional abundances of taxa within a sample, Shannon index was defined as high when greater than or equal to 3, and low when less than 3 based upon the mean value of included studies. Studies reporting any of the primary or secondary outcome variables described below were eligible for inclusion. Operational Taxonomic Units (OTUs) which were significantly enriched or depleted as determined by the study authors were recorded. Abstract-only, non-English language publications and culture-based studies were excluded.


The primary outcome was all-cause in-hospital mortality at day 28, censored at hospital discharge. Secondary outcomes included (i) ICU length of stay, (ii) overall hospital length of stay, (iii) mortality at longest follow-up censored to day 180), (iv) duration of mechanical ventilation (MV) and (v) incidence of infection. Pre-specified subgroups of interest included (i) presentations with sepsis (as defined by the study), (ii) adults (compared with children), (iii) surgical (compared with medical admissions), (iv) severe illness (compared with non-severe illness), (v) patients receiving prior antibiotic therapy and (vi) rectal sampling (compared with stool). Risk of bias assessment was performed using the Cochrane tool for non-interventional studies (ROBINS-I) tool [14] (see Fig. 1).

Fig. 1
figure 1

Cochrane Risk of Bias for non-interventional studies (ROBINS-II) included in systematic review (n = 26). Breakdown per study demonstrated in Additional file 1: Fig. S6

Study selection

MEDLINE and The Cochrane Central Register of Controlled Trials were searched for relevant observational studies with subsequent manual reference search. The search was limited to articles published after 2011 given the recent advent of molecular sequencing methods. Where outcomes were recorded, but not presented suitably for aggregate analysis, data were directly requested from corresponding authors.

The search strategy included terms related to gut microbiome and critical illness:

(("critically ill" OR "critical illness" OR "intensive care")) AND ((gut) OR (gastrointestinal) OR (intestinal)) AND ((microbiome) OR (microbiota) OR (pathobiome) OR (dysbiosis) OR (flora)).

Two authors (T.E. and U.A.) performed the pre-defined literature search, reviewed texts and extracted data independently. Studies were eligible for inclusion in the systematic review where they met all inclusion criteria and none of the exclusion criteria. Discordance was adjudicated by a third reviewer (E.L.).

Data synthesis

For normally distributed continuous data, the mean and standard deviation was used. Where a median and interquartile range (IQR) were reported, the mean and standard deviation (SD) were imputed using validated techniques [15]. An odds ratio with a 95% confidence interval was calculated for binary data. A P-value < 0.05 was considered significant. The effects of age and gender balance on the association between the exposure and the primary outcome were assessed using meta-regression if there is adequate sample size. Heterogeneity was reported using the I2 which describes total variation across studies not due to sampling error. Treatment effect was determined using DerSimonian and Laird random effects model with Egger’s test [16]. Stata was used for all analyses [17].


There were 26 studies eligible for systematic review of which 7 provided moderate certainty evidence [18,19,20,21,22,23,24]. Alpha diversity was significantly predictive of in-hospital mortality in only one [25]. In contrast, pathogen dominance (6 studies) and commensal loss (2 studies) appear to be more highly predictive of in-hospital mortality among other adverse outcomes in adults. In a meta-analysis only 4 studies contributed data, leading to low certainty in the finding of no association between low alpha diversity and in-hospital mortality (odds ratio 0.52, CI 0.12–4.98, I2 = 0.64, Fig. 2).

Fig. 2
figure 2

Forest plot displaying aggregate meta-analysis of binary odds ratio of 28-day all-cause mortality in high versus low alpha diversity groups, represented by Shannon index more than 3 or less than 2.99, respectively

Description of included studies

The PRISMA flow diagram is shown (Fig. 3). A total of 898 studies were screened and 26 studies which reported a total of 1591 patients (142 children and 1503 adults) were included in the systematic review. Most studies were of adults (91.3% of total participants). The median (IQR) age in adult and paediatric studies was 57.3 (52.9–63.8) years and 3.5 years (3.2–4.1), years, respectively. All were prospective observational studies published between 2014 and April 2022 (10, 18, 19, 21–24, 26–39) ranging from 9 to 301 participants (See Additional file 1: Table S1). The majority were single centre (n = 20) and half (n = 13) were mixed medical and surgical ICUs (Table 1). The overall mean (SD) day 28 all-cause in-hospital mortality was 26.4% (15.9) and ranged between 7–48 and 3–9% in adult and child populations, respectively (Table 2). The overall assessment of risk of bias was high in participant selection, missing data and reporting domains, and moderate in the remainder (Fig. 1).

Fig. 3
figure 3

PRISMA flow diagram delineating study selection (n = 26 for systematic review, n = 4 for meta-analysis)

Table 1 Characteristics of cohorts included in systematic review
Table 2 Clinical and infection-related outcomes of study cohorts included in systematic review
Table 3 The microbiome of the critically ill: future research directions

The most common sample for analysis was stool (n = 17), with rectal swabs (n = 4), both (n = 3) and examination glove samples (n = 2) less commonly used. Rectal swab cohorts saw far greater participation and sampling adherence [11, 23]. For measures of alpha-diversity the Shannon index was the most frequently reported (n = 18 studies), followed by Chao richness (n = 6)[40], (inverse) Simpsons (n = 5)[41] and Pielou evenness (n = 2)[42] indices (see Additional file 1: Figs. S1, S2). Beta diversity was reported in most (n = 21). Relative abundance was reported in all studies. Design and methodological differences are summarised in Additional file 1: Table S2 and Fig. S3.

Reduced alpha diversity in acute critical illness was reported in 25 of the 26 systematic review studies. The mean (SD) Shannon index within 72 h of admission reported in 9 studies was 3.58 (1.30), lower than in control groups. The lowest mean alpha diversity (Shannon index) in any study, 2.03 (0.71). Alpha diversity declined over time in all 16 longitudinal studies of adults or children (see Additional file 1: Figs. S1, S2). Moderate changes in the first 24–48 h of ICU admission typically became marked by 72 h (exemplified by [36] including those with elective patient load [24]). No relationship between alpha diversity and admission diagnosis was found in 5 of 6 evaluating studies [29, 35]. No associations were observed with artificial feeding patterns or bowel habit [43]. Reduced early alpha diversity after antibiotic exposure was evident in eight studies [18,19,20,21,22, 24, 28, 43], whereas two reported no difference [34, 35]. One study reported a lower Shannon index in non-survivors compared with survivors [25].

Meta-analysis of alpha diversity and its association with mortality

Although six studies had appropriate data, two studies [20, 43] required exclusion due to two or more categories without mortality events. Meta-analysis of four studies (n = 402 including 85 children) found no significant relationship between high versus low alpha diversity and odds ratio of in-hospital mortality at day 28 (OR 0.59, CI 0.12–4.98), censored to hospital discharge. Heterogeneity was seen in study population, sample type, methodology, and post-processing, which was significant between studies (I2 value was 0.64, p = 0.038). The prespecified primary outcome of in-hospital mortality at day 28 or censored to hospital discharge was reported in 24 of 26 studies yet was infrequently categorised by diversity. Authors were contacted, however data were insufficient for calculation of other outcomes and subgroup comparison.

Outcome and pathogen type

All studies reported taxon abundance measures. Enterococcus genus was enriched in 20 of 26 studies (77%). Composition of up to 70% (29), 99.9% (28), 97.9% (31) and 99% (43) in certain individuals was seen.

Enterococcus spp. Abundance was associated with in-hospital mortality in four of the included studies [11, 19, 26]. Relative abundance was associated with other adverse outcomes including risk of infection [20, 26], elevated inflammatory markers [33] and ICU length of stay [26]. Additional file 1: Table S3 shows overabundance of suspected and confirmed MDR organisms reported by 14 and 3 studies, respectively. In ecological terms, Enteroccocal dominance was associated with significant loss of microbiome richness and diversity [28, 31]. Fontaine and colleagues found colonisation by vancomycin resistant Enterococcus (VRE) but not MDR Gram-negative organisms reduced both richness and evenness [37]. Impact of antibiotic exposure could not be evaluated systematically as only seven studies included drug-specific information.

Commensal organisms may oppose pathogen impact. The only singular commensal taxon to predict survival, Wei et al., showed loss of Bifidobacterium was associated ICU mortality (p = 0.031) [22]. Together with others—Blautia (various species) Faecalibacterium, Collinsella and Streptococcus—Bifidobacterium abundance was associated with survival in one other study [25]. Proportional increase in Actinobacteria was shown to interact favourably with SOFA score [18]. Wan reported this phylum was associated with beneficial metabolites and could be selected for (or against) by certain antibiotics. Prevotella abundance was associated with non-sepsis within a mixed ICU population [19]. Family X1 and Prevotellaceae more abundant in patients who did not acquire MDR organisms [20]. Retention of commensal taxa were associated with reduced risk of carbapenem-resistant Pseudomonas aeruginosa acquisition, whereas loss of butyrate producers associated with pathogen overgrowth [24].


Beta diversity estimation of interindividual variability of gut microbiome was increased in 20/21 studies reporting this (see Additional file 1: Fig. S4). Longitudinal intra-individual variability was also high in each of 4 reporting studies. Rapid substitution of dominant species was commonly observed only within critically ill populations [31], even at phylum level [18].

There is loss of microbiome distinction between body sites over time. Four studies nested multi-day sampling of forehead/antecubital skin, nares, nasogastric or endotracheal aspirates [23, 31, 35]. Initial biogeography was depleted in richness and evenness with convergence seen in both adults and children. Yeh et al. noted commonality of dominant pathogens in distant body sites [23].

Sepsis subgroup

Infection and sepsis were the most frequently reported subgroups. Hospital-acquired infections were seen in a mean 45.5% of patients of the four reporting studies [20, 23, 26, 39]. Wan [30], Ravi [43], Liu [32], and Du et al. [21] recruited only sepsis or septic shock patients. A higher mortality rate was seen among sepsis patients ranging from 17 to 60% in all but two reporting studies which were those employing selective digestive decontamination [34, 35] (Table 2).

Alpha diversity further decreased in sepsis relative to both non-sepsis patients and healthy controls where evaluated [19, 21, 30, 34, 35]. Dysbiosis onset may be delayed: no initial change in alpha diversity was seen in a cohort which later clustered to both sepsis status and source [34]. A longitudinal model noted that sepsis exerted an independent effect on microbiome structure, alongside novel parameters of housing and ICU site [19].

Enrichment of sepsis-specific taxa was apparent, especially Proteobacteria. Ojima et al. reported higher relative abundance of this genus than other categories [18], and sixfold more frequent in sepsis patients elsewhere [30]. Lankelma et al. reported increased Proteobacteria within abdominal sepsis only [34]. Du et al. noted Moraxellaceae was enriched (frequencies are given in Additional file 1: Fig. S6) [21].

Box 1. Summary of results

There are universal, clinically meaningful patterns of early gut dysbiosis in critically ill adults and children (see Box 2)

Early pathogen dominance and commensal depletion can predict mortality where meta-analysis of alpha diversity did not in adults

Lack of methodology and nomenclature standardisation, small sample size, and inadequate reporting confer low certainty in meta-analysis result


Although alpha diversity was globally and markedly reduced by 72 h of ICU admission, it did not predict all-cause mortality at day 28 by meta-analysis, in line with 25 of 26 studies in the systematic review [10, 18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35, 37,38,39, 43, 44]. Confidence in this result is constrained by data availability for meta-analysis, the overall sample size, heterogeneity and moderately high risk of bias.

The novel finding of this systematic review is that overabundance and depletion of selected pathogenic and commensal taxa, respectively, are more frequently predictive of mortality and other clinical outcomes in adults. Alpha diversity metrics do not acknowledge emergence of opportunists, disproportionate loss of obligate anaerobes, infrequent or rare species or detection of syntrophic relationships [45,46,47]. It is plausible these traditional ecological measures of richness and evenness are too ‘blunt’ to show meaningful compositional change and predict outcomes in acute critical illness.

Whilst interactions between organisms across kingdoms may be best observed in ultra-low diversity conditions, they are also not captured by alpha diversity or common sequencing approaches [6]. 16S rRNA amplification was the primary focus of 24 studies yet does not represent fungal species commonly observed by culture, 18S/ITS and whole genome methods. Prevalence studies report Candida spp. in up to 80% of patients 1 week into ICU stay [48]. Ravi noted ESKAPE group and C. albicans constituted up to 75% of some microbiomes [43] similar to the pairings of a single bacterial OTU with C. albicans seen by Zaborin [28]. Dynamic interplay with fastidious bacterial species may govern the behaviour of Candida spp. as an invader, bystander or benefactor given clinical infection is only observed in 5% of those colonised [48] and fungi can recapitulate commensal bacterial functions [49].

Robust and replicable disease-specific patterns were rare [35]. It is plausible that pathogen expansion is largely patient-specific with determination of invasive behaviour by interdependent metabolic, ecological and pharmacologic factors, a hypothesis supported by animal studies [26, 50].

In this regard, antibiotic therapy was seen to be associated with reduced alpha diversity [18,19,20,21,22, 24, 28, 43], particularly carbapenems [38]. Ojima et al. noted the proportion of phylum Actinobacteria was severely reduced by this class [18] and was specifically associated with Pseudomonas abundance (p < 0.01) [19]. Antibiotic usage beyond ICU discharge correlated with sustained loss of alpha diversity in short stay cardiothoracic patients [24].

Antibiotic duration correlated to Acinetobacter spp. and Staphylococcus aureus abundance [21]. Exposure to more than 6 antibiotics also correlated to development of MDR [28]. Emergence of typical MDR organisms were reported in 17 of 26 studies (see Additional file 1: Fig. S5). Garcia et al. noted a median 8-day stay in ICU before acquisition of resistance if this was to occur [20], the same timeframe where diversity is most depleted [51] and where recovery of commensals began, if at all [18]. Circumstantially, these observations suggest day 7–8 in ICU and/or ICU discharge may be imperative junctures for antibiotic rationalisation [18, 20, 24].

There are universal features of early gut dysbiosis amongst the critically ill

Dysbiosis is complex and dynamic within critically ill patients, yet there were a series of commonalities. These patterns were consistent between studies of adults and children, noting paediatric cohorts were smaller. Taxonomic clustering shows separation of microbiota from that of endemic volunteers, distant healthy controls and other hospitalised patients [52, 53]. Loss of biogeographic regionalisation across distant body sites was distinctive of critical illness. Inter- and intra-individual variability seen amongst ICU patients is much greater than that seen in healthy controls [10]. Each of these changes poorly explained by diagnosis, syndrome, or univariate clinical parameters and are amplified in patients with sepsis [21]. A minor contribution is made by demographic factors and disease severity [19] and heritability may be minimal [45], rendering the microbiome an attractive target for ICU interventions.

Regarding composition, richness and evenness is progressively depleted over the first 72 h of ICU stay, with added loss attributable to broad-spectrum antibiotics. As many as 70 of 73 taxa were decreased in this setting [43], although this does not linearly correlate with clinical outcomes. Next-generation approaches have built upon culture-based studies showing disproportionate loss of commensal organisms from both major families, Bacteroidetes and Firmicutes, which are typically dominant with concentrations of 1012–1014 CFU/ml. Disappearance of certain genera among critically ill patients was consistently noted, especially Faecalibacterium, Blautia, and Ruminococcus, and the functional implications are yet to be elucidated from multi-’omic work.

Pathogenic species are thought to encroach on ecological niches created by microbiome disruption [26, 43]. In all reporting studies, organisms which reached metagenomic dominance by any metric, at any body site, predict both colonisation and infectious outcomes. Dominance can predate colonisation by as many as 41 days, and nosocomial infection by 3–21 days [35]. By contrast, in no healthy control samples did one pathogen constitute more than 70% of abundant taxa [10].

In critical illness, coincident biodiversity loss and pathogen dominance was first reported by Iapachino who found a massive presence of Enterococcus by denatured gel electrophoresis and selective PCR [54]. A decade later, Freedberg et al. demonstrated Enterococcus dominance by 16S analysis (and colonisation culture) at ICU admission was an independent predictor of infection and death [26]. In this review, up to 80% of studies featured high abundance of Enterococcus, associated with in-hospital mortality in 3 studies—where alpha diversity was not—as well as other unfavourable clinical and ecological sequelae (Additional file 1: Table S3).

Moreover, Enterococcus has been associated with emergence of drug resistance. An epidemiological study of VRE colonisation found this related to high relative abundance of the genus, and in turn, worst outcomes including mortality within 30 days [55]. While posited as maladaptive ‘overgrowth’, the mechanism by which Enterococcus moderates clinical outcomes is elusive and may be antibiotic dependent. Within a cohort exhibiting Pseudomonas colonies resistant to carbapenems, piperacillin–tazobactam abolished potentially protective taxa and increased risk of Enterococcus dominance [56].

Heterogeneity and microbiome nomenclature is a barrier to consensus

A major challenge for this meta-analysis is the lack of consensus definitions for dysbiosis its core indices for intra- and inter-individual comparison. The recent Strengthening the Organising and Reporting of Microbiome Studies (STORMS) guidelines represent a major step forward in the reporting domain [57].

Heterogeneity is layered in study population, design, sampling and processing methodology, reporting and visualisation. Included studies examine patients admitted to diverse or subspecialist intensive care units and several stratified their results by biomarker, colonisation or other status. Amplicon-based approaches deliver structural taxonomy—conventionally to genus level—upon which function can be inferred. Shotgun metagenomic sequencing delivers subspecies granularity and functional information, however this may obscure hierarchical patterns, with attendant cost and time constraints [9, 58]. Validated methods to adjust for these differences at each pipeline stage are nascent [59, 60].

Common alpha diversity metrics are semi-redundant, and the superior choice of these is not established. Selection of metrics utilised was non-systematic introducing high risk of reporting bias, and in cases only pairwise p values were available. Eligible studies traversed clinical, ecological, ‘resistome’, or methodological research domains. Without standardisation, quality concerns regarding presentation of exposure and outcome variables will persist.

Beta diversity is analysed by taxonomically weighted or unweighted measures of similarity or difference, usually with PERMANOVA comparison and visualisation by Principal Component Orthogonal Analysis plot (see Additional file 3: Fig. S7). This has been qualitatively assessed as comparative clustering is not readily leant to aggregation. As authors are increasingly publishing their data to open-access platforms, pooled or network analyses warrant evaluation.


The included studies were small, the largest registering 301 participants [26]. The varied clinical implications of dysbiosis are infrequently reported leading to exclusion of several high-quality ecological studies. A consensus on the optimal methods for interpretation of metagenomic data is still maturing and the major heterogeneity seen in this meta-analysis will not be resolved by larger studies alone. There is an attendant risk of type II error.

Several modulators were not interrogated by this systematic review due to lack of available data and inability to aggregate or transform various exposures and outcomes. The effect of vasopressors, renal replacement therapy, feeding, medications such as proton pump inhibitors, selective digestive decontamination, and mechanical ventilation on gut dysbiosis is unknown.


This review prompts collection of species or strain-resolved genomes in future studies, given the frequent mortality association with pathogen overabundance. Large, rigorous longitudinal studies are needed to answer the diversity–disease question in this population, and to interrogate exposures such as antibiotics and artificial feeding (and route). These could be embedded using existing MDR screening procedures or as part of novel trials, such as a salvage FMT for refractory septic shock. Data are currently insufficient to target interventions to subclinical groups, however a randomised adaptive trial design could simultaneously evaluate multiple candidate gut therapies with established safety profiles. An array of research directions are laid out in Table 3.


This systematic review of early gut dysbiosis identified that larger, harmonised metagenomic studies in both adult and child cohorts are needed. There are few large, rigorous observational studies within the critically ill population with high heterogeneity. Globally, alpha diversity was dynamically reduced in early ICU admission and was not associated with in-hospital mortality. The abundance of taxa such as Enterococcus spp. appears to offer greater predictive capacity for important clinical and ecological outcomes.

Take-home message

  1. 1.

    Alpha diversity was reduced in early ICU admission in adults and children but was not associated with in-hospital mortality (low-certainty evidence).

  2. 2.

    The abundance of taxa such as Enterococcus spp. appears to offer greater predictive capacity for important clinical and ecological outcomes.

Availability of data and materials

Not applicable.


  1. Rinninella E, Raoul P, Cintoni M, Franceschi F, Miggiano GAD, Gasbarrini A et al (2019) What is the healthy gut microbiota composition? A changing ecosystem across age, environment, diet, and diseases. Microorganisms.

    Article  Google Scholar 

  2. Berg G, Rybakova D, Fischer D, Cernava T, Vergès M-CC, Charles T et al (2020) Microbiome definition re-visited: old concepts and new challenges. Microbiome. 8(1):1–22

    Google Scholar 

  3. Whittaker RH (1972) Evolution and measurement of species diversity. Taxon 21(2–3):213–251

    Article  Google Scholar 

  4. Bae M, Cassilly CD, Liu X, Park SM, Tusi BK, Chen X et al (2022) Akkermansia muciniphila phospholipid induces homeostatic immune responses. Nature 608(7921):168–173

    Article  CAS  Google Scholar 

  5. Hernández M, De Frutos M, Rodríguez-Lázaro D, López-Urrutia L, Quijada NM, Eiros JM (2019) Fecal microbiota of toxigenic Clostridioides difficile-associated diarrhea. Front Microbiol 9:3331

    Article  Google Scholar 

  6. Sharon G, Garg N, Debelius J, Knight R, Dorrestein PC, Mazmanian SK (2014) Specialized metabolites from the microbiome in health and disease. Cell Metab 20(5):719–730

    Article  CAS  Google Scholar 

  7. Hou K, Wu Z-X, Chen X-Y, Wang J-Q, Zhang D, Xiao C et al (2022) Microbiota in health and diseases. Signal Transduct Target Therapy 7(1):1–28

    Article  Google Scholar 

  8. Litton E, Anstey M, Broadhurst D, Chapman A, Currie A, Ferrier J et al (2021) Early and sustained Lactobacillus plantarum probiotic therapy in critical illness: the randomised, placebo-controlled, restoration of gut microflora in critical illness trial (ROCIT). Intensive Care Med 47(3):307–315

    Article  CAS  Google Scholar 

  9. Wozniak H, Beckmann TS, Fröhlich L, Soccorsi T, Le Terrier C, de Watteville A et al (2022) The central and biodynamic role of gut microbiota in critically ill patients. Crit Care.

    Article  Google Scholar 

  10. McDonald D, Ackermann G, Khailova L, Baird C, Heyland D, Kozar R et al (2016) Extreme dysbiosis of the microbiome in critical illness. mSphere.

    Article  Google Scholar 

  11. Bansal S, Nguyen JP, Coburn B (2018) Rectal and naris swabs: practical and informative samples for analyzing the microbiota of critically ill patients. mSphere. 3(3):e00219-e318

    Google Scholar 

  12. Budding AE, Grasman ME, Eck A, Bogaards JA, Vandenbroucke-Grauls CM, van Bodegraven AA et al (2014) Rectal swabs for analysis of the intestinal microbiota. PLoS ONE 9(7):e101344

    Article  Google Scholar 

  13. Bassis CM, Moore NM, Lolans K, Seekatz AM, Weinstein RA, Young VB et al (2017) Comparison of stool versus rectal swab samples and storage conditions on bacterial community profiles. BMC Microbiol 17(1):78

    Article  Google Scholar 

  14. Sterne JA, Hernán MA, Reeves BC, Savović J, Berkman ND, Viswanathan M et al (2016) ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ.

    Article  Google Scholar 

  15. McGrath S, Zhao X, Steele R, Thombs BD, Benedetti A, Collaboration DSD (2020) Estimating the sample mean and standard deviation from commonly reported quantiles in meta-analysis. Stat Methods Med Res 29(9):2520–2537

    Article  Google Scholar 

  16. Egger M, Smith GD, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315(7109):629–634

    Article  CAS  Google Scholar 

  17. StataCorp L (2021) Stata statistical software: release 17. StataCorp LP, College Station, TX

    Google Scholar 

  18. Ojima M, Motooka D, Shimizu K, Gotoh K, Shintani A, Yoshiya K et al (2016) Metagenomic analysis reveals dynamic changes of whole gut microbiota in the acute phase of intensive care unit patients. Dig Dis Sci 61(6):1628–1634

    Article  Google Scholar 

  19. Agudelo-Ochoa GM, Valdes-Duque BE, Giraldo-Giraldo NA, Jaillier-Ramirez AM, Giraldo-Villa A, Acevedo-Castano I et al (2020) Gut microbiota profiles in critically ill patients, potential biomarkers and risk variables for sepsis. Gut Microbes 12(1):1707610

    Article  Google Scholar 

  20. Garcia ER, Vergara A, Aziz F, Narvaez S, Cuesta G, Hernandez M et al (2022) Changes in the gut microbiota and risk of colonization by multidrug-resistant bacteria, infection, and death in critical care patients. Clin Microbiol Infect 28(7):975–982

    Article  Google Scholar 

  21. Du B, Shen N, Tao Y, Sun S, Zhang F, Ren H et al (2021) Analysis of gut microbiota alteration and application as an auxiliary prognostic marker for sepsis in children: a pilot study. Transl Pediatr 10(6):1647–1657

    Article  Google Scholar 

  22. Wei R, Chen X, Hu L, He Z, Chen Q (2021) Dysbiosis of intestinal microbiota in critically ill patients and risk of in-hospital mortality. Am J Transl Res. 3(13):1548–1557

    Google Scholar 

  23. Yeh A, Rogers MB, Firek B, Neal MD, Zuckerbraun BS, Morowitz MJ (2016) Dysbiosis across multiple body sites in critically ill adult surgical patients. Shock 46(6):649–654

    Article  Google Scholar 

  24. Aardema H, Lisotto P, Kurilshikov A, Diepeveen JRJ, Friedrich AW, Sinha B et al (2019) Marked changes in gut microbiota in cardio-surgical intensive care patients: a longitudinal cohort study. Front Cell Infect Microbiol 9:467

    Article  CAS  Google Scholar 

  25. Prevel R, Enaud R, Orieux A, Camino A, Berger P, Boyer A et al (2022) Gut bacteriobiota and mycobiota are both associated with Day-28 mortality among critically ill patients. Crit Care 26(1):105

    Article  Google Scholar 

  26. Freedberg DE, Zhou MJ, Cohen ME, Annavajhala MK, Khan S, Moscoso DI et al (2018) Pathogen colonization of the gastrointestinal microbiome at intensive care unit admission and risk for subsequent death or infection. Intensive Care Med 44(8):1203–1211

    Article  Google Scholar 

  27. Kuo S, Dettmer K, Medini A, Chond D, Uhlemann A, Abrams J, Oefner P, Freedberg D (2021) Associations between urinary 3-indoxyl sulfate, a gut microbiome-derived biomarker, and patient outcomes after intensive care unit admission. J Crit Care 63(June):15–21

    Article  CAS  Google Scholar 

  28. Zaborin A, Smith D, Garfield K, Quensen J, Shakhsheer B, Kade M et al (2014) Membership and behavior of ultra-low-diversity pathogen communities present in the gut of humans during prolonged critical illness. MBio 5(5):e01361-e1414

    Article  Google Scholar 

  29. Xu R, Tan C, Zhu J, Zeng X, Gao X, Wu Q et al (2019) Dysbiosis of the intestinal microbiota in neurocritically ill patients and the risk for death. Crit Care 23(1):195

    Article  Google Scholar 

  30. Wan YD, Zhu RX, Wu ZQ, Lyu SY, Zhao LX, Du ZJ et al (2018) Gut microbiota disruption in septic shock patients: a pilot study. Med Sci Monit 24:8639–8646

    Article  CAS  Google Scholar 

  31. Rogers MB, Firek B, Shi M, Yeh A, Brower-Sinning R, Aveson V et al (2016) Disruption of the microbiota across multiple body sites in critically ill children. Microbiome 4(1):66

    Article  Google Scholar 

  32. Liu J, Wang M, Chen W, Ma J, Peng Y, Zhang M et al (2021) Altered gut microbiota taxonomic compositions of patients with sepsis in a pediatric intensive care unit. Front Pediatr 9:645060

    Article  Google Scholar 

  33. Liu W, Cheng M, Li J, Zhang P, Fan H, Hu Q et al (2020) Classification of the gut microbiota of patients in intensive care units during development of sepsis and septic shock. Genomics Proteom Bioinform 18(6):696–707

    Article  Google Scholar 

  34. Lankelma JM, van Vught LA, Belzer C, Schultz MJ, van der Poll T, de Vos WM et al (2017) Critically ill patients demonstrate large interpersonal variation in intestinal microbiota dysregulation: a pilot study. Intensive Care Med 43(1):59–68

    Article  CAS  Google Scholar 

  35. Lamarche D, Johnstone J, Zytaruk N, Clarke F, Hand L, Loukov D et al (2018) Microbial dysbiosis and mortality during mechanical ventilation: a prospective observational study. Respir Res 19(1):245

    Article  CAS  Google Scholar 

  36. Howard BM, Kornblith LZ, Christie SA, Conroy AS, Nelson MF, Campion EM et al (2017) Characterizing the gut microbiome in trauma: significant changes in microbial diversity occur early after severe injury. Trauma Surg Acute Care Open 2(1):e000108

    Article  Google Scholar 

  37. Fontaine C, Armand-Lefevre L, Magnan M, Nazimoudine A, Timsit JF, Ruppe E (2020) Relationship between the composition of the intestinal microbiota and the tracheal and intestinal colonization by opportunistic pathogens in intensive care patients. PLoS ONE 15(8):e0237260

    Article  CAS  Google Scholar 

  38. Chernevskaya E, Klimenko N, Pautova A, Buyakova I, Tyakht A, Beloborodova N (2021) Host–microbiome interactions mediated by phenolic metabolites in chronically critically ill patients. Metabolites.

    Article  Google Scholar 

  39. Burmeister DM, Johnson TR, Lai Z, Scroggins SR, DeRosa M, Jonas RB et al (2020) The gut microbiome distinguishes mortality in trauma patients upon admission to the emergency department. J Trauma Acute Care Surg 88(5):579–587

    Article  CAS  Google Scholar 

  40. Chao A. Nonparametric estimation of the number of classes in a population. Scand J Stat. 1984;11(4):265–70.

    Google Scholar 

  41. Simpson EH (1949) Measurement of diversity. Nature 163(4148):688

    Article  Google Scholar 

  42. Pielou E (1967) The measurement of diversity in different types of biological collections. J Theor Biol. 15(1):177

    Article  Google Scholar 

  43. Ravi A, Halstead FD, Bamford A, Casey A, Thomson NM, van Schaik W, et al. Loss of microbial diversity and pathogen domination of the gut microbiota in critically ill patients. Microb Genom. 2019;5(9).

  44. Wijeyesekera A, Wagner J, De Goffau M, Thurston S, Rodrigues Sabino A, Zaher S et al (2019) Multi-compartment profiling of bacterial and host metabolites identifies intestinal dysbiosis and its functional consequences in the critically ill child. Crit Care Med 47(9):e727–e734

    Article  CAS  Google Scholar 

  45. Ley RE, Peterson DA, Gordon JI (2006) Ecological and evolutionary forces shaping microbial diversity in the human intestine. Cell 124(4):837–848

    Article  CAS  Google Scholar 

  46. Lozupone CA, Stombaugh JI, Gordon JI, Jansson JK, Knight R (2012) Diversity, stability and resilience of the human gut microbiota. Nature 489(7415):220–230

    Article  CAS  Google Scholar 

  47. Wintermute EH, Silver PA (2010) Emergent cooperation in microbial metabolism. Mol Syst Biol.

    Article  Google Scholar 

  48. Eggimann P, Garbino J, Pittet D (2003) Epidemiology of Candida species infections in critically ill non-immunosuppressed patients. Lancet Infect Dis 3(11):685–702

    Article  Google Scholar 

  49. Jiang TT, Shao TY, Ang WXG, Kinder JM, Turner LH, Pham G et al (2017) Commensal fungi recapitulate the protective benefits of intestinal bacteria. Cell Host Microbe 22(6):809–16 e4

    Article  CAS  Google Scholar 

  50. Mu S, Xiang H, Wang Y, Wei W, Long X, Han Y et al (2022) The pathogens of secondary infection in septic patients share a similar genotype to those that predominate in the gut. Crit Care 26(1):1–15

    Article  Google Scholar 

  51. Kean IRL, Wagner J, Wijeyesekera A, De Goffau M, Thurston S, Clark JA et al (2022) Profiling gut microbiota and bile acid metabolism in critically ill children. Sci Rep 12(1):1–12

    Article  Google Scholar 

  52. Peterson J, Garges S, Giovanni M, McInnes P, Wang L, Schloss JA et al (2009) The NIH human microbiome project. Genome Res 19(12):2317–2323

    Article  Google Scholar 

  53. McDonald D, Hyde E, Debelius JW, Morton JT, Gonzalez A, Ackermann G et al (2018) American gut: an open platform for citizen science microbiome research. Msystems 3(3):e00031-e118

    Article  CAS  Google Scholar 

  54. Iapichino G, Callegari ML, Marzorati S, Cigada M, Corbella D, Ferrari S et al (2008) Impact of antibiotics on the gut microbiota of critically ill patients. J Med Microbiol 57(Pt 8):1007–1014

    Article  CAS  Google Scholar 

  55. Collingwood A, Blostein F, Seekatz AM, Wobus CE, Woods RJ, Foxman B et al (2020) Epidemiological and microbiome associations between Klebsiella pneumoniae and vancomycin-resistant enterococcus colonization in intensive care unit patients. Open Forum Infect Dis 7(1):ofaa012

    Article  Google Scholar 

  56. Pettigrew MM, Gent JF, Kong Y, Halpin AL, Pineles L, Harris AD et al (2019) Gastrointestinal microbiota disruption and risk of colonization with carbapenem-resistant Pseudomonas aeruginosa in intensive care unit patients. Clin Infect Dis 69(4):604–613

    Article  CAS  Google Scholar 

  57. Mirzayi C, Renson A, Zohra F, Elsafoury S, Geistlinger L, Kasselman LJ et al (2021) Reporting guidelines for human microbiome research: the STORMS checklist. Nat Med 27(11):1885–1892

    Article  CAS  Google Scholar 

  58. Qian X-B, Chen T, Xu Y-P, Chen L, Sun F-X, Lu M-P et al (2020) A guide to human microbiome research: study design, sample collection, and bioinformatics analysis. Chin Med J 133(15):1844–1855

    Article  Google Scholar 

  59. Nearing JT, Comeau AM, Langille MG (2021) Identifying biases and their potential solutions in human microbiome studies. Microbiome 9(1):1–22

    Article  Google Scholar 

  60. Poulsen CS, Ekstrøm CT, Aarestrup FM, Pamp SJ (2022) Library preparation and sequencing platform introduce bias in metagenomic-based characterizations of microbiomes. Microbiol Spectr 10(2):e00090-e122

    Article  Google Scholar 

Download references


The authors acknowledge those who kindly shared additional raw for the meta-analysis: Prof Freedberg, Dr Ravi, A/Prof Rubio, Prof Prevel, and Dr Du.


No funding to declare.

Author information

Authors and Affiliations



TE was responsible for conception, primary search, data collation and analysis, generation of graphics, drafting and revision of manuscript. UA was responsible for primary search and data collation. EL and LM contributed significantly to design, manuscript drafting and revision. EL also contributed to search adjudication and data analysis RA contributed significantly to design and methodology. ER contributed significantly to design, drafting and revision.

Corresponding author

Correspondence to Tess Evans.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare no competing or conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Table S1.

Demographic composition, control group and disease severity measures of studies included in systematic review. Table S2. Design and Methodology features of included studies in the systematic review. Table S3. Number of studies reporting adverse outcomes associated with increased relative abundance and/or dominance of Enterococcus genus. Figure S1. Longitudinal intra-individual trend of early gut microbiome samples, stratified by alpha diversity reporting metric. Described relative to earliest same from same patient. X axis represents number of studies reporting that alpha diversity metric, Y axis lists metrics reported. Bar in blue represents a decrease in alpha diversity. Bar in Orange represents no detectable change in diversity. Grey Bar represents studies which did not report alpha diversity direction of change, if any. Figure S2. Alpha diversity lower gut samples at earliest timepoint in ICU admission described relative to non-critically ill controls, stratified by reporting metric. X axis represents percentage of included of studies reporting alpha diversity, Y axis lists metrics reported. Bar in blue represents a decrease in alpha diversity. Bar in Orange represents no detectable change in diversity. Figure S3. Distribution of hypervariable regions of interest among 25 of 26 systematic review studies employing amplicon based sequencing. Figure S4. Proportion of included studies (%) reporting increased versus decreased variability in microbiome between individuals or over course of ICU admission. Figure S5. Proportion of included studies reporting accrual of multi-drug resistant organisms over course of ICU admission. Wedge in dark blue represents confirmed progression of resistance, light blue indicates studies where characteristically multi-resistant organisms increased in relative abundance but resistance was not tested in vitro. Grey represents studies which did not report this phenomenon. Figure S6. Taxonomic flux in sepsis. Radar plot representing the number of studies in systematic review reporting enrichment of a typically pathogenic genus among critically ill patients with sepsis or septic shock.

Additional file 2.

Glossary of Microbiome-related terms.

Additional file 3. Fig. S7.

Commonly reported metrics of the microbiome in clinical studies. Top left panel represents alpha diversity, a composite index representing richness, eveness and dominance of species within a community. Beta diversity (pink banner panels) represents compositional change between communities within an individual over time or between body sites, or between individuals. Relative, or proportional, and absolute abundance of taxa are depicted in the lower middle panel, together with visualisation of log scale change of species within effect size panel (bottom right).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Evans, T., Ali, U., Anderton, R. et al. Lower gut dysbiosis and mortality in acute critical illness: a systematic review and meta-analysis. ICMx 11, 6 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: