1.1 Background

Ventilator-associated pneumonia (VAP) and hospital-acquired pneumonia (HAP) are important causes of  mortality and morbidity in intensive care units (ICU)¹ with gram-negative bacteria often being implicated ². The global increase in multidrug-resistant (MDR) organisms is seen, which propels the usage of colistin regularly as a last-line resort³.

1.2 Rationale

Colistin resistance is being increasingly reported⁴, which makes the ICU-specific respiratory surveillance data very important in guiding therapy⁵ and  the reported proportion of isolates with MIC ≤2 µg/mL to colistin vary among the various geographic areas and methodologically⁶.

1.3 Objective

Review the available reports of MIC ≤2 µg/mL to colistin among the Gram-negative bacteria of adult ICU patients from respiratory samples.

METHODS

2.1 Protocol

PRISMA 2020 guidelines was used for this review⁷,we registered the study in open science framework (DOI:10.17605/OSF.IO/W3AUH).

2.2 Eligibility evaluation

Inclusion:

• Observational studies.
• Adult ICU patients.
• Respiratory samples.
• Isolates with MIC ≤2 µg/mL to colistin.
• Gram-negative bacilli.

Exclusion:

• Studies with children.
• Case reports.
• Studies having no organism-specific data.
• Other than respiratory infections
• Disc diffusion methods of colistin sensitivity

2.3 Information Sources and Search Strategy

Search Strategy

Databases

• PubMed
• Ovid Embase
• Google Scholar

Core Search Concepts

• ICU / critical care
• Respiratory sources
• Gram-negative bacteria
• Colistin susceptibility or resistance

Analysis

All the retrieved records were then imported into Rayyan and analysis was done. The flow of Rayyan which is duplicates screening, Title and abstract screening and full text articles were followed. All the records were independently screened by four reviewers. Discrepancies were resolved by consensus.

2.4 Study Selection

Titles and abstracts were reviewed independently. Full texts were reviewed for eligibility.

2.5 Data Extraction

Data extracted included:

• Study characteristics
• Sample size
• Organisms isolated
• Colistin susceptibility (numerator/denominator)
• Testing method
• Breakpoint criteria

2.6 Risk of Bias Assessment

Risk of bias was done using the JBI Critical Appraisal Checklist for Prevalence Studies⁸.

2.7 Data Synthesis

Because of heterogeneity in study populations and testing methodologies, pooled prevalence estimates and confidence intervals were not done. There was  heterogeneity in breakpoint criteria (CLSI vs EUCAST)^9,10, inclusion of MDR-only populations in some studies, variation in susceptibility testing methods (automated systems vs broth microdilution), and differences in organism distribution and sample size across studies, hence a structured narrative synthesis was conducted.

RESULTS

Study Selection

The database search and study selection process are summarized in Figure 1.

Figure:1 PRISMA 2020 flow diagram

Study Characteristics

The 10 studies that were included were reported between 2013 and 2025 and they from Vietnam (n=2), India (n=2), Turkey (n=3), Saudi Arabia (n=1), Egypt (n=1), and Cuba (n=1).

All studies were observational in design, predominantly retrospective laboratory-based surveillance studies conducted in adult ICUs and  the total number of Gram-negative bacterial (GNB) isolates analyzed across studies ranged from 66 to 1,325 (duplicate isolates from the same patient were excluded in all the studies) . Most studies used automated systems to determine Colistin susceptibility  (e.g., VITEK 2), with interpretation based on CLSI or EUCAST criteria. Evidence of  broth microdilution was not reported in most of the studies.

One study which reported only multidrug-resistant (MDR) isolates, and one study focused exclusively on Escherichia coli isolates.

Risk of Bias Assessment

Risk of bias was performed by  using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Prevalence Studies and it was calculated across 9 JBI checklist items. Overall methodological concern was determined based on domain-level assessment.

Five studies were judged to have low overall risk of bias,  appropriate sampling frames, adequate sample sizes, standardized microbiological identification methods, and sufficient statistical reporting, four studies were assessed as having moderate methodological concerns, primarily due to small sample sizes, unclear sampling strategies, or limited statistical reporting, one study was considered high risk for prevalence estimation because it included only MDR isolates, limiting the representativeness of the general ICU population.

Microbiological identification and susceptibility testing methods were generally standardized and appropriate, but reporting of sample size calculations and confidence intervals was uncommon.

Formal assessment of publication bias (e.g., funnel plot or Egger’s regression) was not performed because a quantitative meta-analysis was not conducted.

Colistin Susceptibility Patterns

Acinetobacter baumannii       

Colistin was reported by nine studies for Acinetobacter baumannii. Isolates with MIC ≤2 µg/mL to colistin ranged from 100% to 87.5%.^11,12 The lowest proportion of  MIC ≤2 µg/mL was reported in a Vietnamese study (87.5%), ¹¹ several Turkish and Middle Eastern studies reported rates exceeding 97%.^13,14. MIC ≥2 µg/mL was not common but was present in certain regions.

Pseudomonas aeruginosa    

Susceptibility for Pseudomonas aeruginosa was given by nine studies for colistin. Isolates with MIC ≤2 µg/mL to colistin had ranged from 80% to 100%^12,14,16. Most studies reported rates above 90%. Lower proportion of isolates with MIC ≤2 µg/mL to colistin was observed in one regional study that included MDR isolates¹⁶. The MDR only study had low proportion of isolates with colistin MIC ≥2 µg/mL despite resistance to other classes.¹⁶

Klebsiella pneumoniae        

The susceptibility data for Klebsiella pneumoniae was reported by nine studies. Variability was observed,  isolates with MIC ≤2 µg/mL to colistin ranging from 13% to 98.4%^17,18. The lowest rates were reported in a study from Saudi Arabia¹⁷, whereas studies from Turkey¹³ and India reported rates above 85%¹². This organism demonstrated the greatest heterogeneity in susceptibility across geographic regions. Indian study also reported  variable susceptibility.¹⁹

Escherichia coli         

Escherichia coli susceptibility was provided by eight studies. Isolates with MIC ≤2 µg/mL to colistin was more compared to other organisms and ranged from 85% to 100%^18,20. One Egypt-based study focusing exclusively on Escherichia coli reported an 85%  of isolates with MIC ≤2 µg/mL to colistin. Escherichia coli is less frequently isolated compared to other organisms.^12,14

Geographic Variation

Regional differences in colistin susceptibility were evident. Studies from Turkey consistently reported high proportion of isolates with MIC ≤2 µg/mL to colistin  across all major Gram-negative organisms. Vietnamese studies demonstrated moderately lower proportion of isolates with MIC ≤2 µg/mL to colistin, particularly for Acinetobacter baumannii. Lower rater was reported by  Saudi Arabian study for Klebsiella pneumoniae. These differences may reflect different antimicrobial usage or infection control practices, and local resistance epidemiology.

Methodological Heterogeneity

We observed heterogeneity across studies; hence we did not perform meta-analysis. Examples of heterogeneity include,

• Specimen types (ETA vs BAL)
• Inclusion of MDR-only isolates
• Testing standards (CLSI vs EUCAST)
• Sample size
• Reporting of statistical precision

DISCUSSION

This systematic review presents available evidence regarding in vitro activity of colistin among Gram-negative pathogens isolated from adult ICU patients with respiratory complaints from 10 observational studies from different regions, colistin showed good  in vitro activity against most major Gram-negative pathogens¹⁴, particularly Acinetobacter baumannii and Pseudomonas aeruginosa. Notable regional variability was observed, especially for Klebsiella pneumoniae.

Principal Findings

The some of the most similar finding across studies include; high activity rates of Acinetobacter baumannii to colistin, with reported rates ranging from 87.5% to 100%. Due to the presence of  widespread carbapenem resistance in numerous ICUs, colistin continues to be used as a last resort therapeutic option for this pathogen¹³.

Pseudomonas aeruginosa also demonstrated high proportion of isolates with MIC ≤2 µg/mL to colistin in most studies, generally exceeding 90%¹²,which is consistent with current understanding that colistin remains active against many non-fermenting Gram-negative organisms, although sporadic isolates with  MIC ≥2 µg/mL for colistin has been reported globally.

In contrast, proportion of  Klebsiella pneumoniae isolates that exhibited MIC ≤2 µg/mL for colistin, ranged from as low as 13% ¹⁷ in one regional study to greater than 95%¹² in others. This may reflect differences in local antimicrobial usage, the presence of transferable MCR gene colistin resistance mechanisms⁴, along with regional antimicrobial stewardship practices.

Escherichia coli isolates with MIC ≤2 µg/mL to colistin remained on the higher side across studies,  above 85%²⁰, data regarding this bacteria were limited in some settings, and one study focused exclusively on this bacteria, hence broader interpretation is limited.

All of the above findings reinforce the fact that continued surveillance and local data is the mainstay for treatment of such infections¹⁶.

Geographic and Epidemiological Variation

Marked geographic variation was evident across the included studies. Turkish ICU settings consistently reported high proportion of isolates with MIC ≤2 µg/mL  to colistin across organisms¹³, whereas some studies from Vietnam¹¹ and the Middle East demonstrated emerging proportion of isolates with MIC ≥2 µg/mL, particularly among Enterobacterales. Regional variations in antibiotic usage, institution specific infection control measures, difference in healthcare infrastructures, and multiple molecular resistance epidemiology might be the reason for these differences observed. Lower number of isolates with  MIC ≤2 µg/mL were observed in Middle Eastern cohorts compared to Turkish studies particularly among Klebsiella pneumoniae ¹⁷ .

ICU environments in general are known to create conditions that facilitate emergence and spread of multidrug-resistant pathogens due to prolonged hospitalization, usage of invasive devices and procedures, antibiotic exposure, and severe patient illness²¹.

Methodological Considerations

While most included studies used standardized microbiological identification systems and interpreted susceptibility according to recognized criteria (CLSI or EUCAST)^9,10, heterogeneity in testing methodology was observed. Breakpoint interpretation differs between the major standards, while, CLSI defines isolates with MIC ≤2 µg/mL as intermediate and ≥4 µg/mL as resistant, whereas EUCAST classifies MIC ≤2 mg/L as susceptible and >2 mg/L as resistant. For the purpose of this review, isolates categorized as intermediate or susceptible by CLSI and susceptible by EUCAST were considered colistin-susceptible, reflecting comparable MIC thresholds across both standards. For the sake of harmonization, ‘sensitive’ or ‘intermediate’ were not used; interpretation should be considered as per standards used in each study. Despite broth microdilution being a  reference standard for colistin susceptibility testing, only a few studies used it explicitly, instead many relied on automated systems for MIC determination such as Vitek 2 compact and Microscan^12,13, these systems; while widely used in  routine clinical laboratories, they may occasionally misclassify colistin susceptibility²².

Sample sizes varied considerably across studies, and formal sample size calculations were rarely reported, one study restricted itself to only multidrug-resistant isolates, which would likely underestimate isolates with MIC ≤2 mg/L compared to general ICU populations and another study was restricted to a single organism, which would limit representativeness. The considerable clinical and methodological heterogeneity lead authors not to  perform quantitative pooling.

The higher interpretative breakpoints were higher without mentioning the standard in the MDRO only, which may have lead to underestimation relative to contemporary breakpoints.

Clinical significance

The results of the study confirm the continued utility of colistin as a last-resort therapeutic agent for difficult to treat infections by multidrug-resistant Gram-negative bacteria in ICU,  however, there is variability observed in Klebsiella pneumoniae susceptibility.

Finding of the review implies, routine testing of colistin is important, especially when increasing reports of plasmid-mediated resistance mechanisms exist. Overreliance on colistin without active monitoring may increase the rate of resistance development, potentially compromising one of the only treatment options for extensively drug-resistant infections²³.

Strengths and Limitations of the Review

This review included studies from many continents and provided a review of organism-specific patterns in adult ICU respiratory infections. This review has  methodological transparency due to systematic assessment of risk of bias using the JBI checklist.

Limitations.

Heterogeneity in study design, specimen type, and susceptibility testing methodology, in-frequent reporting of confidence intervals, uniform evaluation of molecular mechanisms of colistin resistance. many studies were single-center and retrospective^15,16,19, which may limit generalizability. Due to non-representation of some geographic regions; generalizability is limited and reflects regional publication patterns more than true epidemiological distribution and due to change in the methodology with the standards many studies were excluded due to disc diffusion methodology.

Potential Publication and Reporting Bias

Even though  a  quantitative meta-analysis was not conducted, there may be  a risk of publication bias. The studies which reported emerging or unusually high colistin resistance may be more likely that they were  published due to clinical concern, which in turn overestimates the prevalence. The centers with persistently low isolates with MIC ≤2 mg/L may be underrepresented and  many of the included studies were single-center and relatively small. Many of these single center studies  tend to have small-study effects more so, if resistance patterns were caused  by outbreaks localized to the region or hospital and geographic representation was uneven. Predominance of studies from some regions and lack of large multicenter datasets, it has limited global generalizability, mindfull ness of factors are needed while interpreting the reported patterns.

Future Research should prioritize

• Testing using broth microdilution¹⁰.
• Reporting of confidence intervals.
• Molecular characterization⁶.
• Multicenter ICU surveillance.
• Prospective studies.

CONCLUSION

Relatively good activity for colistin in vitro against Acinetobacter baumannii, Pseudomonas aeruginosa, and Escherichia coli is noted across settings for adult ICU respiratory infections ¹⁴. However, marked regional variability, especially among Klebsiella pneumoniae, which indicates emerging resistance in certain regions¹⁷. Susceptibility patterns observed across diverse geographic regions are similar other than few variations ^11,14,15,17. Regular local surveillance and careful stewardship are needed to maintain the clinical effectiveness of this last-line agent¹⁸.

DECLARATIONS

Funding

None.

Conflicts of Interest

The authors declare no conflicts of interest.

manuscript.

REFERENCES

1. Kalil AC, Metersky ML, Klompas M, Muscedere J, Sweeney DA, Palmer LB, et al. Management of Adults With Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the American Thoracic Society [Internet]. [cited 2026 Feb 13]. Available from: https://dx.doi.org/10.1093/cid/ciw353
2. Kalanuria AA, Zai W, Mirski M. Ventilator-associated pneumonia in the ICU. Crit Care. 2014;18(2):208. doi:10.1186/cc13775 PubMed PMID: 25029020; PubMed Central PMCID: PMC4056625.
3. Falagas ME, Kasiakou SK. Colistin: the revival of polymyxins for the management of multidrug-resistant gram-negative bacterial infections. Clin Infect Dis. 2005 May 1;40(9):1333–41. doi:10.1086/429323 PubMed PMID: 15825037.
4. Liu YY, Wang Y, Walsh TR, Yi LX, Zhang R, Spencer J, et al. Emergence of plasmid-mediated colistin resistance mechanism MCR-1 in animals and human beings in China: a microbiological and molecular biological study. Lancet Infect Dis. 2016 Feb;16(2):161–8. doi:10.1016/S1473-3099(15)00424-7 PubMed PMID: 26603172.
5. Centers for Disease Control and Prevention. Antibiotic Resistance Threats in the United States, 2019 [Internet]. Atlanta, GA: U.S. Department of Health and Human Services, CDC; 2019. Report No. Available from: https://www.cdc.gov/drugresistance/biggest-threats.html doi:10.15620/cdc:82532
6. Poirel L, Jayol A, Nordmann P. Polymyxins: Antibacterial Activity, Susceptibility Testing, and Resistance Mechanisms Encoded by Plasmids or Chromosomes. Clinical Microbiology Reviews. 2017 Mar 8;30(2):557–96. doi:10.1128/cmr.00064-16
7. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021 Mar 29;372:n71. doi:10.1136/bmj.n71 PubMed PMID: 33782057.
8. Munn Z, Barker TH, Moola S, Tufanaru C, Stern C, McArthur A, et al. The Joanna Briggs Institute Critical Appraisal Tools for Use in JBI Systematic Reviews [Internet]. Adelaide, Australia: Joanna Briggs Institute; 2020. (JBI Manual for Evidence Synthesis). Available from: https://synthesismanual.jbi.global
9. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing [Internet]. Wayne, PA: Clinical and Laboratory Standards Institute; 2024. (CLSI Supplement M100). Available from: https://clsi.org/standards/products/microbiology/documents/m100/
10. European Committee on Antimicrobial Susceptibility Testing. EUCAST Breakpoint Tables for Interpretation of MICs and Zone Diameters [Internet]. 2024. Available from: https://www.eucast.org/clinical_breakpoints/
11. Le HH, Nguyen AV, Vu LH, Nguyen VTH, Pham HQ, Le HV, et al. Antimicrobial Resistance Patterns of Common Gram-Negative Microorganisms Isolated from Patients with Lower Respiratory Tract Infections in a Teaching Hospital in Vietnam. Jpn J Infect Dis. 2024 May 31;77(3):144–54. doi:10.7883/yoken.JJID.2023.260
12. Koneru KS, Joy P, Gangadharan V, Kumar A, Kumar S. Microbiological Profile of Bronchoalveolar Lavage in Patients with Nosocomial Pneumonia in Intensive Care Unit. JCDR. 2018. doi:10.7860/JCDR/2018/36708.11794
13. Tartar A, Ozer A, Ulu R, Akbulut A. Microbiological Evaluation of the Pathogens Isolated From the Endotracheal Aspirate Samples of the Patients Followed in the Intensive Care Units: A One-Year Retrospective Analysis. Klimik Dergisi/Klimik Journal. 2018 Jun 6;56–60. doi:10.5152/kd.2018.14
14. Medell M, Hart M, Duquesne A, Espinosa F, Valdés R. Nosocomial ventilator-associated pneumonia in Cuban intensive care units: bacterial species and antibiotic resistance. MEDICC Rev. 2013 Apr;15(2):26–9. doi:10.37757/MR2013V15.N2.6 PubMed PMID: 23686252.
15. Arici N, Aksaray S, Ankarali H. Gram-Negative Bacteria and Antibiotic Resistances Isolated from Endotracheal Aspirate Samples of Patients Hospitalized in the Intensive Care Unit: 3-Years Retrospective Analysis. International Journal of Scientific and Technological Research. 2020;6(3):255.
16. Vo TPM, Dinh TC, Phan HV, Cao TTM, Duong PT, Nguyen T. Ventilator-Associated Pneumonia Caused by Multidrug-Resistant Gram-Negative Bacteria in Vietnam: Antibiotic Resistance, Treatment Outcomes, and Colistin-Associated Adverse Effects. Healthcare. 2022 Sep 14;10(9):1765. doi:10.3390/healthcare10091765
17. Almouwlid A, Albenasy K, Kamel Y, Abdelmoktader A, Alaidarous M, Abdel-Hadi A. Incidence and Antibiotic Susceptibility of Gram‐Negative Bacteria Associated With Chest Infections in Intensive Care Unit Patients From a Selected Hospital in the Kingdom of Saudi Arabia. Wesley H, editor. BioMed Research International. 2025 Jan;2025(1):5283526. doi:10.1155/bmri/5283526
18. Gürbüz M, Çi̇Mke P, Altunkara H, Altinkaya FU, Şen G. Evaluation of Pathogens and Antibiotic Resistance Profiles Isolated from Lower Respiratory Tract Samples. Kocaeli Üniversitesi Sağlık Bilimleri Dergisi. 2023 Feb 2;9(1):14–9. doi:10.30934/kusbed.1083430
19. Agarwal S, Kakati B, Kishore N, Khanduri S, Singh M. Colistin resistance in organisms causing ventilator-associated pneumonia – Are we going into pre-antibiotic era? Critical Care and Shock. 2018;21(2):78–85.
20. El-Mokhtar MA, Daef E, Mohamed Hussein AAR, Hashem MK, Hassan HM. Emergence of Nosocomial Pneumonia Caused by Colistin-Resistant Escherichia coli in Patients Admitted to Chest Intensive Care Unit. Antibiotics. 2021 Feb 24;10(3):226. doi:10.3390/antibiotics10030226
21. Global Antimicrobial Resistance and Use Surveillance System (GLASS) [Internet]. [cited 2026 Feb 13]. Available from: https://www.who.int/initiatives/glass
22. Humphries RM, Green DA, Schuetz AN, Bergman Y, Lewis S, Yee R, et al. Multicenter Evaluation of Colistin Broth Disk Elution and Colistin Agar Test: a Report from the Clinical and Laboratory Standards Institute. J Clin Microbiol. 2019 Nov;57(11):e01269-19. doi:10.1128/JCM.01269-19 PubMed PMID: 31511331; PubMed Central PMCID: PMC6813006.
23. Nation RL, Velkov T, Li J. Colistin and polymyxin B: peas in a pod, or chalk and cheese? Clin Infect Dis. 2014 Jul 1;59(1):88–94. doi:10.1093/cid/ciu213 PubMed PMID: 24700659; PubMed Central PMCID: PMC4305129.