Scholarlink

Background <p>Infection are the leading cause of intensive care unit (ICU) admission, yet conventional microbiological methods frequently fail to identify the causative pathogen. Metagenomic next-generation sequencing (mNGS) is an emerging, unbiased, pan-pathogen diagnostic tool. However, its real-world microbiological and clinical impact in the ICU remains poorly characterized. This study aimed to assess the microbiological yield and clinical impact of mNGS when implemented in routine ICU practice.</p> Methods <p>This retrospective multicenter study was conducted across ten tertiary-care ICUs in the Greater Paris area between January 2018 and April 2024. All patients for whom an mNGS analysis was requested by clinicians from a microbiological sample were included. Any additional pathogens identified by mNGS were independently classified as causative, possibly causative, or non-causative by two reviewers. The independent reviewers also categorised therapeutic changes attributable to mNGS as escalation, de-escalation, discontinuation, or other decision support. Discrepancies were adjudicated by a third reviewer.</p> Results <p>A total of 144 mNGS analyses were performed in 132 critically ill patients (median age 55 years), 31% of whom were immunocompromised. The number of mNGS analyses requested increased each year. The most common sample types were cerebrospinal fluid (CSF) (n = 60/144, 41.7%) and pleural fluid (n = 21/144, 14.6%). Pathogens were identified by mNGS in 58 samples (40.3%), with a higher yield in pleural fluid (n = 11/21, 52.4%) than in CSF (n = 16/60, 26.6%). Of the 107 pathogens identified, 43 (40.2%) were detected exclusively by mNGS, notably anaerobic bacteria in pleural fluid and abscess samples. mNGS identified an additional pathogen in 34 cases (25.8%) of the 132 patients included, which was deemed causative in 18 cases (13.6%). mNGS findings influenced therapeutic management in seven patients (5.3%) including five cases of antibiotic de-escalation, one appropriate antibiotic escalation, and one case of clinical decision support.</p> Conclusion <p>In this real-life ICU cohort, mNGS identified additional pathogens in 25.8% of patients, deemed causative in 18 cases (13.6%), and a direct therapeutic impact was observed in 5.3% of cases. However, the median turnaround time of 14 days likely limited its clinical impact. Further studies are needed to better define the role of mNGS in the diagnostic management of critically ill patients.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Real-life impact of clinical metagenomics in the intensive care unit: a multicenter retrospective study in greater paris area hospitals

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月8日

查看全文
  • Pierre Bay,
  • Pierre Cappy,
  • Christophe Rodriguez,
  • Nicolas Mongardon,
  • Matthieu Petit,
  • Guillaume Voiriot,
  • Romain Sonneville,
  • Marc Pineton de Chambrun,
  • Tomas Urbina,
  • Taï Pham,
  • Maxens Decavèle,
  • Sarah Benghanem,
  • Damien Contou,
  • Raphaël Lepeule,
  • Giovanna Melica,
  • Nicolas de Prost,
  • Cécile Angebault,
  • Armand Mekontso Dessap,
  • Paul-Louis Woerther,
  • Keyvan Razazi

摘要

Background

Infection are the leading cause of intensive care unit (ICU) admission, yet conventional microbiological methods frequently fail to identify the causative pathogen. Metagenomic next-generation sequencing (mNGS) is an emerging, unbiased, pan-pathogen diagnostic tool. However, its real-world microbiological and clinical impact in the ICU remains poorly characterized. This study aimed to assess the microbiological yield and clinical impact of mNGS when implemented in routine ICU practice.

Methods

This retrospective multicenter study was conducted across ten tertiary-care ICUs in the Greater Paris area between January 2018 and April 2024. All patients for whom an mNGS analysis was requested by clinicians from a microbiological sample were included. Any additional pathogens identified by mNGS were independently classified as causative, possibly causative, or non-causative by two reviewers. The independent reviewers also categorised therapeutic changes attributable to mNGS as escalation, de-escalation, discontinuation, or other decision support. Discrepancies were adjudicated by a third reviewer.

Results

A total of 144 mNGS analyses were performed in 132 critically ill patients (median age 55 years), 31% of whom were immunocompromised. The number of mNGS analyses requested increased each year. The most common sample types were cerebrospinal fluid (CSF) (n = 60/144, 41.7%) and pleural fluid (n = 21/144, 14.6%). Pathogens were identified by mNGS in 58 samples (40.3%), with a higher yield in pleural fluid (n = 11/21, 52.4%) than in CSF (n = 16/60, 26.6%). Of the 107 pathogens identified, 43 (40.2%) were detected exclusively by mNGS, notably anaerobic bacteria in pleural fluid and abscess samples. mNGS identified an additional pathogen in 34 cases (25.8%) of the 132 patients included, which was deemed causative in 18 cases (13.6%). mNGS findings influenced therapeutic management in seven patients (5.3%) including five cases of antibiotic de-escalation, one appropriate antibiotic escalation, and one case of clinical decision support.

Conclusion

In this real-life ICU cohort, mNGS identified additional pathogens in 25.8% of patients, deemed causative in 18 cases (13.6%), and a direct therapeutic impact was observed in 5.3% of cases. However, the median turnaround time of 14 days likely limited its clinical impact. Further studies are needed to better define the role of mNGS in the diagnostic management of critically ill patients.