Optimizing artificial intelligence in sepsis management: Opportunities in the present and looking closely to the future

O'Reilly, Darragh; McGrath, Jennifer; Martin-Loeches, Ignacio

doi:10.1016/j.jointm.2023.10.001

Cited by 4 publications

(2 citation statements)

References 84 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nevertheless, these challenges can be overcome through the utilization of artificial intelligence tools, specifically with the integration of machine learning with clinical data and molecular markers, such as cfDNA. This approach enables the development of clinical decision support tools pertaining to sepsis and septic shock, ultimately enhancing outcomes for clinicians and facilitating real-time optimization of medical resources [32,33].…”

Section: Discussionmentioning

confidence: 99%

Cell-Free Nuclear and Mitochondrial DNA as Potential Biomarkers for Assessing Sepsis Severity

de Miranda,

Claudio,

de Almeida

et al. 2024

Biomedicines

View full text Add to dashboard Cite

Sepsis continues to be a significant public health challenge despite advances in understanding its pathophysiology and management strategies. Therefore, this study evaluated the value of cell-free nuclear DNA (cf-nDNA) and cell-free mitochondrial DNA (cf-mtDNA) for assessing the severity and prognosis of sepsis. Ninety-four patients were divided into three groups: infection (n = 32), sepsis (n = 30), and septic shock (n = 32). Plasma samples were collected at the time of diagnosis, and cfDNA concentrations were determined by qPCR assay. The results showed that plasma cfDNA levels increased with the severity of the disease. To distinguish between patients with infection and those with sepsis, the biomarker L1PA290 achieved the highest AUC of 0.817 (95% CI: 0.725–0.909), demonstrating a sensitivity of 77.0% and a specificity of 79.3%. When cf-nDNA was combined with the SOFA score, there was a significant improvement in the AUC (0.916 (0.853–0.979)), sensitivity (88.1%), and specificity (80.0%). Moreover, patients admitted to the ICU after being diagnosed with sepsis had significantly higher cf-nDNA concentrations. In patients admitted to the ICU, combining cf-nDNA with the SOFA score yielded an AUC of 0.753 (0.622–0.857), with a sensitivity of 95.2% and a specificity of 50.0%. cfDNA can differentiate between patients with infection and those with sepsis. It can also identify patients who are likely to be admitted to the ICU by predicting those with indications for intensive care, suggesting its potential as a biomarker for sepsis.

show abstract

Section: Discussionmentioning

confidence: 99%

Cell-Free Nuclear and Mitochondrial DNA as Potential Biomarkers for Assessing Sepsis Severity

de Miranda,

Claudio,

de Almeida

et al. 2024

Biomedicines

View full text Add to dashboard Cite

show abstract

“…The dosing software model, description, and application are given below [121]: Whilst there is evidence for artificial intelligence (AI) in the recognition and treatment of sepsis in ICU patients, its use in antimicrobial dosing is not yet established [124,125]. Dosing software in critical care has been implemented recently with specialised algorithms to help determine and administer appropriate medication doses for critically ill patients.…”

Section: Therapeutic Drug Monitoring and Dosing Softwarementioning

confidence: 99%

Practical Lessons on Antimicrobial Therapy for Critically Ill Patients

Cusack,

Little,

Martin-Loeches

2024

Antibiotics

View full text Add to dashboard Cite

Sepsis stands as a formidable global health challenge, with persistently elevated mortality rates in recent decades. Each year, sepsis not only contributes to heightened morbidity but also imposes substantial healthcare costs on survivors. This narrative review aims to highlight the targeted measures that can be instituted to alleviate the incidence and impact of sepsis in intensive care. Here we discuss measures to reduce nosocomial infections and the prevention of equipment and patient colonisation by resilient pathogens. The overarching global crisis of bacterial resistance to newly developed antimicrobial agents intensifies the imperative for antimicrobial stewardship and de-escalation. This urgency has been accentuated in recent years, notably during the COVID-19 pandemic, as high-dose steroids and opportunistic infections presented escalating challenges. Ongoing research into airway colonisation’s role in influencing disease outcomes among critically ill patients underscores the importance of tailoring treatments to disease endotypes within heterogeneous populations, which are important lessons for intensivists in training. Looking ahead, the significance of novel antimicrobial delivery systems and drug monitoring is poised to increase. This narrative review delves into the multifaceted barriers and facilitators inherent in effectively treating critically ill patients vulnerable to nosocomial infections. The future trajectory of intensive care medicine hinges on the meticulous implementation of vigilant stewardship programs, robust infection control measures, and the continued exploration of innovative and efficient technological solutions within this demanding healthcare landscape.

show abstract

Investigating computational models for diagnosis and prognosis of sepsis based on clinical parameters: Opportunities, challenges, and future research directions

Gupta,

Majumder,

Sengupta

et al. 2024

Journal of Intensive Medicine

View full text Add to dashboard Cite

Optimizing artificial intelligence in sepsis management: Opportunities in the present and looking closely to the future

Cited by 4 publications

References 84 publications

Cell-Free Nuclear and Mitochondrial DNA as Potential Biomarkers for Assessing Sepsis Severity

Cell-Free Nuclear and Mitochondrial DNA as Potential Biomarkers for Assessing Sepsis Severity

Practical Lessons on Antimicrobial Therapy for Critically Ill Patients

Investigating computational models for diagnosis and prognosis of sepsis based on clinical parameters: Opportunities, challenges, and future research directions

Contact Info

Product

Resources

About