Comparison between quantitative and qualitative culture of tracheal aspirates in the diagnosis of ventilator associated pneumonia

Marco, Fvc De; Camargo, Lfa; Barbas, Csv; Hoelz, C; Pasternack, Jacyr; Bueno, Mas; Martino, Marinês Dalla Valle; Rodrigues, Mário; Caserta, Raquel; Amado, VM; Knobel, E

doi:10.1186/cc1114

Cited by 3 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An alert at or before VAP onset by any of these methods is likely to improve identification of VAP, potentially overcoming limitations in diagnostic criteria that may lead to both under-and over-treatment with antibiotics. [14,44,45] It is worth noting that XGBoost demonstrated decreasing performance over longer data collection windows for the intubation task. This may be due to the fact that, with increasing time from the point at which predictions are made, many of the model inputs lose clinical relevance to the current patient state, decreasing the overall relevance of the inputs.…”

Section: Discussionmentioning

confidence: 99%

“…While the highest overall AUROC was demonstrated by XGBoost using data from the 6-hour window leading up to and including the prediction time, all models met or exceeded the performance of existing VAP identification methods. An alert at or before VAP onset by any of these methods is likely to improve identification of VAP, potentially overcoming limitations in diagnostic criteria that may lead to both under- and over-treatment with antibiotics [14,44,45] . It is worth noting that XGBoost demonstrated decreasing performance over longer data collection windows for the intubation task.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting ventilator-associated pneumonia with machine learning

Giang¹,

Calvert²,

Rahmani³

et al. 2021

Medicine

View full text Add to dashboard Cite

Ventilator-associated pneumonia (VAP) is the most common and fatal nosocomial infection in intensive care units (ICUs). Existing methods for identifying VAP display low accuracy, and their use may delay antimicrobial therapy. VAP diagnostics derived from machine learning (ML) methods that utilize electronic health record (EHR) data have not yet been explored. The objective of this study is to compare the performance of a variety of ML models trained to predict whether VAP will be diagnosed during the patient stay. A retrospective study examined data from 6126 adult ICU encounters lasting at least 48 hours following the initiation of mechanical ventilation. The gold standard was the presence of a diagnostic code for VAP. Five different ML models were trained to predict VAP 48 hours after initiation of mechanical ventilation. Model performance was evaluated with regard to the area under the receiver operating characteristic (AUROC) curve on a 20% hold-out test set. Feature importance was measured in terms of Shapley values. The highest performing model achieved an AUROC value of 0.854. The most important features for the best-performing model were the length of time on mechanical ventilation, the presence of antibiotics, sputum test frequency, and the most recent Glasgow Coma Scale assessment. Supervised ML using patient EHR data is promising for VAP diagnosis and warrants further validation. This tool has the potential to aid the timely diagnosis of VAP.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Predicting ventilator-associated pneumonia with machine learning

Giang¹,

Calvert²,

Rahmani³

et al. 2021

Medicine

View full text Add to dashboard Cite

show abstract

“…Sputum quality scoring of ET secretion was used (Camargo, 2004) [ 9 ]. A smear from the purulent sputum was a Gram-stained smear and observed under a low power (10X) objective.…”

Section: Methodsmentioning

confidence: 99%

Clinico-Microbiological Correlates of Hospital-Acquired Pneumonia: A Hospital-Based Prospective Cohort Study

Agrawal,

Gupta

2023

Cureus

View full text Add to dashboard Cite

Background and objectives: Hospital-acquired pneumonia (HAP) is a life-threatening hospital-acquired infection contributing to poor outcomes and mortality. Though the prevalence is comparable, the burden of comorbidities and malnutrition further worsens the scenario in developing countries. Infective agents responsible for these infections vary between regions due to the variables involved. There is a dearth of data on clinico-microbiological correlates of HAP from Northern India. With this study, we aim to explore the same and add more evidence to fill the gap. Methodology: A hospital-based cohort study was done on ICU patients of the tertiary care center in Northern India including the cohort of patients obeying a strict inclusion criterion. The clinical and microbiological correlates were estimated following an appraisal of quality of study samples. Results: We found that the most common clinical feature in patients with HAP was fever (82%) followed by purulent respiratory secretions (72%), tachycardia (52%), and crepitations on auscultation (38%). Approximately 86% of cases were found to be culture-positive while others were bacteriologically sterile. Gram-negative bacilli were more commonly isolated (83% Gram-negative vs 17% Gram-positive). The most common organisms isolated were Klebsiella pneumoniae, Citrobacter freundii, Escherichia coli, Acinetobacter, and Pseudomonas aeruginosa. Staphylococcus aureus was isolated from eight specimens and all isolates were susceptible to vancomycin, linezolid, teicoplanin, and tigecycline. Seven isolates were resistant to clindamycin and all 8 were resistant to macrolides and quinolones. Five strains had methicillin resistance indicating a rising burden of ‘superbugs’. The most common side involved was the right side and the right middle zone was the most common zone involved. Forty-four percent of cases had a poor outcome and succumbed to the infection. Conclusions: HAP places patients at a heightened risk of mortality and manifests a distinctive clinical-microbiological profile. It is advisable to adopt a proactive stance in averting HAI by adhering to robust prophylaxis and management protocols in alignment with regional data and hospital guidelines. Despite the study's constrained sample size, it contributes significant insights specific to the region. This underscores the necessity for further exploration through analogous studies and audits in the northern part of India. Such endeavors have the potential to tailor treatment approaches for patients, ultimately enhancing overall outcomes.

show abstract

Development of a quality indicator set to measure and improve quality of ICU care in low- and middle-income countries

Pari¹,

Sluijs²,

Lopez³

et al. 2022

Intensive Care Med

View full text Add to dashboard Cite

Purpose: To develop a set of actionable quality indicators for critical care suitable for use in low-or middle-income countries (LMICs).Methods: A list of 84 candidate indicators compiled from a previous literature review and stakeholder recommendations were categorised into three domains (foundation, process, and quality impact). An expert panel (EP) representing stakeholders from critical care and allied specialties in multiple low-, middle-, and high-income countries was convened. In rounds one and two of the Delphi exercise, the EP appraised (Likert scale 1-5) each indicator for validity, feasibility; in round three sensitivity to change, and reliability were additionally appraised. Potential barriers and facilitators to implementation of the quality indicators were also reported in this round. Median score and interquartile range (IQR) were used to determine consensus; indicators with consensus disagreement (median < 4, IQR ≤ 1) were removed, and indicators with consensus agreement (median ≥ 4, IQR ≤ 1) or no consensus were retained. In round four, indicators were prioritised based on their ability to impact cost of care to the provider and recipient, staff wellbeing, patient safety, and patient-centred outcomes.Results: Seventy-one experts from 30 countries (n = 45, 63%, representing critical care) selected 57 indicators to assess quality of care in intensive care unit (ICU) in LMICs: 16 foundation, 27 process, and 14 quality impact indicators after round three. Round 4 resulted in 14 prioritised indicators. Fifty-seven respondents reported barriers and facilitators, of which electronic registry-embedded data collection was the biggest perceived facilitator to implementation (n = 54/57, 95%) Concerns over burden of data collection (n = 53/57, 93%) and variations in definition (n = 45/57, 79%) were perceived as the greatest barrier to implementation. Conclusion:This consensus exercise provides a common set of indicators to support benchmarking and quality improvement programs for critical care populations in LMICs.

show abstract

Comparison between quantitative and qualitative culture of tracheal aspirates in the diagnosis of ventilator associated pneumonia

Abstract: P14 Facial continuous positive airway pressure therapy for cardiogenic pulmonary oedema: a study of its efficacy in an emergency department setting within the UK

Cited by 3 publications

References 0 publications

Predicting ventilator-associated pneumonia with machine learning

Predicting ventilator-associated pneumonia with machine learning

Clinico-Microbiological Correlates of Hospital-Acquired Pneumonia: A Hospital-Based Prospective Cohort Study

Development of a quality indicator set to measure and improve quality of ICU care in low- and middle-income countries

Contact Info

Product

Resources

About