Esophagogastroduodenoscopy Versus Thoracic Computed Tomography in Prediction of Caustic Stricture.

Background/Objectives: Corrosive substance intake remains a significant public health concern due to its potential for severe gastrointestinal (GI) injury and associated morbidity. Early risk stratification is crucial for appropriate management, yet there is a lack of reliable non-invasive predictors of injury severity. This study aimed to evaluate the clinical characteristics of adult patients with corrosive ingestion and to investigate the correlation between inflammatory markers and endoscopic injury severity. Methods: In this retrospective study, 83 adult patients who underwent esophagogastroduodenoscopy (EGD) following corrosive ingestion between January 2017 and January 2021 were analyzed. Endoscopic injuries were graded using the Zargar classification and categorized as mild (grades 0–2a) or severe (grades 2b–4). Demographic, clinical, endoscopic, and laboratory data, including neutrophil-to-lymphocyte ratio (NLR) and C-reactive protein (CRP) levels, were recorded. The correlation between inflammatory markers and injury severity was assessed, and receiver operating characteristic (ROC) analysis was performed to determine diagnostic accuracy. Results: Among the patients, 55.4% were female with a mean age of 41.5 ± 17.3 years. Most ingestions were accidental (74.7%), with bleach being the most common agent (41%). Endoscopic injury was detected in 55.4% of patients, predominantly in the stomach and esophagus. Severe injuries were associated with longer intensive care and hospital stays, increased complication rates, and more frequent organ involvement (p < 0.001). A weak but statistically significant correlation was found between injury severity and both NLR (r = 0.357, p = 0.001) and CRP (r = 0.247, p = 0.024). ROC analysis revealed an NLR cut-off of 2.95 (AUC = 0.804) and CRP cut-off of 2.5 (AUC = 0.706) for predicting severe injury. Conclusions: Early endoscopic evaluation remains essential for assessing corrosive injury severity. However, NLR and CRP may serve as useful, non-invasive indicators in predicting injury severity, potentially aiding clinical decision-making, especially in settings where endoscopy is not readily available or is contraindicated.

Context. Corrosive substance ingestion is a toxicological emergency with relatively high mortality requiring rational surgical decisions. Objective. Evaluate the role of chest and abdominal computed tomography (CT) in assessing the severity of acute corrosive ingestion. Methods. A retrospective study of adults admitted due to corrosive ingestion, who underwent gastrointestinal endoscopy and CT within 48 h of admission. Endoscopy findings were graded as 0, 1, 2a, 2b, 3a, and 3b (Zargar's criteria), CT findings were graded as 0, 1, 2, and 3. For each patient endoscopy and CT grades were compared, and sensitivity and specificity for predicting mortality or emergency laparotomy were calculated. Results. Twenty-three patients were included, aged 18–87 years; seven underwent emergency laparotomy, five died. Endoscopy grading was higher than CT grading in 14 patients (66%). The sensitivities of endoscopy grades 2b and 3 to predict mortality and emergency laparotomy were 1 and 0.8, respectively; the specificities were 0.38 and 0.37, respectively. The sensitivities of CT grade 3 to predict mortality and emergency laparotomy were 0.4 and 0.28, respectively; the specificities were 0.94 and 0.93, respectively. Three patients had pulmonary infiltrates on CT but not on chest X-ray. Discussion. CT tends to underestimate the severity of corrosive ingestion compared with endoscopy. It has lower sensitivity and higher specificity than endoscopy in predicting major outcome. CT can provide important information on lung injury, and when endoscopy cannot be completed. Conclusion. CT should not be the only basis for surgical decisions during the initial phase of acute corrosive ingestions.

Introduction. Corrosive ingestions are common, although most ingestions do not result in clinically significant effects. Limited guidance is available on the role of endoscopy and/or computerised tomography (CT) in the investigation of individuals with corrosive ingestion, and the present data regarding predictors of poor outcome are confusing. Furthermore, whilst there are many case series describing the use of endoscopy in corrosive ingestions, no clear ideal time frame has been established as to when it should be undertaken. More recently, CT has been used to grade injuries, but there are few studies on its role in managing corrosive injuries, and those studies that have been reported are conflicting in their results. Methods. A Medline search was performed with the terms ‘Caustic ingestion’ and ‘Corrosive ingestion’ and a second search by adding the words ‘Endoscopy’, ‘CT’, and ‘Computerised tomography’ as a subject term or keyword. These searches revealed a total of 277 reviews and papers, of which 33 original papers were relevant for analysis. Three further papers were identified during the analysis of these papers and a PubMed search of the same terms added one further paper, bringing the total to 37. There have been no prospective, randomised controlled trials directly comparing endoscopy and CT. Only two retrospective studies compared the use of CT and that of endoscopy. Thirty-five studies examined whether an endoscopy is always needed, and if so, within what time frame this should be done: CT or endoscopy? A review of these studies suggests that the data regarding the use of CT in these circumstances are not yet of sufficient weight to replace endoscopy as the first-line investigation in corrosive ingestion–related injury. Who needs investigation after corrosive ingestion? We believe that signs and symptoms indicate the likelihood of clinically significant injury in adults. Specifically, any evidence of oropharyngeal burns, drooling, vomiting, pain or dysphagia clearly indicates the need for an endoscopy. In children, it appears that an even greater degree of caution is needed. How soon after ingestion should investigation be performed? For whom an endoscopy is required, it is prudent to enable surgery and other specifics regarding management of corrosives to be decided quickly (< 12 h). There are many incidences where endoscopy has been done safely beyond 48 h although this is not needed frequently. Management recommendations Asymptomatic patients, particularly adults with a normal clinical examination and who can eat and drink normally, can be discharged safely without endoscopy. Endoscopy is preferred over CT in the assessment of risk in symptomatic patients with corrosive ingestion. If patients have any oropharyngeal injury and in particular symptoms of drooling, vomiting, dysphagia or pain (retrosternal or otherwise), the risk of having a high-grade injury is higher, and urgent endoscopy should be performed to grade the injury and determine whether surgical intervention is required. Patients who have non-specific symptoms, such as cough, should also undergo endoscopy, but this is less urgent. Conclusions Despite the lack of high-quality clinical trial data, the available evidence and clinical experience support the use of early endoscopy (< 12 h) in patients who are symptomatic after ingestion of a corrosive substance. We propose a clinical guideline that can be used to help plan management of corrosives.

OBJECTIVES:Esophagogastroduodenoscopy is considered the gold standard in assessing the severity of injury to the gastrointestinal tract following corrosive ingestion. Zargar’s endoscopic grading of injury helps in prognostication as well as guiding management. Since the major burden of cases lies in resource-limited settings, the availability of endoscopic evaluation is a limiting factor. Hence, it is prudent to develop bedside tools that can be used as screening tools to identify patients at high risk of mortality and complications so that timely referrals and judicious utilization of resources can be made. Literature in this regard is limited and published studies have shown that clinical features fail to predict the severity of injury. We aimed our study to find the role of Drooling, Reluctance, Oropharynx, Others, and Leukocytosis (DROOL) score as a predictor of mortality and complications following acute corrosive ingestion.METHODS:This was a diagnostic accuracy study conducted in the emergency department (ED) of a tertiary care hospital in North India. We screened all cases of acute corrosive ingestion presented to our ED. We collected the data on demographic profile, clinical features, investigations, endoscopy findings, treatment, and DROOL score. We followed patients for up to 12 weeks for outcomes including mortality and complications.RESULTS:We studied 79 patients of acute corrosive ingestion. The median age was 26 years with a female predominance. Nausea, vomiting, and pain abdomen were the common symptoms. The median DROOL score was 4. The majority of our patients had normal to Zargar grade 1 injury to the stomach and esophagus. Out of 79 patients, 27 patients developed some complications. The overall mortality up to 12 weeks was 10%. The receiver operating characteristics (ROC) analysis was performed, and the area under the ROC (AUROC) curve of Zargar classification in predicting overall complications was 0.909 (96% confidence interval [CI]: 0.842–0.975) and it was 0.775 (95% CI: 0.553–1.000) in predicting mortality. The AUROC of DROOL score in predicting overall complications was 0.932 (95% CI: 0.877–0.987) and the AUROC of DROOL score in predicting mortality was 0.864 (95% CI: 0.758–0.970). The ROC analysis showed that a DROOL score ≤4 has a sensitivity of 96.2% and a specificity of 77.8% in predicting overall complications. Similarly, DROOL score ≤5 has a sensitivity of 81.7% and a specificity of 62.5% in predicting the development of mortality. Delong test showed that there was no statistically significant difference in Zargar versus DROOL score in terms of prediction of mortality and overall complications (P > 0.05).CONCLUSION:DROOL score is comparable to Zargar score in identifying patients at high risk of mortality and complications. Hence, DROOL score can be used for risk stratification of patients presenting with corrosive ingestion.

Quantitative computed tomography (CT) analysis is an important method for diagnosis and severity evaluation of lung diseases. However, the association between CT-derived biomarkers and chronic obstructive pulmonary disease (COPD) exacerbations remains unclear. We aimed to investigate its potential in predicting COPD exacerbations. Patients with COPD were consecutively enrolled, and their data were analyzed in this retrospective study. Body composition and thoracic abnormalities were analyzed from chest CT scans. Logistic regression analysis was performed to identify independent risk factors of exacerbation. Based on 2-year follow-up data, the deep learning system (DLS) was developed to predict future exacerbations. Receiver operating characteristic (ROC) curve analysis was conducted to assess the diagnostic performance. Finally, the survival analysis was performed to further evaluate the potential of the DLS in risk stratification. A total of 1,150 eligible patients were included and followed up for 2 years. Multivariate analysis revealed that CT-derived high affected lung volume/total lung capacity (ALV/TLC) ratio, high visceral adipose tissue area (VAT), and low pectoralis muscle cross-sectional area (CSA) were independent risk factors causing COPD exacerbations. The DLS outperformed exacerbation history and the BMI, airflow obstruction, dyspnea, and exercise capacity (BODE) index, with an area under the ROC (AUC) value of 0.88 (95%CI, 0.82-0.92) in the internal cohort and 0.86 (95%CI, 0.81-0.89) in the external cohort. The DeLong test revealed significance between this system and conventional scores in the test cohorts (p < 0.05). In the survival analysis, patients with higher risk were susceptible to exacerbation events. The DLS could allow accurate prediction of COPD exacerbations. The newly identified CT biomarkers (ALV/TLC ratio, VAT, and pectoralis muscle CSA) could potentially enable investigation into underlying mechanisms responsible for exacerbations.

This study aimed to validate a deep learning model's diagnostic performance in using computed tomography (CT) to diagnose cervical lymph node metastasis (LNM) from thyroid cancer in a large clinical cohort and to evaluate the model's clinical utility for resident training. The performance of eight deep learning models was validated using 3838 axial CT images from 698 consecutive patients with thyroid cancer who underwent preoperative CT imaging between January and August 2018 (3606 and 232 images from benign and malignant lymph nodes, respectively). Six trainees viewed the same patient images (n = 242), and their diagnostic performance and confidence level (5-point scale) were assessed before and after computer-aided diagnosis (CAD) was included. The overall area under the receiver operating characteristics (AUROC) of the eight deep learning algorithms was 0.846 (range 0.784-0.884). The best performing model was Xception, with an AUROC of 0.884. The diagnostic accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of Xception were 82.8%, 80.2%, 83.0%, 83.0%, and 80.2%, respectively. After introducing the CAD system, underperforming trainees received more help from artificial intelligence than the higher performing trainees (p = 0.046), and overall confidence levels significantly increased from 3.90 to 4.30 (p < 0.001). The deep learning-based CAD system used in this study for CT diagnosis of cervical LNM from thyroid cancer was clinically validated with an AUROC of 0.884. This approach may serve as a training tool to help resident physicians to gain confidence in diagnosis. • A deep learning-based CAD system for CT diagnosis of cervical LNM from thyroid cancer was validated using data from a clinical cohort. The AUROC for the eight tested algorithms ranged from 0.784 to 0.884. • Of the eight models, the Xception algorithm was the best performing model for the external validation dataset with 0.884 AUROC. The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were 82.8%, 80.2%, 83.0%, 83.0%, and 80.2%, respectively. • The CAD system exhibited potential to improve diagnostic specificity and accuracy in underperforming trainees (3 of 6 trainees, 50.0%). This approach may have clinical utility as a training tool to help trainees to gain confidence in diagnoses.

Sepsis is a systemic response to infection with a high rate of mortality andcomplex pathophysiology involving inflammation, infection response, hemostasis,endothelium, and platelets. The purpose of this study was to develop an equationincorporating biomarker levels at intensive care unit (ICU) admission to predictmortality in patients with sepsis, based on the hypothesis that a combination ofbiomarkers representative of multiple physiological systems would provideimproved predictive value. Plasma samples and clinical data were collected from103 adult patients with sepsis at the time of ICU admission. Biomarker levelswere measured using commercially available methods. A 28-day mortality was usedas the primary end point. Stepwise linear regression modeling was performed togenerate a predictive equation for mortality. Differences in biomarker levelsbetween survivors were quantified using the Mann-Whitney test and the area underthe receiver operating curve (AUC) was used to describe predictive ability.Significant differences (P < .05) were observed betweensurvivors and nonsurvivors for plasminogen activator inhibitor 1 (AUC = 0.70),procalcitonin (AUC = 0.77), high mobility group box 1 (AUC = 0.67), interleukin(IL) 6 (AUC = 0.70), IL-8 (AUC = 0.70), protein C (AUC = 0.71), angiopoietin-2(AUC = 0.76), endocan (AUC = 0.58), and platelet factor 4 (AUC = 0.70). Apredictive equation for mortality was generated using stepwise linear regressionmodeling, which incorporated procalcitonin, vascular endothelial growth factor,the IL-6:IL-10 ratio, endocan, and platelet factor 4, and demonstrated a betterpredictive value for patient outcome than any individual biomarker (AUC = 0.87).The use of mathematical modeling resulted in the development of a predictiveequation for sepsis-associated mortality with performance than any individualbiomarker or clinical scoring system which incorporated biomarkersrepresentative of multiple systems.

Hematological abnormalities after severe traumatic brain injury (TBI) are common, and are associated with a poor outcome. Whether these abnormalities offer additional prognostic significance over and beyond validated TBI prognostic models is uncertain. This retrospective cohort study compared the ability of admission hematological abnormalities to that of the IMPACT (International Mission for Prognosis and Analysis of Clinical Trials) prognostic model to predict 18-month neurological outcome of 388 patients who required a decompressive craniectomy after severe TBI, between 2004 and 2016, in Western Australia. Area under the receiver operating characteristic (AUROC) curve was used to assess predictors' ability to discriminate between patients with and without an unfavorable outcome of death, vegetative state, or severe disability. Of the 388 patients included in the study, 151 (38.9%) had an unfavorable outcome at 18 months after decompressive craniectomy for severe TBI. Abnormalities in admission hemoglobin (AUROC 0.594, p = 0.002), plasma glucose (AUROC 0.592, p = 0.002), fibrinogen (AUROC 0.563, p = 0.036), international normalized ratio (INR; AUROC 0.645, p = 0.001), activated partial thromboplastin time (AUROC 0.564, p = 0.033), and disseminated intravascular coagulation score (AUROC 0.623, p = 0.001) were all associated with a higher risk of unfavorable outcome at 18 months after severe TBI. As a marker of inflammation, neutrophil to lymphocyte ratio was not significantly associated with the risk of unfavorable outcome (AUROC 0.500, p = 0.998). However, none of these parameters, in addition to the platelet count, were significantly associated with an unfavorable outcome after adjusting for the IMPACT predicted risk (odds ratio [OR] per 10% increment in risk 2.473, 95% confidence interval [CI] 2.061-2.967; p = 0.001). After excluding 8 patients (2.1%) who were treated with warfarin prior to the injury, there was a suggestion that INR was associated with some additional prognostic significance (OR 3.183, 95% CI 0.856-11.833; p = 0.084) after adjusting for the IMPACT predicted risk. In isolation, INR was the best hematological prognostic parameter in severe TBI requiring decompressive craniectomy, especially when patients treated with warfarin were excluded. However, the prognostic significance of admission hematological abnormalities was mostly captured by the IMPACT prognostic model, such that they did not offer any additional prognostic information beyond the IMPACT predicted risk. These results suggest that new prognostic factors for TBI should be evaluated in conjunction with predicted risks of a comprehensive prognostic model that has been validated, such as the IMPACT prognostic model.

Background: There is a high incidence of acute respiratory failure (ARF) in moderate or severe traumatic brain injury (M-STBI), worsening outcomes. This study aimed to design a predictive model for ARF.Methods: Adult patients with M-STBI [3 ≤ Glasgow Coma Scale (GCS) ≤ 12] with a definite history of brain trauma and abnormal head on CT images, obtained from September 2015 to May 2017, were included. Patients with age >80 years or <18 years, multiple injuries with TBI upon admission, or pregnancy (in women) were excluded. Two models based on machine learning extreme gradient boosting (XGBoost) or logistic regression, respectively, were developed for predicting ARF within 48 h upon admission. These models were evaluated by out-of-sample validation. The samples were assigned to the training and test sets at a ratio of 3:1.Results: In total, 312 patients were analyzed including 132 (42.3%) patients who had ARF. The GCS and the Marshall CT score, procalcitonin (PCT), and C-reactive protein (CRP) on admission significantly predicted ARF. The novel machine learning XGBoost model was superior to logistic regression model in predicting ARF [area under the receiver operating characteristic (AUROC) = 0.903, 95% CI, 0.834–0.966 vs. AUROC = 0.798, 95% CI, 0.697–0.899; p < 0.05].Conclusion: The XGBoost model could better predict ARF in comparison with logistic regression-based model. Therefore, machine learning methods could help to develop and validate novel predictive models.

Abdominal Aortic Aneurysm Anatomic Severity Grading Score: Identifying Anatomic Attributes That Best Predict Endovascular Aneurysm Repair Outcomes

Patients suffering from painful spinal bone metastases (PSBMs) often undergo palliative radiation therapy (RT), with an efficacy of approximately two thirds of patients. In this exploratory investigation, we assessed the effectiveness of machine learning (ML) models trained on radiomics, semantic and clinical features to estimate complete pain response. Gross tumour volumes (GTV) and clinical target volumes (CTV) of 261 PSBMs were segmented on planning computed tomography (CT) scans. Radiomics, semantic and clinical features were collected for all patients. Random forest (RFC) and support vector machine (SVM) classifiers were compared using repeated nested cross-validation. The best radiomics classifier was trained on CTV with an area under the receiver-operator curve (AUROC) of 0.62 ± 0.01 (RFC; 95% confidence interval). The semantic model achieved a comparable AUROC of 0.63 ± 0.01 (RFC), significantly below the clinical model (SVM, AUROC: 0.80 ± 0.01); and slightly lower than the spinal instability neoplastic score (SINS; LR, AUROC: 0.65 ± 0.01). A combined model did not improve performance (AUROC: 0,74 ± 0,01). We could demonstrate that radiomics and semantic analyses of planning CTs allowed for limited prediction of therapy response to palliative RT. ML predictions based on established clinical parameters achieved the best results.

Background: Skeletal muscle gauge (SMG) was recently introduced as an imaging indicator of sarcopenia for the prediction of clinical outcomes, including chemotherapy toxicity and prognosis, in patients with cancer. Computed tomography (CT) is essential for measuring SMG; thus, the use of SMG is limited to patients who undergo CT. Objective: We aimed to develop a machine learning algorithm using clinical and inflammatory markers to predict SMG in patients with colorectal cancer (CRC) Methods: The least absolute shrinkage and selection operator (LASSO) regression model was applied for variable selection and predictive signature building in the training set. The predictive accuracy of the LASSO model, defined as LP-SMG was compared using the area under the receiver operating characteristic (AUROC) and decision curve analysis (DCA) in the test set. Results: A total of 1,094 patients with CRC were enrolled and randomly categorized into training (n=656) and test (n=438) sets. Low SMG was identified in 142 (21.6%) and 90 (20.5%) patients in the training and test sets, respectively. According to multivariable analysis of the test sets, LP-SMG was identified as an independent predictor of low SMG (OR: 1329.431, CI: 271.684-7667.996, p&amp;lt;.001). Its predictive performance was similar in the training and test sets (AUROC: 0.846 vs. 0.869, p=.427). In the test set, LP-SMG showed better outcomes in predicting SMG than single clinical variables, such as sex, height, weight, and hemoglobin, as measured by AUROC and DCA. Conclusions: LP-SMG, incorporating clinical variables and serum inflammatory indicators, showed superior performance compared to single variables in predicting low SMG. This machine learning model can be used as a screening tool to detect sarcopenic status without using CT during the treatment period. Applying a machine learning model might be beneficial in reducing the effort, cost, and radiation exposure from conventional CT-based diagnosis. Citation Format: Jeonghyun Kang. Skeletal muscle gauge prediction by a machine learning model in patients with colorectal cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5428.

Background: Cerebral palsy (CP) causes developmental delays, affecting quality of life. Many risk factors are theorized however, no total risks summary exists, nor a CP prediction score for newborns. The result is under surveillance, treatment delays, and non-rectifiable complications. Objective: To establish total risk factors and create a prediction score for assessing CP neonatal risk before discharge. A prediction score has great utility for medical professionals and parents in screening high-risk patients and developing adequate monitoring systems. Materials and Methods: Using a case-controlled retrospect of children aged 0 to 2 years, born at Thammasat University Hospital, Thailand between 2005 and 2014, prenatal, perinatal, and postnatal risks were compared between children without CP as control, and those diagnosed with CP as case, by multivariable logistic regression. Predictors were assessed with area under the receiver operating characteristic (AuROC), odds ratio (OR), 95% confidence interval (CI), p-value, and clinical predisposition. Logistic regression was applied, including calibration, validation, and categorization of risk. Results: Cerebral and non-cerebral malformations, multi-fetal gestation, low birthweight, and neonatal sepsis were found as potential predictors, scoring 3, 1.5, 1, 2, 2.5, respectively, AuROC being 0.86 (95% CI 0.79 to 0.92). Low, moderate, and high-risk groups were set with scores of less than 1, 1.5 to 3, and more than 3.5, respectively. Conclusion: The present predictive CP risks and scoring system shows excellent discrimination power. If newborns were categorized in the highrisk group, close monitoring and surveillance are needed. Keywords: Cerebral palsy, Risk score, Prenatal, Perinatal, Postnatal

Corrosive substance ingestion remains a significant public health issue in children, often associated with insufficient preventive practices. This study aimed to evaluate the clinical presentation, endoscopic findings, clinical course, and treatment outcomes of pediatric patients admitted to our hospital after corrosive ingestion. A total of 311 children who presented to our center with caustic ingestion between January 2007 and December 2024 were included in the study. Clinical, demographic, endoscopic, and treatment outcomes were retrospectively reviewed using medical records. The mean age of the patients was 4.4 ± 4.5years (range, 6months-17years), and 55.9% were male. Accidental ingestion accounted for 92.2% of cases, whereas eight adolescents ingested the substance with suicidal intent. Household cleaning products were the most frequently ingested substances (77.0%), mainly bleach (29.9%) and nitric‑acid‑based descalers (12.2%). The most severe esophagogastric injuries were associated with sodium‑hydroxide drain openers, fat removers, and nitric‑acid descalers. Endoscopic evaluation was performed in 221 children (71.1%), and repeat endoscopy was required in 35 (11.3%). Fifteen children (4.8%) who developed esophageal strictures underwent endoscopic balloon dilatation, and two required surgical gastrostomy with esophageal repair. No mortality occurred, but five patients developed major complications, including pyloric stenosis, esophageal perforation with pneumomediastinum, and brain abscess. Accidental ingestion of household cleaning substances due to unawareness or negligence can lead to severe morbidities in children. Raising parental awareness is crucial to prevent such incidents. •Corrosive substance ingestion in children is a preventable public health problem, most commonly caused by accidental exposure to household cleaning products. •The severity of injury depends on factors such as pH, concentration, physical form of the substance, and time to presentation, but symptoms do not always correlate with endoscopic findings. •In our cohort of 311 children, acid ingestion-though less frequent-was associated with higher rates of symptoms, endoscopic evaluation, and repeat endoscopy compared with alkali ingestion. •Rare but serious complications, such as brain abscess after esophageal perforation, can occur despite the absence of mortality, underlining the need for improved prevention strategies and stricter regulations.

Utility of B-Type Natriuretic Peptide and N-terminal Pro B-Type Natriuretic Peptide in Evaluation of Respiratory Failure in Critically Ill Patients