Development and validation of novel simple prognostic model for predicting mortality in Korean intensive care units using national insurance claims data.

Background/Aims: Intensive care unit (ICU) quality is largely determined by the mortality rate. Therefore, we aimed to develop and validate a novel prognostic model for predicting mortality in Korean ICUs, using national insurance claims data. Methods: Data were obtained from the health insurance claims database maintained by the Health Insurance Review and Assessment Service of South Korea. From patients who underwent the third ICU adequacy evaluation, 42,489 cases were enrolled and randomly divided into the derivation and validation cohorts. Using the models derived from the derivation cohort, we analyzed whether they accurately predicted death in the validation cohort. The models were verified using data from one general and two tertiary hospitals. Results: Two severity correction models were created from the derivation cohort data, by applying variables selected through statistical analysis, through clinical consensus, and from performing multiple logistic regression analysis. Model 1 included six categorical variables (age, sex, Charlson comorbidity index, ventilator use, hemodialysis or continuous renal replacement therapy, and vasopressor use). Model 2 additionally included presence/absence of ICU specialists and nursing grades. In external validation, the performance of models 1 and 2 for predicting in-hospital and ICU mortality was not inferior to that of pre-existing scoring systems. Conclusions: The novel and simple models could predict in-hospital and ICU mortality and were not inferior compared to the pre-existing scoring systems.

Differences of respiratory mechanics in mechanical ventilation of acute respiratory distress syndrome between patients with COVID-19 and Influenza A.

BACKGROUND: Whether COVID-19-induced acute respiratory distress syndrome (ARDS) should be approached differently in terms of mechanical ventilation therapy compared to other virus-induced ARDS is debatable. Therefore, we aimed to ascertain whether the respiratory mechanical characteristics of COVID-19-induced ARDS differ from those of influenza A induced ARDS, in order to establish a rationale for mechanical ventilation therapy in COVID-19-induced ARDS. METHODS: This was a retrospective cohort study comparing patients with COVID-19-induced ARDS and influenza A induced ARDS. We included intensive care unit (ICU) patients with COVID-19 or Influenza A aged ≥ 19, who were diagnosed with ARDS according to the Berlin definition between January 2015 and July 2021. Ventilation parameters for respiratory mechanics were collected at specific times on days one, three, and seven after intubation. RESULTS: The median age of the 87 participants was 71.0 (62.0-78.0) years old, and 63.2% were male. The ratio of partial pressure of oxygen in arterial blood to the fractional of inspiratory oxygen concentration in COVID-19-induced ARDS was lower than that in influenza A induced ARDS during the initial stages of mechanical ventilation (influenza A induced ARDS 216.1 vs. COVID-19-induced ARDS 167.9, p = 0.009, day 1). The positive end expiratory pressure remained consistently higher in the COVID-19 group throughout the follow-up period (7.0 vs. 10.0, p < 0.001, day 1). COVID-19 and influenza A initially showed different directions for peak inspiratory pressure and dynamic compliance; however, after day 3, both groups exhibited similar directions. Dynamic driving pressure exhibited opposite trends between the two groups during mechanical ventilation. CONCLUSIONS: Respiratory mechanics show clear differences between COVID-19-induced ARDS and influenza A induced ARDS. Based on these findings, we can consider future treatment strategies for COVID-19-induced ARDS.

Risk Factors for Unfavorable Outcomes of Tuberculosis in Korea: Implications for Patient-Centered Care in Tuberculosis Treatment.

BACKGROUND: The treatment success rate for tuberculosis (TB) has stagnated at 80-81% in South Korea, indicating unsatisfactory outcomes. Enhancing treatment success rate necessitates the development of individualized treatment approaches for each patient. This study aimed to identify the risk factors associated with unfavorable treatment outcomes to facilitate tailored TB care. METHODS: We retrospectively analyzed the data of patients with active TB between January 2019 and December 2020 at a single tertiary referral center. We classified unfavorable treatment outcomes according to the 2021 World Health Organization guidelines as follows: "lost to follow-up" (LTFU), "not evaluated" (NE), "death," and "treatment failure" (TF). Moreover, we analyzed risk factors for each unfavorable outcome using Cox proportional hazard regression analysis. RESULTS: A total of 659 patients (median age 62 years; male 54.3%) were included in the study. The total unfavorable outcomes were 28.1%: 4.6% LTFU, 9.6% NE, 9.1% deaths, and 4.9% TF. Multivariate analysis showed that a culture-confirmed diagnosis of TB was associated with a lower risk of LTFU (adjusted hazard ratio [aHR], 0.25; 95% confidence interval [CI], 0.10-0.63), whereas the occurrence of adverse drug reactions (ADRs) significantly increased the risk of LTFU (aHR, 6.63; 95% CI, 2.63-16.69). Patients living far from the hospital (aHR, 4.47; 95% CI, 2.50-7.97) and those with chronic kidney disease (aHR, 3.21; 95% CI, 1.33-7.75) were at higher risk of being transferred out to other health institutions (NE). Higher mortality was associated with older age (aHR, 1.06; 95% CI, 1.04-1.09) and comorbidities. The ADRs that occurred during TB treatment were a risk factor for TF (aHR, 6.88; 95% CI, 2.24-21.13). CONCLUSION: Unfavorable outcomes of patients with TB were substantial at a tertiary referral center, and the risk factors for each unfavorable outcome varied. To improve treatment outcomes, close monitoring and the provision of tailored care for patients with TB are necessary.

Impact of age on mortality and transfer to long-term care in patients in an intensive care unit.

BACKGROUND: In the global trend of population aging, age is one of the significant factors to be considered in critically ill patients. However, the impact of age on clinical outcomes and long-term prognosis in this population varies across different studies. METHODS: We conducted a retrospective cohort analysis for patients admitted to the medical intensive care unit (ICU) (30 beds) between January 2017 and December 2020 at the tertiary referral hospital in Korea. Patients were classified into three groups according to age: <65 years, old age (65-79 years), and very old age (≥ 80 years). Subsequently, enrolled patients were analyzed for acute mortality and long-term prognosis. RESULTS: Among the 1584 patients, the median age was 67.0 (57.0-76.0) years, and 65.2% were male. Median ICU length of stay (LOS) (8, 9, and 10 days in < 65, 65-79, and ≥ 80 years, respectively; p = 0.006) and the proportion of patients who were transferred to long-term care hospital at the time of discharge (12.9% vs. 28.3% vs. 39.4%, respectively; p < 0.001) increased with age. Multivariable logistic analysis showed no significant difference in the 28-day mortality in the old age (adjusted odds ratio [aOR] 0.88; 95% confidence interval [CI] 0.65-1.17) and very old age (aOR 1.05; 95% CI 0.71-1.55) groups compared to that in patients with age < 65 years. However, the relevance of the proportion of ICU LOS ≥ 7 days and transfers to other hospitals after discharge increased with age. CONCLUSIONS: Age did not affect acute mortality in critical illness patients. However, surviving older age groups required more long-term care facilities compared to patients younger than 65 years after acute management. These results indicate that in an aging society, the importance of not only acute management but also long-term care facilities may increase for critical illness patients.

Three-dimensional label-free morphology of CD8 + T cells as a sepsis biomarker.

Sepsis is a dysregulated immune response to infection that leads to organ dysfunction and is associated with a high incidence and mortality rate. The lack of reliable biomarkers for diagnosing and prognosis of sepsis is a major challenge in its management. We aimed to investigate the potential of three-dimensional label-free CD8 + T cell morphology as a biomarker for sepsis. This study included three-time points in the sepsis recovery cohort (N = 8) and healthy controls (N = 20). Morphological features and spatial distribution within cells were compared among the patients' statuses. We developed a deep learning model to predict the diagnosis and prognosis of sepsis using the internal cell morphology. Correlation between the morphological features and clinical indices were analysed. Cell morphological features and spatial distribution differed significantly between patients with sepsis and healthy controls and between the survival and non-survival groups. The model for predicting the diagnosis and prognosis of sepsis showed an area under the receiver operating characteristic curve of nearly 100% with only a few cells, and a strong correlation between the morphological features and clinical indices was observed. Our study highlights the potential of three-dimensional label-free CD8 + T cell morphology as a promising biomarker for sepsis. This approach is rapid, requires a minimum amount of blood samples, and has the potential to provide valuable information for the early diagnosis and prognosis of sepsis.

Biomarker-Based Assessment Model for Detecting Sepsis: A Retrospective Cohort Study.

The concept of the quick sequential organ failure assessment (qSOFA) simplifies sepsis detection, and the next SOFA should be analyzed subsequently to diagnose sepsis. However, it does not include the concept of suspected infection. Thus, we simply developed a biomarker-based assessment model for detecting sepsis (BADS). We retrospectively reviewed the electronic health records of patients admitted to the intensive care unit (ICU) of a 2000-bed university tertiary referral hospital in South Korea. A total of 989 patients were enrolled, with 77.4% (n = 765) of them having sepsis. The patients were divided into a ratio of 8:2 and assigned to a training and a validation set. We used logistic regression analysis and the Hosmer-Lemeshow test to derive the BADS and assess the model. BADS was developed by analyzing the variables and then assigning weights to the selected variables: mean arterial pressure, shock index, lactate, and procalcitonin. The area under the curve was 0.754, 0.615, 0.763, and 0.668 for BADS, qSOFA, SOFA, and acute physiology and chronic health evaluation (APACHE) II, respectively, showing that BADS is not inferior in sepsis prediction compared with SOFA. BADS could be a simple scoring method to detect sepsis in critically ill patients quickly at the bedside.

Protective effect of methotrexate on lung function and mortality in rheumatoid arthritis-related interstitial lung disease: a retrospective cohort study.

BACKGROUND: Studies on the risk and protective factors for lung function decline and mortality in rheumatoid arthritis-related interstitial lung disease (RA-ILD) are limited. OBJECTIVES: We aimed to investigate clinical factors and medication uses associated with lung function decline and mortality in RA-ILD. METHODS: This retrospective cohort study examined the medical records of patients with RA-ILD who visited Severance Hospital between January 2006 and December 2019. We selected 170 patients with RA-ILD who had undergone at least one spirometry test and chest computed tomography scan. An absolute decline of ⩾10% in the functional vital capacity (FVC) was defined as significant decline in pulmonary function. Data for analysis were retrieved from electronic medical records. RESULTS: Ninety patients (52.9%) were female; the mean age was 64.0 ± 10.2 years. Multivariate logistic regression showed that a high erythrocyte sediment rate level at baseline [odds ratio (OR) = 3.056; 95% confidence interval (CI) = 1.183-7.890] and methotrexate (MTX) use (OR = 0.269; 95% CI = 0.094-0.769) were risk and protective factors for lung function decline, respectively. Multivariate Cox regression analysis indicated that age ⩾65 years (OR = 2.723; 95% CI = 1.142-6.491), radiologic pattern of usual interstitial pneumonia (UIP) or probable UIP (OR = 3.948; 95% CI = 1.522-10.242), baseline functional vital capacity (FVC) % predicted (OR = 0.971; 95% CI = 0.948-0.994), and MTX use (OR = 0.284; 95% CI = 0.091-0.880) were predictive of mortality. CONCLUSION: We identified risk and protective factors for lung function decline and mortality in patients with RA-ILD. MTX use was associated with favorable outcome in terms of both lung function and mortality in our cohort.

Complications related to extracorporeal membrane oxygenation support as a bridge to lung transplantation and their clinical significance.

BACKGROUND: Incidence of complications related extracorporeal membrane oxygenation (ECMO) support as a bridge to lung transplantation (BTT) and its association with the patient outcome in lung transplantation (LT) has not been well documented in previous studies. OBJECTIVES: We evaluated the incidence of complications related to the use of ECMO support as a BTT, and the association between the occurrence of the complications and patient outcomes in LTs. METHODS: This retrospective cohort study investigated 100 consecutive patients who started ECMO support as a BTT between April 2013 and March 2020. Data for the analyses were retrieved from electronic medical records. RESULTS: Fifty-six percent of the patients experienced at least one complication during the BTT with ECMO. Major bleeding was the most common complication. In multivariate logistic regression analysis, occurrence of oxygenator thromboses (OR 16.438, P = 0.008) and the use of renal replacement therapy (RRT) (OR 32.288, P < 0.001) were associated with a failed BTT. In the subgroup analysis of the LT recipients, intracranial hemorrhages, (OR 13.825, P = 0.021), RRT use, (OR 11.395, P = 0.038), and bloodstream infection occurrence (OR 6.210; P = 0.034) were identified as risk factors for in-hospital mortality. CONCLUSIONS: The occurrence of complications during the use of ECMO support as a BTT was associated with unfavorable outcomes in LTs. Close monitoring and the proper management of these complications may be important to achieve better outcomes in patients using ECMO support as a BTT.

Impairment of Mitochondrial ATP Synthesis Induces RIPK3-dependent Necroptosis in Lung Epithelial Cells During Lung Injury by Lung Inflammation.

Dysfunction of mitochondrial metabolism is implicated in cellular injury and cell death. While mitochondrial dysfunction is associated with lung injury by lung inflammation, the mechanism by which the impairment of mitochondrial ATP synthesis regulates necroptosis during acute lung injury (ALI) by lung inflammation is unclear. Here, we showed that the impairment of mitochondrial ATP synthesis induces receptor interacting serine/threonine kinase 3 (RIPK3)-dependent necroptosis during lung injury by lung inflammation. We found that the impairment of mitochondrial ATP synthesis by oligomycin, an inhibitor of ATP synthase, resulted in increased lung injury and RIPK3 levels in lung tissues during lung inflammation by LPS in mice. The elevated RIPK3 and RIPK3 phosphorylation levels by oligomycin resulted in high mixed lineage kinase domain-like (MLKL) phosphorylation, the terminal molecule in necroptotic cell death pathway, in lung epithelial cells during lung inflammation. Moreover, the levels of protein in bronchoalveolar lavage fluid (BALF) were increased by the activation of necroptosis via oligomycin during lung inflammation. Furthermore, the levels of ATP5A, a catalytic subunit of the mitochondrial ATP synthase complex for ATP synthesis, were reduced in lung epithelial cells of lung tissues from patients with acute respiratory distress syndrome (ARDS), the most severe form of ALI. The levels of RIPK3, RIPK3 phosphorylation and MLKL phosphorylation were elevated in lung epithelial cells in patients with ARDS. Our results suggest that the impairment of mitochondrial ATP synthesis induces RIPK3-dependent necroptosis in lung epithelial cells during lung injury by lung inflammation.

Long-term clinical course and progression of lymphangioleiomyomatosis in a single lung transplant referral centre in Korea.

We aimed to describe the clinical features of lymphangioleiomyomatosis (LAM) in Korean patients and identify factors associated with progressive disease (PD). Clinical features of 54 patients with definite or probable LAM from 2005 to 2018 were retrospectively analysed. Common features were pneumothorax (66.7%) and abdominal lymphadenopathy (50.0%). Twenty-three (42.6%) patients were initially treated with mechanistic target of rapamycin (mTOR) inhibitors. Lung transplantation (LT) was performed in 13 (24.1%) patients. Grouped based on the annual decline in forced expiratory volume in 1 s (FEV1) from baseline and LT, 36 (66.7%) patients exhibited stable disease (SD). All six deaths (11.1%) occurred in PD. Proportion of SD was higher in those treated initially with mTOR inhibitors than in those under observation (p = 0.043). Univariate analysis revealed sirolimus use, and baseline forced vital capacity, FEV1, and diffusing capacity of the lungs for carbon monoxide are associated with PD. Multivariate analysis showed that only sirolimus use (odds ratio 0.141, 95% confidence interval 0.021-0.949, p = 0.044) reduced PD. Kaplan-Meier analysis estimates overall survival of 92.0% and 74.7% at 5 and 10 years, respectively. A considerable proportion of LAM patients remain clinically stable without treatment. LT is an increasingly viable option for patients with severe lung function decline.

NADPH oxidase 4 signaling in a ventilator-induced lung injury mouse model.

BACKGROUND: For patients with acute respiratory distress syndrome, a ventilator is essential to supply oxygen to tissues, but it may also cause lung damage. In this study, we investigated the role of NOX4 using NOX4 knockout (KO) mice and NOX4 inhibitors in a ventilator-induced lung injury (VILI) model. METHODS: Wild-type (WT) male C57BL/6J mice and NOX4 knockout (KO) male mice were divided into five groups: (1) control group; (2) high tidal ventilation (HTV) group: WT mice + HTV ± DMSO; (3) NOX4 KO group; (4) NOX4 KO with HTV group; (5) NOX4 inhibitor group: WT mice + HTV + NOX4 inhibitor. In the VILI model, the supine position was maintained at 24 mL/kg volume, 0 cm H2O PEEP, 100/min respiratory rate, and 0.21 inspired oxygen fraction. In the NOX4 inhibitor group, 50 µL anti-GKT 137831 inhibitor was injected intraperitoneally, 2 h after ventilator use. After 5 h of HTV, mice in the ventilator group were euthanized, and their lung tissues were obtained for further analysis. In addition, the relationship between EphA2 (which is related to lung injury) and NOX4 was investigated using EphA2 KO mice, and NOX4 and EphA2 levels in the bronchoalveolar lavage fluid (BALF) of 38 patients with pneumonia were examined. RESULTS: Cell counts from BALFs were significantly lower in the NOX4 KO with HTV group (p < 0.01) and EphA2 KO with HTV group (p < 0.001) compared to that in the HTV group. In the NOX4 inhibitor group, cell counts and protein concentrations from BALF were significantly lower than those in the HTV group (both, p < 0.001). In the NOX4 KO group and the NOX4 inhibitor group, EphA2 levels were significantly lower than those in the HTV group (p < 0.001). In patients with respiratory disease, NOX4 and EphA2 levels were significantly higher in patients with pneumonia and patients who received ventilator treatment in the intensive care unit. CONCLUSION: In the VILI model with high tidal volume, NOX4 KO, EphA2 KO or monoclonal antibodies attenuated the VILI. NOX4 and EphA2 levels were significantly higher in patients with pneumonia and especially in mechanical ventilated in the ICU. Inhibition of Nox4 is a potential therapeutic target for the prevention and reduction of VILI.

Reference values of skeletal muscle area for diagnosis of sarcopenia using chest computed tomography in Asian general population.

BACKGROUND: Diagnostic cutoff points for sarcopenia in chest computed tomography (CT) have not been established although CT is widely used for investigating skeletal muscles. This study aimed to determine reference values for sarcopenia of thoracic skeletal muscles acquired from chest CT scans and to analyse variables related to sarcopenia using the cutoff values determined in a general Asian population. METHODS: We retrospectively reviewed chest CT scans of 4470 participants (mean age 54.8 ± 9.9 years, 65.8% male) performed at a check-up centre in South Korea (January 2016-August 2017). To determine cutoffs, 335 participants aged 19-39 years (mean age 35.2 ± 3.6 years, 75.2% male) were selected as the healthy and younger reference group, and 4135 participants aged ≥40 years (mean age 56.4 ± 8.4 years, 65.1% male) were selected as the study group. We measured the following: cross-sectional area (CSA) of the pectoralis, intercostalis, paraspinal, serratus, and latissimus muscles at the 4th vertebral region (T4CSA ); T4CSA divided by height2 (T4MI); pectoralis muscle area (PMCSA ); and PMCSA divided by height2 (PMI) at the 4th vertebral region. Sarcopenia cutoff was defined as sex-specific values of less than -2 SD below the mean from the reference group. RESULTS: In the reference group, T4CSA , T4MI, PMCSA , and PMI cutoffs for sarcopenia were 100.06cm2 , 33.69cm2 /m2 , 29.00cm2 , and 10.17cm2 /m2 in male, and 66.93cm2 , 26.01cm2 /m2 , 18.29cm2 , and 7.31cm2 /m2 in female, respectively. The prevalence of sarcopenia in the study group measured with T4CSA , T4MI, PMCSA and PMI cutoffs were 11.4%, 8.7%, 8.5%, and 10.1%, respectively. Correlations were observed between appendicular skeletal mass divided by height2 measured by bioelectrical impedance analysis (BIA) and T4CSA (r = 0.82; P < 0.001)/T4MI (r = 0.68; P < 0.001), and ASM/height2 measured by BIA and PMCSA (r = 0.72; P < 0.001)/PMI (r = 0.63; P < 0.001). In the multivariate logistic regression models, sarcopenia defined by T4CSA /T4MI were related to age [odds ratio (95% confidence interval), P-values: 1.09 (1.07-1.11), <0.001/1.05 (1.04-1.07), <0.001] and diabetes [1.60 (1.14-2.25), 0.007/1.47 (1.01-2.14), 0.043]. Sarcopenia defined by PMCSA /PMI were related to age [1.09 (1.08-1.10), <0.001/1.05 (1.03-1.06), <0.001], male sex [0.23 (0.18-0.30), <0.001/0.47 (0.32-0.71), <0.001], diabetes [2.30 (1.73-3.05), <0.001/1.63 (1.15-2.32), 0.007], history of cancer [2.51 (1.78-3.55), <0.001/1.61 (1.04-2.48), 0.033], and sufficient physical activity [0.67 (0.50-0.89), 0.007/0.74 (0.56-0.99), 0.042]. CONCLUSIONS: The reference cutoff values of a general population reported here will enable sex-specific standardization of thoracic muscle mass quantification and sarcopenia assessment.

Thoracic skeletal muscle quantification using computed tomography and prognosis of elderly ICU patients.

In elderly ICU patients, the prevalence of skeletal muscle loss is high. Longitudinal effect of thoracic muscles, especially in elderly ICU patients, are unclear although skeletal muscle loss is related with the short- and long-term outcomes. This study aimed to evaluate whether pectoralis muscle mass loss could be a predictor of prognosis in elderly ICU patients. We retrospectively evaluated 190 elderly (age > 70 years) patients admitted to the ICU. We measured the cross-sectional area (CSA) of the pectoralis muscle (PMCSA) at the fourth vertebral region. CT scans within two days before ICU admission were used for analysis. Mortality, prolonged mechanical ventilation, and longitudinal changes in Sequential Organ Failure Assessment (SOFA) scores were examined. PMCSA below median was significantly related with prolonged ventilation (odds ratio 2.92) and a higher SOFA scores during the ICU stay (estimated mean = 0.94). PMCSA below median was a significant risk for hospital mortality (hazards ratio 2.06). In elderly ICU patients, a low ICU admission PMCSA was associated with prolonged ventilation, higher SOFA score during the ICU stay, and higher mortality. Adding thoracic skeletal muscle CSA at the time of ICU admission into consideration in deciding the therapeutic intensity in elderly ICU patients may help in making medical decisions.

Association of protein consumption and energy intake on sarcopenia in tuberculosis survivors.

BACKGROUND: Tuberculosis (TB) causes undernutrition, and it has a long recovery time after treatment. It is accompanied by adverse health outcomes, such as sarcopenia. OBJECTIVE: We aimed to evaluate the prevalence of sarcopenia and its association with protein and total energy intakes among Korean TB survivors. METHODS: Data of the population-based Korea National Health and Nutrition Examination Survey (2008-2011) were analyzed, including 9,203 participants aged ⩾ 40 years. We used three definitions for sarcopenia-appendicular skeletal muscle mass (ASM, kg) divided by body mass index (BMI, kg/m2), weight (kg), or height squared (m2). Daily protein and total energy intakes were estimated with a 24-h recall method. Multiple logistic regression was used to evaluate the association between dietary protein/total energy intake and sarcopenia among TB survivors. RESULTS: The prevalence of sarcopenia was 11.2%, 10.7%, and 24.3% among TB survivors with sarcopenia defined by ASM divided by BMI, weight, and height squared, respectively. The prevalence of sarcopenia among TB survivors was higher than among those without TB. After adjusting for age, weight, sex, education level, employment status, smoking status, and drinking status, sufficient protein and total energy intakes were associated with a lower risk of sarcopenia in TB survivors. CONCLUSION: The prevalence of sarcopenia was higher in TB survivors than in those without TB. We suggest consuming sufficient protein intake along with increasing total energy intake in TB survivors.

Event Prediction Model Considering Time and Input Error Using Electronic Medical Records in the Intensive Care Unit: Retrospective Study.

BACKGROUND: In the era of artificial intelligence, event prediction models are abundant. However, considering the limitation of the electronic medical record-based model, including the temporally skewed prediction and the record itself, these models could be delayed or could yield errors. OBJECTIVE: In this study, we aim to develop multiple event prediction models in intensive care units to overcome their temporal skewness and evaluate their robustness against delayed and erroneous input. METHODS: A total of 21,738 patients were included in the development cohort. Three events-death, sepsis, and acute kidney injury-were predicted. To overcome the temporal skewness, we developed three models for each event, which predicted the events in advance of three prespecified timepoints. Additionally, to evaluate the robustness against input error and delays, we added simulated errors and delayed input and calculated changes in the area under the receiver operating characteristic curve (AUROC) values. RESULTS: Most of the AUROC and area under the precision-recall curve values of each model were higher than those of the conventional scores, as well as other machine learning models previously used. In the error input experiment, except for our proposed model, an increase in the noise added to the model lowered the resulting AUROC value. However, the delayed input did not show the performance decreased in this experiment. CONCLUSIONS: For a prediction model that was applicable in the real world, we considered not only performance but also temporal skewness, delayed input, and input error.

Soluble IL-2R Levels Predict in-Hospital Mortality in COVID-19 Patients with Respiratory Failure.

Acute respiratory distress syndrome is the primary cause of death in patients with coronavirus disease 2019 (COVID-19) pneumonia. Our study aims to determine the association between serum markers and mortality in COVID-19 patients with respiratory failure. This retrospective study was conducted in a tertiary care hospital in South Korea. Forty-nine patients with COVID-19, who required high flow nasal cannulation or mechanical ventilation from February 2020 to April 2021, were included. Demographic and laboratory data were analyzed at baseline and on Day 7 of admission. We found that serum creatinine, troponin, procalcitonin, and soluble interleukin-2 receptor (sIL-2R) at baseline were more elevated in the non-survivor group, but were not associated with mechanical ventilator use on Day 7. Older age, PaO2/FiO2 ratio, lymphocyte and platelet counts, lactate dehydrogenase, IL-6, C-reactive protein, and sIL-2R on Day 7 were significantly associated with mortality. Delta sIL-2R (Day 7-Day 0) per standard deviation was significantly higher in the non-survivor group (adjusted hazard ratio 3.225, 95% confidence interval (CI) 1.151-9.037, p = 0.026). Therefore, sIL-2R could predict mortality in COVID-19 patients with respiratory failure. Its sustained elevation suggests a hyper-inflammatory state, and mirrors the severity of COVID-19 in patients with respiratory failure, thereby warranting further attention.

Development and validation of a novel sepsis biomarker based on amino acid profiling.

BACKGROUND & AIMS: Sepsis is a potentially fatal condition influenced by pathogens and host factors. Current sepsis biomarkers such as white blood cell count and C-reactive protein and procalcitonin levels show unsatisfactory performance in terms of diagnostic sensitivity and specificity in clinical practice. Thus, we developed and validated a new sepsis biomarker based on amino acid profiling. METHODS: We used two independent groups. The training and validation groups included 161 and 22 healthy controls, 123 and 50 patients with systemic inflammatory response syndrome, and 115 and 45 patients with sepsis, respectively. Using mass spectrometry, we measured and analyzed serum amino acid levels to select candidate amino acids that could differentiate sepsis from other conditions. Then, several possible multivariate indexes were developed by generating formulae with different combinations of candidate amino acids. The formula showing the best performance was selected and validated further. RESULTS: Kynurenine, tryptophan, phenylalanine, arginine, aspartic acid, glutamic acid, and glutamine were selected as candidate amino acids. Ten possible formulae were generated, and the formula with the highest diagnostic performance, which included kynurenine, tryptophan, phenylalanine, and arginine, was selected. In the validation group, the area under the receiving operating characteristic curve of the selected multivariate index (0.931) was similar to that of procalcitonin (0.945). Moreover, the generated multivariate index showed potential as a prognostic marker. CONCLUSIONS: Serum amino acid composition in patients with sepsis differs significantly from that in healthy individuals and patients with inflammation only. The newly developed multivariate index is expected to be implementable as a sepsis biomarker in clinical practice in the near future.

Corticosteroid responsiveness in patients with acute exacerbation of interstitial lung disease admitted to the emergency department.

Interstitial lung disease (ILD), particularly idiopathic pulmonary fibrosis (IPF), has a poor prognosis. Corticosteroids are widely used in the treatment of acute exacerbation of ILD (AE-ILD). This study aimed to clarify the causes of AE-ILD, determine the efficacy of corticosteroids for treating AE-ILD, and detect differences in the mortality rate among subgroups of ILD. This was an observational retrospective single-center study. Patients with ILD who presented to the emergency department with acute respiratory symptoms from January 1, 2016, to December 31, 2018, were included. Patients with AE-ILD were classified into two groups depending on the prednisolone dose: low dose (0 to 1.0 mg/kg) or high dose (> 1.0 mg/kg). Mortality rates between patients with and without IPF were compared. This study included 182 patients with AE-ILD, including IPF (n = 117) and non-IPF (n = 65). Multivariate Cox regression analysis showed that corticosteroid dose (HR: 0.221, CI: 0.102-0.408, P < 0.001), initial P/F ratio (HR:0.995, CI:0.992-0.999, P = 0.006), and mechanical ventilation within 3 days of hospitalization (HR:4.205, CI:2.059-8.589, P < 0.001) were independent risk factors for mortality in patients with AE-ILD. This study showed that outcomes improve with higher doses of corticosteroids (> 1 mg/kg prednisolone) in patients with AE-non-IPF-ILD. However, this was not the case in patients with AE-IPF.

Characteristics and risk factors associated with missed diagnosis in patients with smear-negative pulmonary tuberculosis.

BACKGROUND/AIMS: Delayed diagnosis and treatment of smear-negative pulmonary tuberculosis (TB) are major concerns for TB control. We evaluated characteristics of patients with smear-negative pulmonary TB who received a delayed diagnosis and identified risk factors that may have contributed to this delay. METHODS: We reviewed medical records of patients with smear-negative culture-positive pulmonary TB treated at a tertiary care hospital in South Korea between January 2017 and December 2018. Patients who initiated anti-TB treatment after positive cultures were included in the missed TB group, and those who initiated empirical treatment before positive cultures were included in the control group. RESULTS: Of 220 patients included, 117 (53.2%) and 103 (46.8%) were in the missed TB and control groups, respectively. Patients in the missed TB group were older (p = 0.001) and had a higher mean body mass index (BMI) (p = 0.019). Comorbidities (66.9% vs. 46.6%, p = 0.003) and immunocompromised patients (33.1% vs. 20.4%, p = 0.035) were more common in the missed TB group than in the control group. Old age (odds ratio [OR], 1.030; 95% confidence interval [CI], 1.012 to 1.048; p = 0.001), high BMI (OR, 1.114; 95% CI, 1.004 to 1.237; p = 0.042), and negative polymerase chain reaction (PCR) results (OR, 9.551; 95% CI, 4.925 to 18.521; p < 0.001) were associated with delayed diagnosis. CONCLUSION: In more than half of patients with smear-negative pulmonary TB, the diagnosis was delayed. Patients with delayed TB diagnosis were older, had higher BMI, and negative PCR results.

Efficacy of low dose pirfenidone in idiopathic pulmonary fibrosis: real world experience from a tertiary university hospital.

Pirfenidone is an antifibrotic agent that has been proven to slow down the progression of idiopathic pulmonary fibrosis (IPF). The aim of this study was to evaluate the efficacy of low-dose pirfenidone (that is, less than 1200 mg/day). We retrospectively reviewed the medical records of patients with IPF. The patients were divided into the following three groups, those who were not treated with pirfenidone (control) and those who were treated with pirfenidone at doses < 1200 mg/day (low-dose group) and ≥ 1200 mg/day (high-dose group). The adjusted mean changes in forced vital capacity (FVC) in 1 year were - 200.7, - 88.4, and - 94.7 mL in the control, low-dose, and high-dose groups (p = 0.021). The FVC declined more significantly in the control group than in the low-dose and high-dose groups. No significant difference in FVC change was observed between the low-dose and high-dose groups. Dyspepsia, anorexia, and nausea were significantly more frequent in the low-dose than in the high-dose group, suggesting that dose reduction is attributed to gastrointestinal tract-related adverse events. Dose reduction may help patients to better control gastrointestinal tract-related adverse events; continuing taking the medication at low doses is also expected to be effective in reducing the FVC decline.