Pneumonia

Performance of various pneumonia severity models for predicting adverse outcomes in elderly inpatients with community-acquired pneumonia

OBJECTIVE

We aimed to assess the performance of common pneumonia severity scores, such as pneumonia severity index (PSI), CURB-65, CRB-65, A-DROP, and SMART-COP, in predicting adverse outcomes in elderly community-acquired pneumonia cohort and to determine the optimal scoring system for specific outcomes of interest.

METHODS

A total of 822 elderly inpatients were included in the retrospective cohort study. Clinical and laboratory results on admission were used to calculate the above scores. The primary outcome was 30-day mortality. Secondary outcomes were in-hospital mortality, need for mechanical ventilation (MV) and ICU admission. Model discrimination was evaluated by the area under receiver operating characteristic curves (AUCs).

RESULTS

The 30-day and in-hospital mortality rates were 6.8% (56/822) and 8.6% (71/822), respectively. One hundred and ninety-eight (24.0%) received MV and 111 (13.5%) were admitted to the ICU. All five scoring systems showed the same trend of increasing rates of each adverse outcome with increasing risk groups (all p < 0.001). PSI had the highest AUC, sensitivity, and negative predictive value (NPV) in predicting 30-day mortality and in-hospital mortality. SMART-COP had the highest AUC for predicting the need for MV and ICU admission, but PSI had the highest sensitivity and NPV for these two outcomes.

DISCUSSION

PSI performed well in identifying elderly patients at risk for 30-day mortality and its high NPV is helpful in excluding patients who are not at risk. Considering their effectiveness and simplicity, SMART-COP and CURB-65 are easier to perform in clinical practice than PSI.

Antibacterial effects of Kampo products against pneumonia causative bacteria

Community-acquired pneumonia is caused primarily by bacterial infection. For years, antibiotic treatment has been the standard of care for patients with bacterial pneumonia, although the emergence of antimicrobial-resistant strains is recognized as a global health issue. The traditional herbal medicine Kampo has a long history of clinical use and is relatively safe in treating various diseases. However, the antimicrobial effects of Kampo products against pneumonia-causative bacteria remain largely uncharacterized. In this study, we investigated the bacteriological efficacy of 11 Kampo products against bacteria commonly associated with pneumonia. Sho-saiko-To (9), Sho-seiryu-To (19), Chikujo-untan-To (91) and Shin'i-seihai-To (104) inhibited the growth of S. pneumoniae serotype 3, a highly virulent strain that causes severe pneumonia. Also, the growth of S. pneumoniae serotype 1, another highly virulent strain, was suppressed by treatment with Sho-saiko-To (9), Chikujo-untan-To (91), and Shin'i-seihai-To (104). Minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) against these strains ranged from 6.25-50 mg/mL and 12.5-25 mg/mL, respectively. Furthermore, Sho-saiko-To (9), Chikujo-untan-To (91), and Shin'i-seihai-To (104) suppressed the growth of antibiotic-resistant S. pneumoniae isolates. Additionally, Sho-saiko-To (9) and Shin'i-seihai-To (104) showed growth inhibition activity against Staphylococcus aureus, another causative agent for pneumonia, with MIC ranging from 6.25-12.5 mg/mL. These results suggest that some Kampo products have antimicrobial effects against S. pneumoniae and S. aureus, and that Sho-saiko-To (9) and Shin'i-seihai-To (104) are promising medicines for treating pneumonia caused by S. pneumoniae and S. aureus infection.

Is There a Role for Bronchoscopy in Aspiration Pneumonia?

Aspiration represents the passage of oropharyngeal content to the lower respiratory tract. The interplay between the host and the aspirate proprieties determines the subsequent aspiration syndrome. A low pH, typical of gastric aspirate, favors chemical pneumonitis, whereas an increased bacterial inoculum causes aspiration pneumonia. About a quarter of patients with aspiration pneumonitis will develop a bacterial superinfection during the course of recovery. While antibiotic therapy is indicated for aspiration pneumonia, supportive care remains the cornerstone of treatment in aspiration pneumonitis. However, the overlapping clinical features of these syndromes lead to initiation of antimicrobial therapy in most cases of aspiration. Bronchoscopy can aid in clinical decision-making by direct airway visualization and also by providing access to a series of emerging biomarkers. Invasive microbiological studies increase diagnostic yield and enable a tailored antibiotic treatment. In conjunction with stewardship programs, invasive sampling and novel molecular diagnostics can decrease the amount of inappropriate antibiotic therapy. In the context of foreign body aspiration, bronchoscopy represents both diagnostic and treatment gold standard.

Predicting Clinical Outcomes in COVID-19 and Pneumonia Patients: A Machine Learning Approach

In the clinical diagnosis of pneumonia, particularly during the COVID-19 pandemic, individuals who progress to a critical stage requiring mechanical ventilation are classified as mechanically ventilated critically ill patients. Accurately predicting the discharge outcomes for this specific cohort, especially those with COVID-19, is of paramount clinical importance. Missing data, a common issue in medical research, can significantly impact the validity of analyses. In this work, we address this challenge by employing two missing data imputation techniques: multiple imputation and missForest, to enhance data completeness. Additionally, we utilize the smoothly clipped absolute deviation (SCAD) penalized logistic regression method to select significant features. Our real data analysis compares the predictive performances of extreme learning machines, random forests, support vector machines, and XGBoost using 10-fold cross-validation. The results consistently show that XGBoost outperforms the other methods in predicting discharge outcomes, making it a reliable tool for clinical decision-making in the treatment of severe pneumonia, including COVID-19 cases. Within this context, the random forest imputation method generally enhances performance, underscoring its effectiveness in managing missing data compared to multiple imputation.

Glycyrrhetinic acid reduces lung inflammation caused by pneumococcal infection by reducing the toxicity of pneumolysin

Objective

In this study, to provide new methods for the treatment of Streptococcus pneumoniae infection, we aimed to describe the anti-inflammatory and antibacterial value of glycyrrhetinic acid on the basis of its inhibitory effect on bacterial growth (without killing the bacteria) and its reduction of the toxicity of S. pneumoniae.

Methods

A mouse model was established via intranasal administration of Streptococcus pneumoniae D39, and glycyrrhetinic acid was subcutaneously injected for treatment. The wet‒dry ratio, bacterial flora content and inflammatory factor levels in the mouse lungs were determined. Cell experiments were used to evaluate glycyrrhetinic acid-mediated inhibition of PLY hemolysis and A549 cell death, and WB was used to measure glycyrrhetinic acid-mediated inhibition of PLY oligomerization.

Results

Glycyrrhetinic acid reduced the levels of inflammatory factors, the dry‒wet ratio, the abundance of S. pneumoniae in the lungs of infected mice, pneumolysin-mediated A549 cell death, erythrocyte hemolysis and PLY oligoplasia.

Conclusion

Glycyrrhetinic acid can reduce the virulence of S. pneumoniae by preventing the oligomerization of PLY.

Diagnostic Value of Metagenomic next-generation sequencing and X-pert in Bronchoalveolar lavage fluid for pneumonia in HIV-infected and HIV-uninfected patients

Background

The pathogens causing unexplained pneumonia in both HIV-infected or HIV-unfected patients are likely to be complex. This retrospective study aimed to characterize the etiology of pneumonia in HIV-infected and HIV-uninfected patients using bronchoalveolar lavage fluid (BALF) analysis with metagenomic next-generation sequencing (mNGS) and X-pert MTB/RIF.

Methods

Between January 2022 and May2024, 141 HIV-infected and 104 HIV-uninfected patients admitted to Nanjing Second Hospital with pneumonia were included. BALF samples were collected and analyzed using mNGS to detect bacteria, fungi, viruses, tuberculosis (TB) and non-tuberculous mycobacteria (NTM), and X-pert for TB detection. Clinical data including CD4 T-cell counts, comorbidities, and ART status were collected and analyzed.

Results

HIV-uninfected patients were found to be older and exhibited a higher prevalence of comorbidities compared to HIV-infected patients. Despite higher median CD4 T-cell counts in HIV-uninfected individuals (412 cells/μL vs. 31 cells/μL in HIV-infected), TB detection rates using X-pert and mNGS were lower than anticipated, particularly in HIV-infected patients. Mixed-pathogen infections were significantly more prevalent in HIV-infected patients, especially those with lower CD4 T-cell counts. ART use showed variable impacts on pathogen diversity, with longer treatment durations associated with reduced infection complexity but persistent immunodeficiency in some cases.In patients with pneumonia, whether HIV-infected or HIV-uninfected, pathogens often exhibit complexity, underscoring the critical role of timely mNGS and X-pert analysis of BALF for early pathogen detection.

Repetitive Acinetobacter baumannii pneumonia induces infection tolerance in mice

Some long-term hospitalized patients with lung infections exhibit pathogen tolerance. To investigate whether long-term chronic infection can induce tolerance, we constructed a mouse model of pneumonia in which mice were infected once, twice, or three times with Acinetobacter baumannii. The results revealed that the inflammatory factor levels decreased in the lung lavage fluid and that pathological damage to the lung tissue was alleviated in the mice infected three times. Flow cytometry and transcriptome analysis of mouse lung tissue revealed that the expression of genes related to T cell activation, differentiation, and regulation and the proportion and number of regulatory T cells and immune suppression-related genes, such as Ctla4, Tigit, Slamf8, ICOS, and IDO1, were increased in mice infected three times. These findings show that repeated A. baumannii infections can induce tolerance, which may be mediated by immune suppression involving regulatory T cells.

Investigating mechanisms of Sophora davidii (Franch.) skeels flower extract in treating LPS-induced acute pneumonia based on network pharmacology

ETHNOPHARMACOLOGICAL RELEVANCE

In TCM opinion, most of pneumonia is related to "lung heat". Sophora davidii (Franch.) Skeels flower was first documented in "Guizhou Herbal Medicine", and was recorded as having functions of clearing heat, detoxifying, and cooling blood. It can be used to treat lung heat cough.

AIM OF THE STUDY

To investigate main mechanisms of Sophora davidii flower extract (SDFE) in Treating LPS-induced acute Pneumonia.

MATERIALS AND METHODS

Acute pneumonia models on BEAS-2B cells and rats were established using LPS. The rat model was used to verified the protective effects of SDFE through HE staining, lung tissue W/D ratio assay, white blood cell count analysis, and ammonia-induced coughing test. Network pharmacology was applied to predict the active compounds, core targets and main pathways of SDFE in treating acute pneumonia. Western Blot and ELISA kits were employed to validate representative proteins in selected pathway in vivo and in vitro.

RESULTS

HE staining, lung tissue W/D ratio assay, white blood cell count analysis, and ammonia-induced coughing test showed SDFE could improve pathological features (leukocyte infiltration, pulmonary edema, lung injury and cough). Network pharmacology indicated MAPK/NF-κB pathway was the most relevant pathway. SDFE could significantly inhibit the expression of Fos and Jun, and the phosphorylation levels of p38, ERK, JNK, NF-κB and IκB. It also down-regulated the expression of pro-inflammatory factors (TNF-α, IL-6 and IL-1β).

CONCLUSIONS

SDFE can exert protective effects against acute pneumonia through the MAPK/NF-κB signaling pathway.

Simultaneously Controlling Inflammation and Infection by Smart Nanomedicine Responding to the Inflammatory Microenvironment

The overactivated immune cells in the infectious lesion may lead to irreversible organ damages under severe infections. However, clinically used immunosuppressive anti-inflammatory drugs will usually disturb immune homeostasis and conversely increase the risk of infections. Regulating the balance between anti-inflammation and anti-infection is thus critical in treating certain infectious diseases. Herein, considering that hydrogen peroxide (H2O2), myeloperoxidase (MPO), and neutrophils are upregulated in the inflammatory microenvironment and closely related to the severity of appendectomy patients, an inflammatory-microenvironment-responsive nanomedicine is designed by using poly(lactic-co-glycolic) acid (PLGA) nanoparticles to load chlorine E6 (Ce6), a photosensitizer, and luminal (Lum), a chemiluminescent agent. The obtained Lum/Ce6@PLGA nanoparticles, being non-toxic within normal physiological environment, can generate cytotoxic single oxygen via bioluminescence resonance energy transfer (BRET) in the inflammatory microenvironment with upregulated H2O2 and MPO, simultaneously killing pathogens and excessive inflammatory immune cells in the lesion, without disturbing immune homeostasis. As evidenced in various clinically relevant bacterial infection models and virus-induced pneumonia, Lum/Ce6@PLGA nanoparticles appeared to be rather effective in controlling both infection and inflammation, resulting in significantly improved animal survival. Therefore, the BRET-based nanoparticles by simultaneously controlling infections and inflammation may be promising nano-therapeutics for treatment of severe infectious diseases.

Efficacy and Safety of Dexmedetomidine in the Prone Position in Elderly Patients with Pneumonia: A Prospective, Double-Blind, Randomized Controlled Study

PURPOSE

We aimed to identify a safe and effective method to assist older adults with pneumonia in tolerating the prone position for a longer duration.

METHODS

This was a randomized, controlled, double-blinded study performed at the Shanghai Fourth People's Hospital. Eighty patients with pneumonia aged ≥ 65 years were included. The patients were able to spontaneous breath in the prone position and were administered intravenous dexmedetomidine or an isotonic sodium chloride solution. The cumulative daily durations of prone positioning for all patients in the two groups were recorded. The primary outcome was the percentage of patients who completed ≥ 9 h/day in the prone position. The secondary outcomes included the incidence of complications in the prone position and patient outcomes.

RESULTS

Eighty patients were included (average age: 79.6 ± 8.9 years). The percentage of patients who completed ≥ 9 h/day in the prone position was significantly higher in the dexmedetomidine group than in the placebo group (P = 0.011). The percentage of patients who completed ≥ 12 h/day in the prone position was also significantly greater in the dexmedetomidine group than in the placebo group (P = 0.008). There were no significant differences in other variables between the two groups.

CONCLUSIONS

The results of this study demonstrate that intravenous dexmedetomidine injection can significantly prolong the duration of spontaneous breathing in the prone position in elderly pneumonia patients without obvious adverse events. We provide a safe and effective method to help patients with pneumonia, especially those with delirium or cognitive impairment, who cannot tolerate the length of time needed for spontaneous breathing in the prone position to be effective.

TRIAL REGISTRATION

The study was registered with the Chinese Clinical Trial Center (registration number: ChiCRT2300067383) on 2023-01-05.

Comparative study on the incidence of non-COVID-19 viral pneumonia before and after the COVID-19 pandemic: A retrospective analysis based on respiratory non-COVID viral nucleic acid results

Background

The impact of the coronavirus disease 2019 (COVID-19) pandemic on the etiology of non-COVID-19 viral pneumonia remains to be identified. We investigated the evolution of non-COVID-19 viral pneumonia in hospitalized patients before and after the COVID-19 pandemic.

Methods

This is a single-center retrospective study. Patients who came to West China Hospital of Sichuan University diagnosed with non-COVID-19 viral pneumonia from January 1, 2016 to December 31, 2021, were included and divided into pre- and post-COVID-19 groups according to the date of the COVID-19 outbreak in China. The results of 13 viral nucleic acid tests were compared between the two groups.

Results

A total of 5937 patients (3954 in the pre-COVID-19 group and 1983 in the post-COVID-19 group) were analyzed. Compared with the pre-COVID-19 group, the proportion of patients tested for respiratory non-COVID-19 viral nucleic acid was significantly increased in the post-COVID-19 group (14.78% vs. 22.79%, P <0.05). However, the non-COVID-19 virus-positive rates decreased from 37.9% to 14.6% after the COVID-19 outbreak (P < 0.001). Notably, non-COVID-19 viral pneumonia caused by the influenza A virus H1N1 (InfAH1N1) (2009) dropped to 0% after the pandemic. The top three viruses were InfAH1N1 (2009) (13.9%), human rhinovirus (7.4%), and human adenovirus (3.4%) in the pre-COVID-19 group, and human rhinovirus (3.8%), human respiratory syncytial virus (2.0%), human parainfluenza virus (1.1%) and InfAH3N2 (1.1%) in the post-COVID-19 group.

Conclusions

The proportion of non-COVID-19 viral pneumonia decreased significantly after the COVID-19 outbreak, among which InfAH1N1 (2009) pneumonia decreased the most dramatically.

Steroids in severe community-acquired pneumonia

There is conflicting evidence regarding the use of steroids in severe community-acquired pneumonia (CAP), with previous randomised controlled trials limited by small sample sizes. ESCAPe and CAPE COD are two recently published large trials on steroids in severe CAP. ESCAPe assessed the initiation of methylprednisolone within 72-96 h of hospital admission, while CAPE COD studied the use of hydrocortisone within 24 h of the development of severe CAP. ESCAPe did not show any differences in all-cause 60-day mortality or any of its secondary outcomes. CAPE COD showed that hydrocortisone improved all-cause 28-day mortality and reduced the risk of intubation or vasopressor-dependent shock. Important differences between the trials included the steroid regimens used, timing of steroid administration and baseline characteristics, with more diabetic patients included in ESCAPe. The results of CAPE COD support the initiation of hydrocortisone within 24 h of developing severe CAP, but more research is needed to evaluate long-term outcomes and optimum dosing regimens for steroids in severe CAP.

PD1+CD4+ T cells promote receptor editing and suppress autoreactivity of CD19+CD21low B cells within the lower respiratory airways in adenovirus pneumonia

Human adenovirus (HAdV) pneumonia poses a major health burden for young children, however, factors that contribute to disease severity remain elusive. We analyzed immune cells from bronchoalveolar lavage (BAL) of children with HAdV pneumonia and found that CD19+CD21low B cells were significantly enriched in the BAL and were associated with increased autoantibody concentrations and disease severity. Myeloid cells, PD-1+CD4+ T helper cells and CD21low B cells formed tertiary lymphoid structures within the respiratory tracts. Myeloid cells promoted autoantibody production by expressing high amounts of B cell activating factor (BAFF). In contrast, PD-1+CD4+ T helper cells induced production of IgG1 and IgG3 antibodies but suppressed autoreactive IgGs by initiating B cell receptor editing. In summary, this study reveals cellular components involved in protective versus autoreactive immune pathways in the respiratory tract, and these findings provide potential therapeutic targets for severe HAdV lower respiratory tract infections.

Effect of the lack of access to handwashing facilities on the global burden of lower respiratory infections, 1990-2019: A systematic analysis from the global burden of disease study 2019

A lack of access to handwashing facilities is a significant risk factor for lower respiratory infections(LRIs). However, no studies have reported epidemiologic changes in the burden of LRIs attributed to the lack of access to handwashing facilities. We conducted an integrated assessment of the burden of LRIs attributable to the lack of handwashing facilities from 1990 to 2019 using data from the Global Burden of Disease Study 2019. In 2019, 270,000 deaths were attributed to LRIs due to a lack of access to handwashing facilities, with DALYs reaching 14.02 million. The age-standardized mortality rate (ASMR) of LRIs caused by a lack of access to handwashing facilities was approximately 3.74, while the age-standardized DALY rate (ASDR) was reported to be 203.55 in 2019. Over the past 30 years, the burden of LRIs attributed to the lack of access to handwashing facilities has shown a global decline. In 2019, this burden was most pronounced in infants under 1 year of age and in those older than 95 years, reflecting the highest DALY (5591.83) and mortality rates (79.43), respectively. The burden of LRIs caused by the lack of access to handwashing facilities was found to be more severe in males and significantly more pronounced in regions with a low sociodemographic index (SDI), such as the Sahara African region. The development of targeted strategies to address the inadequate and unequal distribution of handwashing facilities holds important value in improving the disease burden of LRIs.

The value of surfactant protein a in evaluating the severity and prognosis in community-acquired pneumonia patients

BACKGROUND

Previous research has discovered that surfactant protein A (SP-A) is involved in the pathophysiology processes of certain lung illnesses. However, no definitive clinical studies have delved into the function of SP-A in individuals afflicted with community-acquired pneumonia (CAP). A prospective cohort study was used to investigate the relationships between blood SP-A levels and the severity and prognosis among CAP patients.

MATERIALS AND METHODS

This study included 260 patients with CAP. Clinical traits and demographic data were examined during hospitalization. The concentrations of serum SP-A and serum interleukin-6 (IL-6) were determined by enzyme-linked immunosorbent assay (ELISA). In addition, to evaluate the severity of CAP, a variety of scores, including the CURB-65, PSI, SMART-COP, and APACHE II, were employed.

RESULTS

The serum levels of SP-A at admission exhibited a gradual decline as the severity scores of CAP increased. Through Spearman correlation analysis, we observed an association between serum SP-A and some clinical indicators among CAP patients. Furthermore, results from a multiple linear regression model suggested changes in PSI scores (-17.868 scores, 95% CI: -32.743, -2.993) affect serum SP-A more than CURB-65 (-0.547 scores, 95% CI: -0.964, -0.131), SMART-COP (-1.097 scores, 95% CI: -1.889, -0.304) and APACHE II (-3.475 scores, 95% CI: -5.874, -1.075) with age, hypertension, diabetes mellitus, cerebral infarction, coronary heart disease, and bronchitis adjusted. In addition, the prognosis in CAP patients was monitored. Throughout their hospital stay, higher serum levels of SP-A decreased the risks of mechanical ventilation (RR: 0.315; 95% CI: 0.106, 0.937), vasoactive agents (RR: 0.165; 95% CI: 0.034, 0.790), intensive care unit (ICU) admissions (RR: 0.218; 95% CI: 0.066, 0.717) and longer hospital stays (RR: 0.397; 95% CI: 0.167, 0.945).

CONCLUSION

In CAP patients, inverse dose-response correlations exist between serum SP-A levels with severity scores as well as prognosis at admission, suggesting that SP-A may take part in the CAP pathophysiological processes. Moreover, lower serum SP-A on admission is associated with an elevated prognostic risk of mechanical ventilation, the use of vasoactive agents, longer hospital stays, ICU admission, and mortality. Therefore, as a biomarker, SP-A may have the potential to predict the severity and poor prognosis of CAP patients.

Outcomes of corticosteroid therapy in patients with viral community-acquired pneumonia

AIM

The objective of this study was to assess the therapeutic effects of corticosteroids in adult patients hospitalized with viral community-acquired pneumonia.

METHODS

This is a retrospective analysis of data collected prospectively from November 1996 to June 2024. All adult patients with viral community-acquired pneumonia were enrolled. The primary outcome was 30-day mortality. Secondary outcomes included all-cause in-hospital mortality, ICU admission, length of ICU and hospital stay, mechanical ventilation, and 1-year mortality. Propensity score matching (PSM) was used to obtain balance among the baseline variables in the two groups.

RESULTS

Of the 524 patients with viral pneumonia, 30 (6%) received corticosteroids and 494 (94%) did not. Patients were primarily male (n = 299, 57%), with a median [Q1-Q3] age of 66.9 [55-81] years. The 3:1 propensity matching procedure identified 90 patients not treated with corticosteroid (CS-) as controls. After PSM, no difference in 30-day mortality was found [7% (95%CI 1 to 22%) vs. 4% (95%CI 1 to 11%), p = 0.639]. The risk of death at 30 days did not differ significantly in unmatched and matched cohorts [Hazard Ratio (HR) 1.33 (0.32-5.63), p = 0.695 vs. HR 1.51 (0.28-8.27), p = 0.632, respectively]. Nor were differences found in hospital length of stay, ICU admission and length of stay, or mechanical ventilation requirement and duration between matched and unmatched CS + and CS-.

CONCLUSIONS

There were no significant differences in the primary and secondary outcomes regarding the use of corticosteroids in patients with viral pneumonia.

Transitions in lung microbiota landscape associate with distinct patterns of pneumonia progression

The precise microbial determinants driving clinical outcomes in severe pneumonia are unknown. Competing ecological forces produce dynamic microbiota states in health; infection and treatment effects on microbiota state must be defined to improve pneumonia therapy. Here, we leverage our unique clinical setting, which includes systematic and serial bronchoscopic sampling in patients with suspected pneumonia, to determine lung microbial ecosystem dynamics throughout pneumonia therapy. We combine 16S rRNA gene amplicon, metagenomic, and transcriptomic sequencing with bacterial load quantification to reveal clinically-relevant pneumonia progression drivers. Microbiota states are predictive of pneumonia category and exhibit differential stability and pneumonia therapy response. Disruptive forces, like aspiration, associate with cohesive changes in gene expression and microbial community structure. In summary, we show that host and microbiota landscapes change in unison with clinical phenotypes and that microbiota state dynamics reflect pneumonia progression. We suggest that distinct pathways of lung microbial community succession mediate pneumonia progression.

Inflammatory and Immune Mechanisms in COPD: Current Status and Therapeutic Prospects

Background

Chronic obstructive pulmonary disease (COPD) currently ranks among the top three causes of mortality worldwide, presenting as a prevalent and complex respiratory ailment. Ongoing research has underscored the pivotal role of immune function in the onset and progression of COPD. The immune response in COPD patients exhibits abnormalities, characterized by diminished anti-infection capacity due to immune senescence, heightened activation of neutrophils and macrophages, T cell infiltration, and aberrant B cell activity, collectively contributing to airway inflammation and lung injury in COPD.

Objective

This review aimed to explore the pivotal role of the immune system in COPD and its therapeutic potential.

Methods

We conducted a review of immunity and COPD published within the past decade in the Web of Science and PubMed databases, sorting through and summarizing relevant literature.

Results

This article examines the pivotal roles of the immune system in COPD. Understanding the specific functions and interactions of these immune cells could facilitate the development of novel therapeutic strategies and interventions aimed at controlling inflammation, enhancing immune function, and mitigating the impact of respiratory infections in COPD patients.

Data-driven classification and explainable-AI in the field of lung imaging

Detecting lung diseases in medical images can be quite challenging for radiologists. In some cases, even experienced experts may struggle with accurately diagnosing chest diseases, leading to potential inaccuracies due to complex or unseen biomarkers. This review paper delves into various datasets and machine learning techniques employed in recent research for lung disease classification, focusing on pneumonia analysis using chest X-ray images. We explore conventional machine learning methods, pretrained deep learning models, customized convolutional neural networks (CNNs), and ensemble methods. A comprehensive comparison of different classification approaches is presented, encompassing data acquisition, preprocessing, feature extraction, and classification using machine vision, machine and deep learning, and explainable-AI (XAI). Our analysis highlights the superior performance of transfer learning-based methods using CNNs and ensemble models/features for lung disease classification. In addition, our comprehensive review offers insights for researchers in other medical domains too who utilize radiological images. By providing a thorough overview of various techniques, our work enables the establishment of effective strategies and identification of suitable methods for a wide range of challenges. Currently, beyond traditional evaluation metrics, researchers emphasize the importance of XAI techniques in machine and deep learning models and their applications in classification tasks. This incorporation helps in gaining a deeper understanding of their decision-making processes, leading to improved trust, transparency, and overall clinical decision-making. Our comprehensive review serves as a valuable resource for researchers and practitioners seeking not only to advance the field of lung disease detection using machine learning and XAI but also from other diverse domains.

Development and validation of machine learning-based prediction model for severe pneumonia: A multicenter cohort study

Severe pneumonia (SP) is a prevalent respiratory ailment characterized by high mortality and poor prognosis. Current scoring systems for pneumonia are not only time-consuming but also exhibit limitations in early SP prediction. To address this gap, this study aimed to develop a machine-learning model using inflammatory markers from peripheral blood for early prediction of SP. A total of 204 pneumonia patients from seven medical centers were studied, with 143 (68 SP cases) in the training cohort and 61 (32 SP cases) in the test cohort. Clinical characteristics and laboratory test results were collected at diagnosis. Various models including Logistic Regression, Random Forest, Naïve Bayes, XGBoost, Support Vector Machine, and Decision Tree were built and evaluated. Seven predictors-age, sex, WBC count, T-lymphocyte count, NLR, CRP, TNF-α, IL-4/IFN-γ ratio, IL-6/IL-10 ratio-were selected through LASSO regression and clinical insight. The XGBoost model, exhibiting best performance, achieved an AUC of 0.901 (95 % CI: 0.827 to 0.985) in the test cohort, with an accuracy of 0.803, sensitivity of 0.844, specificity of 0.759, and F1_score of 0.818. Indeed, SHAP analysis emphasized the significance of elevated WBC counts, older age, and elevated CRP as the top predictors. The use of inflammatory biomarkers in this concise predictive model shows significant potential for the rapid assessment of SP risk, thereby facilitating timely preventive interventions.

An open-source framework for end-to-end analysis of electronic health record data

With progressive digitalization of healthcare systems worldwide, large-scale collection of electronic health records (EHRs) has become commonplace. However, an extensible framework for comprehensive exploratory analysis that accounts for data heterogeneity is missing. Here we introduce ehrapy, a modular open-source Python framework designed for exploratory analysis of heterogeneous epidemiology and EHR data. ehrapy incorporates a series of analytical steps, from data extraction and quality control to the generation of low-dimensional representations. Complemented by rich statistical modules, ehrapy facilitates associating patients with disease states, differential comparison between patient clusters, survival analysis, trajectory inference, causal inference and more. Leveraging ontologies, ehrapy further enables data sharing and training EHR deep learning models, paving the way for foundational models in biomedical research. We demonstrate ehrapy's features in six distinct examples. We applied ehrapy to stratify patients affected by unspecified pneumonia into finer-grained phenotypes. Furthermore, we reveal biomarkers for significant differences in survival among these groups. Additionally, we quantify medication-class effects of pneumonia medications on length of stay. We further leveraged ehrapy to analyze cardiovascular risks across different data modalities. We reconstructed disease state trajectories in patients with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) based on imaging data. Finally, we conducted a case study to demonstrate how ehrapy can detect and mitigate biases in EHR data. ehrapy, thus, provides a framework that we envision will standardize analysis pipelines on EHR data and serve as a cornerstone for the community.

Anti-inflammatory Effect of Symplocos prunifolia Extract in an In Vitro Model of Acute Pneumonia

Acute pneumonia is a respiratory disease characterized by inflammation within the lung tissue, exhibiting higher morbidity rates and mortality rates among immunocompromised children and older adults. Symplocos species have been traditionally used as herbal remedies for conditions like dysentery, skin ulcers, diarrhea, and dyspepsia. Contemporary research has employed various Symplocos species in the study of diverse diseases. However, the exact efficacy and mechanisms of action of Symplocos Prunifolia remain unknown. Therefore, this study investigated the anti-inflammatory mechanism of S. prunifolia extract (SPE) in A549 and RAW264.7 cells stimulated by lipopolysaccharide (LPS). SPE significantly reduced nitric oxide (NO) production and the protein expression levels of like inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) in LPS-stimulated RAW 264.7 cells. Furthermore, it reduced the protein expression levels of iNOS, COX-2 and the levels of pro-inflammatory cytokines in LPS-stimulated A549 cells. The mechanism underlying the anti-inflammatory effect of SPE was associated with the inhibition of LPS stimulated the phosphoinositide-3-kinase/protein kinase B (PI3K/Akt) and Mitogen-activated protein kinase (MAPK) phosphorylation. Moreover, we confirmed that SPE decreased the nuclear translocation of nuclear factor-κB (NF-κB)/p65 stimulated by LPS. In conclusion, these results demonstrate that SPE alleviates inflammatory responses by deactivating the PI3K/Akt, MAPK, and NF-κB signaling pathways. Our findings suggest that SPE is a potential candidate for acute pneumonia prevention.

Dynamic changes of respiratory microbiota associated with treatment outcome in drug-sensitive and drug-resistant pulmonary tuberculosis

BACKGROUND

Respiratory microbiota is closely related to tuberculosis (TB) initiation and progression. However, the dynamic changes of respiratory microbiota during treatment and its association with TB progression remains unclear.

METHODS

A total of 16 healthy individuals and 16 TB patients (10 drug-sensitive TB (DS-TB) and 6 drug-resistant TB (DR-TB)) were recruited. Sputum samples were collected at baseline for all anticipants and after anti-TB treatment at Month-6 for TB patients. High throughput 16 S RNA sequencing was used to characterize the respiratory microbiota composition.

RESULTS

Compared to the healthy individuals, TB patients exhibited lower respiratory microbiota diversity (p < 0.05). This disruption was alleviated after anti-TB treatment, especially for DS-TB patients. Parvimonas spp. numbers significantly increased after six months of anti-TB treatment in both DS-TB and DR-TB patients (p < 0.05). Rothia spp. increase during treatment was associated with longer sputum-culture conversion time and worse pulmonary lesion absorption (p < 0.05). Besides, Moraxella spp. prevalence was associated with longer sputum-culture conversion time, while Gemella spp. increase was associated with worsening resolving of pulmonary lesions (p < 0.05).

CONCLUSION

Dynamic changes of respiratory microbiota during anti-TB treatment is closely related to TB progression. The involvement of critical microorganisms, such as Parvimonas spp., Rothia spp., Moraxella, and Gemella spp., appears to be associated with pulmonary inflammatory conditions, particularly among DR-TB. These microorganisms could potentially serve as biomarkers or even as targets for therapeutic intervention to enhance the prognosis of tuberculosis patients.

Topical adjunctive treatment with flagellin augments pulmonary neutrophil responses and reduces bacterial dissemination in multidrug-resistant K. pneumoniae infection

Objective

Antimicrobial resistance is an emerging problem and multi-drug resistant (MDR) Klebsiella pneumoniae (K. pneumoniae) represents an enormous risk of failing therapy in hospital-acquired pneumonia. The current study aimed to determine the immunomodulatory effect of topical flagellin in addition to antibiotic treatment during respiratory infection evoked by hypervirulent antibiotic-susceptible and antibiotic-resistant K. pneumoniae in mice.

Methods

C57BL6 mice were inoculated intranasally with hypervirulent K. pneumoniae (K2:O1) which was either antibiotic-susceptible or multi-drug resistant. Six hours after infection, mice were treated with antibiotics intraperitoneally and flagellin or vehicle intranasally. Mice were sacrificed 24 hours after infection. Samples were analyzed for bacterial loads and for inflammatory and coagulation markers.

Results

Flagellin therapy induced neutrophil influx in the lung during antibiotic-treated pneumonia evoked by either antibiotic-susceptible or -resistant K. pneumoniae. The pulmonary neutrophil response was matched by elevated levels of neutrophil-attracting chemokines, neutrophil degranulation products, and local coagulation activation. The combined therapy of effective antibiotics and flagellin did not impact K. pneumoniae outgrowth in the lung, but decreased bacterial counts in distant organs. Neutrophil depletion abrogated the flagellin-mediated effect on bacterial dissemination and local coagulation responses.

Conclusion

Topical flagellin administration as an adjunctive to antibiotic treatment augments neutrophil responses during pneumonia evoked by MDR-K. pneumoniae, thereby reducing bacterial dissemination to distant organs.

DSEception: a noval neural networks architecture for enhancing pneumonia and tuberculosis diagnosis

Background

Pneumonia and tuberculosis are prevalent pulmonary diseases globally, each demanding specific care measures. However, distinguishing between these two conditions imposes challenges due to the high skill requirements for doctors, the impact of imaging positions and respiratory intensity of patients, and the associated high healthcare costs, emphasizing the imperative need for intelligent and efficient diagnostic methods.

Method

This study aims to develop a highly accurate automatic diagnosis and classification method for various lung diseases (Normal, Pneumonia, and Tuberculosis). We propose a hybrid model, which is based on the InceptionV3 architecture, enhanced by introducing Deepwise Separable Convolution after the Inception modules and incorporating the Squeeze-and-Excitation mechanism. This architecture successfully enables the model to extract richer features without significantly increasing the parameter count and computational workload, thereby markedly improving the performance in predicting and classifying lung diseases. To objectively assess the proposed model, external testing and five-fold cross-validation were conducted. Additionally, widely used baseline models in the scholarly community were constructed for comparison.

Result

In the external testing phase, the our model achieved an average accuracy (ACC) of 90.48% and an F1-score (F1) of 91.44%, which is an approximate 4% improvement over the best-performing baseline model, ResNet. In the five-fold cross-validation, our model's average ACC and F1 reached 88.27% ± 2.76% and 89.29% ± 2.69%, respectively, demonstrating exceptional predictive performance and stability. The results indicate that our model holds promise for deployment in clinical settings to assist in the diagnosis of lung diseases, potentially reducing misdiagnosis rates and patient losses.

Conclusion

Utilizing deep learning for automatic assistance in the diagnosis of pneumonia and tuberculosis holds clinical significance by enhancing diagnostic accuracy, reducing healthcare costs, enabling rapid screening and large-scale detection, and facilitating personalized treatment approaches, thereby contributing to widespread accessibility and improved healthcare services in the future.

Mechanism of KLF2 in young mice with pneumonia induced by Streptococcus pneumoniae

BACKGROUND

Streptococcus pneumoniae (Spn) is a major causative agent of pneumonia, which can disseminate to the bloodstream and brain. Pneumonia remains a leading cause of death among children aged 1-59 months worldwide. This study aims to investigate the role of Kruppel-like factor 2 (KLF2) in lung injury caused by Spn in young mice.

METHODS

Young mice were infected with Spn to induce pneumonia, and the bacterial load in the bronchoalveolar lavage fluid was quantified. KLF2 expression in lung tissues was analyzed using real-time quantitative polymerase chain reaction and Western blotting assays. Following KLF2 overexpression, lung tissues were assessed for lung wet-to-dry weight ratio and Myeloperoxidase activity. The effects of KLF2 on lung injury and inflammation were evaluated through hematoxylin and eosin staining and enzyme-linked immunosorbent assay. Chromatin immunoprecipitation and dual-luciferase assay were conducted to examine the binding of KLF2 to the promoter of microRNA (miR)-222-3p and cyclin-dependent kinase inhibitor 1B (CDKN1B), as well as the binding of miR-222-3p to CDKN1B. Levels of miR-222-3p and CDKN1B in lung tissues were also determined.

RESULTS

In young mice with pneumonia, KLF2 and CDKN1B were downregulated, while miR-222-3p was upregulated in lung tissues. Overexpression of KLF2 reduced lung injury and inflammation, evidenced by decreased bacterial load, reduced lung injury, and lower levels of proinflammatory factors. Co-transfection of miR-222-3p-WT and oe-KLF2 significantly reduced luciferase activity, suggesting that KLF2 binds to the promoter of miR-222-3p and suppresses its expression. Transfection of CDKN1B-WT with miR-222-3p mimics significantly reduced luciferase activity, indicating that miR-222-3p binds to CDKN1B and downregulates its expression. Overexpression of miR-222-3p or downregulation of CDKN1B increased bacterial load in BALF, lung wet/dry weight ratio, MPO activity, and inflammation, thereby reversing the protective effect of KLF2 overexpression on lung injury in young mice with pneumonia.

CONCLUSIONS

KLF2 alleviates lung injury in young mice with Spn-induced pneumonia by transcriptional regulation of the miR-222-3p/CDKN1B axis.

Comparative effectiveness and safety of six antibiotics in treating MRSA infections: A network meta-analysis

OBJECTIVES

This study conducted a network meta-analysis comparing linezolid, teicoplanin, daptomycin, tigecycline, and ceftaroline fosamil with vancomycin for treating MRSA-related diseases, addressing the lack of comprehensive evaluations in existing research on antibiotic therapy for MRSA infections.

METHODS

We systematically searched databases including PubMed, Embase, Web of Science, the Cochrane Librar up to August 22, 2023. All eligible randomized controlled trials of the six antibiotics were included in the NMA, and their effectiveness and safety were compared across various MRSA-related diseases. Categorical data were used for the odds ratio (OR), and continuous data were used for mean difference (SMD). The surface under the cumulative ranking (SUCRA) was employed to evaluate the incidence rate.

RESULTS

According to SUCRA results, daptomycin was the most effective treatment (73.0%) in bloodstream infections. In pulmonary infections and skin and soft tissue infections, linezolid out-performed other antibiotics in effectiveness rate (90.6% and 86.3%), microbial killing rate (93.3% and 93.1%). Vancomycin showed lower adverse reactions than teicoplanin, with less hepatotoxicity compared to linezolid and tigecycline. Linezolid had higher thrombocytopenia risk but lower nephrotoxicity risk than others. Vancomycin was less effective in microbial killing rates than linezolid across various infections.

CONCLUSION

The present research suggests that in pulmonary infections and skin and soft tissue infections, linezolid may be a better option for treating MRSA-related diseases. However, caution is warranted due to the association of linezolid with thrombocytopenia.

TRIAL REGISTRATION

Our study protocol was registered with the International Prospective Register of SystematicReviews (PROSPERO); Registration number: CRD42024535142.

A review of KLF4 and inflammatory disease: Current status and future perspective

Inflammation is the response of the human body to injury, infection, or other abnormal states, which is involved in the development of many diseases. As a member of the Krüppel-like transcription factors (KLFs) family, KLF4 plays a crucial regulatory role in physiological and pathological processes due to its unique dual domain of transcriptional activation and inhibition. A growing body of evidence has demonstrated that KLF4 plays a pivotal role in the pathogenesis of various inflammatory disorders, including inflammatory bowel disease, osteoarthritis, renal inflammation, pneumonia, neuroinflammation, and so on. Consequently, KLF4 has emerged as a promising new therapeutic target for inflammatory diseases. This review systematically generalizes the molecular regulatory network, specific functions, and mechanisms of KLF4 to elucidate its complex roles in inflammatory diseases. An in-depth study on the biological function of KLF4 is anticipated to offer a novel research perspective and potential intervention strategies for inflammatory diseases.

Influenza vaccination in older adults and patients with chronic disorders: A position paper from the Portuguese Society of Pulmonology, the Portuguese Society of Cardiology, the Portuguese Society of Diabetology, the Portuguese Society of Infectious Diseases and Clinical Microbiology, the Portuguese Society of Geriatrics and Gerontology, and the Study Group of Geriatrics of the Portuguese Society of Internal Medicine

Influenza affects millions of people worldwide each year and can lead to severe complications, hospitalizations, and even death, especially among vulnerable populations such as older adults and those with chronic medical conditions. Annual vaccination is considered the most effective measure for preventing influenza and its complications. Despite the widespread availability of influenza vaccines, however, vaccination coverage rates remain suboptimal in several countries. Based on the latest scientific evidence and expert opinions on influenza vaccination in older people and patients with chronic disease, the Portuguese Society of Pulmonology (SPP), the Portuguese Society of Diabetology (SPD), the Portuguese Society of Cardiology (SPC), the Portuguese Society of Geriatrics and Gerontology (SPGG), the Study Group of Geriatrics of the Portuguese Society of Internal Medicine (NEGERMI-SPMI), and the Portuguese Society of Infectious Diseases and Clinical Microbiology (SPDIMC) discussed best practices for promoting vaccination uptake and coverage and drew up several recommendations to mitigate the impact of influenza. These recommendations focus on the efficacy and safety of available vaccines; the impact of influenza vaccination on older adults; patients with chronic medical conditions, namely cardiac and respiratory conditions, diabetes, and immunosuppressive diseases; and health care professionals, optimal vaccination timing, and strategies to increase vaccination uptake and coverage. The resulting position paper highlights the critical role that vaccinations play in promoting public health, raising awareness, and encouraging more people to get vaccinated.

cUMP elicits interendothelial gap formation during Pseudomonas aeruginosa infection

Pseudomonas aeruginosa utilizes a type 3 secretion system to intoxicate host cells with the nucleotidyl cyclase ExoY. After activation by its host cell cofactor, filamentous actin, ExoY produces purine and pyrimidine cyclic nucleotides, including cAMP, cGMP, and cUMP. ExoY-generated cyclic nucleotides promote interendothelial gap formation, impair motility, and arrest cell growth. The disruptive activities of cAMP and cGMP during the P. aeruginosa infection are established; however, little is known about the function of cUMP. Here, we tested the hypothesis that cUMP contributes to endothelial cell barrier disruption during P. aeruginosa infection. Using a membrane permeable cUMP analog, cUMP-AM, we revealed that during infection with catalytically inactive ExoY, cUMP promotes interendothelial gap formation in cultured pulmonary microvascular endothelial cells (PMVECs) and contributes to increased filtration coefficient in the isolated perfused lung. These findings indicate that cUMP contributes to endothelial permeability during P. aeruginosa lung infection.NEW & NOTEWORTHY During pneumonia, bacteria utilize a virulence arsenal to communicate with host cells. The Pseudomonas aeruginosa T3SS directly introduces virulence molecules into the host cell cytoplasm. These molecules are enzymes that trigger interkingdom communication. One of the exoenzymes is a nucleotidyl cyclase that produces noncanonical cyclic nucleotides like cUMP. Little is known about how cUMP acts in the cell. Here we found that cUMP instigates pulmonary edema during Pseudomonas aeruginosa infection of the lung.

Lipidomic and metabolomic changes in community-acquired and COVID-19 pneumonia

This prospective observational study compared the 1H NMR blood lipidomes and metabolomes of 71 patients with community-acquired pneumonia (CAP), 75 patients with COVID-19 pneumonia, and 75 healthy controls (matched by age and sex) to identify potential biomarkers and pathways associated with respiratory infections. Both pneumonia groups had comparable severity indices, including mortality, invasive mechanical ventilation, and intensive care unit admission rates. Patients with COVID-19 pneumonia exhibited more pronounced hypolipidemia, with significantly lower levels of total cholesterol and LDL-c compared to patients with CAP. Atherogenic lipoprotein subclasses (VLDL-cholesterol, IDL-cholesterol, IDL-triglyceride, and LDL-triglyceride/LDL-cholesterol) were significantly increased in severe cases of both pneumonia types, while lower HDL-c and small, dense HDL particles were associated with more severe illness. Both infected groups showed decreased esterified cholesterol and increased triglycerides, along with reduced phosphatidylcholine, lysophosphatidylcholine, PUFA, omega-3 fatty acids, and DHA. Additionally, infected patients had elevated levels of glucose, lactate, 3-hydroxybutyrate, and acetone, which are linked to inflammation, hypoxemia, and sepsis. Increased levels of branched-chain amino acids, alanine, glycine, and creatine, which are involved in energy metabolism and protein catabolism, were also observed. Neurotransmitter synthesis metabolites like histidine and glutamate were higher in infected patients, especially those with COVID-19. Notably, severe infections showed a significant decrease in glutamine, essential for lymphocyte and macrophage energy. The severity of COVID-19 pneumonia was also associated with elevated glycoprotein levels (glycoprotein A, glycoprotein B, and glycoprotein F), indicating an inflammatory state. These findings suggest that metabolomic and lipidomic changes in pneumonia are connected to bioenergetic pathways regulating the immune response.

Clinical utility of metagenomic next-generation sequencing on bronchoalveolar lavage fluid in diagnosis of lower respiratory tract infections

BACKGROUND

In this study, we aimed to evaluate the clinical utility of Metagenomic Next-Generation sequencing (mNGS) on bronchoalveolar lavage fluid (BALF) in diagnosis of Lower Respiratory Tract Infections (LRTIs).

METHODS

In this study, we retrospectively analyzed 186 hospitalized patients who were suspected with LRTIs and performed mNGS (DNA) test of BALF simultaneously at The Fifth Affiliated Hospital of Sun Yat-Sen University from March 2023 to August 2023. Suspected LRTI was based on LRTI related clinical manifestations or imaging examination. Among them, 155 patients had undergone conventional culture and mNGS (DNA) testing simultaneously. Finally, 138 cases (89.03%,138/155) were diagnosed as LRTI and 17 cases (10.97%,17/155) were diagnosed as non-LRTI. Both detecting rate and diagnostic efficacy of mNGS and conventional culture were compared.

RESULTS

The positive detection rates of pathogens between mNGS and conventional culture were significant different (81.29% VS 39.35%, P < 0.05). Compared with paired conventional culture result, the sensitivity of mNGS in diagnosis of LRTIs was more superior (88.41% VS 43.48%; P < 0.05), the specificity was opposite (76.47% VS 94.12%; P > 0.05). Furthermore, 77.54% and 35.51% of LRTI cases were being etiologically diagnosed by mNGS and culture respectively. Importantly, mNGS directly led to a change of treatment regimen in 58 (37.42%) cases, including antibiotic adjustment (29.68%) and ruling out active infection (7.74%). Moreover, treatment regimen remained unchanged in 97 (62.58%) cases, considering the current antibiotic therapy already covered the detected pathogens (36.13%) or empirical treatment was effective (11.61%).

CONCLUSIONS

mNGS can identify a wide range of pathogens in LRTIs, with improved sensitivity and being more superior at diagnosing LRTIs etiologically. mNGS has the potential to enhance clinical outcomes by optimizing the treatment regimens.

Role of lifestyle factors on the development and long-term prognosis of pneumonia and cardiovascular disease in the Chinese population

BACKGROUND

Whether adherence to a healthy lifestyle is associated with a lower risk of developing pneumonia and a better long-term prognosis remains unclear. This study aimed to investigate associations of individual and combined lifestyle factors (LFs) with the incidence risk and long-term prognosis of pneumonia hospitalization.

METHODS

Using data from the China Kadoorie Biobank study, we used the multistate models to investigate the role of five high-risk LFs, including smoking, excessive alcohol drinking, unhealthy dietary habits, physical inactivity, and unhealthy body shape, alone or in combination in the transitions from a generally healthy state at baseline to pneumonia hospitalization or cardiovascular disease (CVD, regarded as a reference outcome), and subsequently to mortality.

RESULTS

Most of the five high-risk LFs were associated with increased risks of transitions from baseline to pneumonia and from pneumonia to death, but with different risk estimates. The greater the number of high-risk LFs, the higher the risk of developing pneumonia and long-term mortality risk after pneumonia, with the strength of associations comparable to that of LFs and CVD. Compared to participants with 0-1 high-risk LF, the adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for transitions from baseline to pneumonia and from pneumonia to death in those with five high-risk LFs were 1.43 (1.28-1.60) and 1.98 (1.61-2.42), respectively. Correspondingly, the respective HRs (95% CIs) for transitions from baseline to CVD and from CVD to death were 2.00 (1.89-2.11) and 1.44 (1.30-1.59), respectively. The risk estimates changed slightly when further adjusting for the presence of major chronic diseases.

CONCLUSION

In this Chinese population, unhealthy LFs were associated with an increased incidence and long-term mortality risk of pneumonia.

Association of residential greenness with incident pneumonia: A prospective cohort study

Residential greenness is considered beneficial to human health, and its association with respiratory function has been found in previous studies. However, its link with pneumonia remains unclear. To explore the association of residential greenness with incident pneumonia, we conducted a prospective cohort study based on participants of the UK Biobank, followed from 2006 to 2010 to the end of 2019. Residential greenness was measured by Normalized Difference Vegetation Index (NDVI) within 500 m and 1000 m buffer. Cox proportional hazard models were conducted to assess the association, and restricted cubic spline models were also constructed to estimate their exposure-response relationship. Results demonstrate that residential greenness was negatively related to the risk of incident pneumonia. An interquartile (IQR) increase in NDVI 500-m buffer was associated with 4 % [HR (95 % CI) =0.96 (0.94, 0.97), P < 0.001] lower risk of incident pneumonia. Compared to the lowest greenness quartile (Q1), the highest quartile (Q4) had a lower risk of incident pneumonia, with the HR (95 % CI) estimated to be 0.91 (0.87, 0.95) (P values <0.001). Analyses based on NDVI 1000-m buffer obtained similar results. Furthermore, a significant effect of modifications by age and income on the linkage between residential greenness and incident pneumonia was found. These findings propose a potential effective prevention of incident pneumonia and provide the scientific basis for promoting the construction of residential greenness.

Rapid Molecular Diagnostics of Pneumonia Caused by Gram-Negative Bacteria: A Clinician's Review

With approximately half a billion events per year, lower respiratory tract infections (LRTIs) represent a major challenge for the global public health. Among LRTI cases, those caused by Gram-negative bacteria (GNB) are associated with a poorer prognostic. Standard-of-care etiologic diagnostics is lengthy and difficult to establish, with more than half of cases remaining microbiologically undocumented. Recently, syndromic molecular diagnostic panels became available, enabling simultaneous detection of tens of pathogen-related and antimicrobial-resistance genetic markers within a few hours. In this narrative review, we summarize the available data on the performance of molecular diagnostics in GNB pneumonia, highlighting the main strengths and limitations of these assays, as well as the main factors influencing their clinical utility. We searched MEDLINE and Web of Science databases for relevant English-language articles. Molecular assays have higher analytical sensitivity than cultural methods, and show good agreement with standard-of-care diagnostics regarding detection of respiratory pathogens, including GNB, and identification of frequent patterns of resistance to antibiotics. Clinical trials reported encouraging results on the usefulness of molecular assays in antibiotic stewardship. By providing early information on the presence of pathogens and their probable resistance phenotypes, these assays assist in the choice of targeted therapy, in shortening the time from sample collection to appropriate antimicrobial treatment, and in reducing unnecessary antibiotic use.

Esculin alleviates lipopolysaccharide (LPS)-induced pneumonia by regulating the USP7/MAPK14 axis

Pneumonia is a serious and life-threatening lung inflammation with high morbidity and mortality. Accumulating evidence has suggested that esculin, a derivative of coumarin, possesses potent anti-inflammatory effects. This study is designed to explore the pharma role and underlying mechanism of esculin against lipopolysaccharides (LPS)-induced pneumonia. TC-1 cells were stimulated by LPS to mimic the inflammatory injury model in vitro. Cell viability, proliferation, and apoptosis were determined using MTT assay, 5-ethynyl-2'-deoxyuridine assay, and flow cytometry. Interleukin-1β and tumor necrosis factor α levels were analyzed using an enzyme-linked immunosorbent assay. Reactive oxygen species and superoxide dismutase were examined using special assay kits. Macrophage polarization was detected using flow cytometry. Mitogen-activated protein kinase 14 (MAPK14) level was detected by real-time quantitative polymerase chain reaction. MAPK14 and ubiquitin-specific protease 7 (USP7) protein levels were determined using western blot assay. After Ubibrowser database prediction, the interaction between USP7 and MAPK14 was verified using a Co-immunoprecipitation assay. The biological role of esculin was verified in LPS-challenged ALI mice in vivo. Here, we found that esculin significantly relieved LPS-induced TC-1 cell proliferation inhibition, and apoptosis, inflammatory response, oxidative stress, and M1-type macrophage polarization promotion. MAPK14 and USP7 expressions were enhanced in LPS-treated TC-1 cells, which was partly abolished by esculin treatment. Overexpressing MAPK14 attenuated the repression of esculin on LPS-triggered TC-1 cell injury. At the molecular level, USP7 interacted with MAPK14 and maintained its stability by removing ubiquitin. Moreover, esculin repressed the progression of pneumonia in vivo by regulating MAPK14. Taken together, esculin exposure could mitigate LPS-induced TC-1 cell injury partly by targeting the USP7/MAPK14 axis, providing a better understanding of the role of esculin in the anti-inflammatory therapeutics for pneumonia.

Ensemble of Deep Learning Architectures with Machine Learning for Pneumonia Classification Using Chest X-rays

Pneumonia is a severe health concern, particularly for vulnerable groups, needing early and correct classification for optimal treatment. This study addresses the use of deep learning combined with machine learning classifiers (DLxMLCs) for pneumonia classification from chest X-ray (CXR) images. We deployed modified VGG19, ResNet50V2, and DenseNet121 models for feature extraction, followed by five machine learning classifiers (logistic regression, support vector machine, decision tree, random forest, artificial neural network). The approach we suggested displayed remarkable accuracy, with VGG19 and DenseNet121 models obtaining 99.98% accuracy when combined with random forest or decision tree classifiers. ResNet50V2 achieved 99.25% accuracy with random forest. These results illustrate the advantages of merging deep learning models with machine learning classifiers in boosting the speedy and accurate identification of pneumonia. The study underlines the potential of DLxMLC systems in enhancing diagnostic accuracy and efficiency. By integrating these models into clinical practice, healthcare practitioners could greatly boost patient care and results. Future research should focus on refining these models and exploring their application to other medical imaging tasks, as well as including explainability methodologies to better understand their decision-making processes and build trust in their clinical use. This technique promises promising breakthroughs in medical imaging and patient management.

Impact of early myocardial injury on patients with severe pneumonia

Pneumonia often causes myocardial damage. This study sought to understand how early myocardial injury affects severe pneumonia patients' prognoses. This multi-center prospective cohort study from March 2020 to October 2023 comprised severe pneumonia patients. Binary logistic regression analysis examined how myocardial damage affects cardiac complications and acute renal injury (AKI). We used Spearman correlation analysis to examine the relationship between troponin I levels and the vasoactive inotropic score (VIS) in shock patients with myocardial injury. We used the Kaplan-Meier survival curve to evaluate the impact of myocardial injury on 30-day and 1-year survival rates. Mediation investigations examined how AKI and cardiac complications mediate myocardial injury and death. This study included 363 severe pneumonia patients, of whom 204 (56.2%) developed myocardial damage, 132 (36.4%) had cardiac problems, and 146 (40.2%) had AKI. Myocardial damage independently elevated the incidence of cardiac complications (OR = 2.548, 95% CI = 1.404-4.303, P = 0.002) and AKI (OR = 1.946, 95% CI = 1.177-3.219, P = 0.009). There was a positive link between troponin I and VIS in myocardial injury and shock patients (r = 0.43, P < 0.001). COX regression found myocardial injury to be a death risk (HR = 1.472, 95% CI = 1.043-2.077, P = 0.028). Adjusted Kaplan-Meier survival analysis showed significantly decreased short-term and long-term survival rates with myocardial injury (log-rank test P < 0.05). The mediation study showed that cardiac complications and AKI mediated myocardial injury and death by 19.30% and 17.18%, respectively. Early myocardial injury in severe pneumonia patients raises the likelihood of cardiac problems, AKI, and refractory shock, reducing short- and long-term survival.

The diagnostic value of bronchoalveolar lavage fluid metagenomic next-generation sequencing in critically ill patients with respiratory tract infections

Metagenomic next-generation sequencing (mNGS) is an unbiased and rapid method for detecting pathogens. This study enrolled 145 suspected severe pneumonia patients who were admitted to the Affiliated Hospital of Jining Medical University. This study primarily aimed to determine the diagnostic performance of mNGS and conventional microbiological tests (CMTs) using bronchoalveolar lavage fluid samples for detecting pathogens. Our findings indicated that mNGS performed significantly higher sensitivity (97.54% vs 28.68%, P < 0.001), coincidence (90.34% vs 35.17%, P < 0.001), and negative predictive value (80.00% vs 13.21%, P < 0.001) but performed lower specificity than CMTs (52.17% vs 87.5%, P < 0.001). Streptococcus pneumoniae as the most common bacterial pathogen had the largest proportion (22.90%, 30/131) in this study. In addition to bacteria, fungi, and virus, mNGS can detect a variety of atypical pathogens such as Mycobacterium tuberculosis and non-tuberculous. Mixed infections were common in patients with severe pneumonia, and bacterial-fungal-viral-atypical pathogens were the most complicated infection. After adjustments of antibiotics based on mNGS and CMTs, the clinical manifestation improved in 139 (95.86%, 139/145) patients. Our data demonstrated that mNGS had significant advantage in diagnosing respiratory tract infections, especially atypical pathogens and fungal infections. Pathogens were detected timely and comprehensively, contributing to the adjustments of antibiotic treatments timely and accurately, improving patient prognosis and decreasing mortality potentially.IMPORTANCEMetagenomic next-generation sequencing using bronchoalveolar lavage fluid can provide more comprehensive and accurate pathogens for respiratory tract infections, especially when considering the previous usage of empirical antibiotics before admission or complicated clinical presentation. This technology is expected to play an important role in the precise application of antimicrobial drugs in the future.

Causal relationship between air pollution and infections: a two-sample Mendelian randomization study

Background

Traditional observational studies exploring the association between air pollution and infections have been limited by small sample sizes and potential confounding factors. To address these limitations, we applied Mendelian randomization (MR) to investigate the potential causal relationships between particulate matter (PM2.5, PM2.5-10, and PM10), nitrogen dioxide, and nitrogen oxide and the risks of infections.

Methods

Single nucleotide polymorphisms (SNPs) related to air pollution were selected from the genome-wide association study (GWAS) of the UK Biobank. Publicly available summary data for infections were obtained from the FinnGen Biobank and the COVID-19 Host Genetics Initiative. The inverse variance weighted (IVW) meta-analysis was used as the primary method for obtaining the Mendelian randomization (MR) estimates. Complementary analyses were performed using the weighted median method, MR-Egger method, and MR Pleiotropy Residual Sum and Outlier (MR-PRESSO) test.

Results

The fixed-effect IVW estimate showed that PM2.5, PM2.5-10 and Nitrogen oxides were suggestively associated with COVID-19 [for PM2.5: IVW (fe): OR 3.573(1.218,5.288), PIVW(fe) = 0.021; for PM2.5-10: IVW (fe): OR 2.940(1.385,6.239), PIVW(fe) = 0.005; for Nitrogen oxides, IVW (fe): OR 1.898(1.318,2.472), PIVW(fe) = 0.010]. PM2.5, PM2.5-10, PM10, and Nitrogen oxides were suggestively associated with bacterial pneumonia [for PM2.5: IVW(fe): OR 1.720 (1.007, 2.937), PIVW(fe) = 0.047; for PM2.5-10: IVW(fe): OR 1.752 (1.111, 2.767), P IVW(fe) = 0.016; for PM10: IVW(fe): OR 2.097 (1.045, 4.208), PIVW(fe) = 0.037; for Nitrogen oxides, IVW(fe): OR 3.907 (1.209, 5.987), PIVW(fe) = 0.023]. Furthermore, Nitrogen dioxide was suggestively associated with the risk of acute upper respiratory infections, while all air pollution were not associated with intestinal infections.

Conclusions

Our results support a role of related air pollution in the Corona Virus Disease 2019, bacterial pneumonia and acute upper respiratory infections. More work is need for policy formulation to reduce the air pollution and the emission of toxic and of harmful gas.

A review of the value of point-of-care testing for community-acquired pneumonia

INTRODUCTION

Community-acquired pneumonia (CAP) is an infectious disease associated with high mortality worldwide. Although Streptococcus pneumoniae remains the most frequent pathogen in CAP, data from recent studies using molecular tests have shown that respiratory viruses play a key role in adults with pneumonia. The impact of difficult-to-treat pathogens on the outcomes of pneumonia is also important even though they represent only a small proportion of overall cases. Despite improvements in the microbiological diagnosis of CAP in recent decades, the identification of the causative pathogen is often delayed because of difficulties in obtaining good-quality sputum samples, issues in transporting samples, and slow laboratory processes. Therefore, the initial treatment of CAP is usually empirical. Point-of-care testing (POCT) was introduced to avoid treatment delays and reduce reliance on empirical antibiotics.

AREAS COVERED

This review summarizes the main scientific evidence on the role of POCT in the diagnosis and management of patients with CAP. The authors searched for articles on POCT in pneumonia on PubMed from inception to 20 January 2024. The references in the identified articles were also searched.

EXPERT OPINION

POCT involves rapid diagnostic assays that can be performed at the bedside especially in cases of severe CAP and immunocompromised patients. These tests can produce results that could help guide initial therapy and management.

Prediction of new-onset atrial fibrillation with the C2HEST score in patients admitted with community-acquired pneumonia

PURPOSE

Patients hospitalized for community-acquired pneumonia (CAP) may have a higher risk of new-onset atrial fibrillation (NOAF). The C2HEST score was developed to evaluate the NOAF risk in the general population. Data on the value of the C2HEST score in acute patients admitted with CAP are lacking. We want to establish the predictive value of C2HEST score for NOAF in patients with CAP.

METHODS

Patients with CAP enrolled in the SIXTUS cohort were enrolled. C2HEST score was calculated at baseline. In-hospital NOAF was recorded. Receiver-operating Characteristic (ROC) curve and multivariable Cox proportional hazard regression analysis were performed.

RESULTS

We enrolled 473 patients (36% women, mean age 70.6 ± 16.5 years), and 54 NOAF occurred. Patients with NOAF were elderly, more frequently affected by hypertension, heart failure, previous stroke/transient ischemic attack, peripheral artery disease and hyperthyroidism. NOAF patients had also higher CURB-65, PSI class and CHA2DS2-VASc score. The C-index of C2HEST score for NOAF was 0.747 (95% confidence interval [95%CI] 0.705-0.786), higher compared to CURB-65 (0.611, 95%CI 0.566-0.655, p = 0.0016), PSI (0.665, 95%CI 0.621-0.708, p = 0.0199) and CHA2DS2-VASc score (0.696, 95%CI 0.652-0.737, p = 0.0762). The best combination of sensitivity (67%) and specificity (70%) was observed with a C2HEST score ≥ 4. This result was confirmed by the multivariable Cox analysis (Hazard Ratio [HR] for C2HEST score ≥ 4 was 10.7, 95%CI 2.0-57.9; p = 0.006), independently from the severity of pneumonia.

CONCLUSION

The C2HEST score was a useful predictive tool to identify patients at higher risk for NOAF during hospitalization for CAP.

CLINICAL TRIAL REGISTRATION

www.

CLINICALTRIALS

gov (NCT01773863).

Ventilator-associated pneumonia: pathobiological heterogeneity and diagnostic challenges

Ventilator-associated pneumonia (VAP) affects up to 20% of critically ill patients and induces significant antibiotic prescription pressure, accounting for half of all antibiotic use in the ICU. VAP significantly increases hospital length of stay and healthcare costs yet is also associated with long-term morbidity and mortality. The diagnosis of VAP continues to present challenges and pitfalls for the currently available clinical, radiological and microbiological diagnostic armamentarium. Biomarkers and artificial intelligence offer an innovative potential direction for ongoing future research. In this Review, we summarise the pathobiological heterogeneity and diagnostic challenges associated with VAP.

Unravelling the complexity of ventilator-associated pneumonia: a systematic methodological literature review of diagnostic criteria and definitions used in clinical research

BACKGROUND

Ventilator-associated pneumonia (VAP) is a prevalent and grave hospital-acquired infection that affects mechanically ventilated patients. Diverse diagnostic criteria can significantly affect VAP research by complicating the identification and management of the condition, which may also impact clinical management.

OBJECTIVES

We conducted this review to assess the diagnostic criteria and the definitions of the term "ventilator-associated" used in randomised controlled trials (RCTs) of VAP management.

SEARCH METHODS

Based on the protocol (PROSPERO 2019 CRD42019147411), we conducted a systematic search on MEDLINE/PubMed and Cochrane CENTRAL for RCTs, published or registered between 2010 and 2024.

SELECTION CRITERIA

We included completed and ongoing RCTs that assessed pharmacological or non-pharmacological interventions in adults with VAP.

DATA COLLECTION AND SYNTHESIS

Data were collected using a tested extraction sheet, as endorsed by the Cochrane Collaboration. After cross-checking, data were summarised in a narrative and tabular form.

RESULTS

In total, 7,173 records were identified through the literature search. Following the exclusion of records that did not meet the eligibility criteria, 119 studies were included. Diagnostic criteria were provided in 51.2% of studies, and the term "ventilator-associated" was defined in 52.1% of studies. The most frequently included diagnostic criteria were pulmonary infiltrates (96.7%), fever (86.9%), hypothermia (49.1%), sputum (70.5%), and hypoxia (32.8%). The different criteria were used in 38 combinations across studies. The term "ventilator-associated" was defined in nine different ways.

CONCLUSIONS

When provided, diagnostic criteria and definitions of VAP in RCTs display notable variability. Continuous efforts to harmonise VAP diagnostic criteria in future clinical trials are crucial to improve quality of care, enable accurate epidemiological assessments, and guide effective antimicrobial stewardship.

A multimodal integration pipeline for accurate diagnosis, pathogen identification, and prognosis prediction of pulmonary infections

Pulmonary infections pose formidable challenges in clinical settings with high mortality rates across all age groups worldwide. Accurate diagnosis and early intervention are crucial to improve patient outcomes. Artificial intelligence (AI) has the capability to mine imaging features specific to different pathogens and fuse multimodal features to reach a synergistic diagnosis, enabling more precise investigation and individualized clinical management. In this study, we successfully developed a multimodal integration (MMI) pipeline to differentiate among bacterial, fungal, and viral pneumonia and pulmonary tuberculosis based on a real-world dataset of 24,107 patients. The area under the curve (AUC) of the MMI system comprising clinical text and computed tomography (CT) image scans yielded 0.910 (95% confidence interval [CI]: 0.904-0.916) and 0.887 (95% CI: 0.867-0.909) in the internal and external testing datasets respectively, which were comparable to those of experienced physicians. Furthermore, the MMI system was utilized to rapidly differentiate between viral subtypes with a mean AUC of 0.822 (95% CI: 0.805-0.837) and bacterial subtypes with a mean AUC of 0.803 (95% CI: 0.775-0.830). Here, the MMI system harbors the potential to guide tailored medication recommendations, thus mitigating the risk of antibiotic misuse. Additionally, the integration of multimodal factors in the AI-driven system also provided an evident advantage in predicting risks of developing critical illness, contributing to more informed clinical decision-making. To revolutionize medical care, embracing multimodal AI tools in pulmonary infections will pave the way to further facilitate early intervention and precise management in the foreseeable future.

Effects of lung inflammation and injury on pulmonary tissue penetration of meropenem and vancomycin in a model of unilateral lung injury

OBJECTIVE

The timing and dosing of antimicrobial therapy are key in the treatment of pneumonia in critically ill patients. It is uncertain whether the presence of lung inflammation and injury affects tissue penetration of intravenously administered antimicrobial drugs. The effects of lung inflammation and injury on tissue penetration of two antimicrobial drugs commonly used for pneumonia were determined in an established model of unilateral lung injury.

METHODS

Unilateral lung injury was induced in the left lung of 13 healthy pigs through cyclic rinsing; the right healthy lung served as control. Infusions of meropenem and vancomycin were administered and concentrations of these drugs in lung tissue, blood, and epithelial lining fluid (ELF) were compared over a period of 6 h.

RESULTS

Median vancomycin lung tissue concentrations and penetration ratio were higher in inflamed and injured lungs compared with uninflamed and uninjured lungs (AUC0-6h: P = 0.003 and AUCdialysate/AUCplasma ratio: P = 0.003), resulting in higher AUC0-24/MIC. Median meropenem lung tissue concentrations and penetration ratio in inflamed and injured lungs did not differ from that in uninflamed and uninjured lungs (AUC0-6: P = 0.094 and AUCdialysate/AUCplasma ratio: P = 0.173). The penetration ratio for both vancomycin and meropenem into ELF was similar in injured and uninjured lungs.

CONCLUSION

Vancomycin penetration into lung tissue is enhanced by acute inflammation and injury, a phenomenon barely evident with meropenem. Therefore, inflammation in lung tissue influences the penetration into interstitial lung tissue, depending on the chosen antimicrobial drug. Measurement of ELF levels alone might not identify the impact of inflammation and injury.

Ventilator-associated pneumonia - What price does the public health system pay?

BACKGROUND

Ventilator-associated pneumonia (VAP) is the commonest healthcare-associated infection (HAI) in intensive care units (ICU), especially in trauma patients. VAP imposes a significant cost burden on the healthcare ecosystem. However, there are few data from the developing world.

METHODOLOGY

We conducted this study in the trauma ICU (TICU) of PGIMER, Chandigarh, from October 2021 to December 2022. The incidence, incidence density, and average length of stay (ALOS) of both VAP and non-VAP patients were established. The health system cost was assessed using a mixed (top-down and bottom-up) micro-costing approach. We collected data for all the resources (direct and indirect costs) utilized during service delivery and estimated the health system cost per bed per day.

RESULTS

In this study, 494 patients were admitted to TICU, of which 484 received Mechanical Ventilation (MV) and 47 developed VAP. We included 41 and 44 patients with and without VAP. The VAP incidence rate was 9.7% and the VAP incidence density was 10.79/1000 MV days. The ALOS for VAP patients was 21 days, and for non- VAP patients was 8.2 days. Our study estimated a total health system cost of INR 25,927 per bed per day. The health system cost of treating a VAP patient was INR 544,467 compared to INR 207,416 for a non-VAP patient.

CONCLUSION

Treatment of VAP poses substantial costs for the health system and patients. There is a need to focus on preventing VAP, which would eventually reduce the length of stay and the resultant financial impact on the health system and the patient.

Influence of non-pharmaceutical interventions during the COVID-19 pandemic on respiratory viral infections - a prospective population-based cohort study

Non-pharmaceutical interventions (NPI) have been proven successful in a population-based approach to protect from SARS-CoV-2 transmission during the COVID-19 pandemic. As a consequential-effect, a reduction in the spread of all respiratory viruses has been observed, but the primary factors behind this phenomenon have yet to be identified. We conducted a subgroup analysis of participants from the ELISA study, a prospective longitudinal cohort study on SARS-CoV-2 transmission, at four timepoints from November 2020 - September 2022. The aim was to provide a detailed overview of the circulation of respiratory viruses over 2 years and to identify potential personal risk factors of virus distribution. All participants were screened using qPCR for respiratory viral infections from nasopharyngeal swabs and answered a questionnaire regarding behavioral factors. Several categories of risk factors for the transmission of respiratory viruses were evaluated using a scoring system. In total, 1,124 participants were included in the study, showing high adherence to governmental-introduced NPI. The overall number of respiratory virus infections was low (0-4.9% of participants), with adenovirus (1.7%), rhino-/enterovirus (3.2%) and SARS-CoV-2 (1.2%) being the most abundant. We detected an inverse correlation between the number and intensity of NPI and the number of detected respiratory viruses. More precisely, the attendance of social events and household size was associated with rhino-/enterovirus infection while social contacts were associated with being positive for any virus. NPI introduced during the COVID-19 pandemic reduced the occurrence of seasonal respiratory viruses in our study, showing different risk-factors for enhanced transmission between viruses.

Trial registration

DRKS.de, German Clinical Trials Register (DRKS), Identifier: DRKS00023418, Registered on 28 October 2020.

An explainable machine learning-based model to predict intensive care unit admission among patients with community-acquired pneumonia and connective tissue disease

BACKGROUND

There is no individualized prediction model for intensive care unit (ICU) admission on patients with community-acquired pneumonia (CAP) and connective tissue disease (CTD) so far. In this study, we aimed to establish a machine learning-based model for predicting the need for ICU admission among those patients.

METHODS

This was a retrospective study on patients admitted into a University Hospital in China between November 2008 and November 2021. Patients were included if they were diagnosed with CAP and CTD during admission and hospitalization. Data related to demographics, CTD types, comorbidities, vital signs and laboratory results during the first 24 h of hospitalization were collected. The baseline variables were screened to identify potential predictors via three methods, including univariate analysis, least absolute shrinkage and selection operator (Lasso) regression and Boruta algorithm. Nine supervised machine learning algorithms were used to build prediction models. We evaluated the performances of differentiation, calibration, and clinical utility of all models to determine the optimal model. The Shapley Additive Explanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME) techniques were performed to interpret the optimal model.

RESULTS

The included patients were randomly divided into the training set (1070 patients) and the testing set (459 patients) at a ratio of 70:30. The intersection results of three feature selection approaches yielded 16 predictors. The eXtreme gradient boosting (XGBoost) model achieved the highest area under the receiver operating characteristic curve (AUC) (0.941) and accuracy (0.913) among various models. The calibration curve and decision curve analysis (DCA) both suggested that the XGBoost model outperformed other models. The SHAP summary plots illustrated the top 6 features with the greatest importance, including higher N-terminal pro-B-type natriuretic peptide (NT-proBNP) and C-reactive protein (CRP), lower level of CD4 + T cell, lymphocyte and serum sodium, and positive serum (1,3)-β-D-glucan test (G test).

CONCLUSION

We successfully developed, evaluated and explained a machine learning-based model for predicting ICU admission in patients with CAP and CTD. The XGBoost model could be clinical referenced after external validation and improvement.

CT Characteristics and Clinical Findings of Bronchopneumonia Caused by Pepper Aspiration

Purpose

To explore the computed tomography (CT) features of bronchopneumonia caused by pepper aspiration to improve the diagnosis.

Materials and Methods

28 adult patients diagnosed with obstructive pneumonia caused by pepper aspiration from January 2016 to September 2022 were enrolled. The CT characteristics of bronchial changes and pulmonary lesions caused by pepper were analyzed and summarized.

Results

Among 28 patients, the most common symptom was cough (26, 92.9%), followed by expectoration (23, 82.1%). Bronchoscopy revealed that peppers were mainly found in the bronchus of the right lower lobe (n = 18, 64.3%), followed by the bronchus of the left lower lobe (n = 5, 17.9%). In combination with bronchoscopy results, the pepper in the bronchus manifested as circular or V/U-shaped high-density, localized soft tissue, and flocculent opacification in 8 (28.6%), 16 (57.1%), and 3 (10.7%) cases on CT images, respectively. The bronchial wall around the pepper was thickened with localized occlusion (n = 19, 67.9%) and stenosis (n = 9, 32.1%). Regarding adjacent bronchi without peppers, extensive wall thickening with stenosis and/or occlusion was found in 23 (82.1%) cases. Distal pulmonary lesions frequently involved two or three segments (21, 75.0%) and mainly presented as patchy consolidation or atelectasis (24, 85.7%).

Conclusion

In combination to a history of eating peppers and clinical symptoms, bronchopneumonia caused by pepper should be highly suspected if U/V-shaped and annular high-density or localized soft tissue density is detected in the bronchi of the lower lobes, accompanied by extensive bronchial wall thickening, stenosis, or occlusion, and consolidation or atelectasis in multiple distal lung segments.