1
|
Li J, Zhang Y, He S, Tang Y. Interpretable mortality prediction model for ICU patients with pneumonia: using shapley additive explanation method. BMC Pulm Med 2024; 24:447. [PMID: 39272037 PMCID: PMC11395639 DOI: 10.1186/s12890-024-03252-x] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/15/2023] [Accepted: 08/29/2024] [Indexed: 09/15/2024] Open
Abstract
BACKGROUND Pneumonia, a leading cause of morbidity and mortality worldwide, often necessitates Intensive Care Unit (ICU) admission. Accurate prediction of pneumonia mortality is crucial for tailored prevention and treatment plans. However, existing mortality prediction models face limited adoption in clinical practice due to their lack of interpretability. OBJECTIVE This study aimed to develop an interpretable model for predicting pneumonia mortality in ICUs. Leveraging the Shapley Additive Explanation (SHAP) method, we sought to elucidate the Extreme Gradient Boosting (XGBoost) model and identify prognostic factors for pneumonia. METHODS Conducted as a retrospective cohort study, we utilized electronic health records from the eICU-CRD (2014-2015) for all adult pneumonia patients. The first 24 h of each ICU admission records were considered, with 70% of the dataset allocated for model training and 30% for validation. The XGBoost model was employed, and performance was assessed using the area under the receiver operating characteristic curve (AUC). The SHAP method provided insights into the XGBoost model. RESULTS Among 10,962 pneumonia patients, in-hospital mortality was 16.33%. The XGBoost model demonstrated superior predictive performance (AUC: 0.778 ± 0.016)) compared to traditional scoring systems and other machine learning method, which achieved an improvement of 10% points. SHAP analysis identified Aspartate Aminotransferase (AST) as the most crucial predictor. CONCLUSIONS Interpretable predictive models enhance mortality risk assessment for pneumonia patients in the ICU, fostering transparency. AST emerged as the foremost predictor, followed by patient age, albumin, BMI et al. These insights, rooted in strong correlations with mortality, facilitate improved clinical decision-making and resource allocation.
Collapse
Affiliation(s)
- Jiaxi Li
- Department of Clinical Laboratory Medicine, Jinniu Maternity and Child Health Hospital of Chengdu, Chengdu, China
| | - Yu Zhang
- Information Center, West China Hospital, Sichuan University, Chengdu, China
| | - ShengYang He
- Department of Clinical Laboratory Medicine, Jinniu Maternity and Child Health Hospital of Chengdu, Chengdu, China
| | - Yan Tang
- Department of Clinical Laboratory Medicine, Jinniu Maternity and Child Health Hospital of Chengdu, Chengdu, China.
| |
Collapse
|
2
|
Roedl K, Fuhrmann V. [Liver diseases in the intensive care unit]. Med Klin Intensivmed Notfmed 2024; 119:449-457. [PMID: 38937335 DOI: 10.1007/s00063-024-01157-5] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/22/2024] [Accepted: 06/04/2024] [Indexed: 06/29/2024]
Abstract
The frequency of liver diseases in the intensive care unit has increased significantly in recent years and is now observed in up to 20% of critically ill patients. The occurrence of liver disease is associated with significantly increased morbidity and mortality. Two groups of liver diseases in the intensive care unit can be distinguished. First, the group of "primary hepatic dysfunctions", which includes primary acute liver failure as well as acute-on-chronic liver failure in patients with pre-existing liver cirrhosis. The second group of "secondary or acquired liver diseases" includes cholestatic liver diseases, as well as hypoxic liver injury and mixed forms, as well as other rarer liver diseases. Due to the diversity of liver diseases and the very different triggers, sufficient knowledge of the underlying changes (including hemodynamic changes, inflammatory states or drug-related) is essential. Early recognition, diagnosis, and treatment of the underlying disease are essential for all liver dysfunction in critically ill patients in the intensive care unit. This review article aims to take a closer look at liver diseases in the intensive care unit and provides insight into diagnostics and treatment options.
Collapse
Affiliation(s)
- Kevin Roedl
- Klinik für Intensivmedizin, Universitätsklinikum Hamburg-Eppendorf, Martinistraße 52, 20246, Hamburg, Deutschland.
| | - Valentin Fuhrmann
- Abteilung für Innere Medizin und Gastroenterologie, Heilig-Geist-Krankenhaus, Köln, Deutschland
| |
Collapse
|
3
|
Zheng L, Ling W, Zhu D, Li Z, Li Y, Zhou H, Kong L. Roquin-1 resolves sepsis-associated acute liver injury by regulating inflammatory profiles via miRNA cargo in extracellular vesicles. iScience 2023; 26:107295. [PMID: 37554446 PMCID: PMC10405074 DOI: 10.1016/j.isci.2023.107295] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/29/2022] [Revised: 01/05/2023] [Accepted: 07/03/2023] [Indexed: 08/10/2023] Open
Abstract
Sepsis-associated acute liver injury (SALI) is an independent risk for sepsis-induced death orchestrated by innate and adaptive immune responses. Here, we found that Roquin-1 was decreased during SALI and expressed mainly in monocyte-derived macrophages. Meanwhile, Roquin-1 was correlated with the inflammatory profiles in humans and mice. Mechanically, Roquin-1 in macrophages promoted Ago2-K258-ubiquitination and inhibited Ago2-S387/S828-phosphorylation. Ago2-S387-phosphorylation inhibited Ago2-miRNA's complex location in multivesicular bodies and sorting in macrophages-derived extracellular vesicles (MDEVs), while Ago2-S828-phosphorylation modulated the binding between Ago2 and miRNAs by special miRNAs-motifs. Then, the anti-inflammatory miRNAs in MDEVs decreased TSC22D2 expression directly, upregulated Tregs-differentiation via TSC22D2-STAT3 signaling, and inhibited M1-macrophage-polarization by TSC22D2-AMPKα-mTOR pathway. Furthermore, WT MDEVs in mice alleviated SALI by increasing Tregs ratio and decreasing M1-macrophage frequency synchronously. Our study showed that Roquin-1 in macrophages increased Tregs-differentiation and decreased M1-macrophage-polarization simultaneously via miRNA in MDEVs, suggesting Roquin-1 can be used as a potential tool for SALI treatment and MDEVs engineering.
Collapse
Affiliation(s)
- Lei Zheng
- Hepatobiliary Center/Liver Transplantation Center, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, P.R. China
- Department of General Surgery, Shanghai Ninth People’s Hospital, Shanghai Jiao-tong University School of Medicine, 639 Zhizaoju Road, Shanghai 200011, P.R. China
| | - Wei Ling
- Hepatobiliary Center/Liver Transplantation Center, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, P.R. China
| | - Deming Zhu
- Hepatobiliary Center/Liver Transplantation Center, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, P.R. China
| | - Zhi Li
- Hepatobiliary Center/Liver Transplantation Center, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, P.R. China
| | - Yousheng Li
- Department of General Surgery, Shanghai Ninth People’s Hospital, Shanghai Jiao-tong University School of Medicine, 639 Zhizaoju Road, Shanghai 200011, P.R. China
| | - Haoming Zhou
- Hepatobiliary Center/Liver Transplantation Center, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, P.R. China
| | - Lianbao Kong
- Hepatobiliary Center/Liver Transplantation Center, The First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing 210029, P.R. China
| |
Collapse
|
4
|
Immunomodulation by Hemoadsorption—Changes in Hepatic Biotransformation Capacity in Sepsis and Septic Shock: A Prospective Study. Biomedicines 2022; 10:biomedicines10102340. [PMID: 36289602 PMCID: PMC9598581 DOI: 10.3390/biomedicines10102340] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/30/2022] [Revised: 09/13/2022] [Accepted: 09/14/2022] [Indexed: 11/26/2022] Open
Abstract
Background: Sepsis is often associated with liver dysfunction, which is an indicator of poor outcomes. Specific diagnostic tools that detect hepatic dysfunction in its early stages are scarce. So far, the immune modulatory effects of hemoadsorption with CytoSorb® on liver function are unclear. Method: We assessed the hepatic function by using the dynamic LiMAx® test and biochemical parameters in 21 patients with sepsis or septic shock receiving CytoSorb® in a prospective, observational study. Points of measurement: T1: diagnosis of sepsis or septic shock; T2 and T3: 24 h and 48 h after the start of CytoSorb®; T4: 24 h after termination of CytoSorb®. Results: The hepatic biotransformation capacity measured by LiMAx® was severely impaired in up to 95 % of patients. Despite a rapid shock reversal under CytoSorb®, a significant improvement in LiMAx® values appeared from T3 to T4. This decline and recovery of liver function were not reflected by common parameters of hepatic metabolism that remained mostly within the normal range. Conclusions: Hepatic dysfunction can effectively and safely be diagnosed with LiMAx® in ventilated ICU patients under CytoSorb®. Various static liver parameters are of limited use since they do not adequately reflect hepatic dysfunction and impaired hepatic metabolism.
Collapse
|
5
|
Beyer D, Hoff J, Sommerfeld O, Zipprich A, Gaßler N, Press AT. The liver in sepsis: molecular mechanism of liver failure and their potential for clinical translation. Mol Med 2022; 28:84. [PMID: 35907792 PMCID: PMC9338540 DOI: 10.1186/s10020-022-00510-8] [Citation(s) in RCA: 25] [Impact Index Per Article: 12.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/09/2022] [Accepted: 07/13/2022] [Indexed: 12/25/2022] Open
Abstract
Liver failure is a life-threatening complication of infections restricting the host's response to infection. The pivotal role of the liver in metabolic, synthetic, and immunological pathways enforces limits the host's ability to control the immune response appropriately, making it vulnerable to ineffective pathogen resistance and tissue damage. Deregulated networks of liver diseases are gradually uncovered by high-throughput, single-cell resolved OMICS technologies visualizing an astonishing diversity of cell types and regulatory interaction driving tolerogenic signaling in health and inflammation in disease. Therefore, this review elucidates the effects of the dysregulated host response on the liver, consequences for the immune response, and possible avenues for personalized therapeutics.
Collapse
Affiliation(s)
- Dustin Beyer
- Department of Anesthesiology and Intensive Care Medicine, Jena University Hospital, Am Klinikum 1, 07747, Jena, Germany
| | - Jessica Hoff
- Department of Anesthesiology and Intensive Care Medicine, Jena University Hospital, Am Klinikum 1, 07747, Jena, Germany.,Center for Sepsis Control and Care, Jena University Hospital, Bachstr. 18, 07743, Jena, Germany
| | - Oliver Sommerfeld
- Department of Anesthesiology and Intensive Care Medicine, Jena University Hospital, Am Klinikum 1, 07747, Jena, Germany.,Center for Sepsis Control and Care, Jena University Hospital, Bachstr. 18, 07743, Jena, Germany
| | - Alexander Zipprich
- Department of Internal Medicine IV, Jena University Hospital, Am Klinikum 1, 07747, Jena, Germany
| | - Nikolaus Gaßler
- Pathology, Jena University Hospital, Am Klinikum 1, 07747, Jena, Germany
| | - Adrian T Press
- Department of Anesthesiology and Intensive Care Medicine, Jena University Hospital, Am Klinikum 1, 07747, Jena, Germany. .,Center for Sepsis Control and Care, Jena University Hospital, Bachstr. 18, 07743, Jena, Germany. .,Medical Faculty, Friedrich-Schiller-University Jena, Kastanienstr. 1, 07747, Jena, Germany.
| |
Collapse
|
6
|
Chen Q, Liu L, Ni S. Screening of ferroptosis-related genes in sepsis-induced liver failure and analysis of immune correlation. PeerJ 2022; 10:e13757. [PMID: 35923893 PMCID: PMC9341447 DOI: 10.7717/peerj.13757] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/08/2022] [Accepted: 06/29/2022] [Indexed: 01/17/2023] Open
Abstract
Purpose Sepsis-induced liver failure is a kind of liver injury with a high mortality, and ferroptosis plays a key role in this disease. Our research aims to screen ferroptosis-related genes in sepsis-induced liver failure as targeted therapy for patients with liver failure. Methods Using the limma software, we analyzed the differentially expressed genes (DEGs) in the GSE60088 dataset downloaded from the Gene Expression Omnibus (GEO) database. Clusterprofiler was applied for enrichment analysis of DEGs enrichment function. Then, the ferroptosis-related genes of the mice in the FerrDb database were crossed with DEGs. Sepsis mice model were prepared by cecal ligation and perforation (CLP). ALT and AST in the serum of mice were measured using detection kit. The pathological changes of the liver tissues in mice were observed by hematoxylin-eosin (H & E) staining. We detected the apoptosis of mice liver tissues using TUNEL. The expression of Hmox1, Epas1, Sirt1, Slc3a2, Jun, Plin2 and Zfp36 were detected by qRT-PCR. Results DEGs analysis showed 136 up-regulated and 45 down-regulated DEGs. Meanwhile, we found that the up-regulated DEGs were enriched in pathways including the cytokine biosynthesis process while the down-regulated DEGs were enriched in pathways such as organic hydroxy compound metabolic process. In this study, seven genes (Hmox1, Epas1, Sirt1, Slc3a2, Jun, Plin2 and Zfp36) were obtained through the intersection of FerrDb database and DEGs. However, immune infiltration analysis revealed that ferroptosis-related genes may promote the development of liver failure through B cells and natural killer (NK) cells. Finally, it was confirmed by the construction of septic liver failure mice model that ferroptosis-related genes of Hmox1, Slc3a2, Jun and Zfp36 were significantly correlated with liver failure and were highly expressed. Conclusion The identification of ferroptosis-related genes Hmox1, Slc3a2, Jun and Zfp36 in the present study contribute to our understanding of the molecular mechanism of sepsis-induced liver failure, and provide candidate targets for the diagnosis and treatment of the disease.
Collapse
Affiliation(s)
- Qingli Chen
- Department of Emergency Medicine, Lishui City People’s Hospital, Lishui, Zhejiang Province, China
| | - Luxiang Liu
- Department of Infectious Disease, Lishui City People’s Hospital, Lishui, Zhejiang Province, China
| | - Shuangling Ni
- Department of Infectious Disease, Lishui City People’s Hospital, Lishui, Zhejiang Province, China
| |
Collapse
|
7
|
John S, Riessen R, Karagiannidis C, Janssens U, Busch HJ, Kochanek M, Michels G, Hermes C, Buerke M, Kluge S, Baumgärtel M, Braune S, Erbguth F, Fuhrmann V, Lebiedz P, Mayer K, Müller-Werdan U, Oppert M, Sayk F, Sedding D, Willam C, Werdan K. [Core curriculum Medical intensive care medicine of the German Society of Medical Intensive Care and Emergency Medicine (DGIIN)]. Med Klin Intensivmed Notfmed 2021; 116:1-45. [PMID: 33427907 PMCID: PMC7799161 DOI: 10.1007/s00063-020-00765-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 11/26/2020] [Indexed: 11/25/2022]
Abstract
Medical intensive care medicine treats patients with severe, potentially life-threatening diseases covering the complete spectrum of internal medicine. The qualification in medical intensive care medicine requires a broad spectrum of knowledge and skills in medical intensive care medicine, but also in the general field of internal medicine. Both sides of the coin must be taken into account, the treatment with life-sustaining strategies of the acute illness of the patient and also the treatment of patient's underlying chronic diseases. The indispensable foundation of medical intensive care medicine as described in this curriculum includes basic knowledge and skills (level of competence I-III) as well as of behavior and attitudes. This curriculum is primarily dedicated to the internist in advanced training in medical intensive care medicine. However, this curriculum also intends to reach trainers in intensive care medicine and also the German physician chambers with their examiners, showing them which knowledge, skills as well as behavior and attitudes should be taught to trainees according to the education criteria of the German Society of Medical Intensive Care and Emergency Medicine (DGIIN).
Collapse
Affiliation(s)
- S John
- Klinikum Nürnberg-Süd, Medizinische Klinik 8, Abteilung für Internistische Intensivmedizin, Paracelsus Medizinische Privatuniversität, Nürnberg, Deutschland
| | - R Riessen
- Dept. für Innere Medizin, Internistische Intensivstation, Universitätsklinikum Tübingen, Tübingen, Deutschland
| | - C Karagiannidis
- ARDS und ECMO Zentrum Köln-Merheim, Professur für extrakorporale Lungenersatzverfahren der Universität Witten-Herdecke, Abteilung Pneumologie, Intensiv- und Beatmungsmedizin, Kliniken der Stadt Köln gGmbH, Köln, Deutschland
| | - U Janssens
- Klinik für Innere Medizin und Internistische Intensivmedizin, St.-Antonius-Hospital gGmbH, Akademisches Lehrkrankenhaus der RWTH Aachen, Eschweiler, Deutschland
| | - H-J Busch
- Universitäts-Notfallzentrum Freiburg, Universitätsklinikum Freiburg, Freiburg, Deutschland
| | - M Kochanek
- Klinik I für Innere Medizin (Hämatologie und Onkologie), Schwerpunkt Internistische Intensivmedizin, Universitätsklinikum Köln, Köln, Deutschland
| | - G Michels
- Klinik für Akut- und Notfallmedizin, St.-Antonius-Hospital gGmbH, Akademisches Lehrkrankenhaus der RWTH Aachen, Eschweiler, Deutschland
| | | | - M Buerke
- Medizinische Klinik II, St. Marien-Krankenhaus Siegen, Siegen, Deutschland
| | - S Kluge
- Klinik für Intensivmedizin, Universitätsklinikum Hamburg-Eppendorf, Hamburg, Deutschland
| | - M Baumgärtel
- Klinikum Nürnberg-Nord, Intensivstation 10/II, Klinik für Innere Medizin 3, Schwerpunkt Pneumologie, Paracelsus Medizinische Privatuniversität, Nürnberg, Deutschland
| | - S Braune
- IV. Med. Klinik - Internistische Intensivmedizin und Notaufnahme, Franziskus-Hospital Münster, Münster, Deutschland
| | - F Erbguth
- Klinikum Nürnberg, Universitätsklinik für Neurologie, Paracelsus Medizinische Privatuniversität, Nürnberg, Deutschland
| | - V Fuhrmann
- Klinik für Innere Medizin I, Evangelisches Klinikum Niederrhein, Duisburg, Deutschland
| | - P Lebiedz
- Klinik für Innere Medizin und Internistische Intensivmedizin, Ev. Krankenhaus Oldenburg, Steinweg 13-17, Oldenburg, Deutschland
| | - K Mayer
- Medizinische Klinik 4, Pneumologie und Schlafmedizin, ViDia Kliniken, Karlsruhe, Deutschland
| | - U Müller-Werdan
- Klinik für Geriatrie und Altersmedizin, Charité - Universitätsmedizin Berlin, Berlin, Deutschland
- Evangelisches Geriatriezentrum Berlin (EGZB), Berlin, Deutschland
| | - M Oppert
- Klinik für Notfall- und Intensivmedizin, Klinikum Ernst von Bergmann, Potsdam, Deutschland
| | - F Sayk
- Campus Lübeck, Medizinische Klinik I, Universitätsklinikum Schleswig-Holstein, Lübeck, Deutschland
| | - D Sedding
- Universitätsklinikum Halle (Saale), Klinik und Poliklinik für Innere Medizin III, Martin-Luther-Universität Halle-Wittenberg, Ernst-Grube-Straße 40, 06120, Halle (Saale), Deutschland
| | - C Willam
- Universitätsklinikum Erlangen, Medizinische Klinik 4, Nephrologie und Hypertensiologie, Friedrich-Alexander-Universität Erlangen, Erlangen, Deutschland
| | - K Werdan
- Universitätsklinikum Halle (Saale), Klinik und Poliklinik für Innere Medizin III, Martin-Luther-Universität Halle-Wittenberg, Ernst-Grube-Straße 40, 06120, Halle (Saale), Deutschland.
| |
Collapse
|