Developing a prediction model for in-hospital mortality in sepsis patients with gastrointestinal bleeding using the MIMIC-IV database

Developing a prediction model for in-hospital mortality in sepsis patients with gastrointestinal bleeding using the MIMIC-IV database | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Developing a prediction model for in-hospital mortality in sepsis patients with gastrointestinal bleeding using the MIMIC-IV database Fengwei Yao, Ji Luo, Yue Ming, Zhiqiang Zhao, Luhua Wang, Zhijun He This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5406276/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Sepsis associated with gastrointestinal hemorrhage is a critical condition in ICU patients, significantly impacting mortality rates. This study aimed to develop a predictive model for in-hospital death risk in sepsis patients with gastrointestinal bleeding, improving treatment strategies and resource allocation. Methods: In a retrospective investigation of patients with sepsis and gastrointestinal bleeding, we gathered information from the MIMIC-IV database, including key demographics, comorbidities, laboratory indicators, and therapies. The dataset was split 70:30 for model development and validation. The Least Absolute Shrinkage and Selection Operator (LASSO) approach was used to select features, and multivariate logistic regression was then used to create a prognostic model. A nomogram was created to visualize predictive outcomes. Model performance was evaluated using calibration curve, receiver operating characteristic (ROC) curve, clinical impact curve (CIC), and decision curve analysis (DCA). Results: Nine significant predictors of in-hospital mortality were identified: APS III score, prothrombin time, body temperature, activated partial thromboplastin time, respiratory rate, vasopressor use, acute kidney injury, non-invasive ventilation, and malignancy. Area beneath the ROC curve for the training and testing groups The values are 0.8266 (95% CI: 0.8018-0.8515) and 0.7961 (95% CI: 0.7577-0.8345), respectively. Our model outperformed the APS III score in terms of ROC curve discrimination and demonstrated greater net benefit on the DCA curve. The CIC showed strong concordance between predicted and actual in-hospital death rates when the predicted probability exceeded 70%. Conclusion: We developed a robust predictive framework for assessing in-hospital death risk in sepsis patients with gastrointestinal hemorrhage. Early intervention based on identified risk factors could improve patient survival rates. MIMIC-IV Nomogram Predictive model Sepsis Gastrointestinal bleeding Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 1. Background Sepsis is a life-threatening illness caused by an unregulated immunological response to infection, which results in widespread inflammation, immune dysfunction, and organ failure. Its pathophysiology involves a cascade of immune and metabolic disruptions. Without prompt treatment, sepsis rapidly progresses to multi-organ failure and death. Early detection and intervention are essential to improve survival, especially in critically ill patients. It is a leading cause of death and morbidity globally. The incidence of sepsis among hospitalized patients ranges from 1–3%, while in intensive care unit patients, it can increase to 10–30% ( 1 , 2 ). It is estimated that every year, 31.5 million individuals globally are afflicted by sepsis, with 19.4 million having severe sepsis, resulting in around 5.3 million deaths( 3 ). Gastrointestinal symptoms are prevalent during ICU admission, with up to 62% of patients showing at least one gastrointestinal symptom lasting for a minimum of one day. Gastrointestinal hemorrhage is one of the most common presentations, with a greater incidence and fatality rate than those without gastrointestinal bleeding( 4 , 5 ). The occurrence of gastrointestinal complications is linked to a poor prognosis in critically ill patients; on average, 4% of these patients experience gastrointestinal hemorrhage ( 6 ). 5.4% of patients with septic shock develop gastrointestinal bleeding ( 7 ). The main causes of bleeding include stress ulcers in the esophagus, stomach, or duodenum, gastrointestinal mucosal damage due to antibiotics and inflammatory mediators, anticoagulant therapy, and intestinal ischemia resulting from sepsis( 8 , 9 ). The in-hospital death rate in septic patients with gastrointestinal bleeding is much higher, highlighting the important need for early screening and risk assessment in these patients. For sepsis patients, the SAPS II score, SOFA score and APS III score are the main scoring systems currently utilized. In contrast, scoring systems for gastrointestinal bleeding include the Rockall, Glasgow-Blatchford, and AIMS65 scoring systems, each with its own limitations. Recent studies indicate that the APACHE II score demonstrates excellent prognostic accuracy for estimating the likelihood of death among ICU patients (AUC: 0.87, CI: 0.75–0.99), while the SOFA score shows moderate accuracy (AUC: 0.71, CI: 0.50–0.93). However, no scoring system has been created to predict death in individuals with upper gastrointestinal hemorrhage. All grading methods have limited predictive accuracy for ICU duration of stay( 10 ). Therefore, developing a predictive tool to evaluate the likelihood of death during hospitalization in septic patients with gastrointestinal bleeding is of substantial clinical value. The goal of this study is to develop a model that is more specifically tailored to septic patients with gastrointestinal bleeding, providing a more precise early assessment tool for clinical use, facilitating early intervention based on the identified independent risk factors, and ultimately improving patient outcomes. 2. Methods 2.1 Research data The Medical Information Mart for Intensive Care IV (MIMIC-IV, v. 3.0) database provided the data used in this investigation. It is a huge, publicly available database including clinical data on adult patients (aged 18 and up) admitted to the ICU of a major tertiary hospital in the United States between 2008 and 2019. It has more than 70,000 ICU admission cases ( 11 ). We successfully completed the Collaborative Institutional Training Initiative (CITI) program (Researcher Approval Code: 65073876) and received authorization to use the MIMIC-IV database. The Beth Israel Deaconess Medical Center and Massachusetts Institute of Technology institutional review boards have allowed access to the MIMIC-IV database. All patient data in the database is anonymised, thus informed permission is not necessary. We extracted relevant diagnostic, laboratory, demographic data, and associated treatments from the database using Structured Query Language in Navicat Premium 17, and then linked the extracted data to each patient's unique HADM_ID. Sepsis was diagnosed based on a SOFA score of ≥ 2 ( 12 ). Patients with gastrointestinal bleeding were diagnosed using the International Classification of Diseases, Ninth Revision (ICD-9) and Tenth Revision (ICD-10), with no differentiation between upper and lower gastrointestinal bleeding. 2.3 Exclusion criteria ( 1 ) Under 18 years old; ( 2 ) Missing hospitalization data > 20% ;( 3 ) Hospital stay of less than 24 hours; ( 4 ) For recurrent ICU hospitalizations, only data from the first admission were included. 2.4 Collected indicators To conduct a thorough analysis of the clinical determinants influencing patient outcomes, we systematically extracted the mean values of a wide range of clinical indicators from the MIMIC-IV database, with a focus on data collected within the first 24 hours of ICU admission for eligible patients. These included demographic characteristics (age and sex), vital signs (body temperature [T], respiratory rate [RR], heart rate [HR], blood oxygen saturation [SpO2], systolic blood pressure [SBP], and diastolic blood pressure [DBP]), and laboratory parameters (bicarbonate [HCO3], blood urea nitrogen [BUN], sodium [Na], creatinine [Cr], potassium [K], prothrombin time [PT], activated partial thromboplastin time [APTT], white blood cell count [WBC], glucose [Glu], anion gap [AG], international normalized ratio [INR], mean corpuscular hemoglobin concentration [MCHC], mean corpuscular hemoglobin [MCH], red blood cell count [RBC], platelet count [PLT]). In addition, we documented comorbid conditions (diabetes mellitus [DM], myocardial infarction [MI], congestive heart failure [CHF], acute kidney injury [AKI], cerebrovascular disease [CVD], mild liver disease [MLD], malignancy, chronic kidney disease [CKD], chronic lung disease [CLD], and severe liver disease [SLD]), and key severity scores, such as the Acute Physiology Score III (APS III), Sequential Organ Failure Assessment (SOFA), Glasgow Coma Scale (GCS), Simplified We also gathered data on treatment regimens and pharmaceutical use, such as epinephrine (EPI), dopamine (DA), vasopressin (VP), norepinephrine (NE), invasive mechanical ventilation (IV), non-invasive ventilation (NIV), and continuous renal replacement therapy (CRRT). This cohort study tracked in-hospital mortality, with a final cohort of 1,923 patients, of whom 541 died during hospitalization. 2.5 Statistical analysis To begin, the normality of continuous data was examined using the Shapiro-Wilk test. In the baseline characteristics table, variables with normal distributions were reported as means ± standard deviations. The independent t-test was then used to compare group averages and assess intergroup differences. For data that did not follow a normal distribution, values were given as median and interquartile range, and intergroup comparisons were performed using the Wilcoxon rank-sum test. Categorical variables were examined using the chi-square test. Furthermore, LASSO regression was employed for the preliminary identification of risk factors associated with in-hospital death in patients with gastrointestinal bleeding and sepsis. Finally, multivariable logistic regression was used to determine the final independent variables for model formation, which were represented by a nomogram. Variables with P < 0.05 were considered to have statistical significance. For variables with missing data of less than 20%, the “MICE” package in R software was employed for multiple imputation to predict and fill in the missing values. R software version 4.1.1 was used for all data analyses and visualizations. 3. Results 3.1 Baseline comparison First, 2,146 sepsis complicated by gastrointestinal bleeding (GIB) patients (upper and lower gastrointestinal hemorrhage) were extracted from the MIMIC-IV data base from 2008 to 2019. For patients with multiple admissions, only data from their first hospitalization were included. 215 participants in all were disqualified from the research because they spent less than 24 hours in the intensive care unit, which was deemed insufficient for assessing outcomes related to sepsis and gastrointestinal hemorrhage. Additionally, 8 patients were excluded because of missing data that exceeded 20%, which could compromise the reliability and validity of the analysis. As illustrated in the study flowchart (Fig. 1 ), the final cohort consisted of 1,923 patients, of whom 541 experienced in-hospital mortality. To ensure robust model development and validation, the patients were randomly divided into two cohorts: the training cohort (n = 1,346) for model development, and the validation cohort (n = 577) for independent performance evaluation. This division allows for the assessment of model generalizability and accuracy across different patient populations within the study. The patients' median age in the training cohort was 65.45 years (the range between the 25th and 75th percentiles was 54.88, 78.33), with 824 males (61.21%). I In the validation cohort, the patients' median age was 64.52 years (the range between the 25th and 75th percentiles were 53.91, 77.08), with 357 males (61.87%). In our study cohort grouping, 365 (27.1%) and 176 (30.5%) patients died during hospitalization, respectively, and there was no statistically significant difference in mortality rates between the groups. As shown in Table 1 (The table exceeds one page and should be placed at the end of the text), the baseline characteristics of both cohorts were comparable, indicating similar demographic and clinical profiles. Table 1 :Baseline characteristics of patients with sepsis combined with gastrointestinal bleeding. Variables Total(n = 1923) Validation(n = 577) Training(n = 1346) P Age(years) 65.06(54.53,77.82) 64.53(53.91,77.09) 65.45(54.88,78.33) 0.17 INR 1.45(1.23,1.85) 1.47(1.25,1.90) 1.45(1.20,1.85) 0.607 PT(s) 15.90(13.65,20.20) 16.00(13.85,20.58) 15.86(13.61,20.06) 0.417 PTT(s) 33.25(28.26,43.30) 33.25(28.30,43.30) 33.27(28.20,43.29) 0.892 HCO3(mmol/L) 21.50(18.90,24.50) 21.50(18.00,24.00) 21.67(19.00,24.50) 0.121 BUN(mg/dL) 31.67(19.50,53.55) 30.00(18.50,50.50) 32.00(20.00,54.50) 0.138 K(mmol/L) 4.20(3.80,4.70) 4.14(3.75,4.70) 4.20(3.83,4.70) 0.071 Na(mmol/L) 138.50(135.33,141.67) 138.00(135.00,141.50) 138.50(135.50,142.00) 0.373 Glu(mg/dL) 130.50(107.00,168.00) 129.00(106.25,164.50) 131.00(107.75,168.31) 0.499 AG(mg/dL) 14.50(12.00,17.75) 14.75(12.00,18.00) 14.50(12.00,17.67) 0.265 Cr(mg/dL) 1.30(0.80,2.24) 1.27(0.80,2.33) 1.30(0.83,2.20) 0.492 WBC(×10 9 /L) 11.47(7.65,16.23) 11.75(7.92,16.50) 11.40(7.55,16.20) 0.51 MCHC(g/L) 33.20(32.13,34.20) 33.27(32.20,34.20) 33.20(32.10,34.20) 0.112 RBC(×10 12 /L) 3.10(2.69,3.60) 3.11(2.69,3.61) 3.09(2.69,3.60) 0.963 PLT(×10 9 /L) 144.60(86.06,223.00) 152.00(88.33,221.00) 141.93(84.69,223.44) 0.401 MCH(pg) 30.50(29.07,32.10) 30.68(29.32,32.23) 30.49(28.97,32.03) 0.037 Weight(kg) 78.20(66.00,94.00) 78.90(66.70,93.80) 78.00(65.70,94.00) 0.672 HR(times/min) 88.15(76.88,99.68) 88.04(75.71,99.36) 88.22(77.36,99.76) 0.37 SBP(mmHg) 111.19(103.00,122.64) 110.27(102.64,122.14) 111.60(103.09,122.88) 0.35 DBP(mmHg) 59.96(54.23,67.04) 60.09(54.61,66.55) 59.85(54.11,67.29) 0.725 RR(times/min) 19.20(16.71,22.32) 19.19(16.43,22.09) 19.21(16.79,22.38) 0.501 T(°C) 36.79(36.55,37.10) 36.78(36.55,37.10) 36.80(36.55,37.10) 0.646 Spo2(%) 97.46(95.92,98.75) 97.56(95.88,98.89) 97.42(95.92,98.71) 0.633 APSIII 56.00(43.00,71.50) 55.00(44.00,72.00) 56.00(43.00,71.00) 0.812 LODS 6.00(4.00,9.00) 6.00(4.00,9.00) 6.00(4.00,9.00) 0.768 OASIS 35.00(30.00,41.00) 35.00(30.00,41.00) 35.00(29.00,41.00) 0.885 SOFA 6.00(4.00,9.50) 7.00(4.00,10.00) 6.00(4.00,9.00) 0.086 SAPSII 43.00(34.00,53.00) 44.00(34.00,55.00) 42.00(34.00,53.00) 0.043 GCS 15.00(15.00,15.00) 15.00(15.00,15.00) 15.00(15.00,15.00) 0.02 SIRS 3.00(2.00,3.00) 3.00(2.00,3.00) 3.00(2.00,3.00) 0.897 Death, n(%) 0.13 No 1382(71.87) 401(69.50) 981(72.88) Yes 541(28.13) 176(30.50) 365(27.12) Gender, n(%) 0.787 F 742(38.59) 220(38.13) 522(38.78) M 1181(61.41) 357(61.87) 824(61.22) MI, n(%) 0.884 No 1610(83.72) 482(83.54) 1128(83.80) Yes 313(16.28) 95(16.46) 218(16.20) CHF, n(%) 0.009 No 1392(72.39) 441(76.43) 951(70.65) Yes 531(27.61) 136(23.57) 395(29.35) CVD, n(%) 0.999 No 1693(88.04) 508(88.04) 1185(88.04) Yes 230(11.96) 69(11.96) 161(11.96) CLD, n(%) 0.965 No 1451(75.46) 435(75.39) 1016(75.48) Yes 472(24.54) 142(24.61) 330(24.52) MLD, n(%) 0.358 No 1087(56.53) 317(54.94) 770(57.21) Yes 836(43.47) 260(45.06) 576(42.79) DM, n(%) 0.489 No 1325(68.90) 404(70.02) 921(68.42) Yes 598(31.10) 173(29.98) 425(31.58) Malignancy, n(%) 0.028 No 1623(84.40) 471(81.63) 1152(85.59) Yes 300(15.60) 106(18.37) 194(14.41) RD, n(%) 0.802 No 1409(73.27) 425(73.66) 984(73.11) Yes 514(26.73) 152(26.34) 362(26.89) AKI, n(%) 0.945 No 325(16.90) 97(16.81) 228(16.94) Yes 1598(83.10) 480(83.19) 1118(83.06) SLD, n(%) 0.765 No 1256(65.31) 374(64.82) 882(65.53) Yes 667(34.69) 203(35.18) 464(34.47) EPI:, n(%) 0.076 No 1831(95.22) 557(96.53) 1274(94.65) Yes 92(4.78) 20(3.47) 72(5.35) DA, n(%) 0.701 No 1848(96.10) 553(95.84) 1295(96.21) Yes 75(3.90) 24(4.16) 51(3.79) NE, n(%) 0.403 No 1187(61.73) 348(60.31) 839(62.33) Yes 736(38.27) 229(39.69) 507(37.67) VP, n(%) 0.384 No 1591(82.74) 484(83.88) 1107(82.24) Yes 332(17.26) 93(16.12) 239(17.76) IV, n(%) 0.462 No 851(44.25) 248(42.98) 603(44.80) Yes 1072(55.75) 329(57.02) 743(55.20) NIV, n(%) 0.722 No 611(31.77) 180(31.20) 431(32.02) Yes 1312(68.23) 397(68.80) 915(67.98) CRRT, n(%) 0.244 No 1663(86.48) 507(87.87) 1156(85.88) Yes 260(13.52) 70(12.13) 190(14.12) 3.2 Variable Selection via LASSO Regression In this study, the LASSO regression technique with 10-fold cross-validation was used to assess 48 variables, including demographic characteristics such as age, gender, and various laboratory indicators, to identify factors associated with in-hospital mortality in patients with severe sepsis and gastrointestinal bleeding (GIB). The model found 15 factors as the most important predictors at the optimal lambda value (λ = 0.03319349). The selected variables include: anion gap, activated partial thromboplastin time, prothrombin time, respiratory rate, body temperature, presence of malignancy, SOFA score, APS III score, LODS score, SAPS II score, administration of vasopressors, use of norepinephrine, non-invasive mechanical ventilation, AKI and CRRT. Figure 2 depicts the variations in the LASSO regression coefficients for the 48 variables, with each curve showing the trajectory of the coefficient for each independent variable. Figure 3 outlines the procedure for selecting the optimal parameter value through cross-validation within the LASSO model. 3.3 Independent influencing factors selected using multivariable logistic Regression Based on the fifteen prognostic factors selected by LASSO regression, A multivariable logistic regression was used to find the independent variables linked with prognosis. As presented in Table 2 , nine prognostic factors were ultimately recognized as key contributors to the model. An odds ratio (OR) of 3.02 (95% CI: 1.78–5.42) in our study showed that patients with acute kidney injury (AKI) had a significantly higher chance of dying in the hospital. This finding underscores the significant impact of AKI on the likelihood of mortality during hospitalization, suggesting that AKI serves as a strong independent predictor of adverse outcomes in critically ill patients. This finding underscores the importance of early detection and intervention in managing AKI to mitigate its adverse impact on patient outcomes. Similarly, malignancy was found to be a critical factor influencing in-hospital mortality. Patients with a documented history of cancer had a substantially greater in-hospital death rate than those without cancer, with an odds ratio (OR) of 1.92 (95% CI: 1.29–2.87). This finding underscores the elevated and compounded risk that malignancy confers, particularly in critically ill patients with complex comorbidities such as sepsis and gastrointestinal bleeding. The interaction between cancer and acute illness appears to exacerbate patient vulnerability, highlighting the need for more intensive monitoring and tailored therapeutic approaches in this high-risk population. Malignancy in combination with sepsis and gastrointestinal bleeding significantly worsens prognosis, and clinical management must account for these risks. Hemodynamically unstable patients who required vasopressor support had an in-hospital mortality rate 2.20 times greater (OR = 2.20, 95% CI: 1.44–3.36). This result emphasizes how serious circulatory impairment is and how difficult it is to treat these individuals who are in severe condition. It suggests that the use of vasopressors as a therapeutic intervention is reflective of profound hemodynamic instability, which correlates with worse outcomes. In terms of specific biomarkers and clinical measurements, several factors were identified as independent predictors of in-hospital mortality. These included PT (OR = 1.03, 95% CI: 1.01–1.05), activated partial thromboplastin time (OR = 1.01, 95% CI: 1.00–1.02), RR(OR = 1.06, 95% CI: 1.02–1.10), and the Acute Physiology and Chronic Health Evaluation III (APS III) score (OR = 1.01, 95% CI: 1.00–1.02). These clinical parameters, which reflect coagulopathy, respiratory dysfunction, and the overall severity of illness, were found to be significant contributors to mortality risk. The presence of abnormal values in these indicators can help clinicians identify patients at higher risk for mortality and prioritize interventions accordingly. Interestingly, we also identified protective factors that were inversely associated with mortality. Specifically, temperature (OR = 0.72, 95% CI: 0.55–0.94) and non-invasive mechanical ventilation (OR = 0.33, 95% CI: 0.25–0.45) were associated with lower in-hospital mortality. A higher body temperature, likely indicative of an active immune response or the presence of systemic inflammatory reactions, appeared to confer a protective effect, while non-invasive mechanical ventilation was associated with better outcomes by reducing the need for invasive procedures and supporting respiratory function. These findings suggest that maintaining appropriate body temperature and optimizing respiratory support strategies may improve survival rates for critically ill patients with sepsis and gastrointestinal bleeding. Table 2 Multivariable analysis base on LASSO regression Variables B OR 95%CI P value PT 0.031 1.032 1.012–1.052 0.001 PTT 0.012 1.012 1.003–1.020 0.007 RR 0.059 1.060 1.023–1.098 0.001 temperature -0.326 0.722 0.551–0.946 0.018 Malignant cancer 0.654 1.924 1.290–2.871 0.001 APSIII 0.013 1.013 1.001–1.024 0.031 LODS 0.036 1.037 0.959–1.120 0.356 SAPSII 0.007 1.007 0.990–1.120 0.422 norepinephrine 0.255 1.291 0.887–1.878 0.182 SOFA -0.009 0.991 0.935–1.050 0.755 vasopressin 0.787 2.197 1.440–3.350 < 0.001 Noninvasive ventilator -1.099 0.333 0.246–0.450 < 0.001 CRRT 0.303 1.354 0.895–2.049 0.151 AKI 1.106 3.021 1.736–5.258 < 0.001 AG 0.033 1.033 0.999–1.069 0.058 3.4 Model construction and validation Using the nine most important indicators found by multivariable logistic regression analysis and represented by a nomogram, a prediction model for the likelihood of in-hospital death in intensive care unit patients with sepsis and gastrointestinal bleeding was created. Each independent predictor included in the nomogram was assigned a weighted score, with the cumulative score ranging from 0 to 400 points. This scoring system was designed to reflect the relative contribution of each factor to the risk of in-hospital mortality. The total score derived from the nine key predictive factors is directly correlated with the probability of mortality, with values ranging from 0.1 to 0.9. Specifically, higher cumulative scores indicate an elevated risk of in-hospital mortality, underscoring the importance of these factors in stratifying patients based on their clinical prognosis. The relationship between total score and predicted mortality probability is illustrated in Fig. 4 , which visually demonstrates the gradual increase in mortality risk as the total score rises. We initially computed the area under the receiver operating characteristic curve (AUC), a popular and extensively used metric for evaluating the discriminative ability of diagnostic models, in order to analyze the nomogram's predictive performance. Within the training group, the AUC value of 0.827 (95% CI: 0.8018–0.8515) was obtained, reflecting the model's robust ability to distinguish between patients with varying risks of in-hospital mortality, and indicating a high level of predictive accuracy. With an AUC of 0.7961 (95% CI: 0.7577–0.8345) in the validation cohort, consistent performance across datasets was shown. These findings, presented in Figs. 5 and 6 , further validate the robustness and generalizability of the nomogram in predicting in-hospital mortality risk in ICU patients with gastrointestinal bleeding and sepsis. Our nomogram surpassed the APS III scoring system in predictive accuracy, showing significantly better AUCs across both cohorts. Using the APS III scoring technique, The AUC for the validation group was 0.7308 (95% CI: 0.6869–0.7746), while the training set's was 0.7531 (95% CI: 0.7242–0.7821). This data highlights our model's superior discriminative capacity over the APS III score in predicting in-hospital mortality for patients with sepsis and gastrointestinal bleeding in critical care units. We used decision curve analysis (DCA), which measures the net benefit of using the model at various probability thresholds, to further assess the algorithm's clinical value. To further demonstrate the accuracy and dependability of the model, calibration curves were also created to evaluate the agreement between the observed and anticipated probability of in-hospital mortality. To evaluate the accuracy of the model's predictions, calibration curves were constructed for both the training and validation cohorts (Figs. 7 and 8 ). These curves assess the degree of agreement between the predicted probabilities of in-hospital mortality and the actual observed outcomes. A well-calibrated model should exhibit a close alignment between predicted and observed values across a range of predicted probabilities, which was thoroughly assessed in both cohorts. In these curves, the y-axis represents the actual event probability observed in the study cohort, while the x-axis depicts the model’s predicted probabilities. A perfect model would result in all data points lying along the diagonal line, indicating perfect agreement between predicted and observed values. In our study, the calibration plots for the training and validation groups exhibited a strong alignment with this diagonal line, reflecting a high degree of concordance between the predicted probabilities of in-hospital mortality and the actual observed outcomes. The model's clinical application and reliability in predicting In-hospital death rate for sepsis and gastrointestinal bleeding are further supported by the close correlation between predicted and actual values, which highlights the model's outstanding calibration and exceptional fit. In the DCA curves (Figs. 9 and 10 ), the gray diagonal line represents all patients undergoing intervention, while the black parallel line represents no intervention. The training cohort's best risk threshold for optimizing net benefit was less than 94%, and the DCA curve demonstrated that our prediction model outperformed the APS III score in terms of net benefits over a range of thresholds. Additionally, we constructed clinical impact curve (CIC) to assess the real-world applicability and predictive accuracy of our model. As shown in Figs. 11 and 12 , Our findings indicated that when the threshold probability surpassed 70%, the model’s identification of high-risk patients for in-hospital mortality closely matched the observed clinical outcomes, highlighting the model’s robust predictive power. This result further reinforces the clinical applicability of the predictive model, suggesting its potential to guide decision-making and enhance patient management in critical care environments. The robustness of this model in both the training and validation cohorts reinforces its generalizability and reliability in diverse clinical environments. 4. Discussion Developing a prediction model to estimate the likelihood of in-hospital death in patients with sepsis and gastrointestinal bleeding was the aim of this retrospective cohort study. Validation of this model demonstrated good predictive ability and discriminative power. Patients with sepsis combined with GIB exhibited higher mortality rates, greater disease severity, poorer clinical outcomes, longer hospital stays, and higher hospitalization costs. In this research, we established a framework integrating various clinical parameters to forecast the outcomes of ICU patients with gastrointestinal bleeding (GIB) and sepsis. The model identified nine independent risk factors for mortality: PT, APTT, RR, APS III score, AKI, malignancy, and the use of vasopressors. Notably, non-invasive mechanical ventilation and a body temperature below 40°C were found to be protective factors. The APS III score is a component of the APACHE scoring system, which focuses on acute physiological parameters. It includes various physiological indicators, generating a total score that reflects the severity of the patient's condition. This scoring system is used to assess the physiological status of critically ill patients, with higher scores indicating more severe illness and a worse prognosis. Research has shown that APS III outperforms both the SOFA and SAPS II scores in overall prognostic accuracy( 13 , 14 ). These findings provide strong evidence for the inclusion of the Acute Physiology Score III (APS III) as a key risk factor in predicting patient outcomes. Additionally, our analysis highlights PT and APTT as significant independent predictors of in-hospital mortality, further emphasizing the critical role of coagulation parameters in assessing patient prognosis. Currently, inflammation and coagulopathy are considered key contributors to multiple organ dysfunction, with sepsis exacerbating these disturbances by inducing complex dysregulations across various systems, including the coagulation cascade( 15 ). Coagulopathy is a common complication of sepsis and a critical host response to infection, which may progress to disseminated intravascular coagulation (DIC), thereby increasing mortality. Diagnostic standards for sepsis-induced coagulopathy and overt disseminated intravascular coagulation have been set by the International Society on Thrombosis and Hemostasis (ISTH) ( 16 , 17 ). Overt disseminated intravascular coagulation typically indicates decompensated coagulopathy, while sepsis-induced coagulopathy represents systemic activation of coagulation, with compensatory coagulation function still intact. Once it progresses to overt DIC, the benefits of anticoagulant therapy for disease progression may become minimal. One study suggested that, in sepsis, fibrinolysis is inhibited, and levels of fibrin/fibrinogen and D-dimer do not increase in proportion to disease severity. In contrast, prothrombin time prolongs in a linear fashion with increasing severity, which is consistent with the outcome variables we selected( 18 ). Furthermore, we observed that an elevated respiratory rate could be indicative of early organ dysfunction, particularly reflecting respiratory distress or failure, which is closely linked to adverse mortality outcomes. In our study, the mean respiratory rate recorded within the first 24 hours of ICU admission (OR = 1.008, 95% CI = 1.002–1.015, p = 0.014) was identified as a significant predictor of 30-day all-cause mortality in patients with sepsis complicated by gastrointestinal bleeding (GIB). An elevated respiratory rate generally indicates metabolic imbalance and inadequate oxygenation, resulting in insufficient tissue perfusion and worsening tissue hypoxia. Consequently, this impairs intestinal barrier integrity, heightening the risk of gastrointestinal bleeding and increasing intestinal permeability, which promotes bacterial translocation and inflammatory responses, thereby further increasing mortality risk among sepsis patients. Numerous studies have underscored the negative influence of fever on patient prognosis, particularly in the context of neurological conditions. This has led to the widespread adoption of both pharmacological therapies and physical interventions aimed at controlling elevated body temperature, as effective management of fever is considered crucial in mitigating adverse outcomes( 19 ). Surprisingly, our analysis indicated that an early elevation in body temperature may act as a potential protective factor in patients with sepsis and gastrointestinal bleeding (OR = 1.008, 95% CI: 1.002–1.015, p = 0.014). This finding suggests that a modest rise in temperature during the early stages of illness might be associated with better patient outcomes, potentially reflecting an adaptive immune response or a controlled inflammatory reaction. Research indicates that fever plays a crucial role in enhancing microbial clearance, activating immune responses, and stimulating heat shock protein production. Most human pathogens thrive at temperatures ranging from 35°C to 37°C( 20 ), and a rise in body temperature may inhibit microbial proliferation. Additionally, research has shown that elevated body temperature can strengthen the host’s immune response by increasing antibiotic sensitivity, lowering minimum inhibitory concentrations, enhancing neutrophil mobility, and promoting phagocytosis. It also facilitates T-helper cell adhesion and helps prevent lymphocyte depletion. However, it should be noted that at elevated fever levels (40–41°C), the beneficial immunomodulatory effects may be offset by the negative metabolic and inflammatory consequences of fever ( 21 ). In our nomogram, we chose 40 degrees Celsius as the baseline for the scale, which appears to be a reasonable selection based on these factors. Furthermore, a meta-analysis has demonstrated a negative correlation between body temperature and clinical outcomes in sepsis ( 22 ). Analysis of initial body temperature and prognostic outcomes in specific emergency department admissions further suggests that low body temperature is linked to prolonged initial antibiotic administration, extended hospital stays, higher rates of ICU admission, and elevated mortality( 23 ). Patients with malignancies often experience immune dysfunction, malnutrition, and other comorbidities, which render them more vulnerable to infections, thereby increasing the risk of sepsis. The development of sepsis further worsens the condition of cancer patients, leading to poorer prognoses. Our data demonstrate that septic patients with gastrointestinal bleeding exhibit high mortality rates, irrespective of cancer status; However, the mortality rate is significantly higher in cancer patients when compared with non-cancer patients. This result aligns with the findings from a supplementary study, which included over 19 million hospitalized septic patients from the National Inpatient Sample database between 2008 and 2017. In that study, 20.4% of septic cases were associated with cancer, and the in-hospital mortality rate for cancer-related sepsis was higher than that for non-cancer patients( 24 ). Additionally, a meta-analysis indicated that cancer substantially elevates the mortality risk in septic patients (OR = 2.7, 95% CI: 1.07–6.84). However, the link between cancer and early mortality was not significant (OR = 2.77, 95% CI: 0.88–8.66), whereas the risk of late mortality was markedly higher (OR = 2.46, 95% CI: 1.42–4.25) ( 25 ). Sepsis complicated by acute kidney injury (AKI) is frequently observed in critically ill patients. Research indicates that septic patients suffering from AKI display a marked increase in mortality in comparison with those without kidney impairment. The deterioration of renal function can precipitate dysfunction in other vital organs, including the heart, lungs, and liver. Additionally, fluid management in the context of AKI becomes more challenging; excessive fluid resuscitation may lead to complications such as heart failure or pulmonary edema, while insufficient fluid administration can worsen renal ischemia, thereby amplifying the overall mortality risk in septic patients( 26 , 27 ). The "Save Sepsis" initiative has highlighted the importance of non-invasive ventilation (NIV): the 2021 International Guidelines for the Management of Sepsis and Septic Shock suggest that NIV may offer comparable physiological advantages to invasive positive pressure ventilation, such as enhanced gas exchange and decreased respiratory effort in certain patients, while mitigating complications related to intubation, invasive ventilation, and the need for sedation( 28 ). Several studies further support the notion that failure of non-invasive mechanical ventilation is a risk factor for mortality in this cohort ( 29 , 30 ). Vasopressor agents are essential for sustaining hemodynamic stability; however, their excessive or improper use may result in uneven blood flow distribution, potentially worsening gastrointestinal injury. Currently, norepinephrine is the preferred first-line vasopressor for adults with septic shock, as compared to other vasopressors. Norepinephrine causes vasoconstriction and elevates mean arterial pressure with minimal effect on heart rate( 28 ). Furthermore, there was no discernible difference in 28-day mortality rates (39.3% vs. 35.4%, p = 0.26) in the VASST study, which contrasted norepinephrine alone with norepinephrine + vasopressin (0.01–0.03 U/min)( 31 ) Furthermore, research has demonstrated that while individuals treated with dopamine as the first-line vasopressor and those treated with norepinephrine do not significantly vary in effectiveness, dopamine treatment is associated with a greater rate of adverse events.( 32 ). Therefore, vasopressin may not be the preferred option. Our study has some inherent limitations. First, during data collection, we excluded certain variables with significant missing values, such as height, C-reactive protein levels, and the severity of acute kidney injury. These missing data may contain important influencing factors. Second, while our model demonstrated strong validation results within the MIMIC-IV database, external validation using datasets beyond MIMIC is necessary to confirm its generalizability. Third, as our study is a retrospective cohort analysis, further prospective studies are required before the nomogram can be applied in clinical practice. Conclusion In order to identify high-risk individuals early and promote prompt medical measures, we have created a prediction model that accurately assesses hospital mortality in sepsis patients with gastrointestinal bleeding. Declarations Acknowledgements Not applicable Author contributions Each author made a contribution to the design and conceptualization of the study. In addition to writing and editing the publication, FWY was in charge of developing the protocol and the overall study strategy. YM contributed significantly to the academic process by taking part in data collecting, statistical analysis, and result interpretation. To guarantee the accuracy of the study content, JL helped with the literature search and the arrangement of pertinent background materials. ZQZ ensured the analysis's dependability by managing data and visualizing the findings. LHW contributed clinical knowledge and took part in talks about the trial design. ZJH ensured the manuscript's academic quality and logical consistency by doing the final review and editing. Following a discussion of the study findings, all authors agreed on the manuscript's substance. Funding No particular grant from a governmental, private, or nonprofit funding organization was obtained for this study. Competing interests The writers say they have no conflicting interests. Availability of data and materials The MIMIC-IV 3.0 database contains the datasets created and/or examined during the current investigation (https://physionet.org/content/mimiciv/3.0/). Ethics approval and consent to participate An investigation of anonymised publically accessible datasets with prior institutional review board permission was the study's methodology. Consent for publication Not applicable. References Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). Jama. 2016;315(8):801-10. O'Brien JM, Jr., Ali NA, Aberegg SK, Abraham E. Sepsis. Am J Med. 2007;120(12):1012-22. Fleischmann C, Scherag A, Adhikari NK, Hartog CS, Tsaganos T, Schlattmann P, et al. Assessment of Global Incidence and Mortality of Hospital-treated Sepsis. Current Estimates and Limitations. Am J Respir Crit Care Med. 2016;193(3):259-72. Kate V, Sureshkumar S, Gurushankari B, Kalayarasan R. Acute Upper Non-variceal and Lower Gastrointestinal Bleeding. J Gastrointest Surg. 2022;26(4):932-49. Khamaysi I, Gralnek IM. Acute upper gastrointestinal bleeding (UGIB) - initial evaluation and management. Best Pract Res Clin Gastroenterol. 2013;27(5):633-8. Ye Z, Reintam Blaser A, Lytvyn L, Wang Y, Guyatt GH, Mikita JS, et al. Gastrointestinal bleeding prophylaxis for critically ill patients: a clinical practice guideline. Bmj. 2020;368:l6722. Siddiqui AH, Ahmed M, Khan TMA, Abbasi S, Habib S, Khan HM, et al. Trends and Outcomes of Gastrointestinal Bleeding Among Septic Shock Patients of the United States: A 10-Year Analysis of a Nationwide Inpatient Sample. Cureus. 2020;12(5):e8029. Kamboj AK, Hoversten P, Leggett CL. Upper Gastrointestinal Bleeding: Etiologies and Management. Mayo Clin Proc. 2019;94(4):697-703. Tokar JL, Higa JT. Acute Gastrointestinal Bleeding. Ann Intern Med. 2022;175(2):Itc17-itc32. Lincoln M, Keating N, O'Loughlin C, Tam A, O'Kane M, MacCarthy F, et al. Comparison of risk scoring systems for critical care patients with upper gastrointestinal bleeding: predicting mortality and length of stay. Anaesthesiol Intensive Ther. 2022;54(4):310-4. Johnson AEW, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10(1):1. Shankar-Hari M, Phillips GS, Levy ML, Seymour CW, Liu VX, Deutschman CS, et al. Developing a New Definition and Assessing New Clinical Criteria for Septic Shock: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). Jama. 2016;315(8):775-87. Zhu Y, Zhang R, Ye X, Liu H, Wei J. SAPS III is superior to SOFA for predicting 28-day mortality in sepsis patients based on Sepsis 3.0 criteria. Int J Infect Dis. 2022;114:135-41. Hwang SY, Kim IK, Jeong D, Park JE, Lee GT, Yoo J, et al. Prognostic Performance of Sequential Organ Failure Assessment, Acute Physiology and Chronic Health Evaluation III, and Simplified Acute Physiology Score II Scores in Patients with Suspected Infection According to Intensive Care Unit Type. J Clin Med. 2023;12(19). Cecconi M, Evans L, Levy M, Rhodes A. Sepsis and septic shock. Lancet. 2018;392(10141):75-87. Iba T, Levy JH, Yamakawa K, Thachil J, Warkentin TE, Levi M. Proposal of a two-step process for the diagnosis of sepsis-induced disseminated intravascular coagulation. J Thromb Haemost. 2019;17(8):1265-8. Iba T, Levy JH, Warkentin TE, Thachil J, van der Poll T, Levi M. Diagnosis and management of sepsis-induced coagulopathy and disseminated intravascular coagulation. J Thromb Haemost. 2019;17(11):1989-94. Matsubara T, Yamakawa K, Umemura Y, Gando S, Ogura H, Shiraishi A, et al. Significance of plasma fibrinogen level and antithrombin activity in sepsis: A multicenter cohort study using a cubic spline model. Thromb Res. 2019;181:17-23. Polderman KH. Induced hypothermia and fever control for prevention and treatment of neurological injuries. Lancet. 2008;371(9628):1955-69. Mackowiak PA, Marling-Cason M, Cohen RL. Effects of temperature on antimicrobial susceptibility of bacteria. J Infect Dis. 1982;145(4):550-3. Launey Y, Nesseler N, Mallédant Y, Seguin P. Clinical review: fever in septic ICU patients--friend or foe? Crit Care. 2011;15(3):222. Rumbus Z, Matics R, Hegyi P, Zsiboras C, Szabo I, Illes A, et al. Fever Is Associated with Reduced, Hypothermia with Increased Mortality in Septic Patients: A Meta-Analysis of Clinical Trials. PLoS One. 2017;12(1):e0170152. Khodorkovsky B, Youssef E, Adamakos F, Cina T, Falco A, LaMura L, et al. Does Initial Temperature in the Emergency Department Predict Outcomes in Patients Admitted for Sepsis? J Emerg Med. 2018;55(3):372-7. Hensley MK, Donnelly JP, Carlton EF, Prescott HC. Epidemiology and Outcomes of Cancer-Related Versus Non-Cancer-Related Sepsis Hospitalizations. Crit Care Med. 2019;47(10):1310-6. Xiang MJ, Chen GL. Impact of cancer on mortality rates in patients with sepsis: A meta-analysis and meta-regression of current studies. World J Clin Cases. 2022;10(21):7386-96. Peerapornratana S, Manrique-Caballero CL, Gómez H, Kellum JA. Acute kidney injury from sepsis: current concepts, epidemiology, pathophysiology, prevention and treatment. Kidney Int. 2019;96(5):1083-99. Zarbock A, Nadim MK, Pickkers P, Gomez H, Bell S, Joannidis M, et al. Sepsis-associated acute kidney injury: consensus report of the 28th Acute Disease Quality Initiative workgroup. Nat Rev Nephrol. 2023;19(6):401-17. Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47(11):1181-247. Demoule A, Chevret S, Carlucci A, Kouatchet A, Jaber S, Meziani F, et al. Changing use of noninvasive ventilation in critically ill patients: trends over 15 years in francophone countries. Intensive Care Med. 2016;42(1):82-92. Demoule A, Girou E, Richard JC, Taille S, Brochard L. Benefits and risks of success or failure of noninvasive ventilation. Intensive Care Med. 2006;32(11):1756-65. Russell JA, Walley KR, Singer J, Gordon AC, Hébert PC, Cooper DJ, et al. Vasopressin versus norepinephrine infusion in patients with septic shock. N Engl J Med. 2008;358(9):877-87. De Backer D, Biston P, Devriendt J, Madl C, Chochrad D, Aldecoa C, et al. Comparison of dopamine and norepinephrine in the treatment of shock. N Engl J Med. 2010;362(9):779-89. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5406276","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":375594163,"identity":"612296cc-1431-40dd-b682-229b9c13e908","order_by":0,"name":"Fengwei Yao","email":"","orcid":"","institution":"Renmin Hospital, Hubei University of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Fengwei","middleName":"","lastName":"Yao","suffix":""},{"id":375594164,"identity":"ca8a1e30-bc17-456d-aa6b-9410db756744","order_by":1,"name":"Ji Luo","email":"","orcid":"","institution":"Renmin Hospital, Hubei University of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Ji","middleName":"","lastName":"Luo","suffix":""},{"id":375594165,"identity":"a556f613-aab9-4553-963e-13bb5b8f048a","order_by":2,"name":"Yue Ming","email":"","orcid":"","institution":"Renmin Hospital, Hubei University of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yue","middleName":"","lastName":"Ming","suffix":""},{"id":375594166,"identity":"e7f2618a-5ab9-4e7d-be94-eae8bf6faf0f","order_by":3,"name":"Zhiqiang Zhao","email":"","orcid":"","institution":"Renmin Hospital, Hubei University of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Zhiqiang","middleName":"","lastName":"Zhao","suffix":""},{"id":375594167,"identity":"debcf999-a67f-4e84-9102-fbcc7ae576c4","order_by":4,"name":"Luhua Wang","email":"","orcid":"","institution":"Renmin Hospital, Hubei University of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Luhua","middleName":"","lastName":"Wang","suffix":""},{"id":375594168,"identity":"89f48661-8017-47d5-aa0a-50f32158b080","order_by":5,"name":"Zhijun He","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwklEQVRIiWNgGAWjYBACfvn3Hw5+4LHh4Sdai2RDguFhCZk0GckGYrUYHEgwPsBjc9jG4ADR1hw4kHBAIucwj/Hx5A0MPyq2EdbB2Nhw4EDBmXQeszPPChh7ztwmrIWZmbHhgGSPNY/ZjRwDZsY2IrSwsTEzHOD9x8xjPINYLTw8bAwHeHiceQwkiNUiIcHDcFiCJ41HAuiXg0T5xf4GD/NHYFTa87cnb3zwo4IILUgggYSogWshVccoGAWjYBSMEAAA0+o7RuICgfMAAAAASUVORK5CYII=","orcid":"","institution":"Renmin Hospital, Hubei University of Medicine","correspondingAuthor":true,"prefix":"","firstName":"Zhijun","middleName":"","lastName":"He","suffix":""}],"badges":[],"createdAt":"2024-11-07 03:08:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5406276/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5406276/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":71003714,"identity":"fe54a51a-1f1d-4d20-aa9d-ebeeaf0f1a96","added_by":"auto","created_at":"2024-12-10 06:06:49","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":35958,"visible":true,"origin":"","legend":"\u003cp\u003eThe flowchart of the study\u003c/p\u003e","description":"","filename":"fig1.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/5c710a7c54525c43c3e89b27.png"},{"id":71002500,"identity":"456096e0-66a6-4cef-b87f-39c61961a5b2","added_by":"auto","created_at":"2024-12-10 05:58:49","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":36002,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"fig2.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/f7abad1638cd1bab8f87982c.png"},{"id":71003715,"identity":"f782818b-252e-4051-a478-2e6532cfc3c6","added_by":"auto","created_at":"2024-12-10 06:06:49","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":42039,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"fig3.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/bf59461515520c30f095ad01.png"},{"id":71002510,"identity":"fa43509b-f239-485d-bc5c-4ae75a1e5c11","added_by":"auto","created_at":"2024-12-10 05:58:49","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":19664,"visible":true,"origin":"","legend":"\u003cp\u003eNomogram\u003c/p\u003e","description":"","filename":"fig4.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/9be7ce7eecebefac839dec71.png"},{"id":71002508,"identity":"3d48af5b-0ad5-4c0e-9da1-656e62138e14","added_by":"auto","created_at":"2024-12-10 05:58:49","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":36560,"visible":true,"origin":"","legend":"\u003cp\u003eROC curve of the training group\u003c/p\u003e","description":"","filename":"fig5.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/3b1e6b98183d0b4370b606e1.png"},{"id":71003716,"identity":"d9f852bf-950c-4f82-acd2-7e5025336f11","added_by":"auto","created_at":"2024-12-10 06:06:49","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":36624,"visible":true,"origin":"","legend":"\u003cp\u003eROC curve of the validation group\u003c/p\u003e","description":"","filename":"fig6.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/cd05d78f1d0e1ae3d3ff8a2f.png"},{"id":71003717,"identity":"7307cc03-0f3d-47af-a71e-7c93b3dd2bdd","added_by":"auto","created_at":"2024-12-10 06:06:49","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":23852,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"fig7.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/fe8706f9125b06b8b8c4b673.png"},{"id":71002503,"identity":"c8676aff-49c3-408f-a33e-6173a05a5c03","added_by":"auto","created_at":"2024-12-10 05:58:49","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":25010,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"fig8.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/412cf923455f3a48736712c1.png"},{"id":71002507,"identity":"4c284db1-2ba3-4d1f-9a03-df40cff5ec2e","added_by":"auto","created_at":"2024-12-10 05:58:49","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":27853,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"fig9.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/08ea7d831adf0000ac8a85ae.png"},{"id":71002506,"identity":"efd37665-b2a5-43f7-af30-d65cc48070a6","added_by":"auto","created_at":"2024-12-10 05:58:49","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":27871,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"fig10.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/6d8051c6e9e38dcfcd9400b5.png"},{"id":71002511,"identity":"c42ef9b8-c872-469a-adda-fe6aa8661e97","added_by":"auto","created_at":"2024-12-10 05:58:49","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":29006,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"fig11.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/31fcb19f4215da85ccfd50fa.png"},{"id":71002501,"identity":"a50f9cfc-30a2-4b98-a955-39d31ecf4d47","added_by":"auto","created_at":"2024-12-10 05:58:49","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":29075,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"fig12.png","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/d6da4e950e8cc88708fcf97a.png"},{"id":78605088,"identity":"68931f7f-4fcd-4c87-bcdb-2430ccdc7583","added_by":"auto","created_at":"2025-03-16 11:01:49","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1368505,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5406276/v1/47e38f73-14e6-4acc-8ffe-6c1a41693cf4.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eDeveloping a prediction model for in-hospital mortality in sepsis patients with gastrointestinal bleeding using the MIMIC-IV database\u003c/p\u003e","fulltext":[{"header":"1. Background","content":"\u003cp\u003eSepsis is a life-threatening illness caused by an unregulated immunological response to infection, which results in widespread inflammation, immune dysfunction, and organ failure. Its pathophysiology involves a cascade of immune and metabolic disruptions. Without prompt treatment, sepsis rapidly progresses to multi-organ failure and death. Early detection and intervention are essential to improve survival, especially in critically ill patients. It is a leading cause of death and morbidity globally. The incidence of sepsis among hospitalized patients ranges from 1\u0026ndash;3%, while in intensive care unit patients, it can increase to 10\u0026ndash;30% (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). It is estimated that every year, 31.5\u0026nbsp;million individuals globally are afflicted by sepsis, with 19.4\u0026nbsp;million having severe sepsis, resulting in around 5.3\u0026nbsp;million deaths(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). Gastrointestinal symptoms are prevalent during ICU admission, with up to 62% of patients showing at least one gastrointestinal symptom lasting for a minimum of one day. Gastrointestinal hemorrhage is one of the most common presentations, with a greater incidence and fatality rate than those without gastrointestinal bleeding(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). The occurrence of gastrointestinal complications is linked to a poor prognosis in critically ill patients; on average, 4% of these patients experience gastrointestinal hemorrhage (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). 5.4% of patients with septic shock develop gastrointestinal bleeding (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). The main causes of bleeding include stress ulcers in the esophagus, stomach, or duodenum, gastrointestinal mucosal damage due to antibiotics and inflammatory mediators, anticoagulant therapy, and intestinal ischemia resulting from sepsis(\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). The in-hospital death rate in septic patients with gastrointestinal bleeding is much higher, highlighting the important need for early screening and risk assessment in these patients. For sepsis patients, the SAPS II score, SOFA score and APS III score are the main scoring systems currently utilized. In contrast, scoring systems for gastrointestinal bleeding include the Rockall, Glasgow-Blatchford, and AIMS65 scoring systems, each with its own limitations. Recent studies indicate that the APACHE II score demonstrates excellent prognostic accuracy for estimating the likelihood of death among ICU patients (AUC: 0.87, CI: 0.75\u0026ndash;0.99), while the SOFA score shows moderate accuracy (AUC: 0.71, CI: 0.50\u0026ndash;0.93). However, no scoring system has been created to predict death in individuals with upper gastrointestinal hemorrhage. All grading methods have limited predictive accuracy for ICU duration of stay(\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). Therefore, developing a predictive tool to evaluate the likelihood of death during hospitalization in septic patients with gastrointestinal bleeding is of substantial clinical value. The goal of this study is to develop a model that is more specifically tailored to septic patients with gastrointestinal bleeding, providing a more precise early assessment tool for clinical use, facilitating early intervention based on the identified independent risk factors, and ultimately improving patient outcomes.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e2.1 Research data\u003c/h2\u003e\n \u003cp\u003eThe Medical Information Mart for Intensive Care IV (MIMIC-IV, v. 3.0) database provided the data used in this investigation. It is a huge, publicly available database including clinical data on adult patients (aged 18 and up) admitted to the ICU of a major tertiary hospital in the United States between 2008 and 2019. It has more than 70,000 ICU admission cases (\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e). We successfully completed the Collaborative Institutional Training Initiative (CITI) program (Researcher Approval Code: 65073876) and received authorization to use the MIMIC-IV database. The Beth Israel Deaconess Medical Center and Massachusetts Institute of Technology institutional review boards have allowed access to the MIMIC-IV database. All patient data in the database is anonymised, thus informed permission is not necessary. We extracted relevant diagnostic, laboratory, demographic data, and associated treatments from the database using Structured Query Language in Navicat Premium 17, and then linked the extracted data to each patient\u0026apos;s unique HADM_ID. Sepsis was diagnosed based on a SOFA score of \u0026ge;\u0026thinsp;2 (\u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e). Patients with gastrointestinal bleeding were diagnosed using the International Classification of Diseases, Ninth Revision (ICD-9) and Tenth Revision (ICD-10), with no differentiation between upper and lower gastrointestinal bleeding.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003e2.3 Exclusion criteria\u003c/h2\u003e\n \u003cp\u003e(\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e) Under 18 years old; (\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e) Missing hospitalization data\u0026thinsp;\u0026gt;\u0026thinsp;20% ;(\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e) Hospital stay of less than 24 hours; (\u003cspan class=\"CitationRef\"\u003e4\u003c/span\u003e) For recurrent ICU hospitalizations, only data from the first admission were included.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e2.4 Collected indicators\u003c/h2\u003e\n \u003cp\u003eTo conduct a thorough analysis of the clinical determinants influencing patient outcomes, we systematically extracted the mean values of a wide range of clinical indicators from the MIMIC-IV database, with a focus on data collected within the first 24 hours of ICU admission for eligible patients. These included demographic characteristics (age and sex), vital signs (body temperature [T], respiratory rate [RR], heart rate [HR], blood oxygen saturation [SpO2], systolic blood pressure [SBP], and diastolic blood pressure [DBP]), and laboratory parameters (bicarbonate [HCO3], blood urea nitrogen [BUN], sodium [Na], creatinine [Cr], potassium [K], prothrombin time [PT], activated partial thromboplastin time [APTT], white blood cell count [WBC], glucose [Glu], anion gap [AG], international normalized ratio [INR], mean corpuscular hemoglobin concentration [MCHC], mean corpuscular hemoglobin [MCH], red blood cell count [RBC], platelet count [PLT]). In addition, we documented comorbid conditions (diabetes mellitus [DM], myocardial infarction [MI], congestive heart failure [CHF], acute kidney injury [AKI], cerebrovascular disease [CVD], mild liver disease [MLD], malignancy, chronic kidney disease [CKD], chronic lung disease [CLD], and severe liver disease [SLD]), and key severity scores, such as the Acute Physiology Score III (APS III), Sequential Organ Failure Assessment (SOFA), Glasgow Coma Scale (GCS), Simplified We also gathered data on treatment regimens and pharmaceutical use, such as epinephrine (EPI), dopamine (DA), vasopressin (VP), norepinephrine (NE), invasive mechanical ventilation (IV), non-invasive ventilation (NIV), and continuous renal replacement therapy (CRRT).\u003c/p\u003e\n \u003cp\u003eThis cohort study tracked in-hospital mortality, with a final cohort of 1,923 patients, of whom 541 died during hospitalization.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n \u003ch2\u003e2.5 Statistical analysis\u003c/h2\u003e\n \u003cp\u003eTo begin, the normality of continuous data was examined using the Shapiro-Wilk test. In the baseline characteristics table, variables with normal distributions were reported as means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviations. The independent t-test was then used to compare group averages and assess intergroup differences. For data that did not follow a normal distribution, values were given as median and interquartile range, and intergroup comparisons were performed using the Wilcoxon rank-sum test. Categorical variables were examined using the chi-square test. Furthermore, LASSO regression was employed for the preliminary identification of risk factors associated with in-hospital death in patients with gastrointestinal bleeding and sepsis. Finally, multivariable logistic regression was used to determine the final independent variables for model formation, which were represented by a nomogram. Variables with P\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were considered to have statistical significance. For variables with missing data of less than 20%, the \u0026ldquo;MICE\u0026rdquo; package in R software was employed for multiple imputation to predict and fill in the missing values. R software version 4.1.1 was used for all data analyses and visualizations.\u003c/p\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 Baseline comparison\u003c/h2\u003e\n \u003cp\u003eFirst, 2,146 sepsis complicated by gastrointestinal bleeding (GIB) patients (upper and lower gastrointestinal hemorrhage) were extracted from the MIMIC-IV data base from 2008 to 2019. For patients with multiple admissions, only data from their first hospitalization were included. 215 participants in all were disqualified from the research because they spent less than 24 hours in the intensive care unit, which was deemed insufficient for assessing outcomes related to sepsis and gastrointestinal hemorrhage. Additionally, 8 patients were excluded because of missing data that exceeded 20%, which could compromise the reliability and validity of the analysis. As illustrated in the study flowchart (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e), the final cohort consisted of 1,923 patients, of whom 541 experienced in-hospital mortality. To ensure robust model development and validation, the patients were randomly divided into two cohorts: the training cohort (n\u0026thinsp;=\u0026thinsp;1,346) for model development, and the validation cohort (n\u0026thinsp;=\u0026thinsp;577) for independent performance evaluation. This division allows for the assessment of model generalizability and accuracy across different patient populations within the study. The patients\u0026apos; median age in the training cohort was 65.45 years (the range between the 25th and 75th percentiles was 54.88, 78.33), with 824 males (61.21%). I In the validation cohort, the patients\u0026apos; median age was 64.52 years (the range between the 25th and 75th percentiles were 53.91, 77.08), with 357 males (61.87%). In our study cohort grouping, 365 (27.1%) and 176 (30.5%) patients died during hospitalization, respectively, and there was no statistically significant difference in mortality rates between the groups. As shown in Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e(The table exceeds one page and should be placed at the end of the text), the baseline characteristics of both cohorts were comparable, indicating similar demographic and clinical profiles.\u003c/p\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e:Baseline characteristics of patients with sepsis combined with gastrointestinal bleeding.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVariables\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eTotal(n\u0026thinsp;=\u0026thinsp;1923)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eValidation(n\u0026thinsp;=\u0026thinsp;577)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eTraining(n\u0026thinsp;=\u0026thinsp;1346)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eP\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge(years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e65.06(54.53,77.82)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e64.53(53.91,77.09)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e65.45(54.88,78.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eINR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.45(1.23,1.85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.47(1.25,1.90)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.45(1.20,1.85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.607\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePT(s)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e15.90(13.65,20.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e16.00(13.85,20.58)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e15.86(13.61,20.06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.417\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePTT(s)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e33.25(28.26,43.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e33.25(28.30,43.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e33.27(28.20,43.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.892\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHCO3(mmol/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e21.50(18.90,24.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e21.50(18.00,24.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e21.67(19.00,24.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.121\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBUN(mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e31.67(19.50,53.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e30.00(18.50,50.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e32.00(20.00,54.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.138\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eK(mmol/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e4.20(3.80,4.70)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e4.14(3.75,4.70)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e4.20(3.83,4.70)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.071\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNa(mmol/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e138.50(135.33,141.67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e138.00(135.00,141.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e138.50(135.50,142.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.373\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGlu(mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e130.50(107.00,168.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e129.00(106.25,164.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e131.00(107.75,168.31)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.499\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAG(mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e14.50(12.00,17.75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e14.75(12.00,18.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e14.50(12.00,17.67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.265\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCr(mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.30(0.80,2.24)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.27(0.80,2.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.30(0.83,2.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.492\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWBC(\u0026times;10\u003csup\u003e9\u003c/sup\u003e/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e11.47(7.65,16.23)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e11.75(7.92,16.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e11.40(7.55,16.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.51\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMCHC(g/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e33.20(32.13,34.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e33.27(32.20,34.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e33.20(32.10,34.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.112\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRBC(\u0026times;10\u003csup\u003e12\u003c/sup\u003e/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.10(2.69,3.60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.11(2.69,3.61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.09(2.69,3.60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.963\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePLT(\u0026times;10\u003csup\u003e9\u003c/sup\u003e/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e144.60(86.06,223.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e152.00(88.33,221.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e141.93(84.69,223.44)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.401\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMCH(pg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e30.50(29.07,32.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e30.68(29.32,32.23)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e30.49(28.97,32.03)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.037\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWeight(kg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e78.20(66.00,94.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e78.90(66.70,93.80)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e78.00(65.70,94.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.672\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHR(times/min)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e88.15(76.88,99.68)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e88.04(75.71,99.36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e88.22(77.36,99.76)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.37\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSBP(mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e111.19(103.00,122.64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e110.27(102.64,122.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e111.60(103.09,122.88)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDBP(mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e59.96(54.23,67.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e60.09(54.61,66.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e59.85(54.11,67.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.725\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRR(times/min)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e19.20(16.71,22.32)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e19.19(16.43,22.09)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e19.21(16.79,22.38)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.501\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eT(\u0026deg;C)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e36.79(36.55,37.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e36.78(36.55,37.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e36.80(36.55,37.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.646\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSpo2(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e97.46(95.92,98.75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e97.56(95.88,98.89)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e97.42(95.92,98.71)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.633\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAPSIII\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e56.00(43.00,71.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e55.00(44.00,72.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e56.00(43.00,71.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.812\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLODS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6.00(4.00,9.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6.00(4.00,9.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6.00(4.00,9.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.768\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eOASIS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e35.00(30.00,41.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e35.00(30.00,41.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e35.00(29.00,41.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.885\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSOFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6.00(4.00,9.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e7.00(4.00,10.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6.00(4.00,9.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.086\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSAPSII\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e43.00(34.00,53.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e44.00(34.00,55.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e42.00(34.00,53.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.043\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGCS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e15.00(15.00,15.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e15.00(15.00,15.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e15.00(15.00,15.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.02\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSIRS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.00(2.00,3.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.00(2.00,3.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.00(2.00,3.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.897\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDeath, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1382(71.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e401(69.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e981(72.88)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e541(28.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e176(30.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e365(27.12)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGender, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.787\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e742(38.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e220(38.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e522(38.78)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1181(61.41)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e357(61.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e824(61.22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMI, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.884\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1610(83.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e482(83.54)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1128(83.80)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e313(16.28)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e95(16.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e218(16.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCHF, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1392(72.39)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e441(76.43)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e951(70.65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e531(27.61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e136(23.57)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e395(29.35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCVD, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.999\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1693(88.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e508(88.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1185(88.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e230(11.96)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e69(11.96)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e161(11.96)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCLD, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.965\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1451(75.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e435(75.39)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1016(75.48)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e472(24.54)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e142(24.61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e330(24.52)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMLD, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.358\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1087(56.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e317(54.94)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e770(57.21)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e836(43.47)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e260(45.06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e576(42.79)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDM, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.489\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1325(68.90)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e404(70.02)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e921(68.42)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e598(31.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e173(29.98)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e425(31.58)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMalignancy, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.028\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1623(84.40)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e471(81.63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1152(85.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e300(15.60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e106(18.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e194(14.41)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRD, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.802\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1409(73.27)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e425(73.66)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e984(73.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e514(26.73)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e152(26.34)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e362(26.89)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAKI, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.945\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e325(16.90)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e97(16.81)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e228(16.94)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1598(83.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e480(83.19)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1118(83.06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSLD, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.765\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1256(65.31)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e374(64.82)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e882(65.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e667(34.69)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e203(35.18)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e464(34.47)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEPI:, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.076\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1831(95.22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e557(96.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1274(94.65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e92(4.78)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e20(3.47)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e72(5.35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDA, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.701\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1848(96.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e553(95.84)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1295(96.21)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e75(3.90)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e24(4.16)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e51(3.79)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNE, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.403\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1187(61.73)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e348(60.31)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e839(62.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e736(38.27)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e229(39.69)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e507(37.67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVP, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.384\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1591(82.74)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e484(83.88)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1107(82.24)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e332(17.26)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e93(16.12)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e239(17.76)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIV, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.462\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e851(44.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e248(42.98)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e603(44.80)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1072(55.75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e329(57.02)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e743(55.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNIV, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.722\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e611(31.77)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e180(31.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e431(32.02)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1312(68.23)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e397(68.80)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e915(67.98)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCRRT, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.244\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1663(86.48)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e507(87.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1156(85.88)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e260(13.52)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e70(12.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e190(14.12)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2 Variable Selection via LASSO Regression\u003c/h2\u003e\n \u003cp\u003eIn this study, the LASSO regression technique with 10-fold cross-validation was used to assess 48 variables, including demographic characteristics such as age, gender, and various laboratory indicators, to identify factors associated with in-hospital mortality in patients with severe sepsis and gastrointestinal bleeding (GIB). The model found 15 factors as the most important predictors at the optimal lambda value (\u0026lambda;\u0026thinsp;=\u0026thinsp;0.03319349). The selected variables include: anion gap, activated partial thromboplastin time, prothrombin time, respiratory rate, body temperature, presence of malignancy, SOFA score, APS III score, LODS score, SAPS II score, administration of vasopressors, use of norepinephrine, non-invasive mechanical ventilation, AKI and CRRT. Figure \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e depicts the variations in the LASSO regression coefficients for the 48 variables, with each curve showing the trajectory of the coefficient for each independent variable. Figure \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e outlines the procedure for selecting the optimal parameter value through cross-validation within the LASSO model.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003e3.3 Independent influencing factors selected using multivariable logistic Regression\u003c/h2\u003e\n \u003cp\u003eBased on the fifteen prognostic factors selected by LASSO regression, A multivariable logistic regression was used to find the independent variables linked with prognosis. As presented in Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e, nine prognostic factors were ultimately recognized as key contributors to the model. An odds ratio (OR) of 3.02 (95% CI: 1.78\u0026ndash;5.42) in our study showed that patients with acute kidney injury (AKI) had a significantly higher chance of dying in the hospital. This finding underscores the significant impact of AKI on the likelihood of mortality during hospitalization, suggesting that AKI serves as a strong independent predictor of adverse outcomes in critically ill patients. This finding underscores the importance of early detection and intervention in managing AKI to mitigate its adverse impact on patient outcomes. Similarly, malignancy was found to be a critical factor influencing in-hospital mortality. Patients with a documented history of cancer had a substantially greater in-hospital death rate than those without cancer, with an odds ratio (OR) of 1.92 (95% CI: 1.29\u0026ndash;2.87). This finding underscores the elevated and compounded risk that malignancy confers, particularly in critically ill patients with complex comorbidities such as sepsis and gastrointestinal bleeding. The interaction between cancer and acute illness appears to exacerbate patient vulnerability, highlighting the need for more intensive monitoring and tailored therapeutic approaches in this high-risk population. Malignancy in combination with sepsis and gastrointestinal bleeding significantly worsens prognosis, and clinical management must account for these risks. Hemodynamically unstable patients who required vasopressor support had an in-hospital mortality rate 2.20 times greater (OR\u0026thinsp;=\u0026thinsp;2.20, 95% CI: 1.44\u0026ndash;3.36). This result emphasizes how serious circulatory impairment is and how difficult it is to treat these individuals who are in severe condition. It suggests that the use of vasopressors as a therapeutic intervention is reflective of profound hemodynamic instability, which correlates with worse outcomes. In terms of specific biomarkers and clinical measurements, several factors were identified as independent predictors of in-hospital mortality. These included PT (OR\u0026thinsp;=\u0026thinsp;1.03, 95% CI: 1.01\u0026ndash;1.05), activated partial thromboplastin time (OR\u0026thinsp;=\u0026thinsp;1.01, 95% CI: 1.00\u0026ndash;1.02), RR(OR\u0026thinsp;=\u0026thinsp;1.06, 95% CI: 1.02\u0026ndash;1.10), and the Acute Physiology and Chronic Health Evaluation III (APS III) score (OR\u0026thinsp;=\u0026thinsp;1.01, 95% CI: 1.00\u0026ndash;1.02). These clinical parameters, which reflect coagulopathy, respiratory dysfunction, and the overall severity of illness, were found to be significant contributors to mortality risk. The presence of abnormal values in these indicators can help clinicians identify patients at higher risk for mortality and prioritize interventions accordingly. Interestingly, we also identified protective factors that were inversely associated with mortality. Specifically, temperature (OR\u0026thinsp;=\u0026thinsp;0.72, 95% CI: 0.55\u0026ndash;0.94) and non-invasive mechanical ventilation (OR\u0026thinsp;=\u0026thinsp;0.33, 95% CI: 0.25\u0026ndash;0.45) were associated with lower in-hospital mortality. A higher body temperature, likely indicative of an active immune response or the presence of systemic inflammatory reactions, appeared to confer a protective effect, while non-invasive mechanical ventilation was associated with better outcomes by reducing the need for invasive procedures and supporting respiratory function. These findings suggest that maintaining appropriate body temperature and optimizing respiratory support strategies may improve survival rates for critically ill patients with sepsis and gastrointestinal bleeding.\u003c/p\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eMultivariable analysis base on LASSO regression\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVariables\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eB\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eOR\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e95%CI\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eP value\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.031\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.032\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.012\u0026ndash;1.052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePTT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.003\u0026ndash;1.020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.007\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.059\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.060\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.023\u0026ndash;1.098\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003etemperature\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e-0.326\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.722\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.551\u0026ndash;0.946\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.018\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMalignant cancer\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.654\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.924\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.290\u0026ndash;2.871\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAPSIII\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.001\u0026ndash;1.024\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.031\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLODS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.036\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.037\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.959\u0026ndash;1.120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.356\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSAPSII\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.990\u0026ndash;1.120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.422\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003enorepinephrine\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.255\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.291\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.887\u0026ndash;1.878\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.182\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSOFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e-0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.991\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.935\u0026ndash;1.050\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.755\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003evasopressin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.787\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.197\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.440\u0026ndash;3.350\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNoninvasive ventilator\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e-1.099\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.333\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.246\u0026ndash;0.450\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCRRT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.303\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.354\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.895\u0026ndash;2.049\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.151\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAKI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.106\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.736\u0026ndash;5.258\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.033\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.033\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.999\u0026ndash;1.069\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.058\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e3.4 Model construction and validation\u003c/h2\u003e\n \u003cp\u003eUsing the nine most important indicators found by multivariable logistic regression analysis and represented by a nomogram, a prediction model for the likelihood of in-hospital death in intensive care unit patients with sepsis and gastrointestinal bleeding was created. Each independent predictor included in the nomogram was assigned a weighted score, with the cumulative score ranging from 0 to 400 points. This scoring system was designed to reflect the relative contribution of each factor to the risk of in-hospital mortality. The total score derived from the nine key predictive factors is directly correlated with the probability of mortality, with values ranging from 0.1 to 0.9. Specifically, higher cumulative scores indicate an elevated risk of in-hospital mortality, underscoring the importance of these factors in stratifying patients based on their clinical prognosis. The relationship between total score and predicted mortality probability is illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e, which visually demonstrates the gradual increase in mortality risk as the total score rises.\u003c/p\u003e\n \u003cp\u003eWe initially computed the area under the receiver operating characteristic curve (AUC), a popular and extensively used metric for evaluating the discriminative ability of diagnostic models, in order to analyze the nomogram\u0026apos;s predictive performance. Within the training group, the AUC value of 0.827 (95% CI: 0.8018\u0026ndash;0.8515) was obtained, reflecting the model\u0026apos;s robust ability to distinguish between patients with varying risks of in-hospital mortality, and indicating a high level of predictive accuracy. With an AUC of 0.7961 (95% CI: 0.7577\u0026ndash;0.8345) in the validation cohort, consistent performance across datasets was shown. These findings, presented in Figs. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e and \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e, further validate the robustness and generalizability of the nomogram in predicting in-hospital mortality risk in ICU patients with gastrointestinal bleeding and sepsis.\u003c/p\u003e\n \u003cp\u003eOur nomogram surpassed the APS III scoring system in predictive accuracy, showing significantly better AUCs across both cohorts. Using the APS III scoring technique, The AUC for the validation group was 0.7308 (95% CI: 0.6869\u0026ndash;0.7746), while the training set\u0026apos;s was 0.7531 (95% CI: 0.7242\u0026ndash;0.7821). This data highlights our model\u0026apos;s superior discriminative capacity over the APS III score in predicting in-hospital mortality for patients with sepsis and gastrointestinal bleeding in critical care units.\u003c/p\u003e\n \u003cp\u003eWe used decision curve analysis (DCA), which measures the net benefit of using the model at various probability thresholds, to further assess the algorithm\u0026apos;s clinical value. To further demonstrate the accuracy and dependability of the model, calibration curves were also created to evaluate the agreement between the observed and anticipated probability of in-hospital mortality. To evaluate the accuracy of the model\u0026apos;s predictions, calibration curves were constructed for both the training and validation cohorts (Figs. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e and \u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003e). These curves assess the degree of agreement between the predicted probabilities of in-hospital mortality and the actual observed outcomes. A well-calibrated model should exhibit a close alignment between predicted and observed values across a range of predicted probabilities, which was thoroughly assessed in both cohorts. In these curves, the y-axis represents the actual event probability observed in the study cohort, while the x-axis depicts the model\u0026rsquo;s predicted probabilities. A perfect model would result in all data points lying along the diagonal line, indicating perfect agreement between predicted and observed values. In our study, the calibration plots for the training and validation groups exhibited a strong alignment with this diagonal line, reflecting a high degree of concordance between the predicted probabilities of in-hospital mortality and the actual observed outcomes. The model\u0026apos;s clinical application and reliability in predicting In-hospital death rate for sepsis and gastrointestinal bleeding are further supported by the close correlation between predicted and actual values, which highlights the model\u0026apos;s outstanding calibration and exceptional fit.\u003c/p\u003e\n \u003cp\u003eIn the DCA curves (Figs. \u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003e and \u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003e), the gray diagonal line represents all patients undergoing intervention, while the black parallel line represents no intervention. The training cohort\u0026apos;s best risk threshold for optimizing net benefit was less than 94%, and the DCA curve demonstrated that our prediction model outperformed the APS III score in terms of net benefits over a range of thresholds.\u003c/p\u003e\n \u003cp\u003eAdditionally, we constructed clinical impact curve (CIC) to assess the real-world applicability and predictive accuracy of our model. As shown in Figs. \u003cspan class=\"InternalRef\"\u003e11\u003c/span\u003e and \u003cspan class=\"InternalRef\"\u003e12\u003c/span\u003e, Our findings indicated that when the threshold probability surpassed 70%, the model\u0026rsquo;s identification of high-risk patients for in-hospital mortality closely matched the observed clinical outcomes, highlighting the model\u0026rsquo;s robust predictive power. This result further reinforces the clinical applicability of the predictive model, suggesting its potential to guide decision-making and enhance patient management in critical care environments. The robustness of this model in both the training and validation cohorts reinforces its generalizability and reliability in diverse clinical environments.\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eDeveloping a prediction model to estimate the likelihood of in-hospital death in patients with sepsis and gastrointestinal bleeding was the aim of this retrospective cohort study. Validation of this model demonstrated good predictive ability and discriminative power. Patients with sepsis combined with GIB exhibited higher mortality rates, greater disease severity, poorer clinical outcomes, longer hospital stays, and higher hospitalization costs.\u003c/p\u003e\n\u003cp\u003eIn this research, we established a framework integrating various clinical parameters to forecast the outcomes of ICU patients with gastrointestinal bleeding (GIB) and sepsis. The model identified nine independent risk factors for mortality: PT, APTT, RR, APS III score, AKI, malignancy, and the use of vasopressors. Notably, non-invasive mechanical ventilation and a body temperature below 40\u0026deg;C were found to be protective factors.\u003c/p\u003e\n\u003cp\u003eThe APS III score is a component of the APACHE scoring system, which focuses on acute physiological parameters. It includes various physiological indicators, generating a total score that reflects the severity of the patient\u0026apos;s condition. This scoring system is used to assess the physiological status of critically ill patients, with higher scores indicating more severe illness and a worse prognosis. Research has shown that APS III outperforms both the SOFA and SAPS II scores in overall prognostic accuracy(\u003cspan class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e14\u003c/span\u003e). These findings provide strong evidence for the inclusion of the Acute Physiology Score III (APS III) as a key risk factor in predicting patient outcomes. Additionally, our analysis highlights PT and APTT as significant independent predictors of in-hospital mortality, further emphasizing the critical role of coagulation parameters in assessing patient prognosis. Currently, inflammation and coagulopathy are considered key contributors to multiple organ dysfunction, with sepsis exacerbating these disturbances by inducing complex dysregulations across various systems, including the coagulation cascade(\u003cspan class=\"CitationRef\"\u003e15\u003c/span\u003e). Coagulopathy is a common complication of sepsis and a critical host response to infection, which may progress to disseminated intravascular coagulation (DIC), thereby increasing mortality. Diagnostic standards for sepsis-induced coagulopathy and overt disseminated intravascular coagulation have been set by the International Society on Thrombosis and Hemostasis (ISTH) (\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e). Overt disseminated intravascular coagulation typically indicates decompensated coagulopathy, while sepsis-induced coagulopathy represents systemic activation of coagulation, with compensatory coagulation function still intact. Once it progresses to overt DIC, the benefits of anticoagulant therapy for disease progression may become minimal. One study suggested that, in sepsis, fibrinolysis is inhibited, and levels of fibrin/fibrinogen and D-dimer do not increase in proportion to disease severity. In contrast, prothrombin time prolongs in a linear fashion with increasing severity, which is consistent with the outcome variables we selected(\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eFurthermore, we observed that an elevated respiratory rate could be indicative of early organ dysfunction, particularly reflecting respiratory distress or failure, which is closely linked to adverse mortality outcomes. In our study, the mean respiratory rate recorded within the first 24 hours of ICU admission (OR\u0026thinsp;=\u0026thinsp;1.008, 95% CI\u0026thinsp;=\u0026thinsp;1.002\u0026ndash;1.015, p\u0026thinsp;=\u0026thinsp;0.014) was identified as a significant predictor of 30-day all-cause mortality in patients with sepsis complicated by gastrointestinal bleeding (GIB). An elevated respiratory rate generally indicates metabolic imbalance and inadequate oxygenation, resulting in insufficient tissue perfusion and worsening tissue hypoxia. Consequently, this impairs intestinal barrier integrity, heightening the risk of gastrointestinal bleeding and increasing intestinal permeability, which promotes bacterial translocation and inflammatory responses, thereby further increasing mortality risk among sepsis patients. Numerous studies have underscored the negative influence of fever on patient prognosis, particularly in the context of neurological conditions. This has led to the widespread adoption of both pharmacological therapies and physical interventions aimed at controlling elevated body temperature, as effective management of fever is considered crucial in mitigating adverse outcomes(\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e). Surprisingly, our analysis indicated that an early elevation in body temperature may act as a potential protective factor in patients with sepsis and gastrointestinal bleeding (OR\u0026thinsp;=\u0026thinsp;1.008, 95% CI: 1.002\u0026ndash;1.015, p\u0026thinsp;=\u0026thinsp;0.014). This finding suggests that a modest rise in temperature during the early stages of illness might be associated with better patient outcomes, potentially reflecting an adaptive immune response or a controlled inflammatory reaction. Research indicates that fever plays a crucial role in enhancing microbial clearance, activating immune responses, and stimulating heat shock protein production. Most human pathogens thrive at temperatures ranging from 35\u0026deg;C to 37\u0026deg;C(\u003cspan class=\"CitationRef\"\u003e20\u003c/span\u003e), and a rise in body temperature may inhibit microbial proliferation. Additionally, research has shown that elevated body temperature can strengthen the host\u0026rsquo;s immune response by increasing antibiotic sensitivity, lowering minimum inhibitory concentrations, enhancing neutrophil mobility, and promoting phagocytosis. It also facilitates T-helper cell adhesion and helps prevent lymphocyte depletion. However, it should be noted that at elevated fever levels (40\u0026ndash;41\u0026deg;C), the beneficial immunomodulatory effects may be offset by the negative metabolic and inflammatory consequences of fever (\u003cspan class=\"CitationRef\"\u003e21\u003c/span\u003e). In our nomogram, we chose 40 degrees Celsius as the baseline for the scale, which appears to be a reasonable selection based on these factors. Furthermore, a meta-analysis has demonstrated a negative correlation between body temperature and clinical outcomes in sepsis (\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e). Analysis of initial body temperature and prognostic outcomes in specific emergency department admissions further suggests that low body temperature is linked to prolonged initial antibiotic administration, extended hospital stays, higher rates of ICU admission, and elevated mortality(\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e). Patients with malignancies often experience immune dysfunction, malnutrition, and other comorbidities, which render them more vulnerable to infections, thereby increasing the risk of sepsis. The development of sepsis further worsens the condition of cancer patients, leading to poorer prognoses. Our data demonstrate that septic patients with gastrointestinal bleeding exhibit high mortality rates, irrespective of cancer status; However, the mortality rate is significantly higher in cancer patients when compared with non-cancer patients. This result aligns with the findings from a supplementary study, which included over 19\u0026nbsp;million hospitalized septic patients from the National Inpatient Sample database between 2008 and 2017. In that study, 20.4% of septic cases were associated with cancer, and the in-hospital mortality rate for cancer-related sepsis was higher than that for non-cancer patients(\u003cspan class=\"CitationRef\"\u003e24\u003c/span\u003e). Additionally, a meta-analysis indicated that cancer substantially elevates the mortality risk in septic patients (OR\u0026thinsp;=\u0026thinsp;2.7, 95% CI: 1.07\u0026ndash;6.84). However, the link between cancer and early mortality was not significant (OR\u0026thinsp;=\u0026thinsp;2.77, 95% CI: 0.88\u0026ndash;8.66), whereas the risk of late mortality was markedly higher (OR\u0026thinsp;=\u0026thinsp;2.46, 95% CI: 1.42\u0026ndash;4.25) (\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e). Sepsis complicated by acute kidney injury (AKI) is frequently observed in critically ill patients. Research indicates that septic patients suffering from AKI display a marked increase in mortality in comparison with those without kidney impairment. The deterioration of renal function can precipitate dysfunction in other vital organs, including the heart, lungs, and liver. Additionally, fluid management in the context of AKI becomes more challenging; excessive fluid resuscitation may lead to complications such as heart failure or pulmonary edema, while insufficient fluid administration can worsen renal ischemia, thereby amplifying the overall mortality risk in septic patients(\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e27\u003c/span\u003e). The \u0026quot;Save Sepsis\u0026quot; initiative has highlighted the importance of non-invasive ventilation (NIV): the 2021 International Guidelines for the Management of Sepsis and Septic Shock suggest that NIV may offer comparable physiological advantages to invasive positive pressure ventilation, such as enhanced gas exchange and decreased respiratory effort in certain patients, while mitigating complications related to intubation, invasive ventilation, and the need for sedation(\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e). Several studies further support the notion that failure of non-invasive mechanical ventilation is a risk factor for mortality in this cohort (\u003cspan class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e). Vasopressor agents are essential for sustaining hemodynamic stability; however, their excessive or improper use may result in uneven blood flow distribution, potentially worsening gastrointestinal injury. Currently, norepinephrine is the preferred first-line vasopressor for adults with septic shock, as compared to other vasopressors. Norepinephrine causes vasoconstriction and elevates mean arterial pressure with minimal effect on heart rate(\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e). Furthermore, there was no discernible difference in 28-day mortality rates (39.3% vs. 35.4%, p\u0026thinsp;=\u0026thinsp;0.26) in the VASST study, which contrasted norepinephrine alone with norepinephrine\u0026thinsp;+\u0026thinsp;vasopressin (0.01\u0026ndash;0.03 U/min)(\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e) Furthermore, research has demonstrated that while individuals treated with dopamine as the first-line vasopressor and those treated with norepinephrine do not significantly vary in effectiveness, dopamine treatment is associated with a greater rate of adverse events.(\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e). Therefore, vasopressin may not be the preferred option.\u003c/p\u003e\n\u003cp\u003eOur study has some inherent limitations. First, during data collection, we excluded certain variables with significant missing values, such as height, C-reactive protein levels, and the severity of acute kidney injury. These missing data may contain important influencing factors. Second, while our model demonstrated strong validation results within the MIMIC-IV database, external validation using datasets beyond MIMIC is necessary to confirm its generalizability. Third, as our study is a retrospective cohort analysis, further prospective studies are required before the nomogram can be applied in clinical practice.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn order to identify high-risk individuals early and promote prompt medical measures, we have created a prediction model that accurately assesses hospital mortality in sepsis patients with gastrointestinal bleeding.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEach author made a contribution to the design and conceptualization of the study. In addition to writing and editing the publication, FWY was in charge of developing the protocol and the overall study strategy. YM contributed significantly to the academic process by taking part in data collecting, statistical analysis, and result interpretation. To guarantee the accuracy of the study content, JL helped with the literature search and the arrangement of pertinent background materials. ZQZ ensured the analysis\u0026apos;s dependability by managing data and visualizing the findings. LHW contributed clinical knowledge and took part in talks about the trial design. ZJH ensured the manuscript\u0026apos;s academic quality and logical consistency by doing the final review and editing. Following a discussion of the study findings, all authors agreed on the manuscript\u0026apos;s substance.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNo particular grant from a governmental, private, or nonprofit funding organization was obtained for this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe writers say they have no conflicting interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe MIMIC-IV 3.0 database contains the datasets created and/or examined during the current investigation (https://physionet.org/content/mimiciv/3.0/).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAn investigation of anonymised publically accessible datasets with prior institutional review board permission was the study\u0026apos;s methodology.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSinger M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). Jama. 2016;315(8):801-10.\u003c/li\u003e\n\u003cli\u003eO\u0026apos;Brien JM, Jr., Ali NA, Aberegg SK, Abraham E. Sepsis. Am J Med. 2007;120(12):1012-22.\u003c/li\u003e\n\u003cli\u003eFleischmann C, Scherag A, Adhikari NK, Hartog CS, Tsaganos T, Schlattmann P, et al. Assessment of Global Incidence and Mortality of Hospital-treated Sepsis. Current Estimates and Limitations. Am J Respir Crit Care Med. 2016;193(3):259-72.\u003c/li\u003e\n\u003cli\u003eKate V, Sureshkumar S, Gurushankari B, Kalayarasan R. Acute Upper Non-variceal and Lower Gastrointestinal Bleeding. J Gastrointest Surg. 2022;26(4):932-49.\u003c/li\u003e\n\u003cli\u003eKhamaysi I, Gralnek IM. Acute upper gastrointestinal bleeding (UGIB) - initial evaluation and management. Best Pract Res Clin Gastroenterol. 2013;27(5):633-8.\u003c/li\u003e\n\u003cli\u003eYe Z, Reintam Blaser A, Lytvyn L, Wang Y, Guyatt GH, Mikita JS, et al. Gastrointestinal bleeding prophylaxis for critically ill patients: a clinical practice guideline. Bmj. 2020;368:l6722.\u003c/li\u003e\n\u003cli\u003eSiddiqui AH, Ahmed M, Khan TMA, Abbasi S, Habib S, Khan HM, et al. Trends and Outcomes of Gastrointestinal Bleeding Among Septic Shock Patients of the United States: A 10-Year Analysis of a Nationwide Inpatient Sample. Cureus. 2020;12(5):e8029.\u003c/li\u003e\n\u003cli\u003eKamboj AK, Hoversten P, Leggett CL. Upper Gastrointestinal Bleeding: Etiologies and Management. Mayo Clin Proc. 2019;94(4):697-703.\u003c/li\u003e\n\u003cli\u003eTokar JL, Higa JT. Acute Gastrointestinal Bleeding. Ann Intern Med. 2022;175(2):Itc17-itc32.\u003c/li\u003e\n\u003cli\u003eLincoln M, Keating N, O\u0026apos;Loughlin C, Tam A, O\u0026apos;Kane M, MacCarthy F, et al. Comparison of risk scoring systems for critical care patients with upper gastrointestinal bleeding: predicting mortality and length of stay. Anaesthesiol Intensive Ther. 2022;54(4):310-4.\u003c/li\u003e\n\u003cli\u003eJohnson AEW, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10(1):1.\u003c/li\u003e\n\u003cli\u003eShankar-Hari M, Phillips GS, Levy ML, Seymour CW, Liu VX, Deutschman CS, et al. Developing a New Definition and Assessing New Clinical Criteria for Septic Shock: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). Jama. 2016;315(8):775-87.\u003c/li\u003e\n\u003cli\u003eZhu Y, Zhang R, Ye X, Liu H, Wei J. SAPS III is superior to SOFA for predicting 28-day mortality in sepsis patients based on Sepsis 3.0 criteria. Int J Infect Dis. 2022;114:135-41.\u003c/li\u003e\n\u003cli\u003eHwang SY, Kim IK, Jeong D, Park JE, Lee GT, Yoo J, et al. Prognostic Performance of Sequential Organ Failure Assessment, Acute Physiology and Chronic Health Evaluation III, and Simplified Acute Physiology Score II Scores in Patients with Suspected Infection According to Intensive Care Unit Type. J Clin Med. 2023;12(19).\u003c/li\u003e\n\u003cli\u003eCecconi M, Evans L, Levy M, Rhodes A. Sepsis and septic shock. Lancet. 2018;392(10141):75-87.\u003c/li\u003e\n\u003cli\u003eIba T, Levy JH, Yamakawa K, Thachil J, Warkentin TE, Levi M. Proposal of a two-step process for the diagnosis of sepsis-induced disseminated intravascular coagulation. J Thromb Haemost. 2019;17(8):1265-8.\u003c/li\u003e\n\u003cli\u003eIba T, Levy JH, Warkentin TE, Thachil J, van der Poll T, Levi M. Diagnosis and management of sepsis-induced coagulopathy and disseminated intravascular coagulation. J Thromb Haemost. 2019;17(11):1989-94.\u003c/li\u003e\n\u003cli\u003eMatsubara T, Yamakawa K, Umemura Y, Gando S, Ogura H, Shiraishi A, et al. Significance of plasma fibrinogen level and antithrombin activity in sepsis: A multicenter cohort study using a cubic spline model. Thromb Res. 2019;181:17-23.\u003c/li\u003e\n\u003cli\u003ePolderman KH. Induced hypothermia and fever control for prevention and treatment of neurological injuries. Lancet. 2008;371(9628):1955-69.\u003c/li\u003e\n\u003cli\u003eMackowiak PA, Marling-Cason M, Cohen RL. Effects of temperature on antimicrobial susceptibility of bacteria. J Infect Dis. 1982;145(4):550-3.\u003c/li\u003e\n\u003cli\u003eLauney Y, Nesseler N, Mall\u0026eacute;dant Y, Seguin P. Clinical review: fever in septic ICU patients--friend or foe? Crit Care. 2011;15(3):222.\u003c/li\u003e\n\u003cli\u003eRumbus Z, Matics R, Hegyi P, Zsiboras C, Szabo I, Illes A, et al. Fever Is Associated with Reduced, Hypothermia with Increased Mortality in Septic Patients: A Meta-Analysis of Clinical Trials. PLoS One. 2017;12(1):e0170152.\u003c/li\u003e\n\u003cli\u003eKhodorkovsky B, Youssef E, Adamakos F, Cina T, Falco A, LaMura L, et al. Does Initial Temperature in the Emergency Department Predict Outcomes in Patients Admitted for Sepsis? J Emerg Med. 2018;55(3):372-7.\u003c/li\u003e\n\u003cli\u003eHensley MK, Donnelly JP, Carlton EF, Prescott HC. Epidemiology and Outcomes of Cancer-Related Versus Non-Cancer-Related Sepsis Hospitalizations. Crit Care Med. 2019;47(10):1310-6.\u003c/li\u003e\n\u003cli\u003eXiang MJ, Chen GL. Impact of cancer on mortality rates in patients with sepsis: A meta-analysis and meta-regression of current studies. World J Clin Cases. 2022;10(21):7386-96.\u003c/li\u003e\n\u003cli\u003ePeerapornratana S, Manrique-Caballero CL, G\u0026oacute;mez H, Kellum JA. Acute kidney injury from sepsis: current concepts, epidemiology, pathophysiology, prevention and treatment. Kidney Int. 2019;96(5):1083-99.\u003c/li\u003e\n\u003cli\u003eZarbock A, Nadim MK, Pickkers P, Gomez H, Bell S, Joannidis M, et al. Sepsis-associated acute kidney injury: consensus report of the 28th Acute Disease Quality Initiative workgroup. Nat Rev Nephrol. 2023;19(6):401-17.\u003c/li\u003e\n\u003cli\u003eEvans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47(11):1181-247.\u003c/li\u003e\n\u003cli\u003eDemoule A, Chevret S, Carlucci A, Kouatchet A, Jaber S, Meziani F, et al. Changing use of noninvasive ventilation in critically ill patients: trends over 15 years in francophone countries. Intensive Care Med. 2016;42(1):82-92.\u003c/li\u003e\n\u003cli\u003eDemoule A, Girou E, Richard JC, Taille S, Brochard L. Benefits and risks of success or failure of noninvasive ventilation. Intensive Care Med. 2006;32(11):1756-65.\u003c/li\u003e\n\u003cli\u003eRussell JA, Walley KR, Singer J, Gordon AC, H\u0026eacute;bert PC, Cooper DJ, et al. Vasopressin versus norepinephrine infusion in patients with septic shock. N Engl J Med. 2008;358(9):877-87.\u003c/li\u003e\n\u003cli\u003eDe Backer D, Biston P, Devriendt J, Madl C, Chochrad D, Aldecoa C, et al. Comparison of dopamine and norepinephrine in the treatment of shock. N Engl J Med. 2010;362(9):779-89.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"MIMIC-IV, Nomogram, Predictive model, Sepsis, Gastrointestinal bleeding","lastPublishedDoi":"10.21203/rs.3.rs-5406276/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5406276/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground: \u003c/strong\u003eSepsis associated with gastrointestinal hemorrhage is a critical condition in ICU patients, significantly impacting mortality rates. This study aimed to develop a predictive model for in-hospital death risk in sepsis patients with gastrointestinal bleeding, improving treatment strategies and resource allocation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e In a retrospective investigation of patients with sepsis and gastrointestinal bleeding, we gathered information from the MIMIC-IV database, including key demographics, comorbidities, laboratory indicators, and therapies.\u003c/p\u003e\n\u003cp\u003eThe dataset was split 70:30 for model development and validation. The Least Absolute Shrinkage and Selection Operator (LASSO) approach was used to select features, and multivariate logistic regression was then used to create a prognostic model. A nomogram was created to visualize predictive outcomes. Model performance was evaluated using calibration curve, receiver operating characteristic (ROC) curve, clinical impact curve (CIC), and decision curve analysis (DCA).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eNine significant predictors of in-hospital mortality were identified: APS III score, prothrombin time, body temperature, activated partial thromboplastin time, respiratory rate, vasopressor use, acute kidney injury, non-invasive ventilation, and malignancy. Area beneath the ROC curve for the training and testing groups\u003cbr\u003e\nThe values are 0.8266 (95% CI: 0.8018-0.8515) and 0.7961 (95% CI: 0.7577-0.8345), respectively. Our model outperformed the APS III score in terms of ROC curve discrimination and demonstrated greater net benefit on the DCA curve. The CIC showed strong concordance between predicted and actual in-hospital death rates when the predicted probability exceeded 70%.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion:\u003c/strong\u003e We developed a robust predictive framework for assessing in-hospital death risk in sepsis patients with gastrointestinal hemorrhage. Early intervention based on identified risk factors could improve patient survival rates.\u003c/p\u003e","manuscriptTitle":"Developing a prediction model for in-hospital mortality in sepsis patients with gastrointestinal bleeding using the MIMIC-IV database","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-10 05:58:44","doi":"10.21203/rs.3.rs-5406276/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"40f3b059-d830-45ae-b245-031bc9dc6677","owner":[],"postedDate":"December 10th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-03-16T10:53:41+00:00","versionOfRecord":[],"versionCreatedAt":"2024-12-10 05:58:44","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5406276","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5406276","identity":"rs-5406276","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}