Author Response to "Letter to the Editor": "Critical flaws in the predictive model for bacterial infections in older AoCLD patients".

Critical flaws in the predictive model for bacterial infections in older AoCLD patients.

Infections, Immunity & their Effects on Asthma

Animal models are essential for the development of new anti-infectious drugs. Although some bacterial infection models have been established in rodents, small primate models are rare. Here, we report on two bacterial infection models established in tree shrew (Tupaia belangeri chinensis). A burnt skin infection model was induced by dropping 5×10(6) CFU of Staphylococcus aureus on the surface of a wound after a third degree burn. This dose of S. aureus caused persistent infection for 7 days and obvious inflammatory response was observed 4 days after inoculation. A Dacron graft infection model, 2×10(6) CFU of Pseudomonas aeruginosa also caused persistent infection for 6 days, with large amounts of pus observed 3 days after inoculation. These models were used to evaluate the efficacy of levofloxacin (LEV) and cefoperazone (CPZ), which reduced the viable bacteria in skin to 4log10 and 5log10 CFU/100 mg tissue, respectively. The number of bacteria in graft was significantly reduced by 4log10 CFU/mL treatment compared to the untreated group (P<0.05). These results suggest that two bacterial infection models were successfully established in tree shrew using P. aeruginosa and S. aureus. In addition, tree shrew was susceptible to P. aeruginosa and S. aureus, thus making it an ideal bacterial infection animal model for the evaluation of new antimicrobials.

To investigate the additional value of geriatric parameters such as physical impairment to the quick Sequential Organ Failure Assessment (qSOFA) tool for predicting clinical deterioration in older ED patients with a suspected infection and to validate the final prediction model. Post-hoc multivariable regression analysis from a prospective observational cohort study of adult patients visiting the ED of a university hospital to develop a prediction model. External validation of the prediction model was performed using the prospective data-biobank Acutelines. In older patients, qSOFA (OR 1.47 (95% CI 1.12-1.95)) and physical impairment (OR 1.84 (95% CI 1.20-2.82)) were independently associated with clinical deterioration within 72 hours. This resulted in a prediction model with an area under the curve of 0.62 (95% CI 0.56-0.68) in the derivation cohort, and of 0.62 (95% CI 0.56-0.68) in the validation cohort. Calibration of the model was poor. In older ED patients with a suspected infection, not only disease severity scores, but also presence of physical impairment is independently associated with clinical deterioration.

Early mortality following hemodialysis initiation hinders survival improvement in older patients. This study aimed to develop a clinical risk model for predicting 6-month mortality after dialysis initiation in older Korean hemodialysis patients. We analyzed data from incident hemodialysis patients aged >70 years from the Korean Society of Geriatric Nephrology (KSGN) database. A prediction model was developed using multivariate logistic regression analysis and externally validated with independent datasets. Among 1,751 incident hemodialysis patients, the 6-month mortality rate was 15.5%. Using multivariate logistic analysis, we constructed the KSGN score as an independent risk factor for 6-month mortality, and its components and score are as follows: old age at dialysis initiation (≥85 years, score 2); hypertension and renovascular disease as a primary etiology of end-stage kidney disease (ESKD) (score 1); malignancy history (yes, score 1); low serum albumin (<3.5 g/dL, score 1); hypertension treatment (yes, score -1); prepared vascular access on maintenance dialysis (arteriovenous fistula/arteriovenous graft, score -3). In the development cohort, the area under the curve (AUC) for the KSGN score was significantly higher than the Alberta Wick's score (0.707 vs. 0.683, p = 0.001). In the validation cohort, the KSGN score's performance was comparable to existing models. The KSGN score may be a valuable tool for predicting early mortality after dialysis initiation in older patients with ESKD, aiding in decision-making and management regarding dialysis initiation.

BackgroundOlder patients (≥65 years old) experience high rates of adverse outcomes after an emergency department (ED) visit. Reliable tools to predict adverse outcomes in this population are lacking. This manuscript comprises a study protocol for the Risk Stratification in the Emergency Department in Acutely Ill Older Patients (RISE UP) study that aims to identify predictors of adverse outcome (including triage- and risk stratification scores) and intends to design a feasible prediction model for older patients that can be used in the ED.MethodsThe RISE UP study is a prospective observational multicentre cohort study in older (≥65 years of age) ED patients treated by internists or gastroenterologists in Zuyderland Medical Centre and Maastricht University Medical Centre+ in the Netherlands.After obtaining informed consent, patients characteristics, vital signs, functional status and routine laboratory tests will be retrieved. In addition, disease perception questionnaires will be filled out by patients or their caregivers and clinical impression questionnaires by nurses and physicians. Moreover, both arterial and venous blood samples will be taken in order to determine additional biomarkers. The discriminatory value of triage- and risk stratification scores, clinical impression scores and laboratory tests will be evaluated.Univariable logistic regression will be used to identify predictors of adverse outcomes. With these data we intend to develop a clinical prediction model for 30-day mortality using multivariable logistic regression. This model will be validated in an external cohort.Our primary endpoint is 30-day all-cause mortality. The secondary (composite) endpoint consist of 30-day mortality, length of hospital stay, admission to intensive- or medium care units, readmission and loss of independent living.Patients will be followed up for at least 30 days and, if possible, for one year.DiscussionIn this study, we will retrieve a broad range of data concerning adverse outcomes in older patients visiting the ED with medical problems. We intend to develop a clinical tool for identification of older patients at risk of adverse outcomes that is feasible for use in the ED, in order to improve clinical decision making and medical care.Trial registrationRetrospectively registered on clinicaltrials.gov (NCT02946398; 9/20/2016).

To identify the clinical features and independent predictors of survival in older patients with bone metastasis from prostate cancer (PCa). We retrospectively analysed 205 older patients with bone metastases from PCa between 1997 and 2012. The Kaplan-Meier method was used with the log-rank test for survival rate calculations and to evaluate each variable. Multivariate analysis was performed with the Cox regression model. The chi-squared test was used to compare survival rates between older and younger (n=197) patients. All patients were followed up. The 1-, 2-, 3- and 5-year survival rates were 95.5%, 77.5%, 68.5% and 33.7%, respectively. Gleason score, radiotherapy of the primary tumour, the number of bone metastases, the alkaline phosphatase alkaline phosphatase (ALP) level, organ metastasis and regional lymph node metastasis were associated with the survival rates. Multivariate Cox regression analysis showed that Gleason score at diagnosis of the primary tumour was a significant predictor of overall survival following the diagnosis of bone metastases. In addition, the overall survival rates of older patients were higher compared with younger patients, but older patients who underwent radiotherapy had higher mortality. These data may serve as a guide for creating clinical prediction models in further studies.

Racial disparities in survival after trauma are well described for patients younger than 65 years. Similar information among older patients is lacking because existing trauma databases do not include important patient comorbidity information. To determine whether racial disparities in trauma survival persist in patients 65 years or older. Trauma patients were identified from the Nationwide Inpatient Sample (January 1, 2003, through December 30, 2010) using International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis codes. Injury severity was ascertained by applying the Trauma Mortality Prediction Model, and patient comorbidities were quantified using the Charlson Comorbidity Index. In-hospital mortality after trauma for blacks vs whites for younger (16-64 years of age) and older (≥65 years of age) patients was compared using 3 different statistical methods: univariable logistic regression, multivariable logistic regression with and without clustering for hospital effects, and coarsened exact matching. Model covariates included age, sex, insurance status, type and intent of injury, injury severity, head injury severity, and Charlson Comorbidity Index. A total of 1,073,195 patients were included (502,167 patients 16-64 years of age and 571,028 patients ≥65 years of age). Most older patients were white (547,325 [95.8%]), female (406 158 [71.1%]), and insured (567,361 [99.4%]) and had Charlson Comorbidity Index scores of 1 or higher (323,741 [56.7%]). The unadjusted odds ratios (ORs) for death in blacks vs whites were 1.35 (95% CI, 1.28-1.42) for patients 16 to 64 years of age and 1.00 (95% CI, 0.93-1.08) for patients 65 years or older. After risk adjustment, racial disparities in survival persisted in the younger black group (OR, 1.21; 95% CI, 1.13-1.30) but were reversed in the older group (OR, 0.83; 95% CI, 0.76-0.90). This finding was consistent across all 3 statistical methods. Different racial disparities in survival after trauma exist between white and black patients depending on their age group. Although younger white patients have better outcomes after trauma than younger black patients, older black patients have better outcomes than older white patients. Exploration of this paradoxical finding may lead to a better understanding of the mechanisms that cause disparities in trauma outcomes.

This protocol outlines the steps required to produce a robust model of infectious disease and colitis, as well as the methods used to characterize Citrobacter rodentium infection in mice. C. rodentium is a gram negative, murine specific bacterial pathogen that is closely related to the clinically important human pathogens enteropathogenic E. coli and enterohemorrhagic E. coli. Upon infection with C. rodentium, immunocompetent mice suffer from modest and transient weight loss and diarrhea. Histologically, intestinal crypt elongation, immune cell infiltration, and goblet cell depletion are observed. Clearance of infection is achieved after 3 to 4 weeks. Measurement of intestinal epithelial barrier integrity, bacterial load, and histological damage at different time points after infection, allow the characterization of mouse strains susceptible to infection. The virulence mechanisms by which bacterial pathogens colonize the intestinal tract of their hosts, as well as specific host responses that defend against such infections are poorly understood. Therefore the C. rodentium model of enteric bacterial infection serves as a valuable tool to aid in our understanding of these processes. Enteric bacteria have also been linked to Inflammatory Bowel Diseases (IBDs). It has been hypothesized that the maladaptive chronic inflammatory responses seen in IBD patients develop in genetically susceptible individuals following abnormal exposure of the intestinal mucosal immune system to enteric bacteria. Therefore, the study of models of infectious colitis offers significant potential for defining potentially pathogenic host responses to enteric bacteria. C. rodentium induced colitis is one such rare model that allows for the analysis of host responses to enteric bacteria, furthering our understanding of potential mechanisms of IBD pathogenesis; essential in the development of novel preventative and therapeutic treatments.

Platelet transfusion has been reported to modulate the recipients’ immune system. To date, the precise mechanism(s) driving poor patient outcomes (e.g., increased rate of mortality, morbidity, infectious complications and prolonged hospital stays) following platelet transfusion are largely undefined. To determine the potential for platelet concentrates (PC) to modulate responses of crucial immune regulatory cells, a human in vitro whole blood model of transfusion was established. Maturation and activation of human myeloid dendritic cells (mDC) and the specialized subset blood DC antigen (BDCA)3+ DC were assessed following exposure to buffy-coat derived PC at day (D)2 (fresh) and D5 (date-of-expiry). In parallel, to model recipients with underlying viral or bacterial infection, polyinosinic:polycytidylic acid or lipopolysaccharide was added. Exposure to PC had less of an impact on mDC responses than BDCA3+ DC responses. PC alone downregulated BDCA3+ DC expression of co-stimulatory molecules CD40 and CD80. In the model of viral infection, PC downregulated expression of CD83, and in the bacterial model of infection, PC downregulated CD80, CD83, and CD86. PC alone suppressed mDC production of interleukin (IL)-8, IL-12 and tumor necrosis factor (TNF)-α and BDCA3+ DC production of IL-8, IL-12, and IL-6. In the model of viral infection, production of IL-12 and interferon-gamma inducible protein (IP)-10 was reduced in both DC subsets, and IL-8 was reduced in BDCA3+ DC following PC exposure. When modeling bacterial infection, PC suppressed mDC and BDCA3+ DC production of IL-6 and IL-10 with a reduction in TNF-α evident in mDC. This study assessed the impact of PC “transfusion” on DC surface antigen expression and inflammatory mediator production and provided the first evidence that PC transfusion modulates blood mDC and BDCA3+ DC maturation and activation, particularly in the models of infection. Results of this study suggest that patients who receive PC, particularly those with underlying infectious complications, may fail to establish an appropriate immune response precipitating poor patient outcomes.

Delirium superimposed on dementia (DSD) is a severe complication in older adults with dementia, marked by fluctuating cognition, inattention, and altered consciousness. Detection is challenging due to symptom overlap, yet it contributes to cognitive decline, prolonged hospitalization, and increased mortality. Identifying key risk factors and developing an accurate prediction model is crucial for timely intervention. This study aimed to establish a machine learning-based model to predict delirium risk, focusing on significant predictors to aid clinical decision-making. We prospectively collected clinical data from 636 older dementia patients. Five machine learning algorithms-Extreme Gradient Boosting (XGB), Random Forest (RF), Multilayer Perceptron (MLP), Categorical Boosting (CB), and Logistic Regression (LR)-were used to construct prediction models. Feature importance was analyzed using SHapley Additive exPlanations (SHAP) to identify key risk factors. Data included demographic information, biochemical parameters, comorbidities, medication history, and Visual Analogue Scale (VAS) scores. The final analysis included 636 older dementia patients, with a mean age of 78.2 ± 6.3 years, of whom 187 (29.4%) developed delirium during hospitalization. The XGB model demonstrated the best performance, achieving the highest area under the receiver operating characteristic curve (0.930), accuracy (0.870), F1 score (0.892), and area under the precision-recall curve (0.989). The Brier score for the XGB model was 0.08. The SHAP method identified cerebrovascular disease, sedative drug use, hemoglobin levels, VAS score ≥4, superoxide dismutase, diabetes, hsCRP, hypertension, family presence, and hyperlipidemia as the most significant risk factors for delirium. The top 10 variables were used to construct a compact XGB model, which also exhibited good predictive performance. This study developed a machine learning-based prediction model for delirium risk in older dementia patients, with the XGB model demonstrating the best performance. The identified key risk factors provide insights for early intervention, potentially improving delirium management in clinical practice.

BackgroundPredicting outcomes in older patients with influenza in the emergency department (ED) by machine learning (ML) has never been implemented. Therefore, we conducted this study to clarify the clinical utility of implementing ML.MethodsWe recruited 5508 older ED patients (≥65 years old) in three hospitals between 2009 and 2018. Patients were randomized into a 70%/30% split for model training and testing. Using 10 clinical variables from their electronic health records, a prediction model using the synthetic minority oversampling technique preprocessing algorithm was constructed to predict five outcomes.ResultsThe best areas under the curves of predicting outcomes were: random forest model for hospitalization (0.840), pneumonia (0.765), and sepsis or septic shock (0.857), XGBoost for intensive care unit admission (0.902), and logistic regression for in-hospital mortality (0.889) in the testing data. The predictive model was further applied in the hospital information system to assist physicians’ decisions in real time.ConclusionsML is a promising way to assist physicians in predicting outcomes in older ED patients with influenza in real time. Evaluations of the effectiveness and impact are needed in the future.

Oral mucosa is the target tissue for many microorganisms involved in periodontitis and other infectious diseases affecting the oral cavity. Three-dimensional (3D) in vitro and ex vivo oral mucosa equivalents have been used for oral disease modeling and investigation of the mechanisms of oral bacterial and fungal infections. This review was conducted to analyze different studies using 3D oral mucosa models for the evaluation of the interactions of different microorganisms with oral mucosa. In this study, based on our inclusion criteria, 43 articles were selected and analyzed. Different types of 3D oral mucosa models of bacterial and fungal infections were discussed in terms of the biological system used, culture conditions, method of infection, and the biological endpoints assessed in each study. The critical analysis revealed some contradictory reports in this field of research in the literature. Challenges in recovering bacteria from oral mucosa models were further discussed, suggesting possible future directions in microbiomics, including the use of oral mucosa-on-a-chip. The potential use of these 3D tissue models for the evaluation of the effects of antiseptic agents on bacteria and oral mucosa was also addressed. This review concluded that there were many aspects that would require optimization and standardization with regard to using oral mucosal models for infection by microorganisms. Using new technologies-such as microfluidics and bioreactors-could help to reproduce some of the physiologically relevant conditions and further simulate the clinical situation. Impact statement Tissue-engineered or commercial models of the oral mucosa are very useful for the study of diseases that involve the interaction of microorganisms and oral epithelium. In this review, challenges in recovering bacteria from oral mucosa models, the potential use of these three-dimensional tissue models for the evaluation of the effects of antiseptic agents, and future directions in microbiomics are discussed.

IntroductionBacterial infection is one of the most frequent complications in hepatitis B virus-related acute-on-chronic liver failure (HBV-ACLF), which leads to high mortality. However, a predictive model for bacterial infection in HBV-ACLF has not been well established. This study aimed to establish and validate a predictive model for bacterial infection in two independent patient cohorts.MethodsAdmission data from a prospective cohort of patients with HBV-ACLF without bacterial infection on admission was used for derivation. Bacterial infection development from day 3 to 7 of admission was captured. Independent predictors of bacterial infection development on multivariate logistic regression were used to develop the predictive model. External validation was performed on a separate retrospective cohort.ResultsA total of 377 patients were enrolled into the derivation cohort, including 88 patients (23.3%) who developed bacterial infection from day 3 to 7 of admission. On multivariate regression analysis, admission serum globulin (OR 0.862, 95% CI 0.822–0.904; P < 0.001), interleukin-6 (OR 1.023, 95% CI 1.006–1.040; P = 0.009), and C-reactive protein (OR 1.123, 95% CI 1.081–1.166; P < 0.001) levels were independent predictors for the bacterial infection development, which were adopted as parameters of the predictive model (GIC). In the derivation cohort, the area under the curve (AUC) of GIC was 0.861 (95% CI 0.821–0.902). A total of 230 patients were enrolled into the validation cohort, including 57 patients (24.8%) who developed bacterial infection from day 3 to 7 of admission, and the AUC of GIC was 0.836 (95% CI 0.782–0.881). The Hosmer–Lemeshow test showed a good calibration performance of the predictive model in the two cohorts (P = 0.199, P = 0.746). Decision curve analysis confirmed the clinical utility of the predictive model.ConclusionGIC was established and validated for the prediction of bacterial infection development in HBV-ACLF, which may provide a potential auxiliary solution for the primary complication of HBV-ACLF.Supplementary InformationThe online version contains supplementary material available at 10.1007/s40121-021-00454-2.

AimThis study aims to explore the risk factors for perioperative acute heart failure in older patients with hip fracture and establish a nomogram prediction model.MethodsThe present study was a retrospective study. From January 2020 to December 2021, patients who underwent surgical treatment for hip fracture at the Third Hospital of Hebei Medical University were included. Heart failure was confirmed by discharge diagnosis or medical records. The samples were randomly divided into modeling and validation cohorts in a ratio of 7:3. Relevant demographic and clinic data of patients were collected. IBM SPSS Statistics 26.0 performed univariate and multivariate logistic regression analysis, to obtain the risk factors of acute heart failure. The R software was used to construct the nomogram prediction model.ResultsA total of 751 older patients with hip fracture were enrolled in this study, of which 138 patients (18.37%, 138/751) developed acute heart failure. Heart failure was confirmed by discharge diagnosis or medical records. Respiratory disease (odd ratio 7.68; 95% confidence interval 3.82–15.43; value of P 0.001), history of heart disease (chronic heart failure excluded) (odd ratio 2.21, 95% confidence interval 1.18–4.12; value of P 0.010), ASA ≥ 3 (odd ratio 14.46, 95% confidence interval 7.78–26.87; value of P 0.001), and preoperative waiting time ≤ 2 days (odd ratio 3.32, 95% confidence interval 1.33–8.30; value of P 0.010) were independent risk factors of perioperative acute heart failure in older patients with hip fracture. The area under the curve (AUC) of the prediction model based on these factors was calculated to be 0.877 (95% confidence interval 0.836–0.918). The sensitivity and specificity were 82.8% and 80.9%, respectively, and the fitting degree of the model was good. In the internal validation group, the AUC was 0.910, and the 95% confidence interval was 0.869–0.950.ConclusionsSeveral risk factors are identified for acute heart failure in older patients, based on which pragmatic nomogram prediction model is developed, facilitating detection of patients at risk early.