Admission Of Patients Research Articles

Performance of machine learning versus the national early warning score for predicting patient deterioration risk: a single-site study of emergency admissions

ObjectivesIncreasing operational pressures on emergency departments (ED) make it imperative to quickly and accurately identify patients requiring urgent clinical intervention. The widespread adoption of electronic health records (EHR) makes rich...

Enhancing mortality prediction in patients with spontaneous intracerebral hemorrhage: Radiomics and supervised machine learning on non-contrast computed tomography

PurposeThis study aims to develop a Radiomics-based Supervised Machine-Learning model to predict mortality in patients with spontaneous intracerebral hemorrhage (sICH). MethodsRetrospective analysis of a prospectively collected clinical registry of patients with sICH consecutively admitted at a single academic comprehensive stroke center between January-2016 and April-2018. We conducted an in-depth analysis of 105 radiomic features extracted from 105 patients. Following the identification and handling of missing values, radiomics values were scaled to 0-1 to train different classifiers. The sample was split into 80-20% training-test and validation cohort in a stratified fashion. Random Forest(RF), K-Nearest Neighbor(KNN), and Support Vector Machine(SVM) classifiers were evaluated, along with several feature selection methods and hyperparameter optimization strategies, to classify the binary outcome of mortality or survival during hospital admission. A tenfold stratified cross-validation method was used to train the models, and average metrics were calculated. ResultsRF, KNN, and SVM, with the "DropOut+SelectKBest" feature selection strategy and no hyperparameter optimization, demonstrated the best performances with the least number of radiomic features and the most simplified models, achieving a sensitivity range between 0.90 and 0.95 and AUC range from 0.97 to 1 on the validation dataset. Regarding the confusion matrix, the SVM model did not predict any false negative test (negative predicted value 1). ConclusionRadiomics-based Supervised Machine Learning models can predictmortality during admission in patients with sICH. SVM with the "DropOut+SelectKBest"feature selection strategy and no hyperparameter optimization was the best simplifiedmodel to detect mortality during admission in patients with sICH.

Safety and efficacy of a novel mini-pool intravenous immunoglobulin therapy in children with primary immunodeficiency.

Intravenous polyvalent immunoglobulins (IVIG) for prophylaxis in patients with primary immunodeficiency disorders (PIDs) exposes them to life-threatening infections and debilitating diseases. To improve access to IVIG in lower middle-income countries, the WHO recommends a stepwise approach for the local production of purified and virus-inactivated plasma immunoglobulins by national blood transfusion services using new technologies and medical devices. One new technology relies on single-use sterile medical devices for the purification of plasma immunoglobulin G (IgG), as well as lipid-enveloped virus inactivation from mini-pools of recovered plasma separated from whole blood (mini-pool IVIG [MP-IVIG]). This study aimed to compare the safety and efficacy of MP-IVIG to standard IVIG (STD-IVIG). In this prospective crossover clinical study, we investigated the safety and efficacy of MP-IVIG for STD-IVIG preparations as a replacement therapy in a cohort of 21 paediatric patients with PID. Both MP-IVIG and STD-IVIG were effective in reducing the frequency of severe bacterial infections and hospital admissions in patients with PID. Mild side effects have been observed in seven patients (6.2%) with PID who received MP-IVIG and five patients (5.3%) who received STD-IVIG. No moderate or severe side effects or haemolytic transfusion reactions were reported. The mortality rates were also comparable and were not related to the study products. MP-IVIG presented no safety issues and was as effective as STD-IVIG in IgG replacement in patients with PID. Due to the small numbers, the results have to be addressed with caution.

Minimizing Radiation Exposure in Neonatal Intensive Care Unit: A Quality Improvement Approach on X-Ray Practices

Purpose: Radiographic examinations are frequently performed for diagnostic and therapeutic purposes in neonatal intensive care units (NICUs). However, concerns are emerging regarding the safety of radiation exposure, especially in vulnerable preterm infants in periods of rapid cellular division. This quality improvement (QI) project aimed to reduce radiation hazards in level-IV NICU.Methods: We established an "X-ray prescription protocol" and educated the physicians to ensure that only essential radiographs were obtained. Additionally, we discouraged full-body infantograms and emphasized the prescription of targeted radiographs, such as chest or abdominal radiographs. Furthermore, to reduce the dose-area product (DAP, Gy·cm<sup>2</sup>) values, which act as a surrogate for radiation exposure, we provided training to radiologic technologists on meticulous collimation for each radiography session. We aimed to achieve a 30% reduction in the average monthly cumulative DAP per patient, which was calculated by dividing the total monthly DAP from radiographs in the NICU by the monthly average of patient admissions. Retrospective baseline data were collected 8 months pre-intervention and prospectively for 4 months post-interventions.Results: The average monthly X-ray count per patient was 28.3 in the pre-intervention period (October 2022 to May 2023), which decreased to 25.4 in the post-intervention period (June 2023 to September 2023), reflecting a 10.2% reduction (<i>p</i>=0.109). The average monthly infantogram count per patient showed an 18.0% reduction (25.9% to 21.2%, <i>p</i>=0.016), and the proportion of infantograms in the total X-ray counts decreased from 91.5% to 83.3% (<i>p</i>=0.017). The DAP value per X-ray decreased by 42.6%, from an average of 0.25 to 0.14 (<i>p</i>=0.011). The primary outcome, the average monthly cumulative DAP value per patient, showed a substantial reduction of 48.6%, dropping from 7.00 to 3.60 (<i>p</i>=0.004). The baseline characteristics and short-term morbidities of the patients did not differ significantly between the pre- and post-intervention period.Conclusion: Our QI approach, which included discouraging excessive prescriptions of infantograms and promoting optimal collimation, significantly reduced the average monthly radiation exposure in the NICU, benefiting both patients and healthcare workers.

Evaluating the Compliance to 24-Hour Venous Thromboembolism Risk Re-assessment Following Orthopaedic Surgery at a Tertiary Center: A Closed-Loop Clinical Audit.

Introduction Venous thromboembolism (VTE) is a preventable cause of patient morbidity and mortality among hospitalised patients. VTE events have a high incidence among orthopaedic patients, who routinely receive chemical thromboprophylaxis in the form of heparin, warfarin, antiplatelet agents or direct oral anticoagulants. These can be associated with adverse events, most commonly bleeding or heparin-induced thrombocytopenia. A VTE risk reassessment following 24 hours of admission or a change in clinical condition like surgery is recommended to avoid such complications. We evaluated the compliance to completion of these reassessments following surgery at a tertiary elective orthopaedic hospital. Methods A closed-loop audit was undertaken for all elective orthoapedic specialties. First loop was conducted between 01/07/2022 and 25/05/2023, whilst the second loop was done between 01/01/2024 and 01/02/2024. Insight, an online database was used to collect data on initial VTE assessment forms completed at admission and VTE reassessment forms completed within 24 hours of surgery, as agreed with orthopaedic teams. Audit standards were derived from National Institute for Health and Care Excellence (NICE) guidelines on venous thromboembolism in over 16s. A target compliance was agreed at 100%. Paediatric patients, day case procedures and medical admissions comprised the exclusion criteria. Results For the first audit loop, only 6/4780 (0.13%) patients had completed VTE reassessment forms within 24 hours following surgery. Following implementation of a system prompt on Electronic Patient Medication Administration (EPMA) to serve as a reminder for VTE reassessment completion, the second audit loop found 112/156 (74%) patients had completed forms (p<0.001). Conclusion VTE risk reassessment following surgery is recommended by NICE guidelines in order to assess and potentially minimize complications like bleeding and heparin-induced thrombocytopenia. A simple intervention such as a system reminder can serve to improve compliance. This can be implemented at a large scale given most hospitals use integrated electronic medication administration systems where initial VTE risk assessments are completed as part of the initial patient admission process.

Caring for Patients With Injection Drug Use-Associated Infective Endocarditis at the End of Life: A Modified Photo-Elicitation Phenomenological Study.

Admissions for patients with injection drug use-associated infective endocarditis are increasing. Injection drug use-associated infective endocarditis is a serious and often fatal cardiac infection. Nurses are often frustrated when caring for these patients because these care assignments are physically and emotionally demanding. Nurses feel helpless, as they disposition young patients to a funeral homes, morgue, or hospice, or send them home to die, and yet a dearth of information regarding end-of-life care for injection drug use-associated infective endocarditis patients remains. This study explored the lived experiential impact these experiences have on nurses. The method was interpretive and participatory, a modified photo-elicitation phenomenological design. Twenty-nine nurses recruited by purposive sampling took photographs and wrote reflections, with 5 participating in an optional unstructured interview conducted by the author. Three themes emerged from the data that elucidated these experiences as (1) a "heartbreaking" experience to witness, (2) an "exhausting" experience to endure, and (3) a practice-altering experience that transforms. Data analysis reveals that nurses experience intense pain, exhaustion, and, yet, transformation, both negative and positive. Implications for nurses include the need for end-of-life education, formal debriefing process to mitigate turnover, and future research into nature-based therapies to help nurses process trauma, as a form of self-care.

Interleukin-6 as a prognostic marker in acute kidney injury and its klotho-dependent regulation

Background and objectiveIn acute kidney injury (AKI), a strong inflammatory component is activated in response to the renal damage, and one of the main mediators behind this process is the pro-inflammatory interleukin 6 or IL-6. Beside to this phenomenon, there are also alterations in different components of mineral metabolism, such as those dependent on fibroblast growth factor (FGF)23 and the anti-ageing cofactor klotho. The aim of this work was to explore the association between renal function and systemic levels of IL-6, as well as FGF23 and klotho in the early stages of AKI, analysing the predictive capacity of IL-6 in early mortality associated with AKI. Material and methodsPlasma levels of IL-6, klotho and FGF23 were analysed in samples from 28 patients with AKI and related to renal function on hospital admission, and after 24 and 72 h. In addition, the predictive capacity of IL-6 on AKI-associated mortality was analysed at the three study time points. In an experimental nephrotoxic -AKI mouse model, systemic IL-6 and FGF23 values were also analysed 24 and 72 h after induction of kidney damage, as well as in mice overexpressing the anti-ageing protein, klotho. ResultsSystemic IL-6 levels increased in AKI patients, especially in hospital admission time, and decreased in parallel with improving renal function. At the same time as IL-6 values increased, there was an increase in FGF23 and a decrease in klotho levels, with a significant and positive correlation between IL-6 and FGF23 levels. In addition, we obtained that systemic IL-6 levels were a good predictor of mortality in these patients, with an area under the curve equal to one at 72 h after AKI. In the experimental mouse AKI model, we also observed an increase in plasma levels in both IL-6 and FGF23 after 24 h of kidney damage. Nevertheless, in transgenic mice overexpressing klotho, there was no such increase in any of them. ConclusionsThere is an association between renal damage and increased levels of IL-6 and FGF23 in patients with AKI, especially on hospital admission time. Moreover, IL-6 levels are able to predict mortality in these patients, being a promising prognostic biomarker at any study time with a strong prediction at 72 h after patient admission. Maintaining adequate klotho levels could prevent the IL-6 mediated inflammatory response and therefore also reduce the degree and severity of renal damage after AKI.

Association of Child Opportunity Index with Outcomes in Pediatric Supraventricular Tachycardia.

To assess the relationship between the Child Opportunity Index (COI), a validated measure that uses 29 indicators of social determinants of health linked to the US Census, and pediatric supraventricular tachycardia (SVT) management and outcomes, specifically looking at utilization of ablation. A retrospective, single-center study was performed evaluating pediatric SVT patients from January 2018 to July 2023. Patients diagnosed at < 5years old, with pre-excitation, and with hemodynamically significant congenital heart disease were excluded. Characteristics were compared among low, moderate, and high COI categories using Kruskal-Wallis tests for continuous variables, and chi-square or Fisher's exact tests for categorical variables. Logistic regression and log-normal regression models were used to assess the relationship between COI category and primary outcomes of interest among SVT patients. A cohort of 306 patients met inclusion criteria. 45% had a high COI, 29% a moderate COI, and 26% a low COI. The odds of having one or more SVT-related hospital admissions were 3.88 times greater for the low COI patients compared to high COI patients (95% CI 1.39-10.8, p = 0.010). COI was not found to be associated with any other primary outcomes of interest, including utilization of ablation. Lower COI was associated with greater odds of admission in SVT patients but did not differ in rates of ablation or outcomes. This suggests social factors may affect resource utilization while having limited impact on treatments like ablations.

Effects of closed- versus open-system intensive care units on mortality rates in patients with cancer requiring emergent surgical intervention for acute abdominal complications: a single-center retrospective study in Korea.

In this study, we aimed to compare the in-hospital mortality of patients with cancer who experienced acute abdominal complications that required emergent surgery in open (treatment decisions made by the primary attending physician of the patient's admission department) versus closed (treatment decisions made by intensive care unit [ICU] intensivists) ICUs. This retrospective, single-center study enrolled patients with cancer admitted to the ICU before or after emergency surgery between November 2020 and September 2023. Univariate and logistic regression analyses were conducted to explore the associations between patient characteristics in the open and closed ICUs and in-hospital mortality. Among the 100 patients (open ICU, 49; closed ICU, 51), 23 died during hospitalization. The closed ICU group had higher Acute Physiology and Chronic Health Evaluation (APACHE) II scores, vasopressor use, mechanical ventilation, and preoperative lactate levels and a shorter duration from diagnosis to ICU admission, surgery, and antibiotic administration than the open ICU group. Univariate analysis linked in-hospital mortality and APACHE II score, postoperative lactate levels, continuous renal replacement therapy (CRRT), and mechanical ventilation. Multivariate analysis revealed that in-hospital mortality rate increased with CRRT use and was lower in the closed ICU. Compared to an open ICU, a closed ICU was an independent factor in reducing in-hospital mortality through prompt and appropriate treatment.

Correlation of Hyaluronic Acid (HA), Syndecan-1 (SDC-1), Heparan Sulfate (HS) with early stage ad organ dysfunction in sepsis patients

This study aimed to explore the relationship between the changes in early degradation products of polysaccharide coatings (such as hyaluronic acid (HA), syndecan-1 (SDC-1), and heparan sulfate (HS)) and the development of organ dysfunction in sepsis patients. We conducted a retrospective analysis on 140 sepsis patients admitted from January 2021 to June 2022, who formed the study group; 100 healthy individuals who underwent health checks during the same period were included as the control group. The study found that the expression levels of HA, SDC-1, and HS upon admission and within 24 hours of admission in sepsis patients, as well as the early change rates, were positively correlated with organ dysfunction (P &lt; 0.05). Through receiver operating characteristic (ROC) curve analysis, we discovered that the early change rates of HA, SDC-1, and HS have high predictive value for organ dysfunction in sepsis patients, with the combined predictive value being the most significant. The study conclusion points out that the increased levels of HA, SDC-1, HS, and other degradation products of polysaccharide coatings in the early stage of sepsis are positively associated with the occurrence of organ dysfunction. Clinicians can utilize the early expression changes of these biomarkers to predict the risk of organ dysfunction in sepsis patients, enabling timely implementation of preventive measures that may improve patient outcomes.

Risk factors for intensive care admission in childhood patients with COVID-19 - Results of the German nationwide inpatient sample

BackgroundCOVID-19 pandemic research efforts have mainly focused on adults, but data in paediatric populations are sparse. MethodsWe used the German nationwide-inpatient-sample analysing all hospitalized children ≤18 years with confirmed COVID-19-diagnosis in Germany during the year 2020, stratified for intensive care unit (ICU) admission. ResultsOverall, 3360 hospitalization-cases of children ≤18 years with COVID-19-infection were diagnosed in Germany 2020 (median age 7.0 [IQR 0.0–15.0] years, 49.8 % female); among them 4.3 % were admitted on an ICU. In-hospital death occurred in five patients with and three without ICU admission (3.5 % vs. 0.1 %, P < 0.001) and ICU admission was independently associated with increased case-fatality (OR 21.573 [95%CI 4.191–111.044], P < 0.001).Obesity (OR 3.419 [95%CI 1.300–8.993], P = 0.013), diabetes mellitus (OR 6.929 [95%CI 3.327–14.432], P < 0.001), pneumonia (OR 7.373 [95%CI 4.823–11.271], P < 0.001), acute respiratory distress syndrome (ARDS) (OR 48.058 [95%CI 11.689–197.588], P < 0.001) and multisystem inflammatory syndrome caused by COVID-19 (OR 9.573 [95%CI 3.036–30.191], P < 0.001), heart failure (OR 64.509 [95%CI 24.462–170.121], P < 0.001), myocarditis (OR 4.682 [95%CI 1.278–17.149], P = 0.020), acute and/or chronic kidney failure (OR 7.946 [95%CI 3.606–17.508], P < 0.001), cancer (OR 5.220 [95%CI 2.599–10.485], P < 0.001) and liver diseases (OR 5.501 [95%CI 2.177–13.899], P < 0.001) were associated with an ICU admission. ConclusionProportion of hospitalized paediatric COVID-19-patients admitted on ICU in Germany was low with 4.3 % accompanied by 3.5 % case-fatality rate. Independent factors for ICU admission comprised cardio-vascular risk factors, comorbidities, and complications of COVID-19.

Association Between Multivitamin Use on Admission and Clinical Outcomes in Patients Hospitalised with Community-Acquired Pneumonia: A Case—Cohort Study

Background/Objectives: Community-acquired pneumonia (CAP) is a leading cause of hospitalisations worldwide. Micronutrient deficiencies may influence CAP risk and severity, but their impact on CAP outcomes remains unclear. This study investigated the influence of multivitamin use on hospital length of stay (LOS), intensive care unit (ICU) admission, in-hospital mortality, and 30-day readmissions in hospitalised CAP patients. Methods: This retrospective cohort study included all CAP admissions, identified using ICD-10-AM codes, at two tertiary hospitals in Australia between 2018 and 2023. Pneumonia severity was determined using the CURB65 score, while frailty and nutritional status were assessed using the Hospital Frailty Risk Score (HFRS) and the Malnutrition Universal Screening Tool (MUST). Multivitamin use at admission was identified through the hospital pharmacy database. Propensity score matching (PSM) controlled for 22 confounders and the average treatment effect on the treated (ATET) was determined to evaluate clinical outcomes. Results: The mean (SD) age of the 8162 CAP cases was 75.3 (17.5) years, with 54.7% males. The mean (SD) CURB65 score was 1.9 (1.0), with 29.2% having severe CAP (CURB65 ≥ 3). On admission, 563 patients (6.9%) were on multivitamin supplements. Multivitamin users were younger, had more comorbidities, higher frailty, and higher socioeconomic status than non-users (p &lt; 0.05). The ATET analysis found no significant differences in LOS (aOR 0.14, 95% CI 0.03–5.98, p = 0.307), in-hospital mortality (aOR 1.04, 95% CI 0.97–1.11, p = 0.239), or other outcomes. Conclusions: Multivitamin use was documented in 6.9% of CAP patients and was associated with multimorbidity and frailty but not with improved clinical outcomes. Further research is needed to determine if specific vitamin supplements may offer benefits in this population.

The Effectiveness of Convalescent Plasma Administered on the First Day of Admission for COVID-19 Patients: An Open-label Randomized Controlled Trial

The Effectiveness of Convalescent Plasma Administered on the First Day of Admission for COVID-19 Patients: An Open-label Randomized Controlled Trial

Health Service Procedures for BPJS Participants in the Inpatient Room of Lasinrang General Hospital, Pinrang Regency

Hospital services (RS), are an important public service and are needed in an effort to meet health demands. Services for the poor are one aspect implemented by health service providers such as hospitals through after-sales services by utilizing BPJS funds as operational service activities. The type of research used is a survey system with a descriptive design. The research population is all BPJS participant patients who utilize inpatient services during the study, as many as 101 people who are also research samples based on the exhaustive sampling method where the entire population is used as a sample. The results of the study showed that BPJS participants who utilized health services at Lasinrang General Hospital, Pinrang Regency were mostly in the 20-29 year age group (32.7%), more were women (63.4%) with junior high school education (46.5%) and more worked as farmers (24.8%), more BPJS participant patient admission procedures stated in the good category (69.3%), more doctor services stated in the good category (67.3%), more nurse services stated in the good category (65.3%) and more BPJS participant drug services stated in the good category (72.3%). The suggestion put forward in this study is the need to conduct training on excellent service in order to improve skills in providing services, especially to BPJS participants and treatment planning using the drug procurement method through distributors so that if the drug runs out, the distributor can be contacted and this will support the efficiency and effectiveness of drug procurement that meets the needs of patients, especially BPJS participants who utilize health services at the hospital.

Comparison of Oxidative Stress Markers with Clinical Data in Patients Requiring Anesthesia in an Intensive Care Unit.

Objectives: The aim of this study is to assess the oxidative stress status in patients requiring intensive care unit (ICU) admission before initiating ICU treatment, by measuring the total oxidant level (TOS) and total antioxidant level (TAS) and oxidative stress index (OSI) levels. Additionally, we aim to explore the correlation between these oxidative stress markers and biochemical and hematological parameters. Materials and Methods: A total of 153 patients treated in intensive care units were included in the study. Patients who met the patient admission criteria of the ethics committee of the intensive care medicine association were included in the study. Blood samples were taken at the first moment the patients were admitted to the intensive care unit (before starting treatment). In total, 60 healthy volunteers who were compatible with the patient group in terms of age and gender were included in the study as a control group. Patients who had previously received antioxidant treatment and cancer patients were excluded from the study. Results: The TOS was significantly higher in the patient group (13.4 ± 7.5) compared to controls (1.8 ± 4.4) (p = 0.021). TOS > 12.00 means a "very high oxidant level". OSI was significantly higher in the patient group (689.8 ± 693.9) compared to the control group (521.7 ± 546.6) (p = 0.035). Ferritin levels were significantly higher in the patient group (546.5 ± 440.8 ng/mL) compared to controls (45.5 ± 46.5 ng/mL) (p < 0.001). Patients had significantly higher levels of C-reactive protein (CRP), procalcitonin (PCT), white blood cells (WBCs), immature granulocytes (IGs), zinc, and copper compared to the control group, indicating elevated inflammation and oxidative stress. CRP levels were 76.6 ± 85.9 mg/L in patients versus 5.6 ± 15.1 mg/L in controls (p < 0.001). PCT levels were 15.8 ± 8.6 ng/L in patients versus 2.3 ± 7.2 ng/L in controls (p = 0.012). Zinc and copper were also significantly elevated (p = 0.012 and p = 0.002, respectively). Conclusions: Our study provides valuable insights into the relationship between oxidative stress, inflammation, and trace elements, contributing to the growing understanding of oxidative stress as a prognostic tool in critical care. This could help to tailor therapeutic strategies aimed at reducing oxidative damage in ICU patients, enhancing patient outcomes.

Health Service Procedures for BPJS Participants in the Inpatient Room of Lasinrang General Hospital

Hospital services (RS), are an important public service and are needed in an effort to fulfill health demands. Services for the poor are one aspect implemented by health service providers such as Hospitals through after-sales services by utilizing BPJS funds as operational service activities. The type of research used is a survey system with a descriptive design. The population of the study was all BPJS participant patients who utilized inpatient services during the study, totaling 101 people who were also research samples based on the exhaustive sampling method where the entire population was sampled. The results of the study showed that BPJS participants who utilized health services at Lasinrang General Hospital, Pinrang Regency were mostly in the 20-29 year age group (32.7%), more were women (63.4%) with junior high school education (46.5%) and more worked as farmers (24.8%), more BPJS participant patient admission procedures stated that they were in the good category (69.3%), more doctor services stated that they were in the good category (67.3%), more nurse services stated that they were in the good category (65.3%) and more drug services for BPJS participant patients stated that they were in the good category (72.3%). The suggestion put forward in this study is the need to conduct training on excellent service in order to improve skills in providing services, especially to BPJS participants and treatment planning using the drug procurement method through distributors so that if the drug runs out, the distributor can be contacted and this will support the efficiency and effectiveness of drug procurement that meets the needs of patients, especially BPJS participants who utilize health services at the hospital.

A multicentre prospective registry of one thousand sepsis patients admitted in Indian ICUs: (SEPSIS INDIA) study

BackgroundSepsis is a global health problem with high morbidity and mortality. Low- and middle-income countries have a higher incidence and poorer outcome with sepsis. Large epidemiological studies in sepsis using Sepsis-3 criteria, addressing the process of care and deriving predictors of mortality are scarce in India.MethodA multicentre, prospective sepsis registry was conducted using Sepsis 3 criteria of suspected or confirmed infection and SOFA score of 2 or more in 19 ICUs in India over a period of one year (August 2022–July 2023). All adult patients admitted to the Intensive Care Unit who fulfilled the Sepsis 3 criteria for sepsis and septic shock were included. Patient infected with Covid 19 were excluded. Patients demographics, severity, admission details, initial resuscitation, laboratory and microbiological data and clinical outcome were recorded. Performance improvement programs as recommended by the Surviving Sepsis guideline were noted from the participating centers. Patients were followed till discharge or death while in hospital.ResultsRegistry Data of 1172 patients with sepsis (including 500 patients with septic shock) were analysed. The average age of the study cohort was 65 years, and 61% were male. The average APACHE II and SOFA score was 21 and 6.7 respectively. The majority of patients had community-acquired infections, and lung infections were the most common source. Of all culture positive results, 65% were gram negative organism. Carbapenem-resistance was identified in 50% of the gram negative blood culture isolates. The predominant gram negative organisms were Klebsiella spp (25%), Escherechia coli (24%) and Acinetobacter Spp (11%). Tropical infections (Dengue, Malaria, Typhus) constituted minority (n = 32, 2.2%) of sepsis patients. The observed hospital mortality for the entire cohort (n = 1172) was 36.3%, for those without shock (n = 672) it was 25.6% and for those with shock (n = 500) it was 50.8%. The average length of ICU and hospital stay for the study cohort was 8.64 and 11.9 respectively. In multivariate analysis adequate source control, correct choice of empiric antibiotic and the use of intravenous thiamine were protective.ConclusionThe general demographics of the sepsis population in the Indian Sepsis Registry is comparable to Western population. The mortality of sepsis cohort was higher (36.3%) but septic shock mortality (50.8%) was comparable to Western reports. Gram negative infection was the predominant cause of sepsis with a high incidence of carbapenem resistance. Eschericia coli, Klebsiella Spp and Acinetobacter Spp were the predominant causative organism. Tropical infection constituted a minority of sepsis population with low hospital mortality. The SOFA score on admission was a comparatively better predictor of poor outcome. Sepsis secondary to nosocomial infections had the worst outcomes, while source control, correct empirical antibiotic selection, and intravenous thiamine were protective.CTRI Registration CTRI:2022/07/044516.

Chronic treatment with SGLT-2 inhibitors is associated with ICU admission and disease severity in patients with diabetic ketoacidosis: a propensity score–matched cohort study

Chronic treatment with SGLT-2 inhibitors is associated with ICU admission and disease severity in patients with diabetic ketoacidosis: a propensity score–matched cohort study

Predicting involuntary admission following inpatient psychiatric treatment using machine learning trained on electronic health record data

Abstract Background Involuntary admissions to psychiatric hospitals are on the rise. If patients at elevated risk of involuntary admission could be identified, prevention may be possible. Our aim was to develop and validate a prediction model for involuntary admission of patients receiving care within a psychiatric service system using machine learning trained on routine clinical data from electronic health records (EHRs). Methods EHR data from all adult patients who had been in contact with the Psychiatric Services of the Central Denmark Region between 2013 and 2021 were retrieved. We derived 694 patient predictors (covering e.g. diagnoses, medication, and coercive measures) and 1134 predictors from free text using term frequency-inverse document frequency and sentence transformers. At every voluntary inpatient discharge (prediction time), without an involuntary admission in the 2 years prior, we predicted involuntary admission 180 days ahead. XGBoost and elastic net models were trained on 85% of the dataset. The models with the highest area under the receiver operating characteristic curve (AUROC) were tested on the remaining 15% of the data. Results The model was trained on 50 634 voluntary inpatient discharges among 17 968 patients. The cohort comprised of 1672 voluntary inpatient discharges followed by an involuntary admission. The best XGBoost and elastic net model from the training phase obtained an AUROC of 0.84 and 0.83, respectively, in the test phase. Conclusion A machine learning model using routine clinical EHR data can accurately predict involuntary admission. If implemented as a clinical decision support tool, this model may guide interventions aimed at reducing the risk of involuntary admission.

Predictive risk models for COVID-19 patients using the multi-thresholding meta-algorithm

This study aims to develop a Machine Learning model to assess the risks faced by COVID-19 patients in a hospital setting, focusing specifically on predicting the complications leading to Intensive Care Unit (ICU) admission or mortality, which are minority classes compared to the majority class of discharged patients. We operate within a multiclass framework comprising three distinct classes, and address the challenge of dataset imbalance, a common source of model bias. To effectively manage this, we introduce the Multi-Thresholding meta-algorithm (MTh), an innovative output-level methodology that extends traditional thresholding from binary to multiclass classification. This methodology dynamically adjusts class probabilities using misclassification costs, making it highly effective in imbalanced datasets. Our approach is further enhanced by integrating the simplicity, transparency, and effectiveness of Bayesian networks to create a robust predictive model. Using patient admission data, the model accurately identifies key risk and protective factors for COVID-19 outcomes. Our findings indicate that certain patient characteristics, such as high Charlson Index and pre-existing conditions, significantly influence the risk of ICU admission and mortality. Moreover, we introduce an explanatory model that elucidates the interrelationships among these factors, demonstrating the influence of therapeutic limits on the overall risk assessment of COVID-19 patients. Overall, our research provides a significant contribution to the field of Machine Learning by offering a novel solution for multiclass classification in the context of imbalanced datasets. This model not only enhances predictive accuracy but also supports critical decision-making processes in healthcare, potentially improving patient outcomes and optimizing clinical resource allocation.