Predictors of Severity of Influenza-Related Hospitalizations: Results From the Global Influenza Hospital Surveillance Network (GIHSN).

BACKGROUND: The Global Influenza Hospital Surveillance Network (GIHSN) has since 2012 provided patient-level data on severe influenza-like-illnesses from >100 participating clinical sites worldwide based on a core protocol and consistent case definitions. METHODS: We used multivariable logistic regression to assess the risk of intensive care unit admission, mechanical ventilation, and in-hospital death among hospitalized patients with influenza and explored the role of patient-level covariates and country income level. RESULTS: The data set included 73 121 patients hospitalized with respiratory illness in 22 countries, including 15 660 with laboratory-confirmed influenza. After adjusting for patient-level covariates we found a 7-fold increase in the risk of influenza-related intensive care unit admission in lower middle-income countries (LMICs), compared with high-income countries (P = .01). The risk of mechanical ventilation and in-hospital death also increased by 4-fold in LMICs, though these differences were not statistically significant. We also find that influenza mortality increased significantly with older age and number of comorbid conditions. Across all severity outcomes studied and after controlling for patient characteristics, infection with influenza A/H1N1pdm09 was more severe than with A/H3N2. CONCLUSIONS: Our study provides new information on influenza severity in underresourced populations, particularly those in LMICs.

Clinical outcomes and economic burden of seasonal influenza and other respiratory virus infections in hospitalized adults.

The cost of influenza and other respiratory virus infections should be determined to analyze the real burden of these diseases. We aimed to investigate the clinical outcomes and cost of illness due to respiratory virus infections in hospitalized adult patients. Hospitalized patients who had nasal swab sampling for a suspected viral infection between August 1, 2018 to March 31, 2019 were included. Outcome variables were oxygen requirement, mechanical ventilation need, intensive care unit admission, and cost. At least one viral pathogen was detected in 125 (47.7%) of 262 patients who were included in the study. Fifty-five (20.9%) of the patients were infected with influenza. Influenza-positive patients had higher rates for respiratory support, intensive care unit admission, and mortality compared to all other patients. The average cost of hospitalization per person was 2879.76 USD in the influenza-negative group, while the same cost was 3274.03 USD in the influenza-positive group. Although all of the vaccinated influenza-positive patients needed oxygen support, neither of them required invasive mechanical ventilation or intensive care unit admission. The average hospitalization cost per person was 779.70 USD in the vaccinated group compared to 3762.01 USD in the unvaccinated group. Disease-related direct cost of influenza in the community was estimated as 22 776 075.61 USD in the 18-65 years of age group and 15 756 120.02 USD in the 65 years of age and over group per year. Influenza, compared to other respiratory virus infections, can lead to untoward clinical outcomes and mortality as well as higher direct medical costs in adults.

Efficacy and safety of an inactivated whole-virion SARS-CoV-2 vaccine (CoronaVac): interim results of a double-blind, randomised, placebo-controlled, phase 3 trial in Turkey.

BACKGROUND: CoronaVac, an inactivated whole-virion SARS-CoV-2 vaccine, has been shown to be well tolerated with a good safety profile in individuals aged 18 years and older in phase 1/2 trials, and provided a good humoral response against SARS-CoV-2. We present the interim efficacy and safety results of a phase 3 clinical trial of CoronaVac in Turkey. METHODS: This was a double-blind, randomised, placebo-controlled phase 3 trial. Volunteers aged 18-59 years with no history of COVID-19 and with negative PCR and antibody test results for SARS-CoV-2 were enrolled at 24 centres in Turkey. Exclusion criteria included (but were not limited to) immunosuppressive therapy (including steroids) within the past 6 months, bleeding disorders, asplenia, and receipt of any blood products or immunoglobulins within the past 3 months. The K1 cohort consisted of health-care workers (randomised in a 1:1 ratio), and individuals other than health-care workers were also recruited into the K2 cohort (randomised in a 2:1 ratio) using an interactive web response system. The study vaccine was 3 µg inactivated SARS-CoV-2 virion adsorbed to aluminium hydroxide in a 0·5 mL aqueous suspension. Participants received either vaccine or placebo (consisting of all vaccine components except inactivated virus) intramuscularly on days 0 and 14. The primary efficacy outcome was the prevention of PCR-confirmed symptomatic COVID-19 at least 14 days after the second dose in the per protocol population. Safety analyses were done in the intention-to-treat population. This study is registered with ClinicalTrials.gov (NCT04582344) and is active but no longer recruiting. FINDINGS: Among 11 303 volunteers screened between Sept 14, 2020, and Jan 5, 2021, 10 218 were randomly allocated. After exclusion of four participants from the vaccine group because of protocol deviations, the intention-to-treat group consisted of 10 214 participants (6646 [65·1%] in the vaccine group and 3568 [34·9%] in the placebo group) and the per protocol group consisted of 10 029 participants (6559 [65·4%] and 3470 [34·6%]) who received two doses of vaccine or placebo. During a median follow-up period of 43 days (IQR 36-48), nine cases of PCR-confirmed symptomatic COVID-19 were reported in the vaccine group (31·7 cases [14·6-59·3] per 1000 person-years) and 32 cases were reported in the placebo group (192·3 cases [135·7-261·1] per 1000 person-years) 14 days or more after the second dose, yielding a vaccine efficacy of 83·5% (95% CI 65·4-92·1; p<0·0001). The frequencies of any adverse events were 1259 (18·9%) in the vaccine group and 603 (16·9%) in the placebo group (p=0·0108) with no fatalities or grade 4 adverse events. The most common systemic adverse event was fatigue (546 [8·2%] participants in the vaccine group and 248 [7·0%] the placebo group, p=0·0228). Injection-site pain was the most frequent local adverse event (157 [2·4%] in the vaccine group and 40 [1·1%] in the placebo group, p<0·0001). INTERPRETATION: CoronaVac has high efficacy against PCR-confirmed symptomatic COVID-19 with a good safety and tolerability profile. FUNDING: Turkish Health Institutes Association.

A Quality Improvement Study on Delirium Awareness Day: In Pursuit of Missed Delirium Diagnoses.

The current study aimed to screen for delirium in hospitalized older adults and assess the validity of the Turkish version of the 4A's Test (4AT-TR) as a feasible tool to integrate in routine patient care. The point prevalence of delirium according to clinical evaluation in routine practice was detected among all patients aged ≥60 years in 12 pilot wards. Delirium screening was then conducted by two arms: (a) nurses using the 4AT-TR and (b) geriatricians according to Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) criteria. Prevalence of delirium according to clinical impression was 3.3% (n = 4), whereas prevalence was 12.4% (n = 17) according to DSM-5 criteria and 13.8% (n = 17) according to the 4AT-TR. The 4AT-TR performed by nurses had a sensitivity of 66.6% and specificity of 93.5%. Area under the receiver operating characteristic curve for delirium diagnosis was 0.819 (p < 0.001). Most delirium cases remain undetected unless a routine and formal delirium assessment is integrated in hospital care of high-risk patients. The 4AT-TR performed by nurses seems to be a valid tool for determining delirium in hospitalized older adults. [Journal of Gerontological Nursing, 48(8), 43-51.].

'Anxiety is still ongoing!' Evaluation of the effect of the COVID-19 pandemic on anxiety severity of physicians working in the internal medicine department after 1 year: a collaborative cross-sectional study.

The COVID-19 pandemic has increased anxiety in society and particularly in healthcare workers, as shown with a questionnaire in our centre at the beginning of the pandemic. In this collaborative study, we aimed to evaluate the effects of the pandemic on anxiety 1 year later by applying the same questionnaire to the physicians working in the same department. A total of 77 participants consented to the study. The median age was 28 (interquartile range = 4) years and 55.8% were male. As in the first survey, female gender, having family members over 65 years of age, and having family members with chronic diseases were significantly associated with high anxiety scores and levels. There were no statistically significant differences between the first and second survey participants in any of the anxiety scales, which means anxiety persists.

"COVID-19: The final nail in the coffin for physical examination" Evaluation of the effects of COVID-19 pandemic on physical examination habits of residents in a university hospital: A cross-sectional survey.

AIMS: It is evident that the COVID-19 pandemic has affected the medical practice and training of residents. In this study, we evaluated the physical examination (PE) habits of residents working in a university hospital and how their PE practices did change during the pandemic. METHODS: This single-centre, non-interventional, cross-sectional descriptive study was conducted in a university hospital using an online survey questionnaire between 5 and 20 October 2020. RESULTS: Of the 308 residents who participated in the study, 172 of them (55.8%) were female and the median age was 27 (IQR (3) = Q1 (29)-Q3 (26)). Amongst all, 263 participants (85.4%) declared that they have worked in the areas where suspected/confirmed COVID-19 patients were being served. A total of 262 (85%) residents stated that PE habits have changed generally during the pandemic. There was a significant difference with regards to the change in PE habits between those residents who have worked in the COVID-19 areas (n = 230, 87.5%) and those who have not (n = 32, 71.1%) (P = .004). PE habits of Internal Medicine Residents were changed more than others (P < .001). The main reason for the change in PE habits in general (77.9%) and during the examination of suspected/confirmed COVID-19 patients (89.7%) were "self-protection." Independent factors for limited PE in suspected/confirmed COVID-19 patients were found as "Avoiding performing physical examination to be exposed less/to protect (adjusted ORs = 13.067)," "relying on laboratory and radiological investigations during practice (adjusted ORs = 4.358)," and "not having a thought that reduced physical examination will render the diagnosis and course of COVID-19 (adjusted ORs = 2.244)." CONCLUSIONS: This study clearly demonstrated that the COVID-19 pandemic has had a serious impact on the PE habits of the residents while examining patients in general and with COVID-19.

Quality gap in venous thromboembolism prophylaxis practices in inpatients: Assessment of prophylaxis practices in a University Hospital.

BACKGROUND: Venous thromboembolism (VTE) is a prevalent complication of inpatient care, causing high burden of morbidity and mortality. Prophylaxis reduces the events; thus, these practices are promoted. We aimed to assess the VTE prophylaxis practices in the general internal medicine wards of a tertiary-care university hospital and utilize these data to propose quality improvement projects. OBJECTIVE: Assess the pharmacological venous thromboembolism prophylaxis utilization by using the Padua Prediction Score and demonstrate the main determinants of inappropriate prophylaxis use. METHODS: Electronic medical records were used to collect data cross-sectionally from 1 January 2019 to 30 June 2019. Padua Prediction Score was used to assess the appropriateness of VTE prophylaxis. Underlying health conditions, hospitalization causes, anti-platelets and use of glucocorticoids were also assessed as confounding factors. RESULTS: A total of 295 patients were included. Higher proportion of patients who received VTE prophylaxis were obese and had higher rates of diabetes and coronary artery disease. Prophylaxis was more commonly utilized among patients who were elderly, obese, having an acute infection and/or rheumatologic disorder and prolonged bed rest (P ≤ 0.001, P = 0.01, P = 0.22 and P ≤ 0.001, respectively). Thirty-five patients (11.8%) received VTE prophylaxis, despite 89 patients (30.2%) were high-risk patients. In the high-risk group (n = 89), 27 patients (30.3%) received appropriate pharmacological prophylaxis, while 8 patients (3.9%) received VTE prophylaxis in the low-risk group (total n = 206). A considerable percentage of the high-risk population (69.6%) did not receive VTE prophylaxis. DISCUSSION: Underuse constitutes a major challenge in our hospital but differing from other institutions, overuse is not a significant issue. Further studies are needed whether VTE prophylaxis improvement implementations uniformly increase appropriate VTE use.

Thirty-day readmission rate of COVID-19 patients discharged from a tertiary care university hospital in Turkey: an observational, single-center study.

BACKGROUND: The 30-day readmission rate is an important indicator of patient safety and hospital's quality performance. In this study, we aimed to find out the 30-day readmission rate of mild and moderate severity coronavirus disease of 2019 (COVID-19) patients discharged from a tertiary care university hospital and to demonstrate the possible factors associated with readmission. METHODS: This is an observational, single-center study. Epidemiological and clinical data of patients who were hospitalized with a diagnosis of COVID-19 were retrieved from a research database where patient information was recorded prospectively. Readmission data were sought from the hospital information management system and the National Health Information System to detect if the patients were readmitted to any hospital within 30 days of discharge. Adult patients (≥18 years old) hospitalized in COVID-19 wards with a diagnosis of mild or moderate COVID-19 between 20 March 2020 (when the first case was admitted to our hospital) and 26 April 2020 were included. RESULTS: From 26 March to 1 May, there were 154 mild or moderate severity (non-critical) COVID-19 patients discharged from COVID-19 wards, of which 11 (7.1%) were readmitted. The median time of readmission was 8.1 days (interquartile range [IQR] = 5.2). Two patients (18.1%) were categorized to have mild disease and the remaining 9 (81.9%) as moderate disease. Two patients who were over 65 years of age and had metastatic cancers and hypertension developed sepsis and died in the hospital during the readmission episode. Malignancy (18.7% vs. 2.1%, P = 0.04) and hypertension (45.5% vs. 14%, P = 0.02) were more common in those who were readmitted. CONCLUSIONS: This is one of the first studies to report on 30-day readmission rate of COVID-19 in the literature. More comprehensive studies are needed to reveal the causes and predictors of COVID-19 readmissions.

Pulmonary nocardiosis caused by Nocardia abscessus mimicking pulmonary thromboembolism in a patient with atypical anti-glomerular basement membrane glomerulonephritis.

Nocardia species are opportunistic bacteria that are frequently contagious by inhalation. Recently, Nocardia abscessus has been described as a different species. We report a 54-year-old male who presented with acute pleuritic chest pain, mimicking pulmonary thromboembolism on the 5th day of discharge from the hospital. The patient was receiving immunosuppressive therapy for newly diagnosed atypical anti-glomerular basement membrane disease. Thorax computed tomography revealed a 17x19 mm soft tissue lesion in the lateral segment of the middle lobe of the right lung. After further examinations, a biopsy decision was made and Nocardia abscessus was isolated in the aerobic culture of the aspiration material.

Evaluation of the effect of COVID-19 pandemic on anxiety severity of physicians working in the internal medicine department of a tertiary care hospital: a cross-sectional survey.

BACKGROUND: Internists who have an important role in the global response to the COVID-19 pandemic are under both physical and psychological pressures. AIMS: To assess the anxiety among physicians working in the internal medicine department of a tertiary care hospital who are on the frontline of the COVID-19 pandemic. METHODS: This single-centre, non-intervention, cross-sectional descriptive study was conducted using an online survey questionnaire from 1 April to 14 April 2020. Physicians of the Department of Internal Medicine were invited to participate with a self-administered questionnaire. The degree of symptoms of anxiety was assessed by the Turkish versions of the 7-item Generalised Anxiety Disorder scale and Beck Anxiety Inventory, respectively. RESULTS: A total of 113 participants consented for the study and completed the questionnaire. The median age was 29 (IQR = 5) years and 53.1% were male. A total of 72 internists (63.7%) worked as 'frontline' healthcare workers directly engaged in diagnosing, treating or caring for patients with or suspected to have COVID-19. Female gender was significantly associated with high scores and levels in all scales compared to the male gender (P < 0.005). Having family members over 65 years old and with chronic diseases were significantly associated with high anxiety scores and levels (P < 0.005). CONCLUSIONS: In this survey of internists in a university hospital equipped with clinics, wards and intensive care unit for patients with COVID-19, female gender and having family members over 65 years old and with chronic diseases were associated with increased anxiety levels.

Influenza or other respiratory viruses: does it matter as the cause of acute respiratory failure in the critically-ill patients?

INTRODUCTION: Respiratory virus infections may cause serious respiratory failure requiring intensive care unit (ICU) admission. The objective of this study was to evaluate the clinical features and the outcome in patients with acute respiratory failure (ARF) due to viral infections comparing etiological agents. MATERIALS AND METHODS: ARF patients with positive viral serology were retrospectively recruited. Cohort was evaluated with regard to subgroups as influenza and other respiratory viruses (ORV), as well as survivors and nonsurvivors. RESULT: Out of 938 admitted patients, 319 were followed as ARF and only 149 patients had viral respiratory panel results. In 49 patients with ARF, 52 positive viral results were detected and 47 patients with single positive viral isolates of either influenza or ORV were included. Among them, 62% had ORV with quite similar characteristics with influenza group apart from diabetes mellitus which was encountered more in influenza group (p= 0.02). Overall ICU mortality was 32% and there was no difference between the two groups (p= 0.42). Acute Physiology and Chronic Health Evaluation (APACHE) II score was independently associated with ICU mortality (OR: 1.25; 95% CI: 1.04-1.51; p= 0.02). CONCLUSIONS: This study emphasizes to consider the possibility of other respiratory viruses for the cause of ARF with similar characteristics and mortality as influenza species.

The changing trends of peritoneal dialysis related peritonitis and novel risk factors.

AIM: Continuous ambulatory peritoneal dialysis (PD) has become a treatment modality for end stage renal disease with a peak of its use in 1990 s. The aim of this study was to examine the peritonitis rates, causative organisms and the risk factors of peritonitis in a large group of patients in our center. METHODS: The study was conducted in the Nephrology Department of a University Hospital in Turkey. Patients in the PD programme between January 2000 and January 2006 were included. Cohort-specific and subject specific peritonitis incidence, and peritonitis-free survival were calculated. Causative organisms and risk factors were evaluated. RESULTS: Totally 620 episodes of peritonitis occurred in 440 patients over the six years period. Peritonitis rates showed a decreasing trend through the years (0.79 episodes/patient-year 2000-2003 and 0.46 episodes/patient-year 2003-2006). Cohort-specific peritonitis incidence was 0.62 episodes/patient-years and median subject-specific peritonitis incidence was 0.44 episodes/patient-years. The median peritonitis-free survival was 15.25 months (%95 CI, 9.45-21.06 months). The proportion of gram-negative organisms has increased from 9.8% to 17.3%. There was a significant difference in the percentage of culture negative peritonitis between the first three and the last three years (53.1% vs. 43.2%, respectively). Peritonitis incidence was higher in patients who had been transferred from HD, who had catheter related infection and who had HCV infection without cirrhosis. CONCLUSIONS: Our study showed significant trends in the peritonitis rates, causative organisms and antibiotic resistance. Prior HD therapy, catheter related infections and HCV infection were found to be risk factors for peritonitis.

Urine neutrophil gelatinase-associated lipocalin levels predict acute kidney injury in acute decompensated heart failure patients.

BACKGROUND: Acute heart failure (HF) syndromes are frequently complicated with cardiorenal syndromes. The aim of this study was to evaluate the performance of admission neutrophil gelatinase associated lipocalin (NGAL) levels to predict diuretic dose requirement and to predict the occurrence of acute kidney injury (AKI) in patients presenting with acute decompensated HF. METHODS: Patients admitted with HF symptoms between December 2010 and October 2011 were prospectively enrolled. Samples were obtained for NGAL and brain natriuretic peptide. Patients were followed up until discharge or for three days, whichever happened first. They were grouped either to have AKI according to "Acute Kidney Injury Network" criteria or not ("no-AKI"). RESULTS: One hundred patients were enrolled. Urine NGAL levels were higher in AKI group (median 31.3 vs. 16.2 ng/mL) (p < 0.001). Oral furosemide using rates on admission was 60.5% in AKI group, 31.6% in no-AKI group. More AKI developed in patients using less furosemide orally on admission (p = 0.023). Although the mean furosemide doses were similar on the first day (80 mg), diuretic dose increment was less on the following days in AKI group. Urine NGAL levels with 12 ng/mL cut-off value had sensitivity of 79% and specificity of 67% for predicting AKI. Multiple logistic regression analysis yielded an odds ratio of 10.9 for NGAL levels to predict AKI. CONCLUSION: Urine NGAL level in decompensated HF patients was not a significant predictor of diuretic dose requirement, but was a good marker for predicting AKI at 12 ng/mL cut-off value.

Carotid artery intima-media thickness and erectile dysfunction in patients with metabolic syndrome.

BACKGROUND: Metabolic syndrome (MS) has become a pandemic in Turkey, as is the case globally. Increase in carotid artery intima-media thickness (CIMT) and erectile dysfunction (ED) may be evident before the clinical signs of cardiovascular disease appear. We aimed to investigate the prevalence of increased CIMT and ED as markers of atherosclerotic disease in patients with MS. MATERIAL AND METHODS: Thirty-two patients with MS and 29 healthy controls were included. Anthropometric and biochemical parameters, along with total testosterone (TT), high sensitive C-reactive protein (hs-CRP), were recorded. Carotid artery intima-media thickness was measured. Erectile dysfunction was assessed with International Index of Erectile Function. RESULTS: Patients with MS had higher BMI, fasting plasma glucose, post-prandial plasma glucose, insulin, HOMA-IR, total cholesterol, triglycerides, hs-CRP, and CIMT, whereas TT levels were lower (p<0.0001). The prevalence and severity of erectile dysfunction were higher in patients with MS (p<0.0001). Erectile dysfunction scores correlated inversely with CIMT. MS patients with ED were older and had higher CIMT compared to those without ED. Increase in age and HOMA and decrease in TT increased the risk of ED. When KIMT exceeding the 95th percentile of healthy controls was accepted as a risk factor for CVD, presence of ED was the only determinant for this increase. CONCLUSIONS: Erectile dysfunction was more prevalent and severe in patients with MS and correlated with subclinical endothelial dysfunction. Total testosterone deficiency was prominent among MS patients. Presence of ED points to an increased risk of cardiovascular disease when MS is present.

Low bone mineral density in young individuals is associated with greater gingival inflammation and recession.

Most patients with osteoporosis are postmenopausal women or senile people who are deemed to have primary osteoporosis. However, young women, males, and atypical cases need further work up to evaluate the risk factors for secondary osteoporosis. A growing body of literature has accumulated regarding the role of osteoporosis in the onset and progression of periodontal disease and tooth loss. We hypothesized that secondary/idiopathic osteoporosis in young patients will be associated with worse periodontal status. Patients and controls who were seen in the general internal medicine outpatient clinic and who were less than 47 years of age were recruited between December 2005 and June 2011. Bone mineral density (BMD) was measured by dual-energy X-ray absorptiometry. Blood samples were obtained for bone turnover markers and secondary causes of low BMD. Periodontal variables were assessed. Forty-five women whose mean age was 33.9 ± 7.7 years were enrolled. The osteoporotic group consisted of 12 patients, the osteopenic group 17 patients, and the control group 16 subjects. Significantly higher gingival recession (GR), gingival bleeding time index, and hence gingival inflammation were noted in patients with secondary osteoporosis compared to healthy subjects. In logistic regression analysis, having osteoporosis was determined as the single risk factor for increased bleeding time (b = 0.871, p = 0.008), while having osteoporosis (b = 0.181, p = 0.001) and age (b = 0.010, p < 0.001) were significant parameters with regard to GR. In conclusion, low BMD in young individuals was associated with greater gingival inflammation and recession when compared to those individuals with normal BMD values.

Multimodal data fusion using sparse canonical correlation analysis and cooperative learning: a COVID-19 cohort study.

Through technological innovations, patient cohorts can be examined from multiple views with high-dimensional, multiscale biomedical data to classify clinical phenotypes and predict outcomes. Here, we aim to present our approach for analyzing multimodal data using unsupervised and supervised sparse linear methods in a COVID-19 patient cohort. This prospective cohort study of 149 adult patients was conducted in a tertiary care academic center. First, we used sparse canonical correlation analysis (CCA) to identify and quantify relationships across different data modalities, including viral genome sequencing, imaging, clinical data, and laboratory results. Then, we used cooperative learning to predict the clinical outcome of COVID-19 patients: Intensive care unit admission. We show that serum biomarkers representing severe disease and acute phase response correlate with original and wavelet radiomics features in the LLL frequency channel (cor(Xu1, Zv1) = 0.596, p value < 0.001). Among radiomics features, histogram-based first-order features reporting the skewness, kurtosis, and uniformity have the lowest negative, whereas entropy-related features have the highest positive coefficients. Moreover, unsupervised analysis of clinical data and laboratory results gives insights into distinct clinical phenotypes. Leveraging the availability of global viral genome databases, we demonstrate that the Word2Vec natural language processing model can be used for viral genome encoding. It not only separates major SARS-CoV-2 variants but also allows the preservation of phylogenetic relationships among them. Our quadruple model using Word2Vec encoding achieves better prediction results in the supervised task. The model yields area under the curve (AUC) and accuracy values of 0.87 and 0.77, respectively. Our study illustrates that sparse CCA analysis and cooperative learning are powerful techniques for handling high-dimensional, multimodal data to investigate multivariate associations in unsupervised and supervised tasks.

Evaluation of machine learning algorithms for renin-angiotensin-aldosterone system inhibitors associated renal adverse event prediction.

BACKGROUND: Renin-angiotensin-aldosterone system inhibitors (RAASi) are commonly used medications. Renal adverse events associated with RAASi are hyperkalemia and acute kidney injury. We aimed to evaluate the performance of machine learning (ML) algorithms in order to define event associated features and predict RAASi associated renal adverse events. MATERIALS AND METHODS: Data of patients recruited from five internal medicine and cardiology outpatient clinics were evaluated retrospectively. Clinical, laboratory, and medication data were acquired via electronic medical records. Dataset balancing and feature selection for machine learning algorithms were performed. Random forest (RF), k-nearest neighbor (kNN), naïve Bayes (NB), extreme gradient boosting (xGB), support vector machine (SVM), neural network (NN), and logistic regression (LR) were used to create a prediction model. RESULTS: 409 patients were included, and 50 renal adverse events occurred. The most important features predicting the renal adverse events were the index K and glucose levels, as well as having uncontrolled diabetes mellitus. Thiazides reduced RAASi associated hyperkalemia. kNN, RF, xGB and NN algorithms have the highest and similar AUC (≥ 98%), recall (≥ 94%), specifity (≥ 97%), precision (≥ 92%), accuracy (≥ 96%) and F1 statistics (≥ 94%) performance metrics for prediction. CONCLUSION: RAASi associated renal adverse events can be predicted prior to medication initiation by machine learning algorithms. Further prospective studies with large patient numbers are needed to create scoring systems as well as for their validation.

Multimodal Biomedical Data Fusion Using Sparse Canonical Correlation Analysis and Cooperative Learning: A Cohort Study on COVID-19.

Through technological innovations, patient cohorts can be examined from multiple views with high-dimensional, multiscale biomedical data to classify clinical phenotypes and predict outcomes. Here, we aim to present our approach for analyzing multimodal data using unsupervised and supervised sparse linear methods in a COVID-19 patient cohort. This prospective cohort study of 149 adult patients was conducted in a tertiary care academic center. First, we used sparse canonical correlation analysis (CCA) to identify and quantify relationships across different data modalities, including viral genome sequencing, imaging, clinical data, and laboratory results. Then, we used cooperative learning to predict the clinical outcome of COVID-19 patients. We show that serum biomarkers representing severe disease and acute phase response correlate with original and wavelet radiomics features in the LLL frequency channel (corr(Xu1, Zv1) = 0.596, p-value < 0.001). Among radiomics features, histogram-based first-order features reporting the skewness, kurtosis, and uniformity have the lowest negative, whereas entropy-related features have the highest positive coefficients. Moreover, unsupervised analysis of clinical data and laboratory results gives insights into distinct clinical phenotypes. Leveraging the availability of global viral genome databases, we demonstrate that the Word2Vec natural language processing model can be used for viral genome encoding. It not only separates major SARS-CoV-2 variants but also allows the preservation of phylogenetic relationships among them. Our quadruple model using Word2Vec encoding achieves better prediction results in the supervised task. The model yields area under the curve (AUC) and accuracy values of 0.87 and 0.77, respectively. Our study illustrates that sparse CCA analysis and cooperative learning are powerful techniques for handling high-dimensional, multimodal data to investigate multivariate associations in unsupervised and supervised tasks.

Independent predictors of in-hospital mortality and the need for intensive care in hospitalized non-critical COVID-19 patients: a prospective cohort study.

One of the most helpful strategies to deal with ongoing coronavirus pandemics is to use some prudence when treating patients infected with SARS-CoV-2. We aimed to evaluate the clinical, demographic, and laboratory parameters that might have predictive value for in-hospital mortality and the need for intensive care and build a model based on them. This study was a prospective, observational, single-center study including non-critical patients admitted to COVID-19 wards. Besides classical clinic-demographic features, basic laboratory parameters obtained on admission were tested, and then new models for each outcome were developed built on the most significant variables. Receiver operating characteristics (ROC) analyses were performed by calculating each model's probability. A total of 368 non-critical hospitalized patients were recruited, the need for ICU care was observed in 70 patients (19%). The total number of patients who died in either ICU or wards was 39 (10.6%). The first two models (based on clinical features and demographics) were developed to predict ICU and death, respectively; older age, male sex, active cancer, and low baseline saturation were noted to be independent predictors. The area under the curve values of the first two models were noted 0.878 and 0.882 (p < .001; confidence interval [CI] 95% [0.837-0.919], p < .001; CI 95% [0.844-0.922]). Following two models, the third and fourth were based on laboratory parameters with clinic-demographic features. Initial lower sodium and lower albumin levels were determined as independent factors in predicting the need for ICU care; higher blood urea nitrogen and lower albumin were independent factors in predicting in-hospital mortality. The area under the curve values of the third and fourth model was noted 0.938 and 0.929, respectively (p < .001; CI 95% [0.912-0.965], p < .001; CI 95% [0.895-962]). By integrating the widely available blood tests results with simple clinic demographic data, non-critical patients can be stratified according to their risk level. Such stratification is essential to filter the patients' non-critical underlying diseases and conditions that can obfuscate the physician's predictive capacity.