<strong>Background: </strong>The alteration of lymphocyte subpopulations can help to predict the severity and the prognosis of severe Coronavirus disease 2019 (COVID-19). Our goal was to describe the kinet...<strong>Background: </strong>The alteration of lymphocyte subpopulations can help to predict the severity and the prognosis of severe Coronavirus disease 2019 (COVID-19). Our goal was to describe the kinetics of lymphocyte subsets, and their impact on the severity and mortality in critically ill COVID-19 patients. <strong>Methods: </strong>We collected demographic data, comorbidities, clinical signs on admission, laboratory findings on admission then a follow-up during hospitalization. Lymphocyte subsets including CD3+ T cells, CD4+ T cells, CD8+ T cells, B cells, and natural killer (NK) cells were counted by flow cytometer. <strong>Results:</strong> On admission, we observed lymphopenia in 57% of cases, decreased CD3+ T cells in 76% of cases, decreased CD4+ T cells in 81% of cases, decreased CD8+ T cells in 62% of cases, decreased B cells in 52% of cases, and decreased natural killer (NK) cells in 33% of cases. After treatment, decreased CD3+ T cells, decreased CD4+ T cells, decreased CD8+ T cells, and decreased natural killer cells were predictor factors of mortality, in the univariable analysis.<strong> Conclusion:</strong> CD3+ T cells, CD4+ T cells, CD8+ T cells, and natural killer cells were predictor factors of severity, ICU mortality, and also a useful tool for predicting disease progression.展开更多
BACKGROUND Diabetic ketoacidosis(DKA)contributes to 94%of diabetes-related hospital admissions,and its incidence is rising.Due to the complexity of its management and the need for rigorous monitoring,many DKA patients...BACKGROUND Diabetic ketoacidosis(DKA)contributes to 94%of diabetes-related hospital admissions,and its incidence is rising.Due to the complexity of its management and the need for rigorous monitoring,many DKA patients are managed in the intensive care unit(ICU).However,studies comparing DKA patients managed in ICU to non-ICU settings show an increase in healthcare costs without significantly affecting patient outcomes.It is,therefore,essential to identify suitable candidates for ICU care in DKA patients.AIM To evaluate factors that predict the requirement for ICU care in DKA patients.METHODS This retrospective study included consecutive patients with index DKA episodes who presented to the emergency department of four general hospitals of Hamad Medical Corporation,Doha,Qatar,between January 2015 and March 2021.All adult patients(>14 years)fulfilling the American Diabetes Association criteria for DKA diagnosis were included.RESULTS We included 922 patients with DKA in the final analysis,of which 229(25%)were managed in the ICU.Compared to non-ICU patients,patients admitted to ICU were older[mean(SD)age of 40.4±13.7 years vs 34.5±14.6 years;P<0.001],had a higher body mass index[median(IQR)of 24.6(21.5-28.4)kg/m2 vs 23.7(20.3-27.9)kg/m2;P<0.030],had T2DM(61.6%)and were predominantly males(69%vs 31%;P<0.020).ICU patients had a higher white blood cell count[median(IQR)of 15.1(10.2-21.2)×103/uL vs 11.2(7.9-15.7)×103/uL,P<0.001],urea[median(IQR)of 6.5(4.6-10.3)mmol/L vs 5.6(4.0-8.0)mmol/L;P<0.001],creatinine[median(IQR)of 99(75-144)mmol/L vs 82(63-144)mmol/L;P<0.001],C-reactive protein[median(IQR)of 27(9-83)mg/L vs 14(5-33)mg/L;P<0.001]and anion gap[median(IQR)of 24.0(19.2-29.0)mEq/L vs 22(17-27)mEq/L;P<0.001];while a lower venous pH[mean(SD)of 7.10±0.15 vs 7.20±0.13;P<0.001]and bicarbonate level[mean(SD)of 9.2±4.1 mmol/L vs 11.6±4.3 mmol/L;P<0.001]at admission than those not requiring ICU management of DKA(P<0.001).Patients in the ICU group had a longer LOS[median(IQR)of 4.2(2.7-7.1)d vs 2.0(1.0-3.9)d;P<0.001]and DKA duration[median(IQR)of 24(13-37)h vs 15(19-24)h,P<0.001]than those not requiring ICU admission.In the multivariate logistic regression analysis model,age,Asian ethnicity,concurrent coronavirus disease 2019(COVID-19)infection,DKA severity,DKA trigger,and NSTEMI were the main predicting factors for ICU admission.CONCLUSION In the largest tertiary center in Qatar,25%of all DKA patients required ICU admission.Older age,T2DM,newly onset DM,an infectious trigger of DKA,moderate-severe DKA,concurrent NSTEMI,and COVID-19 infection are some factors that predict ICU requirement in a DKA patient.展开更多
BACKGROUND Intensive care unit(ICU)patients demand continuous monitoring of several clinical and laboratory parameters that directly influence their medical progress and the staff’s decision-making.Those data are vit...BACKGROUND Intensive care unit(ICU)patients demand continuous monitoring of several clinical and laboratory parameters that directly influence their medical progress and the staff’s decision-making.Those data are vital in the assistance of these patients,being already used by several scoring systems.In this context,machine learning approaches have been used for medical predictions based on clinical data,which includes patient outcomes.AIM To develop a binary classifier for the outcome of death in ICU patients based on clinical and laboratory parameters,a set formed by 1087 instances and 50 variables from ICU patients admitted to the emergency department was obtained in the“WiDS(Women in Data Science)Datathon 2020:ICU Mortality Prediction”dataset.METHODS For categorical variables,frequencies and risk ratios were calculated.Numerical variables were computed as means and standard deviations and Mann-Whitney U tests were performed.We then divided the data into a training(80%)and test(20%)set.The training set was used to train a predictive model based on the Random Forest algorithm and the test set was used to evaluate the predictive effectiveness of the model.RESULTS A statistically significant association was identified between need for intubation,as well predominant systemic cardiovascular involvement,and hospital death.A number of the numerical variables analyzed(for instance Glasgow Coma Score punctuations,mean arterial pressure,temperature,pH,and lactate,creatinine,albumin and bilirubin values)were also significantly associated with death outcome.The proposed binary Random Forest classifier obtained on the test set(n=218)had an accuracy of 80.28%,sensitivity of 81.82%,specificity of 79.43%,positive predictive value of 73.26%,negative predictive value of 84.85%,F1 score of 0.74,and area under the curve score of 0.85.The predictive variables of the greatest importance were the maximum and minimum lactate values,adding up to a predictive importance of 15.54%.CONCLUSION We demonstrated the efficacy of a Random Forest machine learning algorithm for handling clinical and laboratory data from patients under intensive monitoring.Therefore,we endorse the emerging notion that machine learning has great potential to provide us support to critically question existing methodologies,allowing improvements that reduce mortality.展开更多
文摘<strong>Background: </strong>The alteration of lymphocyte subpopulations can help to predict the severity and the prognosis of severe Coronavirus disease 2019 (COVID-19). Our goal was to describe the kinetics of lymphocyte subsets, and their impact on the severity and mortality in critically ill COVID-19 patients. <strong>Methods: </strong>We collected demographic data, comorbidities, clinical signs on admission, laboratory findings on admission then a follow-up during hospitalization. Lymphocyte subsets including CD3+ T cells, CD4+ T cells, CD8+ T cells, B cells, and natural killer (NK) cells were counted by flow cytometer. <strong>Results:</strong> On admission, we observed lymphopenia in 57% of cases, decreased CD3+ T cells in 76% of cases, decreased CD4+ T cells in 81% of cases, decreased CD8+ T cells in 62% of cases, decreased B cells in 52% of cases, and decreased natural killer (NK) cells in 33% of cases. After treatment, decreased CD3+ T cells, decreased CD4+ T cells, decreased CD8+ T cells, and decreased natural killer cells were predictor factors of mortality, in the univariable analysis.<strong> Conclusion:</strong> CD3+ T cells, CD4+ T cells, CD8+ T cells, and natural killer cells were predictor factors of severity, ICU mortality, and also a useful tool for predicting disease progression.
文摘BACKGROUND Diabetic ketoacidosis(DKA)contributes to 94%of diabetes-related hospital admissions,and its incidence is rising.Due to the complexity of its management and the need for rigorous monitoring,many DKA patients are managed in the intensive care unit(ICU).However,studies comparing DKA patients managed in ICU to non-ICU settings show an increase in healthcare costs without significantly affecting patient outcomes.It is,therefore,essential to identify suitable candidates for ICU care in DKA patients.AIM To evaluate factors that predict the requirement for ICU care in DKA patients.METHODS This retrospective study included consecutive patients with index DKA episodes who presented to the emergency department of four general hospitals of Hamad Medical Corporation,Doha,Qatar,between January 2015 and March 2021.All adult patients(>14 years)fulfilling the American Diabetes Association criteria for DKA diagnosis were included.RESULTS We included 922 patients with DKA in the final analysis,of which 229(25%)were managed in the ICU.Compared to non-ICU patients,patients admitted to ICU were older[mean(SD)age of 40.4±13.7 years vs 34.5±14.6 years;P<0.001],had a higher body mass index[median(IQR)of 24.6(21.5-28.4)kg/m2 vs 23.7(20.3-27.9)kg/m2;P<0.030],had T2DM(61.6%)and were predominantly males(69%vs 31%;P<0.020).ICU patients had a higher white blood cell count[median(IQR)of 15.1(10.2-21.2)×103/uL vs 11.2(7.9-15.7)×103/uL,P<0.001],urea[median(IQR)of 6.5(4.6-10.3)mmol/L vs 5.6(4.0-8.0)mmol/L;P<0.001],creatinine[median(IQR)of 99(75-144)mmol/L vs 82(63-144)mmol/L;P<0.001],C-reactive protein[median(IQR)of 27(9-83)mg/L vs 14(5-33)mg/L;P<0.001]and anion gap[median(IQR)of 24.0(19.2-29.0)mEq/L vs 22(17-27)mEq/L;P<0.001];while a lower venous pH[mean(SD)of 7.10±0.15 vs 7.20±0.13;P<0.001]and bicarbonate level[mean(SD)of 9.2±4.1 mmol/L vs 11.6±4.3 mmol/L;P<0.001]at admission than those not requiring ICU management of DKA(P<0.001).Patients in the ICU group had a longer LOS[median(IQR)of 4.2(2.7-7.1)d vs 2.0(1.0-3.9)d;P<0.001]and DKA duration[median(IQR)of 24(13-37)h vs 15(19-24)h,P<0.001]than those not requiring ICU admission.In the multivariate logistic regression analysis model,age,Asian ethnicity,concurrent coronavirus disease 2019(COVID-19)infection,DKA severity,DKA trigger,and NSTEMI were the main predicting factors for ICU admission.CONCLUSION In the largest tertiary center in Qatar,25%of all DKA patients required ICU admission.Older age,T2DM,newly onset DM,an infectious trigger of DKA,moderate-severe DKA,concurrent NSTEMI,and COVID-19 infection are some factors that predict ICU requirement in a DKA patient.
文摘BACKGROUND Intensive care unit(ICU)patients demand continuous monitoring of several clinical and laboratory parameters that directly influence their medical progress and the staff’s decision-making.Those data are vital in the assistance of these patients,being already used by several scoring systems.In this context,machine learning approaches have been used for medical predictions based on clinical data,which includes patient outcomes.AIM To develop a binary classifier for the outcome of death in ICU patients based on clinical and laboratory parameters,a set formed by 1087 instances and 50 variables from ICU patients admitted to the emergency department was obtained in the“WiDS(Women in Data Science)Datathon 2020:ICU Mortality Prediction”dataset.METHODS For categorical variables,frequencies and risk ratios were calculated.Numerical variables were computed as means and standard deviations and Mann-Whitney U tests were performed.We then divided the data into a training(80%)and test(20%)set.The training set was used to train a predictive model based on the Random Forest algorithm and the test set was used to evaluate the predictive effectiveness of the model.RESULTS A statistically significant association was identified between need for intubation,as well predominant systemic cardiovascular involvement,and hospital death.A number of the numerical variables analyzed(for instance Glasgow Coma Score punctuations,mean arterial pressure,temperature,pH,and lactate,creatinine,albumin and bilirubin values)were also significantly associated with death outcome.The proposed binary Random Forest classifier obtained on the test set(n=218)had an accuracy of 80.28%,sensitivity of 81.82%,specificity of 79.43%,positive predictive value of 73.26%,negative predictive value of 84.85%,F1 score of 0.74,and area under the curve score of 0.85.The predictive variables of the greatest importance were the maximum and minimum lactate values,adding up to a predictive importance of 15.54%.CONCLUSION We demonstrated the efficacy of a Random Forest machine learning algorithm for handling clinical and laboratory data from patients under intensive monitoring.Therefore,we endorse the emerging notion that machine learning has great potential to provide us support to critically question existing methodologies,allowing improvements that reduce mortality.
