Deep-Learning-Based Survival Prediction of Patients in Coronary Care Units

Background. A survival prediction model based on deep learning has higher accuracy than the CPH model in predicting the survival of CCU patients, and it also has a better discrimination ability. We collected information on patients with various diseases in coronary care units (CCUs) from the Medical Information Mart for Intensive Care III (MIMIC-III) database. The purpose of this study was to use this information to construct a neural-network model based on deep learning to predict the survival probabilities of patients with conditions that are common in CCUs. Method. We collected information on patients in the United States with five common diseases in CCUs from 2001 to 2012. We randomly divided the patients into a training cohort and a testing cohort at a ratio of 7 : 3 and applied a survival prediction method based on deep learning to predict their survival probability. We compared our model with the Cox proportional-hazards regression (CPH) model and used the concordance indexes (C-indexes), receiver operating characteristic (ROC) curve, and calibration plots to evaluate the predictive performance of the model. Results. The 3,388 CCU patients included in the study were randomly divided into 2,371 in the training cohort and 1,017 in the testing cohort. The stepwise regression results showed that the important factors affecting patient survival were the type of disease, age, race, anion gap, glucose, neutrophils, white blood cells, potassium, creatine kinase, and blood urea nitrogen ( P < 0.05 ). We used the training cohort to construct a deep-learning model, for which the C-index was 0.833, or about 5% higher than that for the CPH model (0.786). The C-index of the deep-learning model for the test cohort was 0.822, which was also higher than that for the CPH model (0.782). The areas under the ROC curve for the 28-day, 90-day, and 1-year survival probabilities were 0.875, 0.865, and 0.874, respectively, in the deep-learning model, respectively, and 0.830, 0.843, and 0.806 in the CPH model. These values indicate that the survival analysis model based on deep learning is better than the traditional CPH model in predicting the survival of CCU patients. Conclusion. A survival prediction model based on deep learning has higher accuracy than the CPH model in predicting the survival of CCU patients, and it also has a better discrimination ability.

Analysis of Indications for Electrotherapy Using Classification Trees

Electrotherapy is a dynamically developing method of treatment of sinus node dysfunction and atrioventricular conduction disturbances. It is an extremely important method used in the treatment of heart failure. The aim of this paper was to use classification trees for the differentiation between patients implanted with one of the three electrotherapy devices, i.e. SC-VVI/AAI, DC-DDD, ICD/CRT. The analysed data concerned 2071 patients who underwent implantation or device replacement procedures in the years 2010–2018, hospitalized in a coronary care unit. CART-type classification trees with 5-fold cross-validation were used for the analysis. The decision concerning the choice of a particular electrotherapy device is always made based on the latest guidelines and the patient’s clinical condition. The used classification trees may enable verification of the state of implementation of guidelines in real-life therapeutic decisions.

D-dimer as a prognostic factor in patients admitted to a tertiary care intensive coronary care unit

Introduction D-dimer is a small protein fragment and is a product of fibrinolysis. A high levels of D-dimer have been suggested as a prognostic factor in cancerous and other critically ill patients. We aimed to evaluate D-dimer levels and outcomes of critically ill patients admitted to a tertiary care intensive coronary care unit (ICCU). Material and method All patients admitted to the ICCU at our Medical Center between January 1, 2020 and December 31, 2020 were included in the study. Patients were divided into 2 groups according to their D-dimer level on admission. Low D-dimer level &lt;500 ng/ml, and high D-dimer level ≥500 ng/ml. Survival, in-hospital interventions and complications were compared. Results and discussion Overall 1,082 consecutive patients were included, mean age was 67 (±16), 70% were males. Of them 296 (27.4%) had low D-dimer level and 663 (61.3%) had high D-dimer level. Patients with high D-dimer level were older as compared to patients with low D-dimer level (mean age 70.4±15 and 59±13 years respectively, p=0.004), had significantly higher rate of female gender (35.9% vs 15.9% respectively, p&lt;0.0001) and significantly higher rate of any prior cardiac interventions prior to their admission (26.7% vs 4.4% respectively, p&lt;0.0001). Interestingly, patients with high D-dimer level had significantly lower rate of any acute coronary syndrome (ACS) as compared with the low D-dimer group (25.7 vs 66.4% respectively, p&lt;0.0001) and lower rate of smokers (22.5 vs 45.6% respectively, p&lt;0.0001). All 11 post-COVID-19 patients had high D-dimer level on admission. A multivariate Cox proportional hazards analysis for mortality, adjusted for age, gender, risk factors for cardiovascular disease, ejection fraction&lt;40 found that high D-dimer level was independently associated with higher mortality rates (HR=5.8; 95% CI; 1.7–19.1; p=0.004) as shown in Figure 1. Conclusion Elevated D-dimer levels on admission in ICCU patients is a poor prognostic factor of in-hospital morbidity and mortality in the first year following hospitalization. FUNDunding Acknowledgement Type of funding sources: None. Cumulative survival according to d-Dimer

Risk Stratification Model for Predicting Coronary Care Unit Readmission

Background: Use of statistical models for assessing the clinical risk of readmission to medical and surgical intensive care units is well established. However, models for predicting risk of coronary care unit (CCU) readmission are rarely reported. Therefore, this study investigated the characteristics and outcomes of patients readmitted to CCU to identify risk factors for CCU readmission and to establish a scoring system for identifying patients at high risk for CCU readmission. Methods: Medical data were collected for 40,187 patients with a history of readmission to the CCU of a single multi‐center healthcare provider in Taiwan during 2010-2019. Characteristics and outcomes were compared between a readmission group and a non-readmission group. Data were segmented at a 9:1 ratio for model building and validation.Results: The number of patients with a CCU readmission history after transfer to a standard care ward was 2397 (5.9%). The twelve factors that had the strongest associations with CCU readmission were used to develop and validate a CCU readmission risk scoring and prediction model. When the model was used to predict CCU readmission, the receiver-operating curve characteristic was 0.7217 for risk score model group and 0.7316 for the validation group. A CCU readmission risk score was assigned to each patient. The patients were then stratified by risk score into low risk (-20-5), moderate risk (6-26) and high risk (27-33) cohorts check scores, which showed that CCU readmission risk significantly differed among the three groups.Conclusions: This study developed a model for estimating CCU readmission risk. By using the proposed model, clinicians can improve CCU patient outcomes and medical care quality.

Correlation of magnesium and calcium in the management of cardiac arrhythmia: Perspectives for better outcome

Objectives: To find correlation between serum Mg, serum Ca, and cardiac arrhythmia. Materials and Methods: The present case–control analytical study includes records of 100 participants; 50 patients (both male and female average age: 47 ± 12 years, mean ± SD) admitted during the period of March 2019–March 2020 into the Coronary Care Unit of LG Hospital, AMCMET Medical College who were clinically diagnosed as arrhythmia and 50 subjects for control group from OPD patients coming to the same institution for health check-up. Mg was estimated with xylitol blue colorimetric end-point method and Ca was estimated by NM-BAPTA Method by Roche Cobas c311 instrument. Results: In 50 cases, mean Mg value was 1.454 mg/dl and SD 0.2566 while in control group, mean value was 2.2 mg/dl and SD is 0.3110 with 95% confidence interval of 1.381–1.527 and 2.199–2.375 for cases and controls group, respectively, which was statistically significant (p < 0.0001). In 50 cases, mean Ca value was 8.6426 mg/dl and SD 1.3 mg/dl while in control group, mean value was 9.5 mg/dl and SD 0.47 with 95% confidence interval of 8.268–9.018 and 9.377–9.643 for cases and controls, respectively, which was statistically significant (p < 0.0028) and shows correlation between serum Ca and serum Mg which are low in cardiac arrhythmias. Receiver operating characteristic analysis of Ca: Mg (3.36) ratio showed optimum cutoff in diagnosis of cardiac arrhythmia. Conclusion: We concluded that serum Mg and Ca along with Ca/Mg ratio should be considered as an important parameter for investigation of cardiac disorders, especially for patients of cardiac arrhythmia.

Safety of Early and Late Discharge in Patients with ST Elevation Myocardial Infarction after Primary Percutaneous Coronary Intervention

Background: Primary percutaneous coronary intervention (PPCI) is now a standard treatment procedure for ST elevation myocardial infraction (STEMI) patients. Because of the many STEMI patients, there is a space constraint in coronary care unit, especially in Southeast Asian countries. Therefore, we practitioners should be evaluating if the patients could be safely discharged earlier. The current European Society of Cardiology STEMI 2017 guideline recommended early discharge in stable patients; however, the data are limited, especially in the Asian countries. Objective: To determine the rate of 30-day, 1-year mortality, and readmission of STEMI patients that underwent PPCI and were discharged early within three days of admission, compared with the late discharge of more than three days after admission. Materials and Methods: The present study was a retrospective cohort study at King Chulalongkorn Memorial Hospital. The authors collected consecutive cases of STEMI patients that underwent PPCI and were discharged between January 1999 and December 2015.The patients were divided into two groups as group 1 with early discharge within three days of admission and group 2 with late discharge more than three days of admission. The follow up on the mortality and readmission rates were collected at 30-day and 1-year after discharge. Results: Out of 1,242 STEMI patients, 691 patients (55.6%) were classified in group 1 and 551 patients (44.4%) were in group 2. The 30-day mortality was 0.4% in group 1 compared with 1.3% in group 2 (HR 2.93, p=0.12) and 1-year mortality was 3.9% in group 1 compared with 8.0% in group 2 (HR 2.09, p=0.003). There was no difference in 30-day readmission between both groups at 1.3% versus 2.5% (OR 1.98, p=0.113), but there was a difference in 1-year readmission between the two groups at 4.5% versus 10.6% (OR 2.51, p<0.001). In multivariate analysis, the predictive factors for early discharged STEMI patients were male (adjusted OR 1.78, p=0.007), Killip classification 1, 2, and 3 (adjusted OR 5.85, p=0.001), EF greater than 40% (adjusted OR 2.51, p=0.001), and TIMI flow after PPCI 3 (adjusted OR 1.48, p=0.016). Conclusion: Early discharge in STEMI patients within three days after PPCI is safe in terms of mortality and readmission compared to late discharge, especially in STEMI patients with Killip class I. Early discharge can provide more space for coronary care. Keywords: STEMI; PPCI; Early discharge; Late discharge; Mortality; Readmission; Killip class

Neutrophil Percentage to Albumin Ratio Was Associated with Clinical Outcomes in Coronary Care Unit Patients

Background: Neutrophil percentage to albumin ratio (NPAR) was proved to be correlated with the prognosis of a variety of diseases. The purpose of this study was to explore the effect of NPAR on the prognosis of coronary care unit (CCU) dpatients.Method: All data of this study was extracted from Medical Information Mart for Intensive Care III (MIMIC-III, version1.4) database. All patients were divided into four groups according to NPAR quartiles. Primary outcome was in-hospital mortality and secondary outcomes were 30-day mortality, 365-day mortality, length of CCU stay, length of hospital stay, acute kidney injury, renal replacement therapy. Multivariable binary logistic regression analysis was performed to confirm the independent effect of NPAR. Subgroup analysis was used to determine the effect of NPAR on in-hospital mortality in different subgroups. Receiver-operating characteristic (ROC) curves were applied to evaluate the ability of NPAR to predict in-hospital mortality. Kaplan–Meier curves were built to compare cumulative survival of different groups.Result: 2364 CCU patients were enrolled in this study. In-hospital mortality rate increased significantly as NPAR quartiles increased (P < 0.001). In multivariable logistic regression, NPAR was proved to be independently associated with in-hospital mortality (Quartile 4 vs Quartile 1: OR, 95% CI: 1.80, 1.19-2.72, P=0.005, P for trend = 0.001). Moderate ability of NPAR to predict in-hospital mortality was demonstrated through ROC curves, the AUC of NPAR was 0.653 (P<0.001), which is better than PLR (P<0.001), neutrophil (P<0.001) but lower than SOFA(P=0.046) and SAPS II (P<0.001). Subgroup analysis did not find obvious interaction in most subgroups. Moreover, Kaplan-Meier curves showed that as NPAR quartiles increased, 30-day (Log rank, P<0.001) and 365-day (Log rank, P<0.001) cumulative survival decreased significantly. NPAR was also proved to be independently associated with acute kidney injury (Quartile 4 vs Quartile 1: OR, 95% CI: 1.57, 1.19-2.07, P=0.002, P for trend = 0.001). Length of CCU and hospital stay were prolonged significantly in higher NPAR quartiles.Conclusion: NPAR was an independent risk factor of in-hospital mortality in CCU patients and had a moderate ability to predict in-hospital mortality.