Search

ICU readmissions and post-discharge mortality pose significant challenges. Previous studies used EHRs and machine learning models, but mostly focused on structured data. Nursing records contain crucial unstructured information, but their utilization is challenging. Natural language processing (NLP) can extract structured features from clinical text. This study proposes the Crucial Nursing Description Extractor (CNDE) to predict post-ICU discharge mortality rates and identify high-risk patients for unplanned readmission by analyzing electronic nursing records. Developed a deep neural network (NurnaNet) with the ability to perceive nursing records, combined with a bio-clinical medicine pre-trained language model (BioClinicalBERT) to analyze the electronic health records (EHRs) in the MIMIC III dataset to predict the death of patients within six month and two year risk. A cohort and system development design was used. Based on data extracted from MIMIC-III, a database of critically ill in the US between 2001 and 2012, the results were analyzed. We calculated patients' age using admission time and date of birth information from the MIMIC dataset. Patients under 18 or over 89 years old, or who died in the hospital, were excluded. We analyzed 16,973 nursing records from patients' ICU stays. We have developed a technology called the Crucial Nursing Description Extractor (CNDE), which extracts key content from text. We use the logarithmic likelihood ratio to extract keywords and combine BioClinicalBERT. We predict the survival of discharged patients after six months and two years and evaluate the performance of the model using precision, recall, the F 1 -score, the receiver operating characteristic curve (ROC curve), the area under the curve (AUC), and the precision–recall curve (PR curve). The research findings indicate that NurnaNet achieved good F 1 -scores (0.67030, 0.70874) within six months and two years. Compared to using BioClinicalBERT alone, there was an improvement in performance of 2.05 % and 1.08 % for predictions within six months and two years, respectively. CNDE can effectively reduce long-form records and extract key content. NurnaNet has a good F 1 -score in analyzing the data of nursing records, which helps to identify the risk of death of patients after leaving the hospital and adjust the regular follow-up and treatment plan of relevant medical care as soon as possible. [ABSTRACT FROM AUTHOR]

Background: Among patients in whom childhood cancer was diagnosed in the 1970s and 1980s, 18% of those who survived for 5 years died within the subsequent 25 years. In recent decades, cancer treatments have been modified with the goal of reducing life-threatening late effects.Methods: We evaluated late mortality among 34,033 patients in the Childhood Cancer Survivor Study cohort who survived at least 5 years after childhood cancer (i.e., cancer diagnosed before the age of 21 years) for which treatment was initiated during the period from 1970 through 1999. The median follow-up was 21 years (range, 5 to 38). We evaluated demographic and disease factors that were associated with death from health-related causes (i.e., conditions that exclude recurrence or progression of the original cancer and external causes but include the late effects of cancer therapy) using cumulative incidence and piecewise exponential models to estimate relative rates and 95% confidence intervals.Results: Of the 3958 deaths that occurred during the study period, 1618 (41%) were attributable to health-related causes, including 746 deaths from subsequent neoplasms, 241 from cardiac causes, 137 from pulmonary causes, and 494 from other causes. A reduction in 15-year mortality was observed for death from any cause (from 12.4% in the early 1970s to 6.0% in the 1990s, P<0.001 for trend) and from health-related causes (from 3.5% to 2.1%, P<0.001 for trend). These reductions were attributable to decreases in the rates of death from subsequent neoplasm (P<0.001), cardiac causes (P<0.001), and pulmonary causes (P=0.04). Changes in therapy according to decade included reduced rates of cranial radiotherapy for acute lymphoblastic leukemia (85% in the 1970s, 51% in the 1980s, and 19% in the 1990s), of abdominal radiotherapy for Wilms' tumor (78%, 53%, and 43%, respectively), of chest radiotherapy for Hodgkin's lymphoma (87%, 79%, and 61%, respectively), and of anthracycline exposure. Reduction in treatment exposure was associated with reduced late mortality among survivors of acute lymphoblastic leukemia and Wilms' tumor.Conclusions: The strategy of lowering therapeutic exposure has contributed to an observed decline in late mortality among 5-year survivors of childhood cancer. (Funded by the National Cancer Institute and the American Lebanese-Syrian Associated Charities.). [ABSTRACT FROM AUTHOR]