Atmospheric pollen concentrations and chronic obstructive pulmonary disease (COPD) patients visits in Beijing: time series analysis using a generalized additive model.

To investigate the correlation between the daily visits of chronic obstructive pulmonary disease (COPD) patients in hospital clinic and pollen concentrations in Beijing. We collected daily visits of COPD patients of Beijing Shijitan Hospital from April 1st, 2019 to September 30th, 2019. The relationship between pollen concentrations and COPD patient number was analyzed with meteorological factors, time trend, day of the week effect and holiday effect being controlled by the generalized additive model of time series analysis. R4.1.2 software was applied to generate Spearman correlation coefficient, specific and incremental cumulative effect curves of relative risks as well as the response and three-dimensional diagrams for the exposure lag effect prediction. The fitting models were used to predict the lag relative risk and 95% confidence intervals for specific and incremental cumulative effects of specific pollen concentrations. The number of COPD patients was positively correlated with pollen concentration. When pollen concentration increased by 10 grains/1000 mm2, the peak value of the specific cumulative effect appeared on day0, with the effect gone on day4 and a lag time of 4 days observed, whereas the incremental cumulative effect's peak value was shown on day17, and the effect disappeared on day18, with a lag time of 18 days. The results showed that pollen concentration was not only positively correlated with the number of COPD patients, but also had a bimodal lag effect on COPD visits in the hospital at Beijing.

Effect of atmospheric pollen concentration on daily visits of allergic rhinitis in Beijing: a distributed lag nonlinear model analysis.

To investigate the influence and lag effect of atmospheric pollen concentration on daily visits of patients with allergic rhinitis (AR), we collected the AR data during the pollen seasons from 2018 to 2019 from the outpatient and emergency department of Beijing Shijitan Hospital. The distributed lag non-linear model (DLNM) was used to analyze the correlation and the lag effect between pollen concentration and the incidence of AR. R4.1.2 was used to generate the Spearman correlation coefficients and plot the lag response curves of relative risk specific and incremental cumulative effects. In 2018 and 2019, the number of AR visits was moderately positively correlated with pollen concentration. The peak value of the overall specific cumulative effect for every 10 grains/1000 mm2 increase in atmospheric pollen concentration occurred on day 0 (2018, 2019), and the lag disappearance time was day 6 (2018) and day 7 (2019), and the specific cumulative effect duration was respectively 6 days (2018) and 7 days (2019), with the curve showing a downward trend with time increase. In 2018, the peak value of the overall incremental cumulative effect was on day 7, the lag disappearance time was day 13, and the duration of the incremental cumulative effect was 13 days, forming a curve pattern of rising first and then falling. In 2019, the peak value time of the overall incremental cumulative effect was on day 8, and the curve went down afterwards until it showed the trend of ascending again after day26.

Random forest algorithm for predicting postoperative delirium in older patients.

Objective: In this study, we were aimed to identify important variables via machine learning algorithms and predict postoperative delirium (POD) occurrence in older patients. Methods: This study was to make the secondary analysis of data from a randomized controlled trial. The Boruta function was used to screen relevant basic characteristic variables. Four models including Logistic Regression (LR), K-Nearest Neighbor (KNN), the Classification and Regression Tree (CART), and Random Forest (RF) were established from the data set using repeated cross validation, hyper-parameter optimization, and Smote technique (Synthetic minority over-sampling technique, Smote), with the calculation of confusion matrix parameters and the plotting of Receiver operating characteristic curve (ROC), Precision recall curve (PRC), and partial dependence graph for further analysis and evaluation. Results: The basic characteristic variables resulting from Boruta screening included grouping, preoperative Mini-Mental State Examination(MMSE), CHARLSON score, preoperative HCT, preoperative serum creatinine, intraoperative bleeding volume, intraoperative urine volume, anesthesia duration, operation duration, postoperative morphine dosage, intensive care unit (ICU) duration, tracheal intubation duration, and 7-day postoperative rest and move pain score (median and max; VAS-Rest-M, VAS-Move-M, VAS-Rest-Max, and VAS-Move-Max). And Random Forest (RF) showed the best performance in the testing set among the 4 models with Accuracy: 0.9878; Matthews correlation coefficient (MCC): 0.8763; Area under ROC curve (AUC-ROC): 1.0; Area under the PRC Curve (AUC-PRC): 1.0. Conclusion: A high-performance algorithm was established and verified in this study demonstrating the degree of POD risk changes in perioperative elderly patients. And the major risk factors for the development of POD were CREA and VAS-Move-Max.