Missing data imputation using classification and regression trees.

Background: Missing data are common when analyzing real data. One popular solution is to impute missing data so that one complete dataset can be obtained for subsequent data analysis. In the present study, we focus on missing data imputation using classification and regression trees (CART). Methods: We consider a new perspective on missing data in a CART imputation problem and realize the perspective through some resampling algorithms. Several existing missing data imputation methods using CART are compared through simulation studies, and we aim to investigate the methods with better imputation accuracy under various conditions. Some systematic findings are demonstrated and presented. These imputation methods are further applied to two real datasets: Hepatitis data and Credit approval data for illustration. Results: The method that performs the best strongly depends on the correlation between variables. For imputing missing ordinal categorical variables, the rpart package with surrogate variables is recommended under correlations larger than 0 with missing completely at random (MCAR) and missing at random (MAR) conditions. Under missing not at random (MNAR), chi-squared test methods and the rpart package with surrogate variables are suggested. For imputing missing quantitative variables, the iterative imputation method is most recommended under moderate correlation conditions.

[Expression of pGSK-3α/ß Tyr279/216 and XIAP proteins in cholangiocarcinoma and their clinical significance].

OBJECTIVE: To investigate the expressions of the active form of glycogen synthase kinase-3(GSK-3)-pGSK-3α/ß (Tyr279/216) and its downstream moleculor X-linked inhibitor of apoptosis protein (XIAP) in cholangiocarcinoma and to analyze their correlation with clinicopathological and survival significance. METHODS: Immunohistoehemistry was used to detect the expressions of the active form of GSK-3- pGSK-3α/ß (Tyr279/216) and its downstream moleculor XIAP proteins in 50 cholangiocarcinoma tissues and 20 normal bile duct tissues. RESULTS: The positive rates of pGSK-3α/ß (Tyr279/216) and XIAP were 62.0% and 68.0% in cholangiocarcinoma, and 10.0% and 25.0% in normal bile duct tissues, respectively. The intensity of pGSK-3α/ß (Tyr279/216) and XIAP expressions in cholangiocarcinoma were significantly higher than that in the normal bile duct tissues (P < 0.001), and there was a significant correlation between pGSK-3α/ß (Tyr279/216) and XIAP expressions (r = 0.544, P < 0.001). The expression of pGSK-3α/ß(Tyr279/216) protein in cholangiocarcinoma was associated with TNM stage (P = 0.042), histological grade (P = 0.031), whereas the expression of XIAP protein in cholangiocarcinoma was correlated with CEA level (P = 0.006). Patients with positive expression of pGSK-3α/ß (Tyr279/216) and XIAP demonstrate a significantly worse prognosis than that of patients with negative expression of pGSK-3α/ß (Tyr279/216) and XIAP for overall survival (P = 0.002, P = 0.018). Multivariate survival analysis revealed that positive pGSK-3α/ß (Tyr279/216) expression provided significant independent prognostic value for overall survival (P = 0.002). CONCLUSIONS: The expressions of pGSK-3α/ß(Tyr279/216) and XIAP proteins were significantly associated with the development and progression of cholangiocarcinoma. pGSK-3α/ß(Tyr279/216) may be an important prognostic factor for survival of patients with cholangiocarcinoma.

Do probiotics help prevent ventilator-associated pneumonia in critically ill patients? A systematic review with meta-analysis.

BACKGROUND: Probiotic treatments might contribute to the prevention of ventilator-associated pneumonia (VAP). Due to its unclear clinical effects, here we intend to assess the preventive effect and safety of probiotics on intensive care unit (ICU) patients. METHODS: Eligible randomised controlled trials were selected in databases until 30 September 2019. The characteristics of the studies were extracted, including study design, definition of VAP, probiotics intervention, category of included patients, incidence of VAP, mortality, duration of mechanical ventilation (MV) and ICU stay. Heterogeneity was evaluated by Chi-squared and I2 tests. RESULTS: 15 studies involving 2039 patients were identified for analysis. The pooled analysis suggests significant reduction on VAP (risk ratio, 0.68; 95% Cl, 0.60 to 0.77; p<0.00001) in a fixed-effects model. Subgroup analyses performed on the category of clinical and microbiological criteria both support the above conclusion; however, there were no significant differences in duration of MV or length of ICU stay in a random-effects model. Also, no significant differences in total mortality, overall mortality, 28-day mortality or 90-day mortality were found in the fixed-effects model. CONCLUSIONS: The probiotics helped to prevent VAP without impacting the duration of MV, length of ICU stay or mortality.