Comparison of Bayesian statistics and classical statistics in quantile regression analysis / 军事医学

Objective To compare the Bayesian statistics and the classical statistics in the quantile regression analysis in order to select a more effective method .Methods The large sample data was chosen , and the QUANTREG procedure in SAS was used for the classical statistics and the MCMC procedure in SAS for the Bayesian one , respectively .Using ten-fold cross-validation method , the goodness of fitting of the models was appraised in terms of the fitted effect based on the training dataset and the predicted effect based on the predictive dataset .Results In most cases, the indexes of the quantile regression models in the classical statistics were slightly worse than those of the Bayesian one .In the ten-fold cross-validation of the partial samples as a training dataset , the fitting effect of the lower quartile ( Q1 ) and upper quartile ( Q3 ) of the Bayesian statistics was better than that of the classical one .However , the median ( Q2 ) fitting effect of the Bayesian statistics was slightly worse than that of the classical one .As for the prediction effect , the Bayesian statistical quantile regression model was superior to the classic one .Conclusion To expect high accuracy , such as the predictive effects and fitting effects of each quantile , the Bayesian quantile regression analysis should be chosen .If the major concern is the fitting effect of the median , careful selection from the approaches mentioned above is needed .If time and energy are limited, and the sample size is large enough , the classic statistical quantile regression analysis is a good choice .

Performance comparison between Logistic regression, decision trees, and multilayer perceptron in predicting peripheral neuropathy in type 2 diabetes mellitus / 中华医学杂志(英文版)

<p><b>BACKGROUND</b>Various methods can be applied to build predictive models for the clinical data with binary outcome variable. This research aims to explore the process of constructing common predictive models, Logistic regression (LR), decision tree (DT) and multilayer perceptron (MLP), as well as focus on specific details when applying the methods mentioned above: what preconditions should be satisfied, how to set parameters of the model, how to screen variables and build accuracy models quickly and efficiently, and how to assess the generalization ability (that is, prediction performance) reliably by Monte Carlo method in the case of small sample size.</p><p><b>METHODS</b>All the 274 patients (include 137 type 2 diabetes mellitus with diabetic peripheral neuropathy and 137 type 2 diabetes mellitus without diabetic peripheral neuropathy) from the Metabolic Disease Hospital in Tianjin participated in the study. There were 30 variables such as sex, age, glycosylated hemoglobin, etc. On account of small sample size, the classification and regression tree (CART) with the chi-squared automatic interaction detector tree (CHAID) were combined by means of the 100 times 5-7 fold stratified cross-validation to build DT. The MLP was constructed by Schwarz Bayes Criterion to choose the number of hidden layers and hidden layer units, alone with levenberg-marquardt (L-M) optimization algorithm, weight decay and preliminary training method. Subsequently, LR was applied by the best subset method with the Akaike Information Criterion (AIC) to make the best used of information and avoid overfitting. Eventually, a 10 to 100 times 3-10 fold stratified cross-validation method was used to compare the generalization ability of DT, MLP and LR in view of the areas under the receiver operating characteristic (ROC) curves (AUC).</p><p><b>RESULTS</b>The AUC of DT, MLP and LR were 0.8863, 0.8536 and 0.8802, respectively. As the larger the AUC of a specific prediction model is, the higher diagnostic ability presents, MLP performed optimally, and then followed by LR and DT in terms of 10-100 times 2-10 fold stratified cross-validation in our study. Neural network model is a preferred option for the data. However, the best subset of multiple LR would be a better choice in view of efficiency and accuracy.</p><p><b>CONCLUSION</b>When dealing with data from small size sample, multiple independent variables and a dichotomous outcome variable, more strategies and statistical techniques (such as AIC criteria, L-M optimization algorithm, the best subset, etc.) should be considered to build a forecast model and some available methods (such as cross-validation, AUC, etc.) could be used for evaluation.</p>

How to appropriately choose observed indexes / 中西医结合学报

Observed index is a very important element in a research design, because it is a specific reflection of the effects of research factors on the research subjects and is indispensable in any research. Generally, there are two types of observed indexes: the indexes that reflect natural attributes, habits or states of the research subjects and the indexes that reflect the effects of different drugs or treatments on research subjects. This article mainly introduces the definition, characteristics, selection and observation of research indexes and the major and minor indexes.

Mycoplasma pneumoniae pneumonia in hospitalized children diagnosed at acute stage by paired sera / 中华医学杂志(英文版)

<p><b>BACKGROUND</b>Mycoplasma pneumoniae (M. pneumoniae) is a frequent cause of respiratory tract infections. However, there is deficient knowledge about the clinical manifestations of M. pneumoniae infection. We described the clinical and laboratory findings of M. pneumoniae pneumonia in hospitalized children who were all diagnosed by a ≥ fourfold increase in antibody titer.</p><p><b>METHODS</b>M. pneumoniae antibodies were routinely detected in children admitted with acute respiratory infection during a one-year period. The medical history was re-collected from children whose M. pneumoniae antibody titer increased ≥ fourfold at the bedside by a single person, and their frozen paired serum samples were measured again for the M. pneumoniae antibody titer.</p><p><b>RESULTS</b>Of the 635 children whose sera were detected for the M. pneumoniae antibody, paired sera were obtained from 82 and 29.3% (24/82) showed a ≥ fourfold increase in antibody titer. There were 24 cases, nine boys and 15 girls, aged from two to 14 years, whose second serum samples were taken on day 9 at the earliest after symptom onset; the shortest interval was three days. All children presented with a high fever (≥ 38.5°C) and coughing. Twenty-one had no nasal obstruction or a runny nose, and five had mild headaches which all were associated with the high fever. The disease was comparatively severe if the peak temperature was > 39.5°C. All were diagnosed as having pneumonia through chest X-rays. Four had bilateral or multilobar involvement and their peak temperatures were all ≤ 39.5°C. None of the children had difficulty in breathing and all showed no signs of wheezing.</p><p><b>CONCLUSIONS</b>The second serum sample could be taken on day 9 at the earliest after symptom onset meant that paired sera could be used for the clinical diagnosis of M. pneumoniae pneumonia in children at the acute stage. M. pneumoniae is a lower respiratory tract pathogen. Extrapulmonary complications were rare and minor in our study. High peak temperature (> 39.5°C) is correlated with the severity of M. pneumoniae pneumonia in children.</p>

Typing of hemorrhagic fever with renal syndrome by cluster analysis / 中华流行病学杂志

<p><b>OBJECTIVE</b>To make an inquiry into method of typing of hemorrhagic fever with renal syndrome (HFRS).</p><p><b>METHOD</b>Average monthly rates were calculated on the basis of data from 1995 to 1999, then cluster analysis was carried out to type out endemic areas.</p><p><b>RESULTS</b>Compared with the results of 36 surveillance spots from 1980 to 1992, twenty-four surveillance spots had the same results (66.7%). Twenty-three surveillance spots had the same results with the original data in 1999 (82.1%).</p><p><b>CONCLUSION</b>HFRS incidences increased in spring or in summer, but decreased in autumn or in winter. Cluster analysis seemed to be a supplementary method in distinguishing the epidemic types for HFRS.</p>

Cytomegalovirus infection and disease in allogeneic hematopoietic stem cells transplantation / 军事医学科学院院刊

Objective: To investigate the incidence of CMV infection(CMV-I) and CMV related diseases (CMV-D) after allogeneic hematopoietic stem cells transplantation in 70 consecutive allogeneic hematopoietic stem cells transplantation(allo-HSCT) patients and to search for the optimal prophylactic strategy.Methods: Blood samples were monitored using the CMV pp65 antigenemia assay.Of the 70 patients observed,30 patients with chronic myeloid leukemia［CML:CP(27),AP(2),BC(1)］,12 with acute myeloblastic leukemia(AML),10 with acute lymphoblastic leukemia(ALL)and other cases were NHL(3), AA(5), MDS(7), SCLC with pancytopenia (1),CLL(1), and MF (1). Sixty six patients received HLA - identical siblings transplantation and four received tranplants from their HLA- haploidentical donors. Seventy cases included allo-PBPCT (64 cases) , allo-BMT (4 cases) and allo-PB+BMT (2). Before transplantation, all patients and donors received CMV serological examination except 4 pairs of donors/recepients. All 66 patients (3 cases were CMV IgM positive) and 64/66 donors were CMV IgG positive. Results：After transplantation, 64/70 patients developed CMV viremia during monitoring period. Forty three of 70 patients developed CMV-D.Thirty five of them suffered from CMV-associated interstitial pneumonia(CMV-IP). The high peak levels of CMV antigenemia were associated with development of CMV disease　. Close correlation was found between acute graft vs host disease(GVHD) and CMV disease. The patients were followed up for 2 to 24 months. The patients who received preemptive therapy(group A)had significantly better outcome than CMV disease group(group B, P=0.0001). Conclusions: The results suggest that CMV antigenemia has high predictive value for subsequent CMV disease and CMV pp65 antigenemia -guided early therapy has particular advantage for avoiding morbidity and mortality caused by CMV disease.