Comparison of random forests and SARIMA in Predicting Brucellosis Incidence / 公共卫生与预防医学

Objective To compare the effects of random forest and SARIMA (Seasonal Autoregressive Integrated Moving Average) on predicting incidence rate of brucellosis. Methods Using Brucellosis cases reported in the China Disease Prevention and Control Information System from 2005 to 2017, two models, random forest and SARIMA, were established for training and forecasting, and the forecasting results of the two models were compared. Results The R2 (R Squared) and RMSE (Root Mean Squared Error) of SARIMA model and random forest model are 0.904, 0.034351, 0.927 and 0.03345 respectively. Conclusion Both models have high prediction accuracy and can predict the incidence of brucellosis. Random forest prediction is a little bit better than SARIMA model and has more practical value.

Factors influencing the incidence of bacterial dysentery in parts of southwest China, using data from the geodetector / 中华流行病学杂志

Objective@#To analyze the relationship between the incidence of bacterial dysentery and related meteorological, environmental and socio-economic factors in parts of southwest China, under the geodetector.@*Methods@#Incidence data on bacillary dysentery from 2005 to 2014 was collected from the China infectious disease reporting system and meteorological, terrain environment and socio-economic related data were used for statistical analysis with geodetector also used, in Sichuan province, Yunnan province and the Tibet autonomous region (southwest China).@*Results@#Through geodetector model, results showed that the elevation maximum value of power of determinant (PD) (PD=0.308), ethnicity (PD=0.260), followed by summer mean temperature, topography, elevation standard deviation, slope, population density were noticed. The maximum value of PD of summer average relative humidity, summer average precipitation, and regional GDP were all less than 0.200. In the ecological detector model, value of PD appeared significantly different from multiple factors such as summer average precipitation, regional GDP, average altitude, elevation standard deviation, slope, terrain and ethnicity (P<0.05). In the interactive detector model, after the spatial interaction, factors as: value on PD of summer average relative humidity, population density and regional GDP were larger than the value on the sum of PD of two factors with nonlinear superposition enhancing trend. Results from the risk zone detection showed that areas of moderate average summer temperature, low average precipitation, moderate average summer relative humidity, high average altitude, large elevation standard deviation, high slope, mountain topography, Tibeto-Burmese of ethnicity, low population density, low regional GDP were all related to the high incidence rates of bacterial dysentery.@*Conclusions@#Meteorological conditions and topography environments were related to the bacterial dysentery in southwest China, socio-economic behaviors depends on the level of economic development and sanitary conditions etc. were all directing influencing the transmission and incidence of bacterial dysentery. For developing a better intervention programs on bacillary dysentery, socio-economic means should be used for the control of bacillary dysentery, in the region.

Postpartum pelvic floor rehabilitation on prevention of female pelvic floor dysfunction:a multicenter prospective randomized controlled study / 中华妇产科杂志

Objective To study the postpartum pelvic floor rehabilitation on the improvement of pelvic floor electrical physiological indexes and the prevention of female pelvic floor dysfunction in China. Methods A multicenter prospective randomized controlled study was carried out. From October 2011, postpartum women in five provinces were randomly assigned into treatment group and control group. The women in treatment group received electrical stimulation and biofeedback treatment. The women in control group performed pelvic floor muscle exercise at home. When 6 months and 12 months after delivery, comparing two groups of patients with pelvic floor electrical physiological indexes and pelvic organ prolapse quantitation measurements (POP-Q), to evaluate the effect of postpartum pelvic floor rehabilitation on the prevention of pelvic floor dysfunction. Pelvic floor impact questionnaire short form (PFIQ-7) and pelvic organ prolapse/incontinence sexual questionnaire-12 (PISQ-12) were used to evaluate the influence on quality of life and sexual life. Results Until June 2013, 324 women were participated, 124 in control group, 200 in treatment group. According to the baseline results, there was statistical significance in the results of pelvic floor electrical physiological indexes between the treatment and control groups in postpartum 6 months and 12 months; the proportion above level Ⅲ of type Ⅰ and type Ⅱ muscle fibers strength in the treatment group, it was from 41.5% (83/200) and 40.5% (81/200) to 76.3% (145/190) and 79.5% (151/190) in postpartum 6 weeks and postpartum 6 months, increased to 80.6%(58/72) and 80.6%(58/72) in postpartum 12 months, improved significantly comparing with the control group (P0.05). There was no significant difference in the questionnaires in quality of life and quality of sexual life (P>0.05). Conclusion Neuromuscular electrical stimulation and biofeedback therapy in the early postpartum period could obviously improve pelvic floor electrical physiological indexes, and is beneficial to prevent the pelvic floor dysfunction.