Predicting the incidence of infectious diarrhea with symptom surveillance data using a stacking-based ensembled model.

BACKGROUND: Infectious diarrhea remains a major public health problem worldwide. This study used stacking ensemble to developed a predictive model for the incidence of infectious diarrhea, aiming to achieve better prediction performance. METHODS: Based on the surveillance data of infectious diarrhea cases, relevant symptoms and meteorological factors of Guangzhou from 2016 to 2021, we developed four base prediction models using artificial neural networks (ANN), Long Short-Term Memory networks (LSTM), support vector regression (SVR) and extreme gradient boosting regression trees (XGBoost), which were then ensembled using stacking to obtain the final prediction model. All the models were evaluated with three metrics: mean absolute percentage error (MAPE), root mean square error (RMSE), and mean absolute error (MAE). RESULTS: Base models that incorporated symptom surveillance data and weekly number of infectious diarrhea cases were able to achieve lower RMSEs, MAEs, and MAPEs than models that added meteorological data and weekly number of infectious diarrhea cases. The LSTM had the best prediction performance among the four base models, and its RMSE, MAE, and MAPE were: 84.85, 57.50 and 15.92%, respectively. The stacking ensembled model outperformed the four base models, whose RMSE, MAE, and MAPE were 75.82, 55.93, and 15.70%, respectively. CONCLUSIONS: The incorporation of symptom surveillance data could improve the predictive accuracy of infectious diarrhea prediction models, and symptom surveillance data was more effective than meteorological data in enhancing model performance. Using stacking to combine multiple prediction models were able to alleviate the difficulty in selecting the optimal model, and could obtain a model with better performance than base models.

The risk of injuries during work and its association with precipitation: New insight from a sentinel-based surveillance and a case-crossover design.

Background: Injuries during work are often exogenous and can be easily influenced by environmental factors, especially weather conditions. Precipitation, a crucial weather factor, has been linked to unintentional injuries, yet evidence of its effect on work-related injuries is limited. Therefore, we aimed to clarify the impact of precipitation on injuries during work as well as its variation across numerous vulnerability features. Methods: Records on the work-related injury during 2016-2020 were obtained from four sentinel hospitals in Guangzhou, China, and were matched with the daily weather data during the same period. We applied a time-stratified case-crossover design followed by a conditional logistic regression to evaluate the association between precipitation and work-related injuries. Covariates included wind speed, sunlight, temperature, SO 2, NO 2, and PM 2.5. Results were also stratified by multiple factors to identify the most vulnerable subgroups. Results: Daily precipitation was a positive predictor of work-related injuries, with each 10 mm increase in precipitation being associated with an increase of 1.57% in the rate of injuries on the same day and 1.47-1.14% increase of injuries on subsequent 3 days. The results revealed that precipitation had a higher effect on work-related injuries in winter (4.92%; 95%CI: 1.77-8.17%). The elderly (2.07%; 95%CI: 0.64-3.51%), male (1.81%; 95%CI: 0.96-2.66%) workers or those with lower educational levels (2.58%; 95%CI: 1.59-3.54%) were more likely to suffer from injuries on rainy days. There was a higher risk for work-related injuries caused by falls (2.63%; 95%CI: 0.78-4.52%) or the use of glass products (1.75%; 95%CI: 0.49-3.02%) on rainy days. Conclusions: Precipitation was a prominent risk factor for work-related injury, and its adverse effect might endure for 3 days. Certain sub-groups of workers were more vulnerable to injuries in the rain.

Is it necessary for young patients with recurrent implantation failure to undergo preimplantation genetic testing for aneuploidy?

Objective: To determine whether preimplantation genetic testing for aneuploidy (PGT-A) can improve the pregnancy outcomes of patients aged under 38 years who have a history of recurrent implantation failure(RIF). Design: Retrospective cohort study. Methods: We retrospectively studied the pregnancy outcomes of RIF patients aged under 38 years from January 2017 to December 2021.178 patients were divided into two groups according to whether they underwent PGT-A: the PGT-A group(n=59)and the control group(n=119).In the PGT-A group, we compared the euploidy rate of the different quality and developmental rate blastocysts. In both groups,the patients were the first frozen-thaw single blastocysts transfer after the diagnosis of RIF. Among the pregnancy outcomes, the clinical pregnancy rate was assessed as the primary outcome. The spontaneous abortion rate and ongoing pregnancy rate were the secondry outcomes. The generalized estimation equation was used to adjust for the blastocysts derived from the same patients. Multivariate logistic analysis models were used to compare the pregnancy outcomes between the two groups. Results: In the PGT-A group, 293 blastocysts obtained from59 patients underwent PGT-A. The proportions of euploidy, aneuploidy and mosaic blastocysts were 56.31%, 25.60% and 18.09%, respectively. A comparison of the euploidy rates of different quality blastocysts showed that the rate of good-quality blastocysts was significantly higher than that of poor-quality blastocysts (67.66% vs 46.88%; odds ratio [OR], 2.203; 95%confidence interval[CI], 0.943-3.612; P=0.002). However, no significant difference was observed in the different developmental rates blastocysts. Compared with Day 5 blastocysts, the euploidy rates of Day 6 and Day 7 blastocysts were not significantly different(61.54%vs51.91%; OR,0.945; 95%CI, 0.445-2.010; P=0.884; and 61.54%vs47.37%; OR, 1.106; 95%CI, 0.774-1.578; P=0.581, respectively).As for the pregnancy outcomes, the clinical pregnancy rate was significantly increase after the use of PGT-A compared with the control group(71.19%vs56.30%; OR, 0.538; 95%CI, 0.262-1.104; P=0.039). However, the spontaneous abortion rates and ongoing pregnancy rates were not significantly different between the control and PGT-A groups (21.43% vs 19.40%; aOR,0.727; 95%CI,0.271-1.945; P=0.525; and55.93% vs 45.38%; aOR, 0.649; 95%CI, 0.329-1.283; P = 0.214,respectively). Conclusion: PGT-A improved the clinical pregnancy rate after blastocyst transfer in RIF patients aged under 38 years.

The association between wind speed and the risk of injuries among preschool children: New insight from a sentinel-surveillance-based study.

BACKGROUND: Injuries among preschool children are an important public health concern worldwide. Significant gaps remain in understanding the potential impact of wind speed on injuries among preschoolers. We aimed to clarify the association and its variation across subgroups to capture the vulnerability features. METHODS: Using a case-crossover design and conditional logistic regression model, we compared the exposure to wind speed right before the injury events (case period) with that of control periods to determine the excess rate (ER) of injury on each of 0-3 lag days in Guangzhou, 2016-2020. Results were also stratified by sociodemographic characteristics of patients, basic characteristics of injury events, and clinical features of injuries to identify the most vulnerable subgroups of preschoolers. RESULTS: Higher wind speed was significantly associated with an increased risk of injuries among preschoolers on lag 0, reaching an ER of 2.93 % (95 % confidence interval [CI] = 0.87, 5.03), but not on other lag days. The results of the stratified analyses showed that children under 3-year-old (3.41 %; 95 % CI = 0.36, 6.55), boys (3.66 %; 95 % CI = 1.04, 6.35), and non-locally registered children (3.65; 95 % CI = 0.02, 7.40) were more prone to wind-related injuries. Falls (2.67 %; 95 % CI = 0.11, 5.30) were the main cause of wind-related injuries, and taking transportation was the main activity when injuries occurred (13.16 %; 95 % CI = 4.45, 22.60). Additionally, injuries involving buildings/grounds/obstacles (4.69 %; 95 % CI = 1.66, 7.81) and the occurrence of sprain/strain (7.60 %; 95 % CI = 0.64, 15.04) showed a positive association with wind speed. CONCLUSIONS: Higher wind speed was associated with a significantly elevated rate of injuries among preschoolers without delayed effects, where children under 3-year-old, boys, and non-locally registered subgroups were more susceptible to wind-related injuries. This study may provide new insights for refining the prevention measures against wind-related injuries among preschoolers.