Few studies included objective blood pressure (BP) to construct the predictive model of severe obstructive sleep apnea (OSA). This study used binary logistic regression model (BLRM) and the decision tree method (DTM) to constructed the predictive models for identifying severe OSA, and to compare the prediction capability between the two methods. Totally 499 adult patients with severe OSA and 1421 non-severe OSA controls examined at the Sleep Medicine Center of a tertiary hospital in southern Taiwan between October 2016 and April 2019 were enrolled. OSA was diagnosed through polysomnography. Data on BP, demographic characteristics, anthropometric measurements, comorbidity histories, and sleep questionnaires were collected. BLRM and DTM were separately applied to identify predictors of severe OSA. The performance of risk scores was assessed by area under the receiver operating characteristic curves (AUCs). In BLRM, body mass index (BMI) ≥27 kg/m2, and Snore Outcomes Survey score ≤55 were significant predictors of severe OSA (AUC 0.623). In DTM, mean SpO2 <96%, average systolic BP ≥135 mmHg, and BMI ≥39 kg/m2 were observed to effectively differentiate cases of severe OSA (AUC 0.718). The AUC for the predictive models produced by the DTM was higher in older adults than in younger adults (0.807 vs. 0.723) mainly due to differences in clinical predictive features. In conclusion, DTM, using a different set of predictors, seems more effective in identifying severe OSA than BLRM. Differences in predictors ascertained demonstrated the necessity for separately constructing predictive models for younger and older adults.

OBJECTIVE: Prediction of fetal growth restriction (FGR) and small of gestational age (SGA) infants by using various ultrasound cardiac parameters in a logistic regression model.
METHODS: In this retrospective study we obtained standardized ultrasound images of 357 fetuses between the 20th and 39th week of gestation, 99 of these fetuses were between the 3rd and 10th growth percentile, 61 smaller than 3rd percentile and 197- appropriate for gestational age over the 10th percentile (control group). Several cardiac parameters were studied. The cardiothoracic ratio and sphericity of the ventricles was calculated. A binary logistic regression model was developed for prediction of growth restriction using the cardiac and biometric parameters.
RESULTS: There were noticeable differences between the control and study group in the sphericity of the right ventricle (p = 0.000), left and right longitudinal ventricle length (pright = 0.000, pleft = 0.000), left ventricle transverse length (p = 0.000), heart diameter (p = 0.002), heart circumference (p = 0.000), heart area (p = 0.000), and thoracic diameter limited by the ribs (p = 0.002). There was no difference of the cardiothoracic ratio between groups. The logistic regression model achieved a prediction rate of 79.4 % with a sensitivity of 74.5 % and specificity of 83.2 %.
CONCLUSIONS: The heart of growth restricted infants is characterized by a more globular right ventricle, shorter ventricle length and smaller thorax diameter. These parameters could improve prediction of FGR and SGA.

This study investigates the possibility of utilizing drip as a non-destructive method for assessing the freshness and spoilage of chicken meat. The quality parameters [pH, volatile base nitrogen (VBN), and total aerobic bacterial counts (TAB)] of chicken meat were evaluated over a 13-day storage period in vacuum packaging at 4 °C. Simultaneously, the metabolites in the chicken meat and its drip were measured by nuclear magnetic resonance. Correlation (Pearson\'s and Spearman\'s rank) and pathway analyses were conducted to select the metabolites for model training. Binary logistic regression (model 1 and model 2) and multiple linear regression models (model 3-1 and model 3-2) were trained using selected metabolites, and their performance was evaluated using receiver operating characteristic (ROC) curves. As a result, the chicken meat was spoiled after 7 days of storage, exceeding 20 mg/100 g VBN and 5.7 log CFU/g TAB. The correlation analysis identified one organic acid, eight free amino acids, and five nucleic acids as highly correlated with chicken meat and its drip during storage. Pathway analysis revealed tyrosine and purine metabolism as metabolic pathways highly correlated with spoilage. Based on these findings, specific metabolites were selected for model training: ATP, glutamine, hypoxanthine, IMP, tyrosine, and tyramine. To predict the freshness and spoilage of chicken meat, model 1, trained using tyramine, ATP, tyrosine, and IMP from chicken meat, achieved a 99.9 % accuracy and had an ROC value of 0.884 when validated using drip metabolites. This model 1 was improved by training with tyramine and IMP from both chicken meat and its drip (model 2), which increased the ROC value for drip metabolites from 0.884 to 0.997. Finally, selected two metabolites (tyramine and IMP) can predict TAB and VBN quantitatively through models 3-1 and 3-2, respectively. Therefore, the model developed using metabolic changes in drip demonstrated the capability to non-destructively predict the freshness and spoilage of chicken meat at 4 °C. To make generic predictions, it is necessary to expand the model\'s applicability to various conditions, such as different temperatures, and validate its performance across multiple chicken batches.

Currently, risk stratification is the most difficult problem in prostate cancer (PCa) management. Gleason grading cannot adequately predict cancer progression. This study aimed to identify chromosome-specific segment size alterations that could aid risk stratification and predict metastasis using a retrospective cohort-study strategy. A binary logistic regression model was generated using 16 chromosome-specific segments with size alterations (deletions and amplifications) that showed associations with disease stage (primary versus metastatic). The regression model was trained with the MSKCC PIK3R1 PCa cohort (n = 1417), and validated with the TCGA Firehose Legacy (n = 500), MSKCC Prostate Oncogenome Project (n = 218), and the SU2C/PCF Dream Team (n = 150) PCa cohorts. Furthermore, the capacity of the model to predict metastasis between primary tumours with metastasis (n = 54) and primary tumours without metastasis (n = 54) was tested. The accuracy, sensitivity, and specificity of the model at disease stage stratification ranged from 69.02% to 88.55%, 72.8% to 86.00% and 66.30% to 89.50%, respectively. The model also showed good performance at metastasis prediction with accuracy, sensitivity, and specificity of 57.41%, 62.96% and 51.85%, respectively. The study conclusion was that chromosome-specific segment size alterations can aid risk stratification and metastasis prediction. The significance of the study findings is that in combinations with clinical, biochemical, and histopathological variables, chromosome-specific alterations could improve current risk stratification and prediction models for PCa.

With the constrained topography and road geometry, adverse weather conditions and restricted roadway facilities, mountainous highway crash rates and fatality rates are much higher. Considering the potential influence of driver\'s route familiarity level on driving behavior and fault assignment, this research investigates high- and low- route familiarity level drivers (HRF and LRF drivers) fault assignment in mountainous highway fatal crashes in Yunnan Province of China by examining factors of driver, crash/environment and pre-crash behaviors.
Yunnan Province is famous for its tourism, and tourism can also bring in many drivers with low-route familiarity levels. Spatial distance away from residence-based method is used for identifying route familiarity levels of the drivers in this study. We employed two separate binary logistic regression models to investigate the effects of the explanatory variables on the likelihoods that the HRF or LRF drivers were at fault in the mountainous highway fatal crashes.
The results show that driver under alcohol influence, sharp turn, dawn/dusk and left turning are 4 common factors that significantly influence both HRF and LRF drivers\' fault assignments. Factors including driver age, driver seatbelt use, weather condition, road type, section type, lighting condition and pre-crash behaviors have different or opposite influences on HRF and LRF drivers\' fault assignments. HRF drivers are much easier to be distracted under the conditions that are without the need of extra attention. LRF drivers have much more difficulties in figuring out, responding to and making timely driving behavior adjustment to ensure their driving safety on the high-risk sections like tunnels, continuous long downhills and sharp turns. Street parking/backing and left turning of the LRF drivers are very serious problems on the mountainous highways in China.
There is a large difference of significant factors contributing to the fault assignment of HRF and LRF drivers in mountainous highway fatal crashes. Some more effective and targeted countermeasures are put forward for HRF and LRF drivers and transportation managers to improve mountainous highway traffic safety.

OBJECTIVE: Evidence regarding the relationship between serum uric acid-to-creatinine (SUA/Scr) ratio and non-alcoholic fatty liver disease (NAFLD) in Chinese non-obese people is still limited. Therefore, the present study was undertaken to analyze the association between the SUA/Scr ratio and NAFLD.
METHODS: This study was a cross-sectional study that non-selectively and consecutively collected 182,320 non-obese individuals with a normal range of low-density lipoprotein cholesterol in a Chinese hospital from January 2010 to December 2014. A binary logistic regression model was used to evaluate the independent association between the SUA/Scr ratio and NAFLD. A generalized additive model (GAM) and smooth curve fitting (penalized spline method) was conducted to explore the exact shape of the curve between them. A series of sensitivity analyses were used to ensure the robustness of the results. Moreover, subgroup analyses were conducted. In addition, the diagnostic value of the SUA/Scr ratio for NAFLD was evaluated based on the area under the receiver-operating characteristic curve (AUROC). It was stated that the data had been uploaded to the DATADRYAD website.
RESULTS: The average participants\' age was 40.96 ± 14.05 years old, and 90,305 (49.5%) were male. The prevalence of NAFLD was 13.7%, and the mean SUA/Scr was 3.62 ± 0.92. After adjusting covariates, the results showed that SUA/Scr ratio was positively associated with NAFLD (OR = 1.424, 95%CI: 1.396, 1.454). There was also a non-linear relationship between SUA/Scr ratio and NAFLD in participants with normal kidney function, and the inflection point of the SUA/Scr ratio was 4.425. The effect sizes (OR) on the left and right sides of the inflection point were 1.551 (1.504, 1.599) and 1.234 (1.179, 1.291), respectively. And the sensitive analysis demonstrated the robustness of the results. Subgroup analysis showed a stronger association between SUA/Scr ratio and NAFLD in females and the population with age < 50 years, FPG ≤ 6.1 mmol/L, BMI < 24 kg/m2, and HDL-c ≥ 1 mmol/L. In contrast, the weaker association was probed in males and the population with age ≥ 50 years, BMI ≥ 24 kg/m2, FPG > 6.1 mmol/L, and HDL-c < 1 mmol/L. The SUA/Scr ratio had an AUC of 0.6624 (95% CI 0.6589, 0.6660) for diagnosing NAFLD. Based on the best cut-off value of 3.776, the negative predictive value of the SUA/Scr ratio for identifying NAFLD was 91.0%.
CONCLUSIONS: This study demonstrates an independent positive association between SUA/Scr ratio and NAFLD in Chinese non-obese people with a normal range of low-density lipoprotein cholesterol. There is also a non-linear relationship between the SUA/Scr ratio and NAFLD in participants with normal kidney function, and the SUA/Scr ratio is strongly related to NAFLD when SUA/Scr ratio is less than 4.425. The SUA/Scr ratio has a certain reference value for determining NAFLD. When the SUA/Scr ratio is lower than 3.776, identifying NAFLD patients with low risk is a great reference.

Globally, pedestrians are one of the most vulnerable road-user groups. Their vulnerability increases while crossing the road at controlled intersections during the \"don\'t walk\" phase. Previous literature shows that driver yielding behavior has an association with pedestrian safety at intersections. Though several studies have explored driver yielding behavior towards pedestrians at conflict points, evidence on how pedestrian actions influence driver yielding behavior at intersections is yet to be investigated.
To pursue this end, a binary logistic regression model was developed using the collected data to explore the effect of non-compliant pedestrian characteristics and their road crossing behavior on driver yielding behavior towards pedestrians at six controlled intersections of Dhaka, Bangladesh. The data were collected through videography survey.
Results showed that drivers were more likely to yield to pedestrians who were female, crossing in a group, carrying baggage, not using a mobile, making some hand gesture to the driver, or crossing by rolling gap strategy.
These findings add new insights for transportation planners into the complex interaction between vehicles and pedestrians at busy controlled intersections, and thus would help to make a pedestrian friendly street.

to analyze farmers\' willingness to gather homestead and its influencing factors, so as to provide decision-making basis for the rational layout of rural homestead.
questionnaire, logistic model. The results are as follows. (1) Farmers\' willingness to gather homesteads is highest in dam areas, followed by hilly areas, and is lowest in mountainous areas. (2) The respondents\' age, family support ratio, housing structure, whether the access road is paved, and the distance from the main road have significant negative impacts on farmers\' willingness to gather homesteads, while homesteads being idle, the service life of the house, the type of daily energy use, and whether they are far from relatives after relocation have significant positive impacts on farmers\' willingness to gather homesteads. (3) The main influencing factors of farmers\' homestead agglomeration in dam areas are the idle situation of a homestead, housing structure, the service life of the house, and whether they are satisfied with their current homestead residence. (4) The main influencing factors of farmers\' homestead agglomeration in hilly areas are the age of the respondents, the proportion of family workers, and whether they accept the relocation and are far from relatives. (5) The main influencing factors of farmers\' homestead agglomeration in mountainous areas are the age of the respondents, the ratio of family support, the housing structure, and whether the access road is paved. We conclude that there are significant differences in farmers\' willingness to gather homesteads and the influencing factors in different geomorphic areas. Policy makers should formulate differentiated homestead agglomeration optimization schemes and design the optimization paths of homestead agglomeration on the basis of geomorphic classification and subregion.

UNASSIGNED: Tuberculosis is one of the leading infectious diseases for people living with HIV. Therefore, the purpose of this study was to investigate factors affecting the development of TB among HIV-positive adults under treatment in government hospitals of Amhara Region, Ethiopia.
UNASSIGNED: A hospital-based retrospective study design was conducted among 700 HIV-positive adults under HAART in 17 government hospitals in the Amhara region, Ethiopia.
UNASSIGNED: Age of the patients (AOR = 1.122, 95% CI:1.013, 2.234), baseline CD4 cell count (AOR = 0.888, 95% CI: 0.714, 0.945), patients living without their partner (AOR = 1.212, 95% CI: 1.051, 1.123), females under treatment (AOR = 0.786, 95% CI; 0.564, 0.845), non-opportunistic diseases (AOR = 0.865, 95% CI: 0.731, 0.938), patients not disclosed their HIV status (AOR = 1.241, 95% CI: 1.087, 2.341), rural patients (AOR = 1.135, 95% CI: 1.032, 1.453, patient with no education (AOR = 1.125, 95% CI: 1.056, 1.546), low adherence patients (AOR = 1.225, 95% CI: 1.191, 2.453), bedridden patients (AOR = 1.223, 95% CI: 1.131, 1.521), ambulatory patients (AOR = 1.156, 95% CI:1.091, 1.267), non-smoker patients (AOR = 0.854, 95% CI: 0.686, 0.935) significantly affected on the variable of interest. Similarly, alcohol intake, drug toxicity and baseline clinical WHO stages significantly affected for the development of tuberculosis in HIV-positive patients under treatment.
UNASSIGNED: In this study, baseline CD4 cell count, female patients, non-opportunistic diseases, and non-smoking status were negatively associated with the development of TB, whereas age of patients, living without partners, patients with no education, patients with low adherence, bedridden and ambulatory patients were positively associated to the development of TB in HIV patients. The findings obtained in this study are important for both service providers and patients. More attention should be given to those positively associated variables to response variables. The regional health bureau should open TB/HIV co-infection subsections like ART sections in each hospital.

BACKGROUND: Antenatal care (ANC) is a health care intervention intended to ensure the safety of pregnancy. According to the World Health Organization, at least four ANC visits are recommended for a healthy pregnancy. However, whether this recommended number of visits was followed or not in the rural areas of Southwestern Ethiopia is not known. Therefore, the study aimed to investigate the prevalence of, and the associated factors of ANC utilization by pregnant women in the rural areas of Southwestern Ethiopia.
METHODS: A community-based cross-sectional study design was used in three rural zones. The data were collected from n = 978 women through a structured questionnaire with face-to-face interview. The collected data were analyzed using descriptive statistics and a multiple binary logistic regression model.
RESULTS: The results showed that 56% of women made the recommended minimum number of ANC visits and the remaining 44% of them underutilized the ANC service. The multiple binary logistic regression model identified zone, marital status of the woman, educational level of the husband, occupation of the husband, knowledge of danger signs of pregnancy, birth interval, source of information, timely visits, and transportation problem to be statistically significant factors affecting the prevalence of ANC visit utilization of women. Bench Maji zone had smaller odds ratio of ANC visit prevalence as compared to Kaffa zone. Women who lived in the rural area of Sheko zone are 2.67 times less likely to utilize ANC visit than those who lived in the rural area of Kaffa zone keeping other variables constant.
CONCLUSIONS: The study results highlight the need to increase the number of ANC visits, and the importance of using an appropriate model to determine the important socio-demographic factors that ANC service providers shall focus on to improve the health of the unborn baby and the mother during pregnancy.