Between-group Correlations Research Articles

A variable clustering approach for overdispersed high-dimensional count data using a copula-based mixture model

In this paper, we propose a latent variable model for the analysis and clustering of high-dimensional correlated and overdispersed count data. We use a set of random effects to capture within-group correlations and a copula to address between-group correlations. Since maximum likelihood (ML) estimation is computationally intensive for high-dimensional data, we propose a fast moment-based estimation procedure. Additionally, we implement the proposed estimation procedure into a clustering algorithm, borrowing ideas from the K-means algorithm. Based on a simulation study, we demonstrate that the estimation procedure shows good stability and precision. Furthermore, it is computationally fast, unlike the ML approach which shows serious convergence issues. In addition, the clustering algorithm with our moment-based estimation procedure can identify the simulated grouped data structure. The approach is further illustrated with a high-dimensional RNA-Seq dataset of commercial potatoes.

A Multi-Age-Group Interrupted Time-Series Study for Evaluating the Effectiveness of National Expanded Program on Immunization on Mumps.

The national Expanded Program on Immunization (EPI) in China has covered vaccines for measles, mumps, and rubella, among children aged 18–24 months since September 2008. However, no previous studies have quantified the effectiveness of the EPI on mumps incidence. There are methodological challenges in assessing the effect of an intervention that targets a subpopulation but finally influences the whole population. In this study, monthly data on mumps incidence were collected in Guangzhou, China, during 2005–2019. We proposed a multi-age-group interrupted time-series design, setting the starting time of exerting effect separately for 14 different age groups. A mixed-effects quasi-Poisson regression was applied to analyze the effectiveness of the EPI on mumps incidence, after controlling for long-term and seasonal trends, and meteorological factors. The model also accounted for the first-order autocorrelation within each age group. Between-age-group correlations were expressed using the contact matrix of age groups. We found that 70,682 mumps cases were reported during 2005–2019, with an annual incidence rate of 37.91 cases per 100,000 population. The effect of EPI strengthened over time, resulting in a decrease in the incidence of mumps by 16.6% (EPI-associated excess risk% = −16.6%, 95% CI: −27.0% to −4.7%) in September 2009 to 40.1% (EPI-associated excess risk% = −40.1%, 95% CI: −46.1% to −33.3%) in September 2019. A reverse U-shape pattern was found in age-specific effect estimates, with the largest reduction of 129 cases per 100,000 population (95% CI: 14 to 1173) in those aged 4–5 years. The EPI is effective in reducing the mumps incidence in Guangzhou. The proposed modeling strategy can be applied for simultaneous assessment of the effectiveness of public health interventions across different age groups, with adequate adjustment for within- and between-group correlations.

Group least squares regression for linear models with strongly correlated predictor variables

Traditionally, the least squares regression is mainly concerned with studying the effects of individual predictor variables, but strongly correlated variables generate multicollinearity which makes it difficult to study their effects. Existing methods for handling multicollinearity such as ridge regression are complicated. To resolve the multicollinearity issue without abandoning the simple least squares regression, for situations where predictor variables are in groups with strong within-group correlations but weak between-group correlations, we propose to study the effects of the groups with a group approach to the least squares regression. Using an all positive correlations arrangement of the strongly correlated variables, we first characterize group effects that are meaningful and can be accurately estimated. We then present the group approach with numerical examples and demonstrate its advantages over existing methods for handling multicollinearity. We also address a common misconception about prediction accuracy of the least squares estimated model and discuss through an example similar group effects in generalized linear models.

Structural and Functional Characterization of Gray Matter Alterations in Female Patients With Neuropsychiatric Systemic Lupus.

ObjectiveTo investigate morphological and functional alterations within gray matter (GM) in female patients with neuropsychiatric systemic lupus (NPSLE) and to explore their clinical significance.Methods54 female patients with SLE (30 NPSLE and 24 non-NPSLE) and 32 matched healthy controls were recruited. All subjects received a quantitative MRI scan (FLAIR, 3DT1, resting-state functional MRI). GM volume (GMV), fractional amplitude of low-frequency fluctuation (fALFF), regional homogeneity (ReHo), and degree of centrality (DC) were obtained. Between-group comparison, clinical correlation, and discrimination of NPSLE from non-NPSLE were achieved by voxel-based analysis, cerebellar seed-based functional connectivity analysis, regression analysis, and support vector machine (SVM), respectively.ResultsPatients with NPSLE showed overt subcortical GM atrophy without significantly abnormal brain functions in the same region compared with controls. The dysfunction within the left superior temporal gyri (L-STG) was found precede the GM volumetric loss. The function of the nodes in default mode network (DMN) and salience network (SN) were weakened in NPSLE patients compared to controls. The function of the cerebellar posterior lobes was significantly activated in non-NPSLE patients but attenuated along with GM atrophy and presented higher connectivity with L-STG and DMN in NPSLE patients, while the variation of the functional activities in the sensorimotor network (SMN) was the opposite. These structural and functional alterations were mainly correlated with disease burden and anti-phospholipid antibodies (aPLs) (r ranges from -1.53 to 1.29). The ReHos in the bilateral cerebellar posterior lobes showed high discriminative power in identifying patients with NPSLE with accuracy of 87%.ConclusionPatients with NPSLE exhibit both structural and functional alterations in the GM of the brain, which especially involved the deep GM, the cognitive, and sensorimotor regions, reflecting a reorganization to compensate for the disease damage to the brain which was attenuated along with pathologic burden and cerebral vascular risk factors. The GM within the left temporal lobe may be one of the direct targets of lupus-related inflammatory attack. The function of the cerebellar posterior lobes might play an essential role in compensating for cortical functional disturbances and may contribute to identifying patients with suspected NPSLE in clinical practice.

The Interplay of Demographic Variables and Social Distancing Scores in Deep Prediction of U.S. COVID-19 Cases

With the severity of the COVID-19 outbreak, we characterize the nature of the growth trajectories of counties in the United States using a novel combination of spectral clustering and the correlation matrix. As the United States and the rest of the world are still suffering from the effects of the virus, the importance of assigning growth membership to counties and understanding the determinants of the growth is increasingly evident. For the two communities (faster versus slower growth trajectories) we cluster the counties into, the average between-group correlation is 88.4% whereas the average within-group correlations are 95.0% and 93.8%. The average growth rate for one group is 0.1589 and 0.1704 for the other, further suggesting that our methodology captures meaningful differences between the nature of the growth across various counties. Subsequently, we select the demographic features that are most statistically significant in distinguishing the communities: number of grocery stores, number of bars, Asian population, White population, median household income, number of people with the bachelor’s degrees, and population density. Lastly, we effectively predict the future growth of a given county with a long short-term memory (LSTM) recurrent neural network using three social distancing scores. The best-performing model achieves a median out-of-sample R 2 of 0.6251 for a four-day ahead prediction and we find that the number of communities and social distancing features play an important role in producing a more accurate forecasting. This comprehensive study captures the nature of the counties’ growth in cases at a very micro-level using growth communities, demographic factors, and social distancing performance to help government agencies utilize known information to make appropriate decisions regarding which potential counties to target resources and funding to. Supplementary materials for this article, including a standardized description of the materials available for reproducing the work, are available as an online supplement.

Deep-gKnock: Nonlinear group-feature selection with deep neural networks

Feature selection is central to contemporary high-dimensional data analysis. Group structure among features arises naturally in various scientific problems. Many methods have been proposed to incorporate the group structure information into feature selection. However, these methods are normally restricted to a linear regression setting. To relax the linear constraint, we design a new Deep Neural Network (DNN) architecture and integrating it with the recently proposed knockoff technique to perform nonlinear group-feature selection with controlled group-wise False Discovery Rate (gFDR). Experimental results on high-dimensional synthetic data demonstrate that our method achieves the highest power and accurate gFDR control compared with state-of-the-art methods. The performance of Deep-gKnock is especially superior in the following five situations: (1) nonlinearity relationship; (2) dimension p greater than sample size n; (3) high between-group correlation; (4) high within-group correlation; (5) large number of associated groups. And Deep-gKnock is also demonstrated to be robust to the misspecification of the feature distribution and the change of network architecture. Moreover, Deep-gKnock achieves scientifically meaningful group-feature selection results for cutting-edge real world datasets.

Fully Automated Segmentation Algorithm for Perihematomal Edema Volumetry After Spontaneous Intracerebral Hemorrhage.

Background and Purpose- Perihematomal edema (PHE) is a promising surrogate marker of secondary brain injury in patients with spontaneous intracerebral hemorrhage, but it can be challenging to accurately and rapidly quantify. The aims of this study are to derive and internally validate a fully automated segmentation algorithm for volumetric analysis of PHE. Methods- Inpatient computed tomography scans of 400 consecutive adults with spontaneous, supratentorial intracerebral hemorrhage enrolled in the Intracerebral Hemorrhage Outcomes Project (2009-2018) were separated into training (n=360) and test (n=40) datasets. A fully automated segmentation algorithm was derived from manual segmentations in the training dataset using convolutional neural networks, and its performance was compared with that of manual and semiautomated segmentation methods in the test dataset. Results- The mean volumetric dice similarity coefficients for the fully automated segmentation algorithm were 0.838±0.294 and 0.843±0.293 with manual and semiautomated segmentation methods as reference standards, respectively. PHE volumes derived from the fully automated versus manual (r=0.959; P<0.0001), fully automated versus semiautomated (r=0.960; P<0.0001), and semiautomated versus manual (r=0.961; P<0.0001) segmentation methods had strong between-group correlations. The fully automated segmentation algorithm (mean 18.0±1.8 seconds/scan) quantified PHE volumes at a significantly faster rate than both of the manual (mean 316.4±168.8 seconds/scan; P<0.0001) and semiautomated (mean 480.5±295.3 seconds/scan; P<0.0001) segmentation methods. Conclusions- The fully automated segmentation algorithm accurately quantified PHE volumes from computed tomography scans of supratentorial intracerebral hemorrhage patients with high fidelity and greater efficiency compared with manual and semiautomated segmentation methods. External validation of fully automated segmentation for assessment of PHE is warranted.

Concurrent Determination of Reliability and Validity of Force Platform, Danesh Salar-e Iranian, Using Kistler's Platform in Measuring Static Balance of Men and Women Athletes; A Descriptive-Surveying Study

Introduction : The aim of the current research was to concurrently determine the reliability and validity of force platform device, Danesh Salar-e Iranian, using Kistler's platform device in measuring the static balance of men and women athletes. Materials and Methods: Participants consisted of 50 athletes (25 men and 25 women) from Isfahan City, Iran, who were selected using a purposive convenience sampling method. The research instrument included the force platform device of Danesh Salar-e Iranian model and Kistler's platform device. The subjects tried to keep their static balance on each of the force platforms for 40 seconds with eyes open and eyes closed. After 48-72 hours, test-retest was done under the same conditions as the first test. Each test repeated three times. To measure internal and time reliability, the within-group and between-group correlation tests were used. To measure the concurrent validity, the Pearson correlation test was used. The significance level was set at P < 0.05. Results : Kistler's platform had internal reliability in all open- and closed-eyes parameters; while the Danesh Salar-e Iranian force platform had internal reliability in all closed-eyes parameters. Both the force platform deviceshad time stability in all open-eyes parameters. Kistler's platform had time stability in all closed-eyes parameters (except total oscillation), too. In addition, there was a significant correlation in all variables between the two force platform devices. Conclusion : Danesh Salar-e Iranian force platform is a valid and reliable equipment to measure the static balance in men and women athletes in most parameters. Technical improvements of the equioment can increase the psychometric parameters of this tool, as well as its usability and effectiveness. This can promote national production, and pave the way for economic growth, national efficiency, and emancipation of dependency.

Abstract 78: Fully Automated Segmentation Algorithm for Hematoma Volumetric Analysis in Spontaneous Intracerebral Hemorrhage

Background and Purpose: Hematoma volume measurements influence prognosis and treatment decisions in patients with spontaneous intracerebral hemorrhage (ICH). The aims of this study are to derive and validate a fully automated segmentation algorithm for ICH volumetric analysis using deep learning methods. Methods: Inpatient computed tomography scans of 300 consecutive adults (age ≥18 years) with spontaneous, supratentorial ICH who were enrolled in the Intracerebral Hemorrhage Outcomes Project (2009-2018) were separated into training (n=260) and test (n=40) datasets. A fully automated segmentation algorithm was derived using convolutional neural networks (CNN), and it was trained on manual segmentations from the training dataset. The algorithm’s performance was assessed against manual and semi-automated segmentation methods in the test dataset. Results: The mean volumetric Dice similarity coefficients for the fully automated segmentation algorithm when tested against manual and semi-automated segmentation methods were 0.894±0.264 and 0.905±0.254, respectively. ICH volumes derived from fully automated vs. manual (R 2 =0.981;p&lt;0.0001), fully automated vs. semi-automated (R 2 =0.978;p&lt;0.0001), and semi-automated vs. manual (R 2 =0.990;p&lt;0001) segmentation methods had strong between-group correlations. The fully automated segmentation algorithm (mean 12.0±2.7 seconds/scan) was significantly faster than both of the manual (mean 201.5±92.2 seconds/scan; p&lt;0.001) and semi-automated (mean 288.58±160.3 seconds/scan; p&lt;0.001) segmentation methods. Conclusions: The fully automated segmentation algorithm quantified hematoma volumes from CT scans of supratentorial ICH patients with similar accuracy and substantially greater efficiency compared with manual and semi-automated segmentation methods.

Fully Automated Segmentation Algorithm for Hematoma Volumetric Analysis in Spontaneous Intracerebral Hemorrhage.

Background and Purpose- Hematoma volume measurements influence prognosis and treatment decisions in patients with spontaneous intracerebral hemorrhage (ICH). The aims of this study are to derive and validate a fully automated segmentation algorithm for ICH volumetric analysis using deep learning methods. Methods- In-patient computed tomography scans of 300 consecutive adults (age ≥18 years) with spontaneous, supratentorial ICH who were enrolled in the ICHOP (Intracerebral Hemorrhage Outcomes Project; 2009-2018) were separated into training (n=260) and test (n=40) datasets. A fully automated segmentation algorithm was derived using convolutional neural networks, and it was trained on manual segmentations from the training dataset. The algorithm's performance was assessed against manual and semiautomated segmentation methods in the test dataset. Results- The mean volumetric Dice similarity coefficients for the fully automated segmentation algorithm when tested against manual and semiautomated segmentation methods were 0.894±0.264 and 0.905±0.254, respectively. ICH volumes derived from fully automated versus manual (R2=0.981; P<0.0001), fully automated versus semiautomated (R2=0.978; P<0.0001), and semiautomated versus manual (R2=0.990; P<0001) segmentation methods had strong between-group correlations. The fully automated segmentation algorithm (mean 12.0±2.7 s/scan) was significantly faster than both of the manual (mean 201.5±92.2 s/scan; P<0.001) and semiautomated (mean 288.58±160.3 s/scan; P<0.001) segmentation methods. Conclusions- The fully automated segmentation algorithm quantified hematoma volumes from computed tomography scans of supratentorial ICH patients with similar accuracy and substantially greater efficiency compared with manual and semiautomated segmentation methods. External validation of the fully automated segmentation algorithm is warranted.

Road rage in China: An exploratory study

This study consisted of a stated-preference questionnaire survey to investigate road rage among Chinese drivers. A total of 5157 Chinese drivers completed the survey. The Driving Anger Scale (DAS) and Driving Anger Expression Inventory (DAX) were used to measure road rage under various anger-provoking situations. Meanwhile, the theory of planned behavior (TPB) was adopted to explore respondents’ attitudinal perception towards road rage behaviors. Three TPB variables, i.e., behavioral attitude (BA), subjective norm (SN), and perceived behavioral control (PBC), were used to describe drivers’ perceived anger levels, driving habits, self-efficacy, and perceived controllability of road rage, respectively. Additionally, aggressive reaction (AR) was introduced to describe drivers’ inclination to express road rage. Then, descriptive analyses were performed to examine the connection between individual characteristics on road rage. Results exhibited these variables have different between-group correlation and significant relationships with gender, age group, education, driving years, professional or not, daily driving hours, and vehicle type. Also, outcomes from structural equation models (SEMs) demonstrated that the above four descriptive variables in the modified TPB model could be linked to drivers’ intention to exhibit road rage behaviors. Finally, countermeasures for controlling and curing road rage were proposed based on the findings.

Analysis of Dynamic Correlation of Japanese Stock Returns with Network Clustering

In this paper, the dynamic correlation of Japanese stock returns is estimated by using the dynamic conditional correlation (DCC–GARCH) model to study their correlation dynamics empirically. It is difficult to fit the model to the whole stock market jointly at the same time; therefore, a network-based clustering is applied for the dimensionality reduction of the sample data. Two types correlation structures are estimated: homogeneous groups of stocks in a balanced size are created by clustering to observe within-group correlation, while a single portfolio that comprises group portfolio returns is also created to observe between-group correlation. The estimation result reveals dynamic changes in correlation intensity represented by the largest eigenvalue of the estimated correlation matrix. A higher level of correlation intensity and volatility are observed during the crisis periods, namely after both the Lehman collapse and the Great East Japan Earthquake, for the between- and within-group correlations. It is also confirmed that the pattern of correlation change is significantly different between the groups. The proposed method is useful for monitoring dynamic correlation of asset returns efficiently in a large scale of portfolio.

Transcriptomics and methylomics in chronic periodontitis with tobacco use: a pilot study

BackgroundAccumulating evidence suggests that tobacco smoking affects the susceptibility to and severity of chronic periodontitis. Epigenetics may explain the role of smoking in the development and progress of periodontal disease. In this study, we performed transcriptomic and methylomic analyses of non-periodontitis and periodontitis-affected gingival tissues according to smoking status.MethodsHuman gingival tissues were obtained from 20 patients, including non-smokers with and without periodontitis (n = 5 per group) and smokers with and without periodontitis (n = 5 per group). Total RNA and genomic DNA were isolated, and their quality was validated according to strict standards. The Illumina NextSeq500 sequencing system was used to generate transcriptome and methylome datasets.ResultsComprehensive analysis, including between-group correlation, differential gene expression, DNA methylation, gene set enrichment, and protein-protein interaction, indicated that smoking may change the transcription and methylation states of extracellular matrix (ECM) organization-related genes, which exacerbated the periodontal condition.ConclusionsOur results suggest that smoking-related changes in DNA methylation patterns and subsequent alterations in the expression of genes coding for ECM components may be causally related to the increased susceptibility to periodontitis in smokers as they could influence ECM organization, which in turn may have an effect on disease characteristics.

White matter lesions characterise brain involvement in moderate to severe chronic obstructive pulmonary disease, but cerebral atrophy does not

BackgroundBrain pathology is relatively unexplored in chronic obstructive pulmonary disease (COPD). This study is a comprehensive investigation of grey matter (GM) and white matter (WM) changes and how these relate to disease severity and cognitive function.MethodsT1-weighted and fluid-attenuated inversion recovery images were acquired for 31 stable COPD patients (FEV1 52.1% pred., PaO2 10.1 kPa) and 24 age, gender-matched controls. T1-weighted images were segmented into GM, WM and cerebrospinal fluid (CSF) tissue classes using a semi-automated procedure optimised for use with this cohort. This procedure allows, cohort-specific anatomical features to be captured, white matter lesions (WMLs) to be identified and includes a tissue repair step to correct for misclassification caused by WMLs. Tissue volumes and cortical thickness were calculated from the resulting segmentations. Additionally, a fully-automated pipeline was used to calculate localised cortical surface and gyrification. WM and GM tissue volumes, the tissue volume ratio (indicator of atrophy), average cortical thickness, and the number, size, and volume of white matter lesions (WMLs) were analysed across the whole-brain and regionally – for each anatomical lobe and the deep-GM. The hippocampus was investigated as a region-of-interest. Localised (voxel-wise and vertex-wise) variations in cortical gyrification, GM density and cortical thickness, were also investigated. Statistical models controlling for age and gender were used to test for between-group differences and within-group correlations. Robust statistical approaches ensured the family-wise error rate was controlled in regional and local analyses.ResultsThere were no significant differences in global, regional, or local measures of GM between patients and controls, however, patients had an increased volume (p = 0.02) and size (p = 0.04) of WMLs. In patients, greater normalised hippocampal volume positively correlated with exacerbation frequency (p = 0.04), and greater WML volume was associated with worse episodic memory (p = 0.05). A negative relationship between WML and FEV1 % pred. approached significance (p = 0.06).ConclusionsThere was no evidence of cerebral atrophy within this cohort of stable COPD patients, with moderate airflow obstruction. However, there were indications of WM damage consistent with an ischaemic pathology. It cannot be concluded whether this represents a specific COPD, or smoking-related, effect.

Multiple Mini-Interview for Postgraduate Year Residency Recruitment: First Experience in Taiwan

Background: We investigated the effectiveness of the first hospital-held multiple mini-interview for recruitment into a postgraduate residency program in Taiwan. Methods: We conducted a multiple mini-interview to recruit 37 trainees into the postgraduate general medicine residency in January 2017 of the 37 postgraduate year candidates, 28 had interned at our hospital. Between-group correlations among subgroup analyses using multiple comparisons, Chi square test, and the generalized linear model were conducted to test the validity and discrimination of the multiple mini-interview questions. Correlations between multiple mini-interview scores, objective structured clinical examination scores, and clinical practice examinations were conducted and available for 19 applicants who interned at this hospital to test the effectiveness of the multiple mini-interview as a recruiting tool. Results: The multiple mini-interview process exhibited high test discrimination (P<0.001). Whether an intern studied at our hospital or at a different hospital did not affect the performance of the multiple mini-interview. Significant correlations were observed among performance in the multiple mini-interview, all three objective structured clinical examinations, and in four of six clinical practice departments among the 19 candidates who interned at our hospital (P<0.05). Conclusions: The significant correlation of the multiple mini-interviews with clinical performance scores provides powerful support for multiple mini-interview assessment in recruiting residents suitable for clinical work. Continuing to collect data to study the multiple mini-interview program would improve our understanding of it and provide better support for the postgraduate year recruitment program.

Correlations Among Hospital Quality Measures: What "Hospital Compare" Data Tell Us.

A number of quality rating systems to rank health care providers have been developed over the years with the intention of helping consumers make informed health care purchasing decisions. Many use sets of individual quality measures to calculate a global rating. The utility of a global rating for consumer choice hinges on the relationships among included measures and the extent to which they jointly reflect an underlying dimension of quality. Publicly reported data on 4 quality domains-complication, mortality, readmission, and patient safety-from Centers for Medicare & Medicaid Services' Hospital Compare website were used to examine correlations among individual measures within each measure group (within-group correlations) and correlations between pairs of measures across different measure groups (between-group correlations). Modest within-group correlations were found in only 2 domains (mortality and readmission), and there were no meaningful between-group associations. These findings raise questions about whether consumers can reliably depend on global quality ratings to make informed decisions.

Longitudinal influence of microglial activation and amyloid on neuronal function in Alzheimer's disease.

Amyloid deposition, tangle formation, neuroinflammation and neuronal dysfunction are pathological processes involved in Alzheimer's disease. However, the relative role of these processes in driving disease progression is still unclear. The aim of this positron emission tomography study was to: (i) investigate longitudinal changes of microglial activation, amyloid and glucose metabolism; and (ii) assess the temporospatial relationship between these three processes in Alzheimer's disease. A group of eight patients with a diagnosis of Alzheimer's disease (66 ± 4.8 years) and 14 healthy controls (65 ± 5.5 years) underwent T1 and T2 magnetic resonance imaging, along with (11)C-(R)-PK11195, (11)C-Pittsburgh compound B and (18)F-fluorodeoxyglucose positron emission tomography scans for microglial activation, amyloid deposition and glucose metabolism. All patients were followed-up with repeated magnetic resonance imaging and three positron emission tomography scans after 16 months. Parametric maps were interrogated using region of interest analysis, Statistical Parametric Mapping, and between-group correlation analysis at voxel-level using Biological Parametric Mapping. At baseline, patients with Alzheimer's disease showed significantly increased microglial activation compared to the control subjects. During follow-up, for the first time, we found that while there is a progressive reduction of glucose metabolism, there was a longitudinal increase of microglial activation in the majority of the patients with Alzheimer's disease. Voxel-wise correlation analysis revealed that microglial activation in patients with Alzheimer's disease was positively correlated with amyloid deposition and inversely correlated with regional cerebral metabolic rate at voxel level over time. Even though one of the limitations of this study is the lack of longitudinal follow-up of healthy control subjects, this study demonstrates that there is persistent neuroinflammation throughout the Alzheimer's disease process with associated synaptic dysfunction and reduced glucose metabolism. Voxel-wise correlation analysis suggests that neuroinflammation is associated with localized amyloid deposition and glucose metabolism over time, however, the level of inflammation could also occur independently of amyloid pathology, especially in the later stages of Alzheimer's disease.

Effects of Geodemographic Profiles of Drivers on Their Injury Severity from Traffic Crashes Using Multilevel Mixed-Effects Ordered Logit Model

The purpose of this study was to use a statistical model to examine the effects of various geodemographic factors on levels of driver injury severity. A driver's geodemographic profile with respect to involvement in a traffic crash consists of variables from multiple hierarchical levels such as drivers nested within crashes and crashes clustered within areas. This structure implies that driver-level observations are correlated rather than independent as assumed in many injury severity models. To capture within-group and between-group correlations in observations, a multilevel mixed-effects ordered logit model was employed. The mixed effect allows some variables to vary by observations (i.e., random parameters). The analysis was based on national traffic crash data in the United Kingdom between 2009 and 2011, consisting of 271,654 drivers from 217,523 traffic crashes occurring across 27,773 census areas. Data on area, social deprivation, census, and land use patterns were collected from multiple sources and integrated with a geographic information system framework. Results indicated that the severity of injuries sustained by urban drivers involved in crashes increased if they traveled to rural areas; the level of driver injury severity also increased if traffic crashes occurred in areas with high car ownership per capita; and drivers from more disadvan-taged areas would sustain, if all else is equal, more severe injuries. Findings from this study would be useful to the UK Department for Transport and local authorities in formulating safety policies aimed at enhancing driver education, training, and licensing programs.

Effects of sevoflurane and desflurane on otoacoustic emissions in humans.

Otoacoustic emissions (OAEs) are non-invasive, easy to apply and objective test methods which are widely used to determine the presence of hearing in audiology clinics. Under certain circumstances, the study should be applied under general anesthesia. The aim of this study was to determine the influence of new short-acting inhalation agents, desflurane and sevoflurane, on OAE in humans. These short-acting agents are widely used in general anesthesia. Thirty-one healthy patients who underwent septoplasty and turbinoplasty surgery were included in this study. Unpremedicated patients were anesthetized and monitored by a standard protocol except the inhalation agents. Desflurane and sevoflurane were added to the inhaled gas mixture at ~1MAC, 5-6 % and 1.5-2 %, respectively. Transient evoked otoacoustic emissions and distortion product otoacoustic emissions measured in both ears of each patient preoperatively in the operating room before induction, 5 min after induction, after the completion of surgical procedure while the anesthetic agents are still given and 1 h after surgery in the ward. Between-group and within-group comparisons and correlations with hemodynamic parameters were performed for statistical analysis. The measurements of 26 ears in desflurane group and 28 ears in sevoflurane group were evaluated. There were no differences in initial measurements between groups (p > 0.05). Both groups presented significant decrease in intraoperative measurements and changes in time were statistically significant (p < 0.05). The changes in OAEs were similar to changes in systemic blood pressures. Correlation between OAEs and systemic blood pressures were significant (p < 0.05). In conclusion, sevoflurane and desflurane decreased OAEs around 2-3 dB; OAEs are still measurable under inhalation agents. This provides some findings about the OAE status of patient, but the evaluations should be done with the impact of anesthetic agents in mind.

Relation Between Mixed Venous Oxygen Saturation and Cerebral Oxygen Saturation Measured by Absolute and Relative Near-Infrared Spectroscopy During Off-pump Coronary Artery Bypass Grafting

Background We hypothesized that previously reported contradictory results regarding the equivalence of mixed venous (SmvO2) and cerebral (rScO2) oxygen saturation might be related to time delay issues and to measurement technology. In order to explore these two factors, we designed a prospective clinical study comparing SmvO2 with relative (INVOS®) and absolute (Foresight®) rScO2 measurements. Methods Forty-two consenting patients undergoing elective off-pump coronary artery bypass grafting were included. Two INVOS and two Foresight sensors continuously registered rScO2. SmvO2 was measured continuously via a pulmonary artery catheter. Data were assessed by within- and between-group comparisons and correlation analysis. Results A similar time delay of 19 (4) and 18 (4) s was found for SmvO2 compared with rScO2 measurements by Foresight and INVOS, respectively, during haemodynamic changes. After adjusting for this time delay, the correlation between SmvO2 and rScO2 increased from r=0.25 to 0.75 (P Conclusions Considering the important time delay with SmvO2, rScO2 seems to reflect more appropriately acute haemodynamic alterations. This might suggest its use as a valid alternative to invasive monitoring of tissue oxygen saturation. Relative and absolute rScO2 measurements demonstrated significant differences in measured rScO2 values and in the magnitude of rScO2 changes during haemodynamic alterations.