Linked Lives: Does Disability and Marital Quality Influence Risk of Marital Dissolution among Older Couples?

Using fourteen waves of data from the Health and Retirement Study (HRS), a longitudinal panel survey with respondents in the United States, this research explores whether marital quality—as measured by reports of enjoyment of time together—influences risk of divorce or separation when either spouse acquires basic care disability. Discrete-time event history models with multiple competing events were estimated using multinomial logistic regression. Respondents were followed until they experienced the focal event (i.e., divorce or separation) or right-hand censoring (i.e., a competing event or were still married at the end of observation). Disability among wives was predictive of divorce/separation in the main effects model. Low levels of marital quality (i.e., enjoy time together) were associated with marital dissolution. An interaction between marital quality and disability yielded a significant association among couples where at least one spouse acquired basic care disability. For couples who acquired disability, those who reported low enjoyment were more likely to divorce/separate than those with high enjoyment; however, the group with the highest predicted probability were couples with low enjoyment, but no acquired disability.

Tagging terms in text

As with many tasks in natural language processing, automatic term extraction (ATE) is increasingly approached as a machine learning problem. So far, most machine learning approaches to ATE broadly follow the traditional hybrid methodology, by first extracting a list of unique candidate terms, and classifying these candidates based on the predicted probability that they are valid terms. However, with the rise of neural networks and word embeddings, the next development in ATE might be towards sequential approaches, i.e., classifying each occurrence of each token within its original context. To test the validity of such approaches for ATE, two sequential methodologies were developed, evaluated, and compared: one feature-based conditional random fields classifier and one embedding-based recurrent neural network. An additional comparison was added with a machine learning interpretation of the traditional approach. All systems were trained and evaluated on identical data in multiple languages and domains to identify their respective strengths and weaknesses. The sequential methodologies were proven to be valid approaches to ATE, and the neural network even outperformed the more traditional approach. Interestingly, a combination of multiple approaches can outperform all of them separately, showing new ways to push the state-of-the-art in ATE.

Banking disclosure and banking crises in Africa: does board gender diversity play a role?

The paper aims to investigate the impact of board gender diversity in explaining the relationship between bank disclosure and the predicted probability of banking crises in Africa. The study employs robust panel estimates based on an aggregate dataset of banks in 42 African countries over the 2006–2018 periods. From the study, board gender diversity (more women on boards and the presence of women on boards) has a positive impact on information disclosure of banks. We find that board gender diversity and bank disclosure have the possibility of reducing a banking crisis. We observe that board gender diversity enhances the reductive effect of bank disclosure on a predicted probability of a banking crisis. The implication is that women on boards provide prudent decisions on financial information disclosure that significantly reduce the possibility of a banking crises in order to ensure stable banking systems.

CT-Assisted Improvements in the Accuracy of the Intraoperative Frozen Section Examination of Ground-Glass Density Nodules

Objectives. The intraoperative frozen section examination (IFSE) of pulmonary ground-glass density nodules (GGNs) is a great challenge. In the present study, through comparing the correlation between the computed tomography (CT) findings and pathological diagnosis of GGNs, the CT features as independent risk factors affecting the examination were defined, and their value in the rapid intraoperative examination of GGNs was explored. Methods. The relevant clinical data of 90 patients with GGNs on CT were collected, and all CT findings of GGNs, including the maximum transverse diameter, average CT value, spiculation, solid component, vascular sign, air sign, bronchus sign, lobulation, and pleural indentation, were recorded. All the cases received thoracoscopic surgery, and final pathological results were obtained. The cases were divided into three groups on the basis of pathological diagnosis: benign/atypical adenomatous hyperplasia (AAH), adenocarcinoma in situ (AIS)/microinvasive adenocarcinoma (MIA), and invasive adenocarcinoma (IAC). The CT findings were analyzed statistically, the independent risk factors were identified through the intergroup bivariate logistic regression analysis on variables with statistically significant differences, and a receiver operating curve (ROC) was plotted to establish a logistic regression model for diagnosing GGNs. A retrospective analysis was conducted on the coincidence rate of the rapid intraoperative and routine postoperative pathological examinations of the 90 cases with GGNs. The relevant clinical data of 49 cases with GGNs were collected. Conventional rapid intraoperative examination and CT-assisted rapid intraoperative examination were performed, and their coincidence rates with routine postoperative pathological examinations were compared. Results. No statistical differences in the onset age, gender, smoking history, and family history of malignant tumors were found among cases with GGNs in the identification of benign/AAH, AIS/MIA, and IAC ( P = 0.158 , P = 0.947 , P = 0.746 , P = 0.566 ). No statistically significant difference was found among the three groups in terms of CT findings, such as lobulation, bronchus sign, pleural indentation, spiculation, vascular sign, and solid component ( P > 0.05 ). The air sign, the maximum transverse diameter of GGNs, and average CT value showed statistically significant differences among the groups ( P < 0.001 , P < 0.05 , P < 0.001 ). Bivariate logistic regression analysis was performed on three risk factors, and the predicted probability value was obtained. A ROC curve was plotted by using the maximum transverse diameter as a predictor for analysis between the groups with benign/AAH and AIS/MIA, and the results demonstrated that the area under the curve (AUC) was 0.692. A ROC curve was plotted by using the predicted probability value, maximum transverse diameter, and average CT value as predictors for distinguishing between the groups with AIS/MIA and IAC, and the results showed that the AUC values of the predicted probability value, maximum transverse diameter, and CT value were 0.920, 0.816, and 0.772, respectively. A regression model Logit P = 2.304 − 2.689 X 1 + 0.302 X 2 + 0.011 X 3 was established to identify GGNs as IAC, obtaining AUC values of up to 0.920 for the groups with AIS/MIA and IAC, the sensitivity of 0.821, and the specificity of 0.894. The coincidence rate of rapid intraoperative and routine postoperative pathological examinations taken for modeling was 79.3%, that of conventional IFSE and postoperative pathological examination in prospective studies was 83.7%, and that of CT-assisted rapid intraoperative and postoperative pathological examinations was 98.0%. The former two were statistically different from the last one ( P = 0.003 and P = 0.031 , respectively). Conclusion. The air sign, maximum transverse diameter, and average CT value of the CT findings of GGNs had superior capabilities to enhance the pathologic classification of GGNs. The auxiliary function of the comprehensive multifactor analysis of GGNs was better than that of single-factor analysis. CT-assisted diagnosis can improve the accuracy of rapid intraoperative examination, thereby increasing the accuracy of the selection of operative approaches in clinical practice.

Decision Support Tool Identifies Ulcerative Colitis Patients Most Likely to Achieve Remission With Vedolizumab vs Adalimumab

Background & Aims We have previously validated a clinical decision support tool (CDST) (vedolizumab CDST [VDZ-CDST]) for clinical and endoscopic remission with VDZ in ulcerative colitis (UC). We aim to expand the validation for predicting histoendoscopic mucosal improvement (HEMI) with VDZ vs adalimumab (ADA). Methods In a post hoc analysis of a clinical trial for VDZ vs ADA in moderate to severe UC (VARSITY trial; NCT02497469), comparative accuracy was evaluated for the VDZ-CDST among an external validation cohort of VDZ- and ADA-treated patients for week 52 HEMI (Mayo endoscopic subscore 0-1 and Geboes score &lt;3.2). Comparative effectiveness of VDZ and ADA was assessed after stratifying the cohort by baseline probability of response to VDZ using the VDZ-CDST. Results A total of 419 patients were included. The majority of patients enrolled in the VARSITY trial had a high (61%) or intermediate (29%) baseline predicted probability of response to VDZ. The baseline VDZ-CDST score was significantly more likely to predict week 52 HEMI for VDZ (area under the curve , 0.712; 95% confidence interval, 0.636-0.787) relative to ADA-treated patients (area under the curve, 0.538; 95% confidence interval, 0.377-0.700; P &lt; .001 for AUC comparison). A significant (P &lt; .001) association was observed between the VDZ-CDST and measured VDZ drug exposure over 52 weeks. Superiority of VDZ to ADA was only observed in patients with a high baseline predicted probability of response to VDZ. Conclusions Superiority of VDZ to ADA is dependent on baseline probability of response, and a VDZ-CDST is capable of identifying UC patients most appropriate for VDZ vs ADA.

Spatial and temporal effects on SARS-CoV-2 contamination of the healthcare environment

Background: The spatial and temporal extent of SARS-CoV-2 environmental contamination has not been precisely defined. We sought to elucidate contamination of different surface types and how contamination changes over time. Methods: We sampled surfaces longitudinally within COVID-19 patient rooms, performed quantitative RT-PCR for the detection of SARS-CoV-2 RNA, and modeled distance, time, and severity of illness on the probability of detecting SARS-CoV-2 using a mixed-effects binomial model. Results: The probability of detecting SARS-CoV-2 RNA in a patient room did not vary with distance. However, we found that surface type predicted probability of detection, with floors and high-touch surfaces having the highest probability of detection (floors odds ratio (OR) 67.8 (95% CrI 36.3 to 131); high-touch elevated OR 7.39 (95% CrI 4.31 to 13.1)). Increased surface contamination was observed in room where patients required high-flow oxygen, positive airway pressure, or mechanical ventilation (OR 1.6 (95% CrI 1.03 to 2.53)). The probability of elevated surface contamination decayed with prolonged hospitalization, but the probability of floor detection increased with duration of the local pandemic wave. Conclusions: Distance from patient’s bed did not predict SARS-CoV-2 RNA deposition in patient rooms, but surface type, severity of illness, and time from local pandemic wave predicted surface deposition.

A Prediction Model to Discriminate Small Choroidal Melanoma from Choroidal Nevus

Objective: To develop a validated machine learning model to diagnose small choroidal melanoma. Design: Cohort study Subjects, Participants, and/or Controls: The training data included 123 patients diagnosed as small choroidal melanocytic tumor (5.0-16.0 mm in largest basal diameter and 1.0 mm to 2.5 mm in height; Collaborative Ocular Melanoma Study criteria). Those diagnosed as melanoma (n=61) had either documented growth or pathologic confirmation. 62 patients with stable lesions classified as choroidal nevus, were used as negative controls. The external validation data set included 240 patients managed at a different tertiary clinic, also with small choroidal melanocytic tumor, observed for malignant growth. Methods: In the training data, lasso logistic regression was used to select variables for inclusion in the final model for the association with melanoma versus choroidal nevus. Internal and external validation were performed to assess model performance. Main Outcome Measures: Predicted probability of small choroidal melanoma Results: Distance to optic disc ≥3mm and drusen were associated with decreased odds of melanoma whereas male versus female sex, increased height, subretinal fluid, and orange pigment were associated with increased odds of choroidal melanoma. The area under the receiver operating characteristic (AUROC) “discrimination value” for this model was 0.880. The top four variables that were most frequently selected for inclusion in the model on internal validation, implying their importance as predictors of melanoma, were subretinal fluid, height, distance to optic disc, and orange pigment. When tested against the validation data, the prediction model could distinguish between choroidal nevus and melanoma with high discrimination of 0.861. The final prediction model was converted into an online calculator to generate predicted probability of melanoma. Conclusions: To minimize diagnostic uncertainty, a machine learning based diagnostic prediction calculator can be readily applied for decision making and counselling patients with small choroidal melanoma.

Workplace Contact Patterns in England during the COVID-19 Pandemic: Analysis of the Virus Watch prospective cohort study

Background: Workplaces are an important potential source of SARS-CoV-2 exposure; however, investigation into workplace contact patterns is lacking. This study aimed to investigate how workplace attendance and features of contact varied between occupations and over time during the COVID-19 pandemic in England. Methods: Data were obtained from electronic contact diaries submitted between November 2020 and November 2021 by employed/self-employed prospective cohort study participants (n=4,616). We used mixed models to investigate the main effects and potential interactions between occupation and time for: workplace attendance, number of people in shared workspace, time spent sharing workspace, number of close contacts, and usage of face coverings. Findings: Workplace attendance and contact patterns varied across occupations and time. The predicted probability of intense space sharing during the day was highest for healthcare (78% [95% CI: 75-81%]) and education workers (64% [59%-69%]), who also had the highest probabilities for larger numbers of close contacts (36% [32%-40%] and 38% [33%-43%] respectively). Education workers also demonstrated relatively low predicted probability (51% [44%-57%]) of wearing a face covering during close contact. Across all occupational groups, levels of workspace sharing and close contact were higher and usage of face coverings at work lower in later phases of the pandemic compared to earlier phases. Interpretation: Major variations in patterns of workplace contact and mask use are likely to contribute to differential COVID-19 risk. Across occupations, increasing workplace contact and reduced usage of face coverings presents an area of concern given ongoing high levels of community transmission and emergence of variants.

Predicting the Outcome of Open Type IIIC Tibial Fractures Regarding Union, Osteomyelitis, and Amputation

Backgrounds: This study aimed to assess factors that affect union time and complications in Gustilo IIIC tibial fractures.Methods: Patients who presented to our center with IIIC open tibial fractures from January 2000 to October 2020 were eligible for this retrospective analysis. Patient demographics, fracture characteristics, timing, numbers, and type of surgical intervention were documented. Outcomes of interest included union time, occurrence of osteomyelitis, and amputation. Results: Fifty-eight patients were enrolled and grouped by fracture type; eight union on time (13.8%); 27 late union (46.6%); eight delayed union (13.8%); three nonunion (5.2%); and 12 amputation (20.7%). Nine fractures (15.5%) were complicated by osteomyelitis. Union time was prolonged in cases of triple arterial injury, distal third fractures, multiple trauma with Injury Severity Score (ISS) ≥ 16 points, and increased length of bone defect. Additionally, a bone gap > 50 mm, diabetes mellitus, low body mass index, and triple arterial injury in the lower leg were significant risk factors for amputation. A time from injury to definitive soft tissue coverage of more than 22 days was the major risk factor for osteomyelitis. A scoring system to predict union time was devised and the predicted probability of union within two years was stratified based on this score. Conclusions: IIIC tibial fractures involving the distal third of the tibia, fractures with bone defects, triple arterial injury, and multiple trauma with ISS ≥ 16 points demonstrated delayed union, and an effective prediction system for union time was introduced in this study. Early soft tissue coverage can reduce the risk of osteomyelitis. Finally, diabetes and severe bone and soft tissue defects pose a higher risk of amputation.

Development of a new score for early mortality prediction in trauma ICU patients: RETRASCORE

Background Severity scores are commonly used for outcome adjustment and benchmarking of trauma care provided. No specific models performed only with critically ill patients are available. Our objective was to develop a new score for early mortality prediction in trauma ICU patients. Methods This is a retrospective study using the Spanish Trauma ICU registry (RETRAUCI) 2015–2019. Patients were divided and analysed into the derivation (2015–2017) and validation sets (2018–2019). We used as candidate variables to be associated with mortality those available in RETRAUCI that could be collected in the first 24 h after ICU admission. Using logistic regression methodology, a simple score (RETRASCORE) was created with points assigned to each selected variable. The performance of the model was carried out according to global measures, discrimination and calibration. Results The analysis included 9465 patients: derivation set 5976 and validation set 3489. Thirty-day mortality was 12.2%. The predicted probability of 30-day mortality was determined by the following equation: 1/(1 + exp (− y)), where y = 0.598 (Age 50–65) + 1.239 (Age 66–75) + 2.198 (Age > 75) + 0.349 (PRECOAG) + 0.336 (Pre-hospital intubation) + 0.662 (High-risk mechanism) + 0.950 (unilateral mydriasis) + 3.217 (bilateral mydriasis) + 0.841 (Glasgow ≤ 8) + 0.495 (MAIS-Head) − 0.271 (MAIS-Thorax) + 1.148 (Haemodynamic failure) + 0.708 (Respiratory failure) + 0.567 (Coagulopathy) + 0.580 (Mechanical ventilation) + 0.452 (Massive haemorrhage) − 5.432. The AUROC was 0.913 (0.903–0.923) in the derivation set and 0.929 (0.918–0.940) in the validation set. Conclusions The newly developed RETRASCORE is an early, easy-to-calculate and specific score to predict in-hospital mortality in trauma ICU patients. Although it has achieved adequate internal validation, it must be externally validated.