A Review of Hot Deck Imputation for Survey Non-response

Is there an association between a history of military service and cancer diagnosis? Results from a US national-level study of self-reported outcomes

PURPOSE

To examine cancer prevalence in men with and without military service history, using national-level self-reported outcomes.

METHODS

A cross-sectional survey-based US study, including men aged 18 and above from the Health Information National Trends Survey database between 2011 and 2014. The primary endpoint was self-reported cancer prevalence. Multivariable logistic regression analyses assessed the association of various covariates with the prevalence of cancer.

RESULTS

A total of 4,527 men were analyzed, with 1,352 (29.9%) reporting a history of military service. Compared to men with no military service history, men with a military service history were older (median of 65 [IQR 56, 74] vs. 53 [IQR 41, 62] years, p < 0.0001), more commonly Caucasian (71.4% vs. 61.4%, p < 0.0001), born in the US (95.6% vs. 79.5%, p < 0.0001), attained higher education level and annual household income (p < 0.0001), and consisted of more smokers(58.3% vs. 44.5%, p < 0.0001). The age-adjusted comparison demonstrated a higher cancer prevalence in men with military service history (20.5% vs. 7.6%, p < 0.0001). Specifically, genitourinary, dermatological, gastrointestinal, and hematological cancers were generally more prevalent. Adjusting for all available confounders, multivariable models showed that military service history was associated with 1.56 (95% CI 1.20-2.03), and 1.57 (95% CI 1.07-2.31) increased odds of having any cancer, and specifically genitourinary cancer, respectively.

CONCLUSIONS

Further research is needed to ascertain whether the association between military service and increased cancer diagnosis results from better screening programs or increased exposure to risk factors during military service.

Everyday Discrimination Typologies Among Older African Americans: Gender and Socioeconomic Status

OBJECTIVES

Discrimination is associated with several negative social, economic, and health consequences. Past research focuses on the impact of discrimination while less is known about both the type and correlates of discrimination, particularly among older adults.

METHODS

Using the National Survey of American Life, we used latent class analysis to identify discrimination typologies (frequency and type) among African Americans aged 55 and older. We then used multinomial logistic regression to identify demographic correlates of discrimination types, including a statistical interaction between gender and educational attainment.

RESULTS

We identified three discrimination typologies. Increasing age was associated with lower probability of belonging to the high discrimination and disrespect and condescension subtypes. Men and non-Southern residents were most likely to belong in the high discrimination subtype. Higher levels of education increased the probability of belonging in the high discrimination and disrespect and condescension subtypes for older men, but not women.

DISCUSSION

Older African American men, particularly those with more education, are vulnerable to both high-frequency discrimination and discrimination characterized by disrespect and condescension. This finding suggests that, for men with more years of education, increased exposure to discrimination reflects efforts to maintain social hierarchies (male target hypothesis).

A Novel Multiplex Based Platform for Osteoarthritis Drug Candidate Evaluation

Osteoarthritis (OA) is characterized by irreversible cartilage degradation with very limited therapeutic interventions. Drug candidates targeted at prototypic players had limited success until now and systems based approaches might be necessary. Consequently, drug evaluation platforms should consider the biological complexity looking beyond well-known contributors of OA. In this study an ex vivo model of cartilage degradation, combined with measuring releases of 27 proteins, was utilized to study 9 drug candidates. After an initial single drug evaluation step the 3 most promising compounds were selected and employed in an exhaustive combinatorial experiment. The resulting most and least promising treatment candidates were selected and validated in an independent study. This included estimation of mechanical properties via finite element modelling (FEM) and quantification of cartilage degradation as glycosaminoglycan (GAG) release. The most promising candidate showed increase of Young's modulus, decrease of hydraulic permeability and decrease of GAG release. The least promising candidate exhibited the opposite behaviour. The study shows the potential of a novel drug evaluation platform in identifying treatments that might reduce cartilage degradation. It also demonstrates the promise of exhaustive combination experiments and a connection between chondrocyte responses at the molecular level with changes of biomechanical properties at the tissue level.

Metabolic recovery after weight loss surgery is reflected in serum microRNAs

INTRODUCTION

Bariatric surgery offers the most effective treatment for obesity, ameliorating or even reverting associated metabolic disorders, such as type 2 diabetes. We sought to determine the effects of bariatric surgery on circulating microRNAs (miRNAs) that have been implicated in the metabolic cross talk between the liver and adipose tissue.

RESEARCH DESIGN AND METHODS

We measured 30 miRNAs in 155 morbidly obese patients and 47 controls and defined associations between miRNAs and metabolic parameters. Patients were followed up for 12 months after bariatric surgery. Key findings were replicated in a separate cohort of bariatric surgery patients with up to 18 months of follow-up.

RESULTS

Higher circulating levels of liver-related miRNAs, such as miR-122, miR-885-5 p or miR-192 were observed in morbidly obese patients. The levels of these miRNAs were positively correlated with body mass index, percentage fat mass, blood glucose levels and liver transaminases. Elevated levels of circulating liver-derived miRNAs were reversed to levels of non-obese controls within 3 months after bariatric surgery. In contrast, putative adipose tissue-derived miRNAs remained unchanged (miR-99b) or increased (miR-221, miR-222) after bariatric surgery, suggesting a minor contribution of white adipose tissue to circulating miRNA levels. Circulating levels of liver-derived miRNAs normalized along with the endocrine and metabolic recovery of bariatric surgery, independent of the fat percentage reduction.

CONCLUSIONS

Since liver miRNAs play a crucial role in the regulation of hepatic biochemical processes, future studies are warranted to assess whether they may serve as determinants or mediators of metabolic risk in morbidly obese patients.

Training Load and Its Role in Injury Prevention, Part 2: Conceptual and Methodologic Pitfalls

In part 2 of this clinical commentary, we highlight the conceptual and methodologic pitfalls evident in current training-load-injury research. These limitations make these studies unsuitable for determining how to use new metrics such as acute workload, chronic workload, and their ratio for reducing injury risk. The main overarching concerns are the lack of a conceptual framework and reference models that do not allow for appropriate interpretation of the results to define a causal structure. The lack of any conceptual framework also gives investigators too many degrees of freedom, which can dramatically increase the risk of false discoveries and confirmation bias by forcing the interpretation of results toward common beliefs and accepted training principles. Specifically, we underline methodologic concerns relating to (1) measure of exposures, (2) pitfalls of using ratios, (3) training-load measures, (4) time windows, (5) discretization and reference category, (6) injury definitions, (7) unclear analyses, (8) sample size and generalizability, (9) missing data, and (10) standards and quality of reporting. Given the pitfalls of previous studies, we need to return to our practices before this research influx began, when practitioners relied on traditional training principles (eg, overload progression) and adjusted training loads based on athletes' responses. Training-load measures cannot tell us whether the variations are increasing or decreasing the injury risk; we recommend that practitioners still rely on their expert knowledge and experience.

Exploiting mutual information for the imputation of static and dynamic mixed-type clinical data with an adaptive k-nearest neighbours approach

BACKGROUND

Clinical registers constitute an invaluable resource in the medical data-driven decision making context. Accurate machine learning and data mining approaches on these data can lead to faster diagnosis, definition of tailored interventions, and improved outcome prediction. A typical issue when implementing such approaches is the almost unavoidable presence of missing values in the collected data. In this work, we propose an imputation algorithm based on a mutual information-weighted k-nearest neighbours approach, able to handle the simultaneous presence of missing information in different types of variables. We developed and validated the method on a clinical register, constituted by the information collected over subsequent screening visits of a cohort of patients affected by amyotrophic lateral sclerosis.

METHODS

For each subject with missing data to be imputed, we create a feature vector constituted by the information collected over his/her first three months of visits. This vector is used as sample in a k-nearest neighbours procedure, in order to select, among the other patients, the ones with the most similar temporal evolution of the disease over time. An ad hoc similarity metric was implemented for the sample comparison, capable of handling the different nature of the data, the presence of multiple missing values and include the cross-information among features captured by the mutual information statistic.

RESULTS

We validated the proposed imputation method on an independent test set, comparing its performance with those of three state-of-the-art competitors, resulting in better performance. We further assessed the validity of our algorithm by comparing the performance of a survival classifier built on the data imputed with our method versus the one built on the data imputed with the best-performing competitor.

CONCLUSIONS

Imputation of missing data is a crucial -and often mandatory- step when working with real-world datasets. The algorithm proposed in this work could effectively impute an amyotrophic lateral sclerosis clinical dataset, by handling the temporal and the mixed-type nature of the data and by exploiting the cross-information among features. We also showed how the imputation quality can affect a machine learning task.

The FUTUREPAIN study: Validating a questionnaire to predict the probability of having chronic pain 7-10 years into the future

OBJECTIVES

The FUTUREPAIN study develops a short general-purpose questionnaire, based on the biopsychosocial model, to predict the probability of developing or maintaining moderate-to-severe chronic pain 7-10 years into the future.

METHODS

This is a retrospective cohort study. Two-thirds of participants in the National Survey of Midlife Development in the United States were randomly assigned to a training cohort used to train a predictive machine learning model based on the least absolute shrinkage and selection operator (LASSO) algorithm, which produces a model with minimal covariates. Out-of-sample predictions from this model were then estimated using the remaining one-third testing cohort to determine the area under the receiver operating characteristic curve (AUROC). An optimal cut-point that maximized sensitivity and specificity was determined.

RESULTS

The LASSO model using 82 variables in the training cohort, yielded an 18-variable model with an out-of-sample AUROC of 0.85 (95% Confidence Interval (CI): 0.80, 0.91) in the testing cohort. The sum of sensitivity (0.88) and specificity (0.76) was maximized at a cut-point of 17 (95% CI: 15, 18) on a 0-100 scale where the AUROC was 0.82.

DISCUSSION

We developed a short general-purpose questionnaire that predicts the probability of an adult having moderate-to-severe chronic pain in 7-to-10 years. It has diagnostic ability greater than 80% and can be used regardless of whether a patient is currently experiencing chronic pain. Knowing which patients are likely to have moderate-to-severe chronic pain in the future allows clinicians to target preventive treatment.

Investigating Parallel Analysis in the Context of Missing Data: A Simulation Study Comparing Six Missing Data Methods

Exploratory factor analysis is a statistical method commonly used in psychological research to investigate latent variables and to develop questionnaires. Although such self-report questionnaires are prone to missing values, there is not much literature on this topic with regard to exploratory factor analysis-and especially the process of factor retention. Determining the correct number of factors is crucial for the analysis, yet little is known about how to deal with missingness in this process. Therefore, in a simulation study, six missing data methods (an expectation-maximization algorithm, predictive mean matching, Bayesian regression, random forest imputation, complete case analysis, and pairwise complete observations) were compared with respect to the accuracy of the parallel analysis chosen as retention criterion. Data were simulated for correlated and uncorrelated factor structures with two, four, or six factors; 12, 24, or 48 variables; 250, 500, or 1,000 observations and three different missing data mechanisms. Two different procedures combining multiply imputed data sets were tested. The results showed that no missing data method was always superior, yet random forest imputation performed best for the majority of conditions-in particular when parallel analysis was applied to the averaged correlation matrix rather than to each imputed data set separately. Complete case analysis and pairwise complete observations were often inferior to multiple imputation.

Examining the Relationship Between Mental Skills and Grit in Senior Olympic Athletes

Given the challenge of consistent physical activity among aging adults, it is important to find ways to increase physical activity levels in this population. Participants in the Senior Olympic Games may extend their sport participation through the use of mental skills. This study examined the relationship between mental skills use by Senior Olympic Games participants and their grit, or passion and perseverance, toward a long-term goal. The participants in the Arizona Senior Olympic Games (n = 304) completed an online survey of mental skills use (Athletic Coping Skills Inventory) and grit (Grit Scale-Short). Based on the ongoing validity and reliability issues of the Grit Scale-Short, two regression models were examined, with consistency of interests (passion) and perseverance of effort (perseverance) as dependent variables. After controlling for age and sex, mental skills accounted for 15.2% of the variance in consistency of interests and 13.1% of the variability in perseverance of effort. The results are discussed in light of the findings.

Missing Value Imputation in Stature Estimation by Learning Algorithms Using Anthropometric Data: A Comparative Study

Estimating stature is essential in the process of personal identification. Because it is difficult to find human remains intact at crime scenes and disaster sites, for instance, methods are needed for estimating stature based on different body parts. For instance, the upper and lower limbs may vary depending on ancestry and sex, and it is of great importance to design adequate methodology for incorporating these in estimating stature. In addition, it is necessary to use machine learning rather than simple linear regression to improve the accuracy of stature estimation. In this study, the accuracy of statures estimated based on anthropometric data was compared using three imputation methods. In addition, by comparing the accuracy among linear and nonlinear classification methods, the best method was derived for estimating stature based on anthropometric data. For both sexes, multiple imputation was superior when the missing data ratio was low, and mean imputation performed well when the ratio was high. The support vector machine recorded the highest accuracy in all ratios of missing data. The findings of this study showed appropriate imputation methods for estimating stature with missing anthropometric data. In particular, the machine learning algorithms can be effectively used for estimating stature in humans.

Multinomial logistic regression with missing outcome data: An application to cancer subtypes

Many diseases such as cancer and heart diseases are heterogeneous and it is of great interest to study the disease risk specific to the subtypes in relation to genetic and environmental risk factors. However, due to logistic and cost reasons, the subtype information for the disease is missing for some subjects. In this article, we investigate methods for multinomial logistic regression with missing outcome data, including a bootstrap hot deck multiple imputation (BHMI), simple inverse probability weighted (SIPW), augmented inverse probability weighted (AIPW), and expected estimating equation (EEE) estimators. These methods are important approaches for missing data regression. The BHMI modifies the standard hot deck multiple imputation method such that it can provide valid confidence interval estimation. Under the situation when the covariates are discrete, the SIPW, AIPW, and EEE estimators are numerically identical. When the covariates are continuous, nonparametric smoothers can be applied to estimate the selection probabilities and the estimating scores. These methods perform similarly. Extensive simulations show that all of these methods yield unbiased estimators while the complete-case (CC) analysis can be biased if the missingness depends on the observed data. Our simulations also demonstrate that these methods can gain substantial efficiency compared with the CC analysis. The methods are applied to a colorectal cancer study in which cancer subtype data are missing among some study individuals.

Amoxicillin for 3 or 5 Days for Chest-Indrawing Pneumonia in Malawian Children

BACKGROUND

Evidence regarding the appropriate duration of treatment with antibiotic agents in children with pneumonia in low-resource settings in Africa is lacking.

METHODS

We conducted a double-blind, randomized, controlled, noninferiority trial in Lilongwe, Malawi, to determine whether treatment with amoxicillin for 3 days is less effective than treatment for 5 days in children with chest-indrawing pneumonia (cough lasting <14 days or difficulty breathing, along with visible indrawing of the chest wall with or without fast breathing for age). Children not infected with human immunodeficiency virus (HIV) who were 2 to 59 months of age and had chest-indrawing pneumonia were randomly assigned to receive amoxicillin twice daily for either 3 days or 5 days. Children were followed for 14 days. The primary outcome was treatment failure by day 6; noninferiority of the 3-day regimen to the 5-day regimen would be shown if the percentage of children with treatment failure in the 3-day group was no more than 1.5 times that in the 5-day group. Prespecified secondary analyses included assessment of treatment failure or relapse by day 14.

RESULTS

From March 29, 2016, to April 1, 2019, a total of 3000 children underwent randomization: 1497 children were assigned to the 3-day group, and 1503 to the 5-day group. Among children with day 6 data available, treatment failure had occurred in 5.9% in the 3-day group (85 of 1442 children) and in 5.2% (75 of 1456) in the 5-day group (adjusted difference, 0.7 percentage points; 95% confidence interval [CI], -0.9 to 2.4) - a result that satisfied the criterion for noninferiority of the 3-day regimen to the 5-day regimen. Among children with day 14 data available, 176 of 1411 children (12.5%) in the 3-day group and 154 of 1429 (10.8%) in the 5-day group had had treatment failure by day 6 or relapse by day 14 (between-group difference, 1.7 percentage points; 95% CI, -0.7 to 4.1). The percentage of children with serious adverse events was similar in the two groups (9.8% in the 3-day group and 8.8% in the 5-day group).

CONCLUSIONS

In HIV-uninfected Malawian children, treatment with amoxicillin for chest-indrawing pneumonia for 3 days was noninferior to treatment for 5 days. (Funded by the Bill and Melinda Gates Foundation; ClinicalTrials.gov number, NCT02678195.).

Approximate Bayesian Bootstrap procedures to estimate multilevel treatment effects in observational studies with application to type 2 diabetes treatment regimens

Randomized clinical trials are considered as the gold standard for estimating causal effects. Nevertheless, in studies that are aimed at examining adverse effects of interventions, randomized trials are often impractical because of ethical and financial considerations. In observational studies, matching on the generalized propensity scores was proposed as a possible solution to estimate the treatment effects of multiple interventions. However, the derivation of point and interval estimates for these matching procedures can become complex with non-continuous or censored outcomes. We propose a novel Approximate Bayesian Bootstrap algorithm that results in statistically valid point and interval estimates of the treatment effects with categorical outcomes. The procedure relies on the estimated generalized propensity scores and multiply imputes the unobserved potential outcomes for each unit. In addition, we describe a corresponding interpretable sensitivity analysis to examine the unconfoundedness assumption. We apply this approach to examine the cardiovascular safety of common, real-world anti-diabetic treatment regimens for type 2 diabetes mellitus in a large observational database.

[Predictive Mean Matching as an alternative imputation method to hot deck in Vigitel]

This study aimed to describe the estimated means for weight, height, and body mass index (BMI) according to two imputation methods, using data from Vigitel (Risk and Protective Factors Surveillance System for Chronic Non-Communicable Diseases Through Telephone Interview). This was a cross-sectional study that used secondary data from the Vigitel survey from 2006 to 2017. The two imputation methods used in the study were hot deck and Predictive Mean Matching (PMM). The weight and height variables imputed by hot deck were provided by Vigitel. Two models were conducted with PMM: (i) explanatory variables - city, sex, age in years, race/color, and schooling; (ii) explanatory variables - city, sex, and age in years. Weight and height were the outcome variables in the two models. PMM combines linear regression and random selection of the value for imputation. Linear prediction is used as a measure of distance between the missing value and the possible donors, thereby creating the virtual space with the candidate cases for yielding the value for imputation. One of the candidates from the pool is randomly selected, and its value is assigned to the missing unit. BMI was calculated by dividing weight in kilograms by height squared. The result shows the means and standard deviations for weight, height, and BMI according to imputation method and year. The estimates used the survey module from Stata, which considers the sampling effects. The mean values for weight, height, and BMI estimated by hot deck and PMM were similar. The results with the Vigitel data suggest the applicability of PMM to the set of health surveys.

Additive integer-valued data envelopment analysis with missing data: A multi-criteria evaluation approach

Traditional data envelopment analysis (DEA) models assume that all the inputs and outputs data are available. However, missing data is a common problem in data analysis. Although several scholars have developed techniques to conduct DEA with missing data, these techniques have some disadvantages. A multi-criteria evaluation approach is proposed to measure the efficiency of decision making units (DMUs) with missing data. In this approach, analysts first estimate the upper and lower bounds of DMUs’ efficiency using the proposed I-addIDEA-U models (interval additive integer-valued DEA models with undesirable outputs) that can be applied to address integer-valued variables and undesirable outputs. Then, DMUs’ “relative” efficiency is evaluated using the proposed “Halo + Hot deck” DEA method (if there is no correlation between variables) or regression DEA techniques (if there is a correlation between variables). Finally, the multi-index comprehensive evaluation method is applied to determine which scenario (the lower bound of efficiency, the “relative” efficiency, or the upper bound of efficiency) should be selected. With a case study, it is shown that the proposed multi-criteria evaluation approach is more effective than traditional approaches such as the mean imputation DEA method, the deletion DEA method, and the dummy entries DEA method.

Improvements in Uninsurance Estimates for Fully Imputed Cases in the Current Population Survey Annual Social and Economic Supplement

In 2019, the Current Population Survey Annual Social and Economic Supplement introduced updates to data processing, including to the imputation of health insurance for cases with no reported health insurance information. This article examines the impact on health insurance estimates of modernized imputation procedures that were part of a redesign of the Current Population Survey Annual Social and Economic Supplement. We use descriptive analysis and multinomial logistic regression to examine whether imputation biases estimates of health insurance coverage using data from the 2017 Current Population Survey Annual Social and Economic Supplement, which used legacy methods, and the 2017 Current Population Survey Annual Social and Economic Supplement Research File, which debuted the processing redesign. We find that cases with all of their health insurance information imputed using legacy methods were more likely to be uninsured or to be covered by multiple insurance types after adjusting for factors associated with having missing data. With the processing updates, fully imputed cases do not differ from other cases in their likelihood of being uninsured, having private coverage, having public coverage, or in having private and public coverage. Processing updates in the Current Population Survey Annual Social and Economic Supplement improved data quality by increasing the percent of people with any health insurance coverage and decreasing the percent of people with multiple types of coverage, especially among fully imputed cases.

Predictors of prostate-specific antigen testing in men aged ≥55 years: A cross-sectional study based on patient-reported outcomes

OBJECTIVES

To examine the predictors of prostate-specific antigen discussion with a physician and prostate-specific antigen testing in men aged ≥55 years.

METHODS

Utilizing the USA Health Information National Trends Survey, 4th Ed., a cross-sectional study from 2011 to 2014 was carried out to analyze the factors predicting prostate-specific antigen testing and discussion in men ≥55 years. Associations between each covariate and prostate-specific antigen discussion/testing were determined. Multivariable logistic regression models were used to determine clinically relevant predictors of prostate-specific antigen discussion/testing. Due to multiple comparisons, the Bonferroni correction was used.

RESULTS

A total of 2731 men included in the Health Information National Trends Survey were analyzed. Several socioeconomic parameters were found to increase the likelihood of men aged ≥55 years to undergo prostate-specific antigen testing: living with a spouse, a higher level of education (college graduate or above), a higher income (>$50 000 annually) and previous history of any cancer. In contrast, current smokers were less likely to undergo prostate-specific antigen testing. Having a prostate-specific antigen discussion with a physician was more likely for men surveyed in 2014, for men who were living with a spouse, who had a higher annual income (>$50 000 annually) and those with a history of any cancer.

CONCLUSIONS

Significant inequalities in prostate-specific antigen testing and discussion exist among men in the USA, mainly driven by socioeconomic factors. Ideally, prostate-specific antigen testing and discussion should be based on relevant clinical factors with a shared decision-making approach for every man. Therefore, a better understanding of the socioeconomic factors influencing prostate-specific antigen testing/discussions can inform strategies to reduce existing gaps in care.

Changes in Health Care Access, Behaviors, and Self-reported Health Among Low-income US Adults Through the Fourth Year of the Affordable Care Act

BACKGROUND

Health care access increased for low-income Americans under the Affordable Care Act (ACA). It is unknown whether these changes in access were associated with improved self-reported health.

OBJECTIVE

Determine changes in health care access, health behaviors, and self-reported health among low-income Americans over the first 4 years of the ACA, stratified by state Medicaid expansion status.

DESIGN

Interrupted time series and difference-in-differences analysis.

SUBJECTS

Noninstitutionalized US adults (18-64 y) in low income households (<138% federal poverty level) interviewed in the Behavioral Risk Factor Surveillance System 2011-2017 (N=505,824).

MEASURES

Self-reported insurance coverage, access to a primary care physician, avoided care due to cost; self-reported general health, days of poor physical health, days of poor mental health days, and days when poor health limited usual activities; self-reported health behaviors, use of preventive services, and diagnoses.

RESULTS

Despite increases in access, the ACA was not associated with improved physical or general health among low-income adults during the first 4 years of implementation. However, Medicaid expansion was associated with fewer days spent in poor mental health (-1.1 d/mo, 95% confidence interval: -2.1 to -0.5). There were significant changes in specific health behaviors, preventive service use, and diagnosis patterns during the same time period which may mediate the relationship between the ACA rollout and self-reported health.

CONCLUSION

In nationally-representative survey data, we observed improvements in mental but not physical self-reported health among low-income Americans after 4 years of full ACA implementation.

Comparability of modern contraceptive use estimates between a face-to-face survey and a cellphone survey among women in Burkina Faso

Introduction

The proliferation of cell phone ownership in Sub-Saharan Africa (SSA) presents the opportunity to collect public health indicators at a lower cost compared to face-to-face (FTF) surveys. This analysis assesses the equivalence of modern contraceptive prevalence estimates between a nationally representative FTF survey and a cell phone survey using random digit dialing (RDD) among women of reproductive age in Burkina Faso.

Methods

We analyzed data from two surveys conducted in Burkina Faso between December 2017 and May 2018. The FTF survey conducted by Performance Monitoring and Accountability (PMA2020) comprised a nationally representative sample of 3,556 women of reproductive age (15–49 years). The RDD survey was conducted using computer-assisted telephone interviewing and included 2,379 women of reproductive age.

Results

Compared to FTF respondents, women in the RDD sample were younger, were more likely to have a secondary degree and to speak French. RDD respondents were more likely to report using modern contraceptive use (40%) compared to FTF respondents (26%) and the difference remained unchanged after applying post-stratification weights to the RDD sample (39%). This difference surpassed the equivalence margin of 4%. The RDD sample also produced higher estimates of contraceptive use than the subsample of women who owned a phone in the FTF sample (32%). After adjusting for women’s sociodemographic factors, the odds of contraceptive use were 1.9 times higher (95% CI: 1.6–2.2) in the RDD survey compared to the FTF survey and 1.6 times higher (95% CI: 1.3–1.8) compared to FTF phone owners.

Conclusions

Modern contraceptive prevalence in Burkina Faso is over-estimated when using a cell phone RDD survey, even after adjusting for a number of sociodemographic factors. Further research should explore causes of differential estimates of modern contraceptive use by survey modes.

Perceptions of Facilitators and Barriers to Measuring and Improving Quality in Palliative Care Programs

OBJECTIVE

To examine perceptions of facilitators and barriers to quality measurement and improvement in palliative care programs and differences by professional and leadership roles.

METHODS

We surveyed team members in diverse US and Canadian palliative care programs using a validated survey addressing teamwork and communication and constructs for educational support and training, leadership, infrastructure, and prioritization for quality measurement and improvement. We defined key facilitators as constructs rated ≥4 (agree) and key barriers as those ≤3 (disagree) on 1 to 5 scales. We conducted multivariable linear regressions for associations between key facilitators and barriers and (1) professional and (2) leadership roles, controlling for key program and respondent factors and clustering by program.

RESULTS

We surveyed 103 respondents in 11 programs; 45.6% were physicians and 50% had leadership roles. Key facilitators across sites included teamwork, communication, the implementation climate (or environment), and program focus on quality improvement. Key barriers included educational support and incentives, particularly for quality measurement, and quality improvement infrastructure such as strategies, systems, and skilled staff. In multivariable analyses, perceptions did not differ by leadership role, but physicians and nurse practitioners/nurses/physician assistants rated most constructs statistically significantly more negatively than other team members, especially for quality improvement (6 of the 7 key constructs).

CONCLUSIONS

Although participants rated quality improvement focus and environment highly, key barriers included lack of infrastructure, especially for quality measurement. Building on these facilitators and measuring and addressing these barriers might help programs enhance palliative care quality initiatives' acceptability, particularly for physicians and nurses.

Accounting for missing data in statistical analyses: multiple imputation is not always the answer

Background

Missing data are unavoidable in epidemiological research, potentially leading to bias and loss of precision. Multiple imputation (MI) is widely advocated as an improvement over complete case analysis (CCA). However, contrary to widespread belief, CCA is preferable to MI in some situations.

Methods

We provide guidance on choice of analysis when data are incomplete. Using causal diagrams to depict missingness mechanisms, we describe when CCA will not be biased by missing data and compare MI and CCA, with respect to bias and efficiency, in a range of missing data situations. We illustrate selection of an appropriate method in practice.

Results

For most regression models, CCA gives unbiased results when the chance of being a complete case does not depend on the outcome after taking the covariates into consideration, which includes situations where data are missing not at random. Consequently, there are situations in which CCA analyses are unbiased while MI analyses, assuming missing at random (MAR), are biased. By contrast MI, unlike CCA, is valid for all MAR situations and has the potential to use information contained in the incomplete cases and auxiliary variables to reduce bias and/or improve precision. For this reason, MI was preferred over CCA in our real data example.

Conclusions

Choice of method for dealing with missing data is crucial for validity of conclusions, and should be based on careful consideration of the reasons for the missing data, missing data patterns and the availability of auxiliary information.

Association between the price of ultra-processed foods and obesity in Brazil

BACKGROUND AND AIMS

To estimate the relationship between the price of ultra-processed foods and prevalence of obesity in Brazil and examine whether the relationship differed according to socioeconomic status.

METHODS AND RESULTS

Data from the national Household Budget Survey from 2008/09 (n = 55 570 households, divided in 550 strata) were used. Weight and height of all individuals were used. Weight was measured by using portable electronic scales (maximum capacity of 150 kg). Height (or length) was measured using portable stadiometers (maximum capacity: 200 cm long) or infant anthropometers (maximum capacity: 105 cm long). Multivariate regression models (log-log) were used to estimate price elasticity. An inverse association was found between the price of ultra-processed foods (per kg) and the prevalence of overweight (Body mass index (BMI) ≥25 kg/m²) and obesity (BMI ≥30 kg/m²) in Brazil. The price elasticity for ultra-processed foods was -0.33 (95% CI: -0.46; -0.20) for overweight and -0.59 (95% CI: -0.83; -0.36) for obesity. This indicated that a 1.00% increase in the price of ultra-processed foods would lead to a decrease in the prevalence of overweight and obesity of 0.33% and 0.59%, respectively. For the lower income group, the price elasticity for price of ultra-processed foods was -0.34 (95% CI: -0.50; -0.18) for overweight and -0.63 (95% CI: -0.91; -0.36) for obesity.

CONCLUSION

The price of ultra-processed foods was inversely associated with the prevalence of overweight and obesity in Brazil, mainly in the lowest socioeconomic status population. Therefore, the taxation of ultra-processed foods emerges as a prominent tool in the control of obesity.

Variability of cost trajectories over the last year of life in patients with advanced breast cancer in the Netherlands

Objective

In breast cancer patients, treatment at the end of life accounts for a major share of medical spending. However, little is known about the variability of cost trajectories between patients. This study aims to identify underlying latent groups of advanced breast cancer patients with similar cost trajectories over the last year before death.

Methods

Data from deceased advanced breast cancer patients, diagnosed between 2010 and 2017, were retrieved from the Southeast Netherlands Advanced Breast Cancer (SONABRE) Registry. Costs of hospital care over the last twelve months before death were analyzed, and the variability of longitudinal patterns between patients were explored using group-based trajectory modeling. Descriptive statistics and multinomial logistic regression were applied to investigate differences between the identified latent groups.

Results

We included 558 patients. Over the last twelve months before death, mean hospital costs were €2,255 (SD = €492) per month. Costs increased over the last five months and reached a maximum of €3,614 in the last month of life, driven by hospital admissions, while spending for medication declined over the last three months of life. Based on patients’ individual cost trajectories, we identified six latent groups with distinct longitudinal patterns, of which only two showed a marked increase in costs over the last twelve months before death. Latent groups were constituted of heterogeneous patients, and clinical characteristics explained membership only to a limited extent.

Conclusions

The average costs of advanced breast cancer patients increased towards the end of life. However, we uncovered several latent groups of patients with divergent cost trajectories, which did not reflect the overall increasing trend. The mechanisms underlying the variability in cost trajectories warrants further research.

Cumulative cortisol exposure increases during the academic term: Links to performance-related and social-evaluative stressors

OBJECTIVE

To examine whether cumulative cortisol production changes during a period of increased demands when cortisol and stress are assessed concurrently. The study also compared stress perceptions vs. cumulative stressful events on their respective association with cortisol output. Finally, it explored whether certain types of stressful events, those involving school/job performance or social-evaluative threat, were linked to cortisol levels across multiple weeks.

METHOD

The current study assessed cumulative cortisol production via hair sample in 56 undergraduates (88 % female) during both lower stress (summer break) and higher stress (academic term) periods. During the latter, both negative events (checklist) and stress perceptions were assessed weekly, and these reports were aggregated across the 10-weeks to minimize retrospective bias.

RESULTS

Cortisol levels in hair samples were significantly higher (d = 0.84) during the academic term (M = 14.24 pg/mg, SD = 11.36) compared to summer break (M = 8.00 pg/mg, SD = 4.14), suggesting greater cumulative exposure to cortisol. Although perceived stress was not associated with cortisol levels (r_partial(53) = .10, p = 0.46), exposure to more stressful events (r_partial(53) = .27, p = 0.047), particularly events involving academic demands (r_partial(53) = .37, p = 0.006), or negative evaluation/social rejection (r_partial(53) = .27, p = 0.045), was positively associated with cumulative cortisol exposure.

CONCLUSIONS

This study demonstrates that cortisol levels in hair may be linked to cumulative exposure to stressors when measured concurrently (3 months), and that stressful events, rather than perceptions, are reflected in HPA axis activity. Real-world stressors involving performance demands and social-evaluative threat accumulate to enhance cortisol production, consistent with their acute HPA effects in the lab. Hair samples may provide a window into the past by allowing researchers to feasibly assess cortisol production before, during, and after the onset of a chronic stressor.

Longitudinal pain and pain interference in long-term survivors of childhood cancer: A report from the Childhood Cancer Survivor Study

BACKGROUND

The objective of this study was to characterize the prevalence and risk of pain, pain interference, and recurrent pain in adult survivors of childhood cancer in comparison with siblings.

METHODS

This study analyzed longitudinal data from survivors (n = 10,012; 48.7% female; median age, 31 years [range, 17-57 years]; median time since diagnosis, 23 years) and siblings (n = 3173) from the Childhood Cancer Survivor Study. Survivors were diagnosed between 1970 and 1986 at 1 of 26 participating sites. Associations between risk factors (demographics, cancer-related factors, and psychological symptoms) and pain, pain interference, and recurrent pain (5 years apart) were assessed with multinomial logistic regression. Path analyses examined cross-sectional associations between risk factors and pain outcomes.

RESULTS

Twenty-nine percent of survivors reported moderate to severe pain, 20% reported moderate to extreme pain interference, and 9% reported moderate to severe recurrent pain. Female sex, a sarcoma/bone tumor diagnosis, and severe/life-threatening chronic medical conditions were associated with recurrent pain. Depression and anxiety were associated with increased risk for all pain outcomes. Poor vitality mediated the effects of anxiety on high pain and pain interference (root mean square error of approximation, 0.002).

CONCLUSIONS

A large proportion of adult survivors report moderate to severe pain and pain interference more than 20 years after their diagnosis. Increased screening and early intervention for pain interference and recurrent pain are warranted.

Modern Soft-Sensing Modeling Methods for Fermentation Processes

For effective monitoring and control of the fermentation process, an accurate real-time measurement of important variables is necessary. These variables are very hard to measure in real-time due to constraints such as the time-varying, nonlinearity, strong coupling, and complex mechanism of the fermentation process. Constructing soft sensors with outstanding performance and robustness has become a core issue in industrial procedures. In this paper, a comprehensive review of existing data pre-processing approaches, variable selection methods, data-driven (black-box) soft-sensing modeling methods and optimization techniques was carried out. The data-driven methods used for the soft-sensing modeling such as support vector machine, multiple least square support vector machine, neural network, deep learning, fuzzy logic, probabilistic latent variable models are reviewed in detail. The optimization techniques used for the estimation of model parameters such as particle swarm optimization algorithm, ant colony optimization, artificial bee colony, cuckoo search algorithm, and genetic algorithm, are also discussed. A comprehensive analysis of various soft-sensing models is presented in tabular form which highlights the important methods used in the field of fermentation. More than 70 research publications on soft-sensing modeling methods for the estimation of variables have been examined and listed for quick reference. This review paper may be regarded as a useful source as a reference point for researchers to explore the opportunities for further enhancement in the field of soft-sensing modeling.

A new approach for handling missing correlation values for meta-analytic structural equation modeling: Corboundary R package

With increased use of multivariate meta-analysis in numerous disciplines, where structural relationships among multiple variables are examined, researchers often encounter a particular challenge due to missing information. The current research concerns missing correlations (rs) in the correlation matrix of m variables (R m × m ) and establish more informative and empirical prior distributions for missing rs in R m × m . In particular, the method for deriving mathematically/analytically boundaries for missing rs in relation to other adjacent rs in R m × m , while satisfying conditions for a valid R m × m (i.e., a symmetric and positive semidefinite correlation matrix containing real numbers between -1 and 1) is first discussed. Then, a user-defined R package for constructing the empirical distributions of boundaries for rs in R m × m is demonstrated with an example. Furthermore, the applicability of constructing empirical boundaries for rs in R m × m beyond multivariate meta-analysis is discussed.

Parental Perceptions of Culturally Sensitive Care and Well-Child Visit Quality

OBJECTIVE

Incorporating culturally sensitive care into well-child visits may help address pediatric preventive care disparities faced by racial and ethnic minorities, families with limited English proficiency, and immigrants. We explored parents' perspectives about the extent to which their children's pediatric care is culturally sensitive and potential associations between culturally sensitive care and well-child visit quality.

METHODS

We conducted cross-sectional surveys with parents attending a well-child visit for a child ages 3 to 48 months. To measure culturally sensitive care, we created a composite score by averaging 8 subscales from an adapted version of the Clinicians' Cultural Sensitivity Survey. We assessed well-child visit quality through the Promoting Healthy Development Survey. Multivariate linear regression was used to understand associations between demographic characteristics and parent-reported culturally sensitive care. We used multivariate logistic regression to examine associations between culturally sensitive care and well-child visit quality.

RESULTS

Two hundred twelve parents (71% of those approached) completed the survey. Parents born abroad, compared with those born in the United States, reported significantly higher culturally sensitive care scores (+0.21; confidence interval [CI]: 0.004, 0.43). Haitian parents reported significantly lower culturally sensitive care scores compared with non-Hispanic white parents (-0.49; CI: -0.89, -0.09). Parent-reported culturally sensitive care was significantly associated with higher odds of well-child visit quality including receipt of anticipatory guidance (adjusted odds ratio: 2.68; CI: 1.62, 4.62) and overall well-child visit quality (adjusted odds ratio: 2.54; CI: 1.59, 4.22).

CONCLUSIONS

Consistent with prior research of adult patients, this study demonstrates an association between parent-reported culturally sensitive care and well-child visit quality. Future research should explore best practices to integrating culturally sensitive care in pediatric preventive health care settings.

Changes In Health Insurance Coverage, Access To Care, And Income-Based Disparities Among US Adults, 2011-17

The Affordable Care Act increased insurance coverage and access to care, according to numerous national studies. However, the administration of President Donald Trump implemented several policies that may have affected the act's effectiveness. It is unknown what effect these changes had on access to care. We used survey data for 2011-17 from the Behavioral Risk Factor Surveillance System to assess changes access to care among nonelderly adults from before to after the change in administration in 2017. We found that the proportion of adults who were uninsured or avoided care because of cost increased by 1.2 percentage points and 1.0 percentage points, respectively, during 2017. These changes were greater among respondents who had household incomes below 138 percent of the federal poverty level, resided in states that did not expand eligibility for Medicaid, or both. At the population level, our findings imply that approximately two million additional US adults experienced these outcomes at the end of 2017, compared to the end of 2016.

Fish Consumption for the Adult Population of Hawai'i, Collected with a Self-Reported Household Survey

The population of Hawai'i has traditionally been high in average fish consumption when compared to the national average. However, information is lacking on patterns of fish consumption among subpopulations. Data on fish consumption in the last 30 days from 11,293 adults was collected with the use of the Hawai'i Health Survey (an annual telephone survey of households and household members) during the years 2007 and 2008 and weighted to represent the adult population of Hawai'i. The US Department of Agriculture's, Environmental Protection Agency, and the United States Food and Drug Administration, recommend 8-12 ounces of fish per week for associated health benefits. Present estimates of fish consumption were skewed to the right (mean 10.5 and median 7.9 ounces) with some adults eating large amounts of fish per day and frequently. It may be of concern, given high amounts of methylmercury in select fish, that 13.7% of adults were eating fish 20 or more times per month. In addition, the serving size increased with increasing number of times per week fish was eaten. The subpopulation variables examined included age, sex, marital status, education, ethnicity, poverty, and demographics representing the adults of Hawai'i. The prevalence of adults consuming 8 or more ounces was highest for other than Honolulu counties, men, ages 18-74, married, with at least one year of college, Filipino and Native Hawaiian ethnicities (White, Japanese, Chinese, and All Others were also examined), employed, and adults living with higher income levels (lower poverty). However, only an estimated 46.9% of adults in Hawai'i were eating the eight or more ounces of fish weekly. Prevalence was lowest for women, Chinese, age >74 years, and the City and County of Honolulu. Adults who rated their general health better (excellent/good compared to fair/poor) were more likely to consume 8 or more ounces of fish per week, significantly for men. Men who rated their physical health higher and their mental health higher (via Optum SF™ Health Surveys) were associated with higher prevalence of consuming 8 or more ounces of fish per week. Higher consumption of the beneficial omega-3 fatty acids associated with fish low in methylmercury needs to be encouraged for those not meeting the recommended ounces of fish per week, particularly women, Chinese, older age groups, and adults living below the poverty level.

R-Ensembler: A greedy rough set based ensemble attribute selection algorithm with kNN imputation for classification of medical data

BACKGROUND AND OBJECTIVE

Retrieving meaningful information from high dimensional dataset is an important and challenging task. Normally, medical dataset suffers from several issues such as curse of dimensionality problem, uncertainty, presence of missing values, non-relevant and redundant attributes, etc. Any machine learning technique applied on such data (without any preprocessing) by and large takes a considerable amount of computational time and may degrade the performance of the model.

METHODS

In this article, R-Ensembler, a parameter free greedy ensemble attribute selection method is proposed adopting the concept of rough set theory by using the attribute-class, attribute-significance and attribute-attribute relevance measures to select a subset of attributes which are most relevant, significant and non-redundant from a pool of different attribute subsets in order to predict the presence or absence of different diseases in medical dataset. The main role of the proposed ensembler is to combine multiple subsets of attributes produced by different rough set filters and to produce an optimal subset of attributes for subsequent classification task. A novel n number of set intersection method is also proposed to reduce the biasness during the time of attribute selection process. Before selecting the minimal attribute set from a given data by the proposed R-Ensembler method, the dataset is preprocessed by the k nearest neighbour (kNN) imputation method for missing value treatment.

RESULTS

Experiments are carried out on seven benchmark medical datasets collected from University of California at Irvine (UCI) repository. The performance of the proposed ensemble method is compared with five state-of-the-art attribute selection algorithms, results of which are measured using three benchmark classifiers viz., Naïve Bayes, decision trees and random forest. Experimental results clearly justify the superiority of the proposed R-Ensembler method over other attribute selection algorithms. Results of paired t-test performed on average accuracies produced by different classifiers simulated on the reduced data sets achieved by the proposed and counter part attribute selection methods confirm the statistical significance of the better reduced attribute subsets achieved by the proposed R-Ensembler method compared to others.

CONCLUSION

The proposed ensemble method turned out to be very effective for selecting high relevant, high significant and less redundant attributes from a pool of different subsets of attributes.

Changes Over Time in Disparities in Health Behaviors and Outcomes by Race in Allegheny County: 2009 to 2015

Racial/ethnic disparities in health behaviors and disease outcomes on the national level have persisted over time despite overall improvements in public health. To better understand the changes over time in racial/ethnic health disparities at the county level, we examined the Allegheny County Health Survey (ACHS) for Pittsburgh, PA and the surrounding area, which was conducted in 2009/2010 and 2015/2016 using random digit dialing of residents aged 18 and older. The prevalence rates and rate ratios at each time period were calculated using survey weights and general linear models. The change in prevalence over time was calculated using race-time interaction terms. The results showed a significant improvement in asthma, stroke, cholesterol, and fair or poor health disparities as well as persistent disparities in diabetes and hypertension after adjustment for socioeconomic factors. The change over time in the prevalence of fair or poor health in black compared to white respondents was significant, with absolute improvement of approximately 5% versus < 1%, respectively (p = 0.01). These findings demonstrate that some disparities improved while others persist, noting the importance of monitoring the changes over time at the local health department level.

Early Second Trimester Maternal Serum Steroid-Related Biomarkers Associated with Autism Spectrum Disorder

Epidemiologic studies link increased autism spectrum disorder (ASD) risk to obstetrical conditions associated with inflammation and steroid dysregulation, referred to as prenatal metabolic syndrome (PNMS). This pilot study measured steroid-related biomarkers in early second trimester maternal serum collected during the first and second trimester evaluation of risk study. ASD case and PNMS exposure status of index offspring were determined through linkage with autism registries and birth certificate records. ASD case (N = 53) and control (N = 19) groups were enriched for PNMS exposure. Higher estradiol and lower sex hormone binding globulin (SHBG) were significantly associated with increased ASD risk. Study findings provide preliminary evidence to link greater placental estradiol activity with ASD and support future investigations of the prenatal steroid environment in ASD.

Changes in Insurance Coverage and Access to Care for Young Adults in 2017

PURPOSE

Previous work has demonstrated the Affordable Care Act (ACA) increased young adults' health care access during its first years. However, it is unclear if these trends continued through 2017; recent policies enacted by the Trump administration may have decreased the ACA's effectiveness. Our purpose was to determine changes in young adults' health care access during the transition from Obama to Trump administrations.

METHODS

Data on noninstitutionalized U.S. young adults (18-24 years) was obtained from the Behavioral Risk Factor Surveillance System 2011-2017 (N = 173,848). We used interrupted time series and difference-in-differences analysis to quantify changes in self-reported insurance coverage, access to a primary care physician, and unmet care because of cost from 2013 to 2017.

RESULTS

Young adults' health care access continued to improve through 2016; for instance, the percentage of respondents experiencing uninsurance declined by 8.7 points from 2013 to 2016 (95% confidence interval [CI] -9.4 to -8.0). However, these trends began to reverse and from 2016 to 2017, the percentage of young adults who experienced uninsurance increased by 1.4 points (95% CI .6-2.1), not having a personal doctor increased by 1.1 points (95% CI .2-2.0), and unmet care because of cost increased by 1.0 points (95% CI .3-1.7). The 2017 declines in access were concentrated in states which did not expand Medicaid and in households earning above 138% of federal poverty level.

CONCLUSIONS

Health care access declined for young adults in 2017, after several years of improvements. These changes correspond with recent policy actions, which may have weakened the ACA's reforms.

Association between facility volume and mortality of patients with classic Hodgkin lymphoma

BACKGROUND

Prior studies in oncology have shown that a higher annual facility patient volume is associated with reduced mortality. Because classic Hodgkin lymphoma is uncommon but highly curable, this study used the National Cancer Database (2003-2014) to analyze whether such a relationship exists for this disease.

METHODS

The facilities were classified by quartiles, and random intercepts were used to account for clustering of patients within facilities. A Cox regression model was used to determine the volume-outcome relationship.

RESULTS

There were 47,633 patients with classic Hodgkin lymphoma treated at 1310 facilities. The first quartile (Q1), which included 58.4% of the facilities, treated 3 or fewer patients per year, whereas the fourth quartile (Q4), which included 5.9% of the facilities, treated more than 9 patients per year. Compared with the patients treated at Q4 facilities, those treated at lower quartile facilities had a higher risk of death (hazard ratio for the third quartile [HR], 1.19; 95% confidence interval [CI], 1.1-1.29; HR for the second quartile, 1.28; 95% CI, 1.19-1.38; HR for Q1, 1.29; 95% CI, 1.2-1.39) after adjustments for all other factors (P < .0001). Compared with facilities treating 10 patients per year, facilities treating 40 patients per year had approximately 27% lower overall mortality rates.

CONCLUSIONS

Patients with classic Hodgkin lymphoma treated at high-volume centers had lower overall mortality than those treated at lower volume centers. Because this is a highly curable malignancy, such differences may suggest a benefit from referral to higher volume facilities or the emulation of their care models.

Prospective association between several dietary scores and risk of cardiovascular diseases: Is the Mediterranean diet equally associated to cardiovascular diseases compared to National Nutritional Scores?

BACKGROUND

Mediterranean diet has been consistently negatively associated with cardiovascular diseases (CVD) but the superiority compared to official nutritional guidelines has not been tested yet. Our objective was to prospectively investigate the association between several nutritional scores and incidence of cardiovascular diseases.

METHODS AND FINDINGS

A total of 94,113 participants from the NutriNet-Santé cohort were followed between 2009 and 2018. The participants have completed at least three 24 h dietary records during the first two-years of follow-up to compute nutritional scores reflecting adherence to the Mediterranean diet (MEDI-LITE), American dietary guidelines (AHEI-2010) and French dietary guidelines (mPNNS-GS). Sex-specific quartiles (Q) of scores were computed. Multivariable Cox proportional hazards models were used to estimate the associations between scores and incidence of CVD, documented using Hazard Ratio (HR) and 95% confidence intervals (95%CI). Thus, 1399 incident CVD events occurred during the follow-up (mean follow-up = 5.4 years). Comparing Q4 versus Q1 quartile, HR for the MEDI-LITE and AHEI-2010 were 0.79 (95% CI: 0.67-0.93, P-trend = .004) and 0.75 (95% CI: 0.63-0.89, P-trend = .002) respectively. These associations remained similar when removing early cases of CVD, when analyses were restricted to participants with >6 dietary records and when considering transient ischemic attacks. In this last case, association between CVD' risk and mPNNS-GS become significant.

CONCLUSIONS

A better nutritional quality of diet is overall associated with lower risk of CVD. The future version of the PNNS-GS, based on the updated version of the French dietary guidelines, should strengthen the CVD protective effect of French recommendations.

An ex vivo tissue model of cartilage degradation suggests that cartilage state can be determined from secreted key protein patterns

The pathophysiology of osteoarthritis (OA) involves dysregulation of anabolic and catabolic processes associated with a broad panel of proteins that ultimately lead to cartilage degradation. An increased understanding about these protein interactions with systematic in vitro analyses may give new ideas regarding candidates for treatment of OA related cartilage degradation. Therefore, an ex vivo tissue model of cartilage degradation was established by culturing tissue explants with bacterial collagenase II. Responses of healthy and degrading cartilage were analyzed through protein abundance in tissue supernatant with a 26-multiplex protein profiling assay, after exposing the samples to a panel of 55 protein stimulations present in synovial joints of OA patients. Multivariate data analysis including exhaustive pairwise variable subset selection identified the most outstanding changes in measured protein secretions. MMP9 response to stimulation was outstandingly low in degrading cartilage and there were several protein pairs like IFNG and MMP9 that can be used for successful discrimination between degrading and healthy samples. The discovered changes in protein responses seem promising for accurate detection of degrading cartilage. The ex vivo model seems interesting for drug discovery projects related to cartilage degradation, for example when trying to uncover the unknown interactions between secreted proteins in healthy and degrading tissues.

Correlation-Based Ensemble Feature Selection Using Bioinspired Algorithms and Classification Using Backpropagation Neural Network

A framework for clinical diagnosis which uses bioinspired algorithms for feature selection and gradient descendant backpropagation neural network for classification has been designed and implemented. The clinical data are subjected to data preprocessing, feature selection, and classification. Hot deck imputation has been used for handling missing values and min-max normalization is used for data transformation. Wrapper approach that employs bioinspired algorithms, namely, Differential Evolution, Lion Optimization, and Glowworm Swarm Optimization with accuracy of AdaBoostSVM classifier as fitness function has been used for feature selection. Each bioinspired algorithm selects a subset of features yielding three feature subsets. Correlation-based ensemble feature selection is performed to select the optimal features from the three feature subsets. The optimal features selected through correlation-based ensemble feature selection are used to train a gradient descendant backpropagation neural network. Ten-fold cross-validation technique has been used to train and test the performance of the classifier. Hepatitis dataset and Wisconsin Diagnostic Breast Cancer (WDBC) dataset from University of California Irvine (UCI) Machine Learning repository have been used to evaluate the classification accuracy. An accuracy of 98.47% is obtained for Wisconsin Diagnostic Breast Cancer dataset, and 95.51% is obtained for Hepatitis dataset. The proposed framework can be tailored to develop clinical decision-making systems for any health disorders to assist physicians in clinical diagnosis.

The Association Between Child Abuse and Attempted Suicide

Abstract.Background: Child abuse and suicide among the young population is a serious and prevalent problem. Many studies have demonstrated that people with adverse childhood experiences, such as child abuse, are likely to develop suicidal behavior. This study evaluates the connection between child abuse and suicidal behavior in the Hong Kong community where incidents of child abuse have been on the rise over the past decade. Aims: To determine the association between child abuse and attempted suicide in the child population of Hong Kong using hospital electronic medical records system. Method: From January 1, 1995 to July 31, 2016, patients with admission age < 18 years with the diagnosis of child abuse or influenza infection (comparison group) were included in this study (n = 54,256). In secondary data analysis, an association was found between children who had experienced child abuse and the outcome measure of hospital admission for attempted suicide compared with influenza infection. Results: The adjusted hazard ratio of attempting suicide in children who experienced sexual abuse and physical abuse compared with the influenza-infected group was 6.48 (95% CI [4.56, 9.19]) and 4.83 (95% CI [3.67, 6.34]). The age at onset of adverse incidents was negatively associated with the attempted suicide timing. Female patients had a 1.64 higher risk of repeating attempted suicide. In addition, nearly 5% of children who had experienced child abuse attempted suicide in the 10 years after their admission, and more than 36% of patients had a record of repeated suicide attempt in the 20 years after the initial admission. Limitations: The accuracy of the diagnosis, selection bias, insufficiency of study period, Berkson's bias, incomplete socioeconomic status, as well as the absence of psychiatric diagnosis are the limitations. Conclusion: Our results indicate that there is a significant association between child abuse and suicide attempts in Hong Kong. If confirmed, the study (a) demonstrates that hospital admission records are a critical source for identifying children with a high risk for suicidal behavior; (b) may inform policy makers that additional and long-term intervention programs should be provided to children so as to reduce subsequent suicide attempts.

Multiple-Partner Fertility and Cohort Change in the Prevalence of Half-Siblings

The transformation of the American family under the second demographic transition has created more opportunities for parents to have children with multiple partners, but data limitations have hampered prevalence estimates of multiple-partner fertility from the perspective of children. This study uses nationally representative data from the 1979 and 1997 cohorts of the National Longitudinal Surveys of Youth to examine cohort change in children's exposure to multiple-partner fertility. We find that one in five children in the 1979 cohort had at least one half-sibling by their 18th birthday, and the prevalence grew to more than one in four children by the 1997 cohort. A strong educational gradient in exposure to half-siblings persists across both cohorts, but large racial/ethnic disparities have narrowed over time. Using demographic decomposition techniques, we find that change in the racial/ethnic and socioeconomic composition of the U.S. population cannot explain the growth in exposure to half-siblings. We conclude by discussing the shifting patterns of fertility and family formation associated with sibling complexity and considering the implications for child development and social stratification.