Generative adversarial networks for imputing missing data for big data clinical research.

BACKGROUND: Missing data is a pervasive problem in clinical research. Generative adversarial imputation nets (GAIN), a novel machine learning data imputation approach, has the potential to substitute missing data accurately and efficiently but has not yet been evaluated in empirical big clinical datasets. OBJECTIVES: This study aimed to evaluate the accuracy of GAIN in imputing missing values in large real-world clinical datasets with mixed-type variables. The computation efficiency of GAIN was also evaluated. The performance of GAIN was compared with other commonly used methods, MICE and missForest. METHODS: Two real world clinical datasets were used. The first was that of a cohort study on the long-term outcomes of patients with diabetes (50,000 complete cases), and the second was of a cohort study on the effectiveness of a risk assessment and management programme for patients with hypertension (10,000 complete cases). Missing data (missing at random) to independent variables were simulated at different missingness rates (20, 50%). The normalized root mean square error (NRMSE) between imputed values and real values for continuous variables and the proportion of falsely classified (PFC) for categorical variables were used to measure imputation accuracy. Computation time per imputation for each method was recorded. The differences in accuracy of different imputation methods were compared using ANOVA or non-parametric test. RESULTS: Both missForest and GAIN were more accurate than MICE. GAIN showed similar accuracy as missForest when the simulated missingness rate was 20%, but was more accurate when the simulated missingness rate was 50%. GAIN was the most accurate for the imputation of skewed continuous and imbalanced categorical variables at both missingness rates. GAIN had a much higher computation speed (32 min on PC) comparing to that of missForest (1300 min) when the sample size is 50,000. CONCLUSION: GAIN showed better accuracy as an imputation method for missing data in large real-world clinical datasets compared to MICE and missForest, and was more resistant to high missingness rate (50%). The high computation speed is an added advantage of GAIN in big clinical data research. It holds potential as an accurate and efficient method for missing data imputation in future big data clinical research. TRIAL REGISTRATION: ClinicalTrials.gov ID: NCT03299010 ; Unique Protocol ID: HKUCTR-2232.

A pilot study on the validity and psychometric properties of the electronic EQ-5D-5L in routine clinical practice.

BACKGROUND: Electronic measurement of health-related quality of life (HRQOL) may facilitate timely and regular assessments in routine clinical practice. This study evaluated the validity and psychometric properties of an electronic version of the EQ-5D-5L (e-EQ-5D-5L) in Chinese patients with chronic knee and/or back problems. METHODS: 151 Chinese subjects completed an electronic version of the Chinese (Hong Kong) EQ-5D-5L when they attended a primary care or orthopedics specialist out-patient clinic in Hong Kong. They also completed the Chinese Western Ontario and McMaster University Osteoarthritis Index (WOMAC), a Pain Rating Scale, and a structured questionnaire on socio-demographics, co-morbidities and health service utilization. 32 subjects repeated the e-EQ-5D-5L two weeks after the baseline. 102 subjects completed e-EQ-5D-5L and 99 completed the Global Rating on Change Scale at three-month clinic follow up. Construct validity was assessed by the association of EQ-5D-5L scores with external criterion of WOMAC scores. We tested mean differences of WOMAC scores between adjacent response levels of the EQ-5D-5L dimensions by one-way ANOVA, test-retest reliability by intra-class correlation, sensitivity by known group comparisons and responsiveness by changes in EQ-5D-5L scores over 3 months. RESULTS: There was an association between EQ-5D-5L and WOMAC scores. Mean WOMAC scores increased with the increase in adjacent response levels of EQ-5D-5L dimensions. Test-retest intraclass correlation coefficient (ICC) of EQ-5D-5L utility and EQ-VAS scores were 0.76 and 0.83, respectively, indicating good reliability. There were significant differences in the proportions reporting limitations in the EQ-5D-5L dimensions, the utility and VAS scores between the mild and severe pain groups (utility = 0.28, p = 0.001; VAS = 11.46, p < 0.001), and between primary care and specialist out-patient clinic patients (utility = 0.15, p = 0.001; VAS = 10.21, p < 0.001), supporting sensitivity. Among those reporting 'better' global health at three-months, their EQ-5D-5L utility and EQ-VAS scores were significantly increased from baseline (utility = 0.18, p < 0.001; VAS = 10.75, p = 0.005). CONCLUSIONS: The electronic version of the EQ-5D-5L is valid, reliable, sensitive and responsive in the measurement of HRQOL in Chinese patients with chronic knee or back pain in routine clinical practice.

The validity, reliability, sensitivity and responsiveness of a modified Patient Enablement Instrument (PEI-2) as a tool for serial measurements of health enablement.

BACKGROUND: Patient enablement is a core tenet of patient-centred and holistic primary care. The Patient Enablement Instrument (PEI) is a transitional measure limited in its ability to measure changes over time. A modified version, PEI-2, has been developed to measure enablement at a given time-point without comparison to a recalled baseline. OBJECTIVE: To assess the validity, reliability, sensitivity and responsiveness of PEI-2. METHODS: PEI-2 was modified from the Chinese PEI to assess enablement over 4 weeks in a prospective cohort study nested within a community support programme [Trekkers Family Enhancement Scheme (TFES)] in Hong Kong. Construct validity was assessed by factor analysis and convergent validity by Spearman's correlations with health-related quality of life and depressive symptoms. Internal reliability was assessed using Cronbach's alpha. Test-retest reliability was assessed by intraclass correlation (ICC), responsiveness by 12-24-month change in PEI-2 score and sensitivity by differences in change of PEI-2 score between TFES participants and a control group. RESULTS: PEI-2 demonstrated construct validity with all items loading on one factor (factor loadings >0.7). Convergent validity was confirmed by significant correlations with 12-item Short Form Questionnaire, version 2 (r = 0.1089-0.1919) and Patient Health Questionnaire-9 (r = -0.2030). Internal reliability was high (Cronbach's alpha = 0.9095) and test-retest reliability moderate (ICC = 0.520, P = 0.506). Significant improvements in PEI-2 scores among the TFES group suggested good responsiveness (P < 0.001). The difference in change of PEI-2 scores between TFES and control was significant (P = 0.008), indicating good sensitivity. CONCLUSIONS: This study supports the validity, reliability, sensitivity and responsiveness of PEI-2 in measuring changes in enablement, making it a promising tool for evaluating enablement in cohort and intervention studies.

Trend in health-related quality of life and health utility and their decrements due to non-communicable diseases and risk factors: analysis of four population-based surveys between 1998 and 2015.

PURPOSE: To revisit the population norms of health-related quality of life (HRQoL) and health utility for the Hong Kong general population, compare these scores over past health surveys, and assess the association of scores with non-communicable diseases (NCDs) and their risk factors. METHODS: HRQoL data measured by the standard Short Form 12 Health Survey-version 2 (SF-12v2) were extracted from the surveys in 1998, 2003/2004, 2008/2009 and 2014/2015. SF-12v2 data were mapped to the Short-form 6-dimension (SF-6D) preference-based measure to generate the health utility scores. Population weighting based on the sex and age in the second quarter of 2015 was applied when generating population normative values. Linear regression models were fitted to assess the effect of the number of NCDs and modifiable lifestyle factors on HRQoL and health utility. RESULTS: The general population mean scores of SF-12v2 domains and SF-6D in 2014/15 were higher compared to past surveys. Linear increases in General Health, Vitality and Mental Health domains were observed from 1998 to 2014/15. More doctor-diagnosed NCDs, insufficient physical activity and fruit/vegetable consumption, poor sleep quality and insufficient or excessive amount of sleep (< 6/≥ 10 h) were all associated with worse physical- and mental-related HRQoL and health utility. CONCLUSION: This study compared HRQoL and health utility in the Hong Kong general population derived from multiple surveys and found an improving trend over twenty years. More NCDs were associated with worse HRQoL. It is suggested that promoting adequate physical activity, consumption of fruit/vegetable and 6-9 h of sleep could improve health.

Improving medicines management for people with dementia in primary care: a qualitative study of healthcare professionals to develop a theory-informed intervention.

BACKGROUND: People with dementia (PwD) face unique challenges with medicines management, yet little is known about these challenges from the perspectives of primary healthcare professionals, particularly general practitioners (GPs) and community pharmacists. Few medicines management interventions have been developed which are aimed at community-dwelling PwD. This study sought to develop an intervention to improve medicines management for PwD in primary care using a theory-informed approach. METHODS: Semi-structured interviews were conducted with GPs (n = 15) and community pharmacists (n = 15) to explore participants' views and experiences of medicines management for PwD, and their perceptions of barriers and facilitators to successful medicines management for PwD. The 14-domain Theoretical Domains Framework was the underpinning theoretical guide, allowing key theoretical domains to be identified and mapped to behaviour change techniques (BCTs) which are considered the 'active ingredients' of an intervention. Draft interventions were developed to operationalise selected BCTs and were presented to GPs and community pharmacists during task groups. Final selection of an intervention for feasibility testing was guided by feedback provided during these task groups and through application of the APEASE (Affordability, Practicability, Effectiveness/cost-effectiveness, Acceptability, Side-effects/safety, Equity) criteria. RESULTS: Participants expressed a number of concerns about medicines management for PwD, particularly monitoring adherence to medication regimens and conducting medication review. Two draft interventions comprising selected BCTs ('Modelling or demonstration of behaviour'; 'Salience of consequences'; 'Health consequences'; 'Social and environmental consequences'; 'Action planning'; Social support or encouragement', 'Self-monitoring of behaviour') were developed, each targeting GPs and community pharmacists. Following the task groups and discussions within the research team, the community pharmacy-based intervention was selected for future feasibility testing. The intervention will target community pharmacists to conduct a medication review (incorporating an adherence check) with a PwD, delivered as an online video demonstrating key behaviours. The video will include feedback emphasising positive outcomes of performing the behaviours. Action planning and a quick reference guide will be used as complementary intervention components. CONCLUSIONS: A community pharmacist-based intervention has been developed targeting medicines management for PwD in primary care using a systematic, theory-informed approach. Future work will determine the usability and acceptability of implementing this intervention in clinical practice.

Lung cancer screening: does pulmonary nodule detection affect a range of smoking behaviours?

BACKGROUND: Lung cancer screening can reduce lung cancer mortality by 20%. Screen-detected abnormalities may provide teachable moments for smoking cessation. This study assesses impact of pulmonary nodule detection on smoking behaviours within the first UK trial of a novel auto-antibody test, followed by chest x-ray and serial CT scanning for early detection of lung cancer (Early Cancer Detection Test-Lung Cancer Scotland Study). METHODS: Test-positive participants completed questionnaires on smoking behaviours at baseline, 1, 3 and 6 months. Logistic regression compared outcomes between nodule (n = 95) and normal CT groups (n = 174) at 3 and 6 months follow-up. RESULTS: No significant differences were found between the nodule and normal CT groups for any smoking behaviours and odds ratios comparing the nodule and normal CT groups did not vary significantly between 3 and 6 months. There was some evidence the nodule group were more likely to report significant others wanted them to stop smoking than the normal CT group (OR across 3- and 6-month time points: 3.04, 95% CI: 0.95, 9.73; P = 0.06). CONCLUSION: Pulmonary nodule detection during lung cancer screening has little impact on smoking behaviours. Further work should explore whether lung cancer screening can impact on perceived social pressure and promote smoking cessation.

Detection in blood of autoantibodies to tumour antigens as a case-finding method in lung cancer using the EarlyCDT®-Lung Test (ECLS): study protocol for a randomized controlled trial.

BACKGROUND: Lung cancer is the most common cause of cancer related death worldwide. The majority of cases are detected at a late stage when prognosis is poor. The EarlyCDT®-Lung Test detects autoantibodies to abnormal cell surface proteins in the earliest stages of the disease which may allow tumour detection at an earlier stage thus altering prognosis. The primary research question is: Does using the EarlyCDT®-Lung Test to identify those at high risk of lung cancer, followed by X-ray and computed tomography (CT) scanning, reduce the incidence of patients with late-stage lung cancer (III & IV) or unclassified presentation (U) at diagnosis, compared to standard practice? METHODS: A randomised controlled trial of 12 000 participants in areas of Scotland targeting general practices serving patients in the most deprived quintile of the Scottish Index of Multiple Deprivation. Adults aged 50-75 who are at high risk of lung cancer and healthy enough to undergo potentially curative therapy (Performance Status 0-2) are eligible to participate. The intervention is the EarlyCDT®-Lung Test, followed by X-ray and CT in those with a positive result. The comparator is standard clinical practice in the UK. The primary outcome is the difference, after 24 months, between the rates of patients with stage III, IV or unclassified lung cancer at diagnosis. The secondary outcomes include: all-cause mortality; disease specific mortality; a range of morbidity outcomes; cost-effectiveness and measures examining the psychological and behavioural consequences of screening. Participants with a positive test result but for whom the CT scan does not lead to a lung cancer diagnosis will be offered 6 monthly thoracic CTs for 24 months. An initial chest X-ray will be used to determine the speed and the need for contrast in the first screening CT. Participants who are found to have lung cancer will be followed-up to assess both time to diagnosis and stage of disease at diagnosis. DISCUSSION: The study will determine the clinical and cost effectiveness of EarlyCDT®-Lung Test for early lung cancer detection and assess its suitability for a large-scale, accredited screening service. The study will also assess the potential psychological and behavioural harms arising from false positive or false negative results, as well as the potential benefits to patients of true negative EarlyCDT lung test results. A cost-effectiveness model of lung cancer screening based on the results of the EarlyCDT Lung Test study will be developed. TRIAL REGISTRATION: NCT01925625 . August 19, 2013.

Recalibration of a Non-Laboratory-Based Risk Model to Estimate Pre-Diabetes/Diabetes Mellitus Risk in Primary Care in Hong Kong.

INTRODUCTION/OBJECTIVES: A non-laboratory-based pre-diabetes/diabetes mellitus (pre-DM/DM) risk prediction model developed from the Hong Kong Chinese population showed good external discrimination in a primary care (PC) population, but the estimated risk level was significantly lower than the observed incidence, indicating poor calibration. This study explored whether recalibrating/updating methods could improve the model's accuracy in estimating individuals' risks in PC. METHODS: We performed a secondary analysis on the model's predictors and blood test results of 919 Chinese adults with no prior DM diagnosis recruited from PC clinics from April 2021 to January 2022 in HK. The dataset was randomly split in half into a training set and a test set. The model was recalibrated/updated based on a seven-step methodology, including model recalibrating, revising and extending methods. The primary outcome was the calibration of the recalibrated/updated models, indicated by calibration plots. The models' discrimination, indicated by the area under the receiver operating characteristic curves (AUC-ROC), was also evaluated. RESULTS: Recalibrating the model's regression constant, with no change to the predictors' coefficients, improved the model's accuracy (calibration plot intercept: -0.01, slope: 0.69). More extensive methods could not improve any further. All recalibrated/updated models had similar AUC-ROCs to the original model. CONCLUSION: The simple recalibration method can adapt the HK Chinese pre-DM/DM model to PC populations with different pre-test probabilities. The recalibrated model can be used as a first-step screening tool and as a measure to monitor changes in pre-DM/DM risks over time or after interventions.

External validation of the Hong Kong Chinese non-laboratory risk models and scoring algorithm for case finding of prediabetes and diabetes mellitus in primary care.

AIMS/INTRODUCTION: Two Hong Kong Chinese non-laboratory-based prediabetes/diabetes mellitus (pre-DM/DM) risk models were developed using logistic regression (LR) and machine learning, respectively. We aimed to evaluate the models' validity in case finding of pre-DM/DM in a Chinese primary care (PC) population. We also evaluated the validity of a risk-scoring algorithm derived from the LR model. MATERIALS AND METHODS: This was a cross-sectional external validation study on Chinese adults, without a prior DM diagnosis, who were recruited from public/private PC clinics in Hong Kong. A total of 1,237 participants completed a questionnaire on the models' predictors. Of that, 919 underwent blood glucose testing. The primary outcome was the models' and the algorithm's sensitivity in finding pre-DM/DM cases. The secondary outcomes were the models' and the algorithm's specificity, positive/negative predictive values, discrimination and calibration. RESULTS: The models' sensitivity were 0.70 (machine learning) and 0.72 (LR). Both showed good external discrimination (area under the receiver operating characteristic curve: machine learning 0.744, LR 0.739). The risks estimated by the models were lower than the observed incidence, indicating poor calibration. Both models were more effective among participants with lower pretest probabilities; that is, age 18-44 years. The algorithm's sensitivity was 0.77 at the cut-off score of ≥16 out of 41. CONCLUSION: This study showed the validity of the models and the algorithm for finding pre-DM/DM cases in a Chinese PC population in Hong Kong. They can facilitate more cost-effective identification of high-risk individuals for blood testing to diagnose pre-DM/DM in PC. Further studies should recalibrate the models for more precise risk estimation in PC populations.

Non-Laboratory-Based Risk Prediction Tools for Undiagnosed Pre-Diabetes: A Systematic Review.

Early detection of pre-diabetes (pre-DM) can prevent DM and related complications. This review examined studies on non-laboratory-based pre-DM risk prediction tools to identify important predictors and evaluate their performance. PubMed, Embase, MEDLINE, CINAHL were searched in February 2023. Studies that developed tools with: (1) pre-DM as a prediction outcome, (2) fasting/post-prandial blood glucose/HbA1c as outcome measures, and (3) non-laboratory predictors only were included. The studies' quality was assessed using the CASP Clinical Prediction Rule Checklist. Data on pre-DM definitions, predictors, validation methods, performances of the tools were extracted for narrative synthesis. A total of 6398 titles were identified and screened. Twenty-four studies were included with satisfactory quality. Eight studies (33.3%) developed pre-DM risk tools and sixteen studies (66.7%) focused on pre-DM and DM risks. Age, family history of DM, diagnosed hypertension and obesity measured by BMI and/or WC were the most common non-laboratory predictors. Existing tools showed satisfactory internal discrimination (AUROC: 0.68-0.82), sensitivity (0.60-0.89), and specificity (0.50-0.74). Only twelve studies (50.0%) had validated their tools externally, with a variance in the external discrimination (AUROC: 0.31-0.79) and sensitivity (0.31-0.92). Most non-laboratory-based risk tools for pre-DM detection showed satisfactory performance in their study populations. The generalisability of these tools was unclear since most lacked external validation.

The Impact of a Health Empowerment Program on Self-Care Enablement and Mental Health among Low-Income Families: Evidence from a 5 Year Cohort Study in Hong Kong.

Health empowerment can be an effective way to reduce health inequities. This prospective cohort study evaluated the 5 year impact of a health empowerment program (HEP) on health outcomes among adults from low-income families. The Patient Enablement Instrument version 2 (PEI-2), Depression, Anxiety and Stress Scale 21 (DASS-21), and 12 item Short-Form Health Survey version 2 (SF-12v2) were administered at baseline and follow-up for both intervention and comparison groups. A total of 289 participants (n = 162 for intervention group, n = 127 for comparison group) were included in the analysis. Most of the participants were female (72.32%), and aged from 26 to 66 years old (M = 41.63, SD = 6.91). Linear regressions weighted by inverse probability weighting using the propensity score showed that, after follow-up of 5 years, the intervention group demonstrated significantly greater increases in all items and total scores for the PEI-2 (all B > 0.59, p < 0.001), greater decreases in the DASS depression score (B = -1.98 p = 0.001), and greater increases in the Mental Component Summary score of the SF-12v2 (B = 2.99, p = 0.027) than the comparison group. The HEP may be an effective intervention enabling adults from low-income families to manage their health-related issues and improve their mental health, as evidenced by our study.

Who reports insufficient and disturbed sleep? Results from a representative population-based health survey in Hong Kong.

OBJECTIVES: To highlight the prevalence of sleep problems and identify associated risk factors among a representative sample recruited from the general population of Hong Kong. DESIGN, SETTING AND PARTICIPANTS: Participants included 12 022 individuals (aged 15 or above) who took part in the Population Health Survey 2014/15, a territory-wide survey conducted by the Department of Health of the Government of the Hong Kong Special Administrative Region. PRIMARY AND SECONDARY OUTCOME MEASURES: Outcomes were the prevalence of (1) insufficient sleep (<6 hours sleep per day) and (2) any sleep disturbance (difficulty initiating sleep, intermittent awakenings, early awakening) ≥3 times per week in the past 30 days. Multivariable logistic regression identified associations between sleep problems and sociodemographic, clinical and lifestyle factors. RESULTS: 9.7% of respondents reported insufficient sleep and 10.5% reported sleep disturbances ≥3 times a week. Female gender, monthly household income <$12 250 (Hong Kong dollar), lower education level, mental health condition and physical health condition were significantly associated with both insufficient and disturbed sleep (all p<0.05). Unemployment, homemaker, insufficient physical activity, current/former smoking status and harmful alcohol consumption were associated with sleep disturbances only (all p<0.01). CONCLUSIONS: Sleep problems are highly prevalent in Hong Kong. As such problems are associated with a range of health conditions, it is important to facilitate improvements in sleep. Our results show that harmful alcohol consumption, insufficient physical activity and current smoking are modifiable risk factors for sleep disturbances. Public health campaigns should focus on these risk factors in order to promote a healthy lifestyle and ultimately reduce sleep disturbances. Targeted interventions for high-risk groups may also be warranted, particularly for those with doctor-diagnosed physical and mental health conditions.

Development and validation of a diabetes mellitus and prediabetes risk prediction function for case finding in primary care in Hong Kong: a cross-sectional study and a prospective study protocol paper.

INTRODUCTION: Diabetes mellitus (DM) is a major non-communicable disease with an increasing prevalence. Undiagnosed DM is not uncommon and can lead to severe complications and mortality. Identifying high-risk individuals at an earlier disease stage, that is, pre-diabetes (pre-DM), is crucial in delaying progression. Existing risk models mainly rely on non-modifiable factors to predict only the DM risk, and few apply to Chinese people. This study aims to develop and validate a risk prediction function that incorporates modifiable lifestyle factors to detect DM and pre-DM in Chinese adults in primary care. METHODS AND ANALYSIS: A cross-sectional study to develop DM/Pre-DM risk prediction functions using data from the Hong Kong's Population Health Survey (PHS) 2014/2015 and a 12-month prospective study to validate the functions in case finding of individuals with DM/pre-DM. Data of 1857 Chinese adults without self-reported DM/Pre-DM will be extracted from the PHS 2014/2015 to develop DM/Pre-DM risk models using logistic regression and machine learning methods. 1014 Chinese adults without a known history of DM/Pre-DM will be recruited from public and private primary care clinics in Hong Kong. They will complete a questionnaire on relevant risk factors and blood tests on Oral Glucose Tolerance Test (OGTT) and haemoglobin A1C (HbA1c) on recruitment and, if the first blood test is negative, at 12 months. A positive case is DM/pre-DM defined by OGTT or HbA1c in any blood test. Area under receiver operating characteristic curve, sensitivity, specificity, positive predictive value and negative predictive value of the models in detecting DM/pre-DM will be calculated. ETHICS AND DISSEMINATION: Ethics approval has been received from The University of Hong Kong/Hong Kong Hospital Authority Hong Kong West Cluster (UW19-831) and Hong Kong Hospital Authority Kowloon Central/Kowloon East Cluster (REC(KC/KE)-21-0042/ER-3). The study results will be submitted for publication in a peer-reviewed journal. TRIAL REGISTRATION NUMBER: US ClinicalTrial.gov: NCT04881383; HKU clinical trials registry: HKUCTR-2808; Pre-results.

Non-laboratory-based risk assessment model for case detection of diabetes mellitus and pre-diabetes in primary care.

INTRODUCTION: More than half of diabetes mellitus (DM) and pre-diabetes (pre-DM) cases remain undiagnosed, while existing risk assessment models are limited by focusing on diabetes mellitus only (omitting pre-DM) and often lack lifestyle factors such as sleep. This study aimed to develop a non-laboratory risk assessment model to detect undiagnosed diabetes mellitus and pre-diabetes mellitus in Chinese adults. METHODS: Based on a population-representative dataset, 1,857 participants aged 18-84 years without self-reported diabetes mellitus, pre-diabetes mellitus, and other major chronic diseases were included. The outcome was defined as a newly detected diabetes mellitus or pre-diabetes by a blood test. The risk models were developed using logistic regression (LR) and interpretable machine learning (ML) methods. Models were validated using area under the receiver-operating characteristic curve (AUC-ROC), precision-recall curve (AUC-PR), and calibration plots. Two existing diabetes mellitus risk models were included for comparison. RESULTS: The prevalence of newly diagnosed diabetes mellitus and pre-diabetes mellitus was 15.08%. In addition to known risk factors (age, BMI, WHR, SBP, waist circumference, and smoking status), we found that sleep duration, and vigorous recreational activity time were also significant risk factors of diabetes mellitus and pre-diabetes mellitus. Both LR (AUC-ROC = 0.812, AUC-PR = 0.448) and ML models (AUC-ROC = 0.822, AUC-PR = 0.496) performed well in the validation sample with the ML model showing better discrimination and calibration. The performance of the models was better than the two existing models. CONCLUSIONS: Sleep duration and vigorous recreational activity time are modifiable risk factors of diabetes mellitus and pre-diabetes in Chinese adults. Non-laboratory-based risk assessment models that incorporate these lifestyle factors can enhance case detection of diabetes mellitus and pre-diabetes.

How Does the Family Influence Adolescent Eating Habits in Terms of Knowledge, Attitudes and Practices? A Global Systematic Review of Qualitative Studies.

Promoting healthy eating habits can prevent adolescent obesity in which family may play a significant role. This review synthesized findings from qualitative studies to identify family barriers and facilitators of adolescent healthy eating in terms of knowledge, attitudes, and practices (KAP). A literature search of four databases was completed on 31 July 2020; qualitative studies that explored family factors of adolescent (aged 10 to 19 years) eating habits were included. A total of 48 studies were identified, with the majority being from North America and sampled from a single source. Ten themes on how family influences adolescent dietary KAP were found: Knowledge-(1) parental education, (2) parenting style, and (3) family illness experience; Attitudes-(4) family health, (5) cultivation of preference, and (6) family motivation; Practices-(7) home meals and food availability, (8) time and cost, (9) parenting style, and (10) parental practical knowledge and attitudes. This review highlights five parental characteristics underlying food parenting practices which affect adolescents' KAP on healthy eating. Adolescents with working parents and who are living in low-income families are more vulnerable to unhealthy eating. There is a need to explore cultural-specific family influences on adolescents' KAP, especially regarding attitudes and food choices in Asian families.

Unintentional Injury Burden in Hong Kong: Results from a Representative Population-Based Survey.

Unintentional injuries are major causes of mortality and morbidity. Although generally perceived as accidents, it is possible to identify those at higher risk and implement appropriate prevention measures. This study aims to investigate the common causes of unintentional injuries and their associated risk factors among a large representative sample. Data of 12,022 individuals who completed the Hong Kong Population Health Survey 2014/15 were extracted. The primary outcome was the prevalence of having unintentional injury(-ies) in the previous 12 months that was severe enough to limit daily activities. Multivariable logistic regression analyses were conducted to identify associations between injuries and sociodemographic, clinical and lifestyle factors. 14.5% of respondents reported episode(s) of unintentional injury in the past 12 months in the population level. The main causes of top three most severe unintentional injuries were sprains (24.0%), falls (19.9%) and being hit/struck (19.6%). 13.2% injury episodes were work-related among the most severe episode. Factors independently associated with significantly higher risks of injury included currently employed, homemaker or student, born in Hong Kong (as compared with immigrants), doctor-diagnosed chronic conditions, harmful alcohol consumption, insufficient sleep, and disturbed sleep. To summarize, unintentional injuries are highly prevalent and associated with harmful drinking, insufficient sleep, and disturbed sleep, which are potential modifiable risk factors for prevention.

Screening for chronic conditions with reproductive factors using a machine learning based approach.

A large proportion of cases with chronic conditions including diabetes or pre-diabetes, hypertension and dyslipidemia remain undiagnosed. To include reproductive factors (RF) might be able to improve current screening guidelines by providing extra effectiveness. The objective is to study the relationships between RFs and chronic conditions' biomarkers. A cross-sectional study was conducted. Demographics, RFs and metabolic biomarkers were collected. The relationship of the metabolic biomarkers were shown by correlation analysis. Principal component analysis (PCA) and autoencoder were compared by cross-validation. The better one was adopted to extract a single marker, the general chronic condition (GCC), to represent the body's chronic conditions. Multivariate linear regression was performed to explore the relationship between GCC and RFs. In total, 1,656 postmenopausal females were included. A multi-layer autoencoder outperformed PCA in the dimensionality reduction performance. The extracted variable by autoencoder, GCC, was verified to be representative of three chronic conditions (AUC for patoglycemia, hypertension and dyslipidemia were 0.844, 0.824 and 0.805 respectively). Linear regression showed that earlier age at menarche (OR = 0.9976) and shorter reproductive life span (OR = 0.9895) were associated with higher GCC. Autoencoder performed well in the dimensionality reduction of clinical metabolic biomarkers. Due to high accessibility and effectiveness, RFs have potential to be included in screening tools for general chronic conditions and could enhance current screening guidelines.

Unconditional and conditional monetary incentives to increase response to mailed questionnaires: A randomized controlled study within a trial (SWAT).

RATIONALE, AIMS, AND OBJECTIVES: High response rates to research questionnaires can help to ensure results are more representative of the population studied and provide increased statistical power, on which the study may have been predicated. Improving speed and quality of response can reduce costs. METHOD: We conducted a randomized study within a trial (SWAT) to assess questionnaire response rates, reminders sent, and data completeness with unconditional compared with conditional monetary incentives. Eligible individuals were mailed a series of psychological questionnaires as a follow-up to a baseline host trial questionnaire. Half received a £5 gift voucher with questionnaires (unconditional), and half were promised the voucher after returning questionnaires (conditional). RESULTS: Of 1079 individuals, response rates to the first follow-up questionnaire were 94.2% and 91.7% in the unconditional and conditional monetary incentive groups, respectively (OR 1.78; 95% CI, 0.85-3.72). There were significantly greater odds of returning repeat questionnaires in the unconditional group at 6 months (OR 2.97; 95% CI, 1.01-8.71; .047) but not at 12 months (OR 1.12; 95% CI, 0.44-2.85). Incentive condition had no impact at any time point on the proportion of sent questionnaires that needed reminders. Odds of incomplete questionnaires were significantly greater at 3 months in the unconditional compared with the conditional incentive group (OR 2.45; 95% CI, 1.32-4.55; .004). CONCLUSIONS: Unconditional monetary incentives can produce a transitory greater likelihood of mailed questionnaire response in a clinical trial participant group, consistent with the direction of effect in other settings. However, this could have been a chance finding. The use of multiple strategies to promote response may have created a ceiling effect. This strategy has potential to reduce administrative and postage costs, weighed against the cost of incentives used, but could risk compromising the completeness of data.

Do Negative Emotional States Play a Role in the Association Between Intimate Partner Violence and Poor Health-Related Quality of Life in Chinese Women From Low-Income Families?

A cross-sectional analysis of a dataset of 156 participants in a health assessment program explored whether negative emotional states mediated the association between intimate partner violence (IPV) and health-related quality of life (HRQoL). Compared with IPV screen-negative participants, those who screened positive had significantly lower HRQoL and significantly higher levels of depression, anxiety, and stress. The inverse associations between the presence of IPV and HRQoL were found to be mediated by depression, anxiety, and stress. Therefore, interventions to alleviate negative emotions in women suffering from IPV have the potential to be useful in improving their HRQoL.

Factors associated with youth gang membership in low- and middle-income countries: a systematic review.

This Campbell systematic review examines the predictors of youth gang membership in low- and middle-income countries. The review summarises findings from eight reports from five countries and the Caribbean region. The lack of available evidence limits the extent to which clear conclusions can be drawn about the factors associated with youth gang membership. The review is based on a very small number of studies, and has significant limitations in coverage. The limited evidence of the correlates of youth gang membership suggests factors that may drive gang membership and suggests areas where interventions may prove promising in the family, school, and community domains, as well as provide a starting point for future studies. Plain language summary: Evidence shows which factors predict gang membership in low- and middle-income countries, but more studies needed: Youth gang membership is associated with delinquency, violent crime and trafficking. A range of individual, peer, family, school and community factors can predict the likelihood of youths getting involved with gangs. Knowledge of these factors can be helpful for reducing gang membership.What is this review about?: Youth gang membership is associated with delinquency, violent crime and trafficking - and gang members are themselves frequently the victims of these offences. Yet youth gangs can also provide a form of social capital, a sense of belonging and purpose to disenfranchised youth.This review identifies the factors associated with young people joining gangs, and the differences between gang-involved and non-gang-involved youth. Understanding these associations is essential to reduce the levels of gang membership and the incidence of related violence.What studies are included?: Studies of youth gangs in in low- and middle-income countries were included, with participants aged 10-29 years. The studies had to assess an individual predictor or correlate of youth gang membership, where the predictor or correlate is a single characteristic, not a conglomeration of multiple constructs. Included studies had designs including data on both gang-involved and non-gang-involved youth, recruited with strategies that were eligible.Nine studies met the eligibility criteria and were included in the review. One of these studies did not report all the required data and so was not included in the analyses. The studies were conducted in Turkey, Trinidad and Tobago, the Caribbean, El Salvador, China and Brazil. Factors associated with gang membership [Table: see text] What are the main results in this review?: The eight studies analysed in the review address the associations between life events and circumstances, and the likelihood of being a youth gang member across five domains: individual, peers, family, school and community. Significant associations were found with factors in each domain.What do the findings of this review mean?: The lack of available evidence limits the extent to which clear conclusions can be drawn about the factors associated with youth gang membership. The review is based on a very small number of studies, and has significant limitations in coverage. The limited evidence of the correlates of youth gang membership suggests factors that may drive gang membership and suggests areas where interventions may prove promising in the family, school, and community domains, as well as provide a starting point for future studies.How up-to-date is this review?: The review authors searched for studies published up to September 2013. This Campbell systematic review was published in December 2018. Summary: Background: Youth gang membership is well documented throughout low- and middle-income countries, and youth gang members are increasingly associated with delinquency, violent crime and trafficking. They are also frequently the victims of these offences, often in disproportionate numbers compared to non-gang youth. Yet youth gangs can also provide a form of social capital, a sense of belonging and purpose to disenfranchised youth.Extensive research, primarily from high-income countries, has categorized five domains of risk and protective factors for youth gang involvement, drawn from the realm of developmental psychology. These domains are: Individual, Peer, Family, School, and Community. Youth gang membership is seen as the culmination of interrelated structural and process factors, which in combination with negative life events may increase the attractiveness of gang membership.This review aimed to identify the factors associated with young people joining gangs, and to identify and quantify the differences between gang-involved and non-gang-involved youth. Understanding these associations is essential to reduce the levels of gang membership and the incidence of related violence.Objectives: This review addresses two key objectives: (1) to synthesize the published and unpublished empirical evidence on the factors associated with membership of youth gangs in low- and middle-income countries;(2) to assess the relative strength of the different factors across the domains of individual, family, school, peer group and community.Search methods: The search was conducted in English, French, Chinese, Arabic, Russian, Spanish and Portuguese. We searched 55 locations including academic databases, journals, and grey literature locations, and located over 54,000 documents to screen.Selection criteria: We included studies meeting the following criteria: Reports on youth gangsAll participants aged 10-29 yearsLocated in low- or middle-income countryAssesses an individual predictor or correlate of youth gang membershipPredictor or correlate is a single characteristicPredictor or correlate is not a conglomeration of multiple constructsEligible recruitment strategy for respondentsStudy design included data on both gang involved and non-gang involved youth. Data collection and analysis: We conducted a broad abstract screening of over 54,000 titles and abstracts, followed by a close abstract screening of 1509 abstracts. We screened the full-text of 98 documents. Nine studies met the eligibility criteria and were included in the review. One of these studies did not report sufficient data to allow the calculation of a standardized effect size, and so was not included in the analyses. A total of 85 independent effect sizes were extracted from the eight studies with sufficient data to create a standardized effect size. We calculated Cohen's d from continuous data and the Log Odds Ratio from dichotomous data. All effects were categorized into the five predictor domains, and further classified into conceptually similar group and risk or protective factors. We synthesized the data using multiple random effects meta-analyses with inverse variance weighting.Results: The eight studies analysed in the review address the associations between life events and circumstances, and the likelihood of being a youth gang member. All studies were based on cross-sectional survey data, using different statistical methods to identify correlations between youth characteristics and the likelihood of being a member of a youth gang. No longitudinal, prospective or retrospective studies were located. The studies were conducted in Turkey, Trinidad and Tobago, the Caribbean, El Salvador, China and Brazil. We organized the analyses according to five domains of factors that may be associated with youth gang involvement and summarize the main findings below: Individual factors Delinquency, alcohol and soft drug use, male gender, risky sexual behaviours, employment, psychological risk factors (low self-control, impulsivity), and victimisation were each associated with significantly higher odds of youth gang membership.Psychological protective factors (empathy, future orientation, belief in moral order) were associated with lower odds.No association between youth gang membership and age, minority ethnicity, or protective behaviours surrounding sexual behaviour or alcohol and soft drug use. Peer factors There is a small amount of evidence that socialising with delinquent peers is associated with increased odds of youth gang membership, but there is no significant relationship demonstrated between socialising with pro-social peers and youth gang membership. Family factors Negative family environments are associated with more youth gang membership, while both a positive family environment and parental monitoring were associated with lower odds.One study showed that youth from middle-income families had greater odds of reporting youth gang membership than those from either high or low-income families.No significant relationship was seen between youth gang membership and parental education or parental attitudes to antisocial behaviour. School factors Low school attachment, exposure to violence at school, educational difficulties and opportunities for prosocial involvement all show significant association with more youth gang membership.No significant relationship was seen between youth gang membership and level of education, school type, or school performance. Community factors There were significantly higher odds of youth gang membership amongst those who reported that they were exposed to violence in their neighbourhood.No demonstrated association between youth gang membership and neighbourhood environment risk or protective factors, or geography. Gang membership is typically viewed as a culmination of interrelated structural and process factors. Understanding the factors associated with youth gang membership can help inform prevention strategies to reduce the levels of youth gang membership and the incidence of youth gang violence. Unfortunately the small number of studies contributing to any analysis limits the conclusions that can be drawn from this study. Moreover, many of the individual and peer associations identified in this review (such as delinquency, drug use, and sexual risk factors) may be as a result of gang membership rather than its cause. However, our results suggest certain family, school, and community level factors associated with gang membership that could be addressed through targeted preventive interventions, particularly family environment, parental monitoring, school attachment, educational difficulties, and exposure to violence in the home, at school, or in the community.Authors' conclusions: The lack of available evidence limits the extent to which we can draw any clear conclusions about the factors associated with youth gang membership. This current review is based on a very small number of studies, and has significant limitations in coverage; however it provides some limited evidence of the correlates of youth gang membership. Specifically, this review suggests factors that may drive gang membership and suggests areas where interventions may prove promising in the family, school, and community domains, as well as provide a starting point for future studies.