School health promotion and fruit and vegetable consumption in secondary schools: a repeated cross-sectional multilevel study

BACKGROUND

Worldwide, recommendations for fruit and vegetable consumption are not met, which can cause chronic diseases. Especially adolescence is an important phase for the development of health behaviours. Therefore, in the Netherlands, the Healthy School program was established to aid schools in promoting healthy lifestyles among their students. We examined to what extent the variation between secondary schools regarding students' fruit and vegetable consumption could be explained by differences between schools regarding Healthy School certification, general school characteristics, and the school population. Additionally, we examined whether Healthy School certification was related to the outcomes, and whether the association differed for subgroups.

METHODS

We performed a repeated cross-sectional multilevel study. We used data from multiple school years from the national Youth Health Monitor on secondary schools (grades 2 and 4, age ranged from approximately 12 to 18 years) of seven Public Health Services, and added data with regard to Healthy School certification, general school characteristics and school population characteristics. We included two outcomes: the number of days a student consumed fruit and vegetables per week. In total, we analysed data on 168,127 students from 256 secondary schools in the Netherlands.

RESULTS

Results indicated that 2.87% of the variation in fruit consumption and 5.57% of the variation in vegetable consumption could be attributed to differences at the school-level. Characteristics related to high parental educational attainment, household income, and educational track of the students explained most of the variance between schools. Additionally, we found a small favourable association between Healthy School certification and the number of days secondary school students consumed fruit and vegetables.

CONCLUSIONS

School population characteristics explained more variation between schools than Healthy School certification and general school characteristics, especially indicators of parental socioeconomic status. Nevertheless, Healthy School certification seemed to be slightly related to fruit and vegetable consumption, and might contribute to healthier dietary intake. We found small differences for some subgroups, but future research should focus on the impact in different school contexts, since we were restricted in the characteristics that could be included in this study.

Computerized adaptive testing to screen pre-school children for emotional and behavioral problems

Questionnaires to detect emotional and behavioral (EB) problems in preventive child healthcare (PCH) should be short; this potentially affects their validity and reliability. Computerized adaptive testing (CAT) could overcome this weakness. The aim of this study was to (1) develop a CAT to measure EB problems among pre-school children and (2) assess the efficiency and validity of this CAT. We used a Dutch national dataset obtained from parents of pre-school children undergoing a well-child care assessment by PCH (n = 2192, response 70%). Data regarded 197 items on EB problems, based on four questionnaires, the Strengths and Difficulties Questionnaire (SDQ), the Child Behavior Checklist (CBCL), the Ages and Stages Questionnaire: Social Emotional (ASQ:SE), and the Brief Infant-Toddler Social and Emotional Assessment (BITSEA). Using 80% of the sample, we calculated item parameters necessary for a CAT and defined a cutoff for EB problems. With the remaining part of the sample, we used simulation techniques to determine the validity and efficiency of this CAT, using as criterion a total clinical score on the CBCL. Item criteria were met by 193 items. This CAT needed, on average, 16 items to identify children with EB problems. Sensitivity and specificity compared to a clinical score on the CBCL were 0.89 and 0.91, respectively, for total problems; 0.80 and 0.93 for emotional problems; and 0.94 and 0.91 for behavioral problems.    Conclusion: A CAT is very promising for the identification of EB problems in pre-school children, as it seems to yield an efficient, yet high-quality identification. This conclusion should be confirmed by real-life administration of this CAT. What is Known: • Studies indicate the validity of using computerized adaptive test (CAT) applications to identify emotional and behavioral problems in school-aged children. • Evidence is as yet limited on whether CAT applications can also be used with pre-school children. What is New: • The results of this study show that a computerized adaptive test is very promising for the identification of emotional and behavior problems in pre-school children, as it appears to yield an efficient and high-quality identification.

Behavior Change Techniques That Prevent or Decrease Obesity in Youth With a Low Socioeconomic Status: A Systematic Review and Meta-Analysis

Background: Interventions, targeting youth, are necessary to prevent obesity later in life. Especially youth with low socioeconomic status (SES) are vulnerable to develop obesity. This meta-analysis examines the effectiveness of behavioral change techniques (BCTs) to prevent or reduce obesity among 0 to 18-year-olds with a low SES in developed countries. Method: Intervention studies were identified from systematic reviews or meta-analyses published between 2010 and 2020 and retrieved from PsycInfo, Cochrane systematic review, and PubMed. The main outcome was body mass index (BMI), and we coded the BCTs. Results: Data from 30 studies were included in the meta-analysis. The pooled postintervention effects of these studies indicated a nonsignificant decrease in BMI for the intervention group. Longer follow-up (≥12 months) showed favorable differences for intervention studies, although that BMI change was small. Subgroup analyses showed larger effects for studies with six or more BCTs. Furthermore, subgroup analyses showed a significant pooled effect in favor of the intervention for the presence of a specific BCT (problem-solving, social support, instruction on how to perform the behavior, identification of self as role model, and demonstration of the behavior), or absence of a specific BCT (information about health consequences). The intervention program duration and age group of the study population did not significantly influence the studies' effect sizes. Conclusions: Generally, the effects of interventions on BMI change among youth with low SES are small to neglectable. Studies with more than six BCTs and/or specific BCTs had a higher likelihood of decreasing BMI of youth with low SES.

A study on the applicability of the Strengths and Difficulties Questionnaire among low- and higher-educated adolescents

Aim

The Strengths and Difficulties Questionnaire self-report (SDQ-SR) is a valid instrument for detection of emotional and behavioral problems. The aim of this study was to compare the psychometric properties of the SDQ-SR for low and higher educated adolescents, and to explore its suitability.

Methods

We included 426 adolescents. We compared internal consistency for low-educated, i.e., at maximum pre-vocational secondary education, and higher educated adolescents and assessed whether the five-factor structure of the SDQ holds across educational levels. We also interviewed 24 low-educated adolescents, and 17 professionals.

Results

On most SDQ subscales the low-educated adolescents had more problematic mean scores than the higher educated adolescents. Findings on the invariance factor analyses were inconsistent, with some measures showing a bad fit of the five factor model, and this occurring relatively more for the low-educated adolescents. Professionals and adolescents reported that the SDQ included difficult wordings.

Discussion

Our findings imply that the scale structure of the SDQ-SR is slightly poorer for low educated adolescents. Given this caveat, psychometric properties of the SDQ-SR are generally sufficient for use, regardless of educational level.

Key developmental milestones helped to identify children with special educational needs and disabilities at an early stage

AIM

Responding to developmental delay promptly is important, as it helps children to reach their full potential. This study investigated how developmental milestones predicted primary school children with special educational needs and disabilities (SEND) at an early stage.

METHODS

We obtained data about 36 milestones between 12 and 45 months using the Dutch Development Instrument. Development, primary school classification and background characteristics were collected from the Dutch Preventive Child Healthcare system in Utrecht from 2008 to 2016. We investigated SEND classifications and the primary schools that the children attended at 4-12 years of age. The findings include area under the curve (AUC) data.

RESULTS

Data on 30 579 children in mainstream schools and 1055 children with SEND were available. Different milestones predicted SEND classifications. Fourteen milestones and parental education predicted attendance at special needs schools with smaller classes (AUC 0.913). Nine milestones, sex, migration background and parental education predicted attendance at schools for severe communication problems (AUC 0.963). Ten milestones and parental education predicted attendance at schools for severe learning difficulties (AUC 0.995). Milestones did not accurately predict attendance at schools for severe behavioural or psychiatric problems.

CONCLUSION

Milestones at 12-45 months predicted most SEND classifications at primary school age, except severe behavioural or psychiatric problems.

Studies on Reliability and Measurement Error of Measurements in Medicine - From Design to Statistics Explained for Medical Researchers

Reliability and measurement error are measurement properties that quantify the influence of specific sources of variation, such as raters, type of machine, or time, on the score of the individual measurement. Several designs can be chosen to assess reliability and measurement error of a measurement. Differences in design are due to specific choices about which sources of variation are varied over the repeated measurements in stable patients, which potential sources of variation are kept stable (ie, restricted), and about whether or not the entire measurement instrument (or measurement protocol) was repeated or only part of it. We explain how these choices determine how intraclass correlation coefficients and standard errors of measurement formulas are built for different designs by using Venn diagrams. Strategies for improving the measurement are explained, and recommendations for reporting the essentials of these studies are described. We hope that this paper will facilitate the understanding and improve the design, analysis, and reporting of future studies on reliability and measurement error of measurements.

School health promotion and the consumption of water and sugar-sweetened beverages in secondary schools: a cross-sectional multilevel study

BACKGROUND

Overweight among adolescents remains a serious concern worldwide and can have major health consequences in later life, such as cardiovascular diseases and cancer. Still, 33% of secondary school adolescents in the Netherlands consume sugar-sweetened beverages daily and over 26% do not consume water every day. The Dutch Healthy School program was developed to support schools in stimulating healthier lifestyles by focusing on health education, school environments, identifying students' health problems, and school policy. We examined the variation between secondary schools regarding the daily consumption of water and sugar-sweetened beverages and whether this variation can be explained by differences between schools regarding Healthy School certification, general school characteristics, and the school population.

METHODS

We performed a cross-sectional multilevel study. We used data from the national Youth Health Monitor of 2019 on secondary schools (grades 8 and 10, age range about 12 to 18 years) of seven Public Health Services and combined these with information regarding Healthy School certification and general school- and school population characteristics. Our outcomes were daily consumption of water and sugar-sweetened beverages. In total, data from 51,901 adolescents from 191 schools were analysed. We calculated the intraclass correlation to examine the variation between schools regarding our outcomes. Thereafter, we examined whether we could explain this variation by the included characteristics.

RESULTS

The school-level explained 4.53% of the variation in the consumption of water and 2.33% of the variation in the consumption of sugar-sweetened beverages. This small variation in water and sugar-sweetened consumption could not be explained by Healthy School certification, yet some general school- and school population characteristics did: the proportion of the school population with at least one parent with high educational attainment, the educational track of the adolescents, urbanicity (only for water consumption) and school type (only for sugar-sweetened beverages consumption).

CONCLUSIONS

The low percentages of explained variation indicate that school-level characteristics in general (including Healthy School certification) do not matter substantially for the daily consumption of water and sugar-sweetened beverages. Future research should examine whether school health promotion can contribute to healthier lifestyles, and if so, under which level of implementation and school conditions.

Measuring Implementation of Health Promoting School (HPS) Programs: Development and Psychometric Evaluation of the HPS Implementation Questionnaire

BACKGROUND

Implementation of Health Promoting School (HPS) programs can be challenging due to the dynamic school context. Navigating between program fidelity and adaptation, as well as integrating the program, is essential for successful implementation, and consequently, for program effects. As part of an evaluation study in the Netherlands, this study aimed to develop a measurement instrument that differentiates schools according to fidelity, adaptation, and integration of HPS implementation.

METHODS

This study presents the development and psychometric evaluation of the 28-item HPS Implementation Questionnaire, covering 7 dimensions: adherence, dose, participant responsiveness, quality of delivery, program differentiation, adaptation, and integration. The questionnaire, to be filled out by school employees, was developed for primary, secondary, secondary vocational, and special needs education, in close collaboration with experts (n = 54) in school health promotion.

RESULTS

Semi-structured interviews aimed at dimension clarification resulted in a list of 58 items. Items were revised, combined, and/or removed based on quantitative and qualitative feedback by the evaluation study's Community of Practice, 2-round expert consultation, and pre-tests. Psychometric evaluation (n = 535 schools), consisting of calculating Cronbach's α and confirmatory factor analysis (CFA), confirmed internal consistency (α > .72) and the 7-dimension framework.

CONCLUSION

The brief yet comprehensive HPS Implementation Questionnaire offers possibilities for research into HPS implementation in various educational sectors and contexts, as well as self-monitoring by individual schools. This study provides first evidence for internal consistency and validity of the questionnaire.

The impact of occupational exposures on infection rates during the COVID-19 pandemic: a test-negative design study with register data of 207 034 Dutch workers

OBJECTIVE

This study aimed to investigate the effects of occupational exposures on the risk of a positive COVID-19 test, and whether this differed across pandemic waves.

METHODS

Data from 207 034 workers from The Netherlands with test data on COVID-19 from June 2020 until August 2021 were available. Occupational exposure was estimated by using the eight dimensions of a COVID-19 job exposure matrix (JEM). Personal characteristics, household composition and residence area were derived from Statistics Netherlands. A test-negative design was applied in which the risk of a positive test was analyzed in a conditional logit model.

RESULTS

All eight dimensions of occupational exposure included in the JEM increased the odds of a positive test for the entire study period and three pandemic waves [OR ranging from 1.09, (95% confidence interval (CI) 1.02-1.17) to 1.77 (95% CI 1.61-1.96)]. Adjusting for a previous positive test and other covariates strongly reduced the odds to be infected, but most dimensions remained at elevated risk. Fully adjusted models showed that contaminated work spaces and face covering were mostly relevant in the first two pandemic waves, whereas income insecurity showed higher odds in the third wave. Several occupations have a higher predicted value for a positive COVID-19 test, with variation over time. Discussion Occupational exposures are associated with a higher risk of a positive test, but variations over time exist in occupations with the highest risks. These findings provide insights for interventions among workers for future pandemic waves of COVID-19 or other respiratory epidemics.

Protocol for validation of the Global Scales for Early Development (GSED) for children under 3 years of age in seven countries

INTRODUCTION

Children's early development is affected by caregiving experiences, with lifelong health and well-being implications. Governments and civil societies need population-based measures to monitor children's early development and ensure that children receive the care needed to thrive. To this end, the WHO developed the Global Scales for Early Development (GSED) to measure children's early development up to 3 years of age. The GSED includes three measures for population and programmatic level measurement: (1) short form (SF) (caregiver report), (2) long form (LF) (direct administration) and (3) psychosocial form (PF) (caregiver report). The primary aim of this protocol is to validate the GSED SF and LF. Secondary aims are to create preliminary reference scores for the GSED SF and LF, validate an adaptive testing algorithm and assess the feasibility and preliminary validity of the GSED PF.

METHODS AND ANALYSIS

We will conduct the validation in seven countries (Bangladesh, Brazil, Côte d'Ivoire, Pakistan, The Netherlands, People's Republic of China, United Republic of Tanzania), varying in geography, language, culture and income through a 1-year prospective design, combining cross-sectional and longitudinal methods with 1248 children per site, stratified by age and sex. The GSED generates an innovative common metric (Developmental Score: D-score) using the Rasch model and a Development for Age Z-score (DAZ). We will evaluate six psychometric properties of the GSED SF and LF: concurrent validity, predictive validity at 6 months, convergent and discriminant validity, and test-retest and inter-rater reliability. We will evaluate measurement invariance by comparing differential item functioning and differential test functioning across sites.

ETHICS AND DISSEMINATION

This study has received ethical approval from the WHO (protocol GSED validation 004583 20.04.2020) and approval in each site. Study results will be disseminated through webinars and publications from WHO, international organisations, academic journals and conference proceedings.

REGISTRATION DETAILS

Open Science Framework https://osf.io/ on 19 November 2021 (DOI 10.17605/OSF.IO/KX5T7; identifier: osf-registrations-kx5t7-v1).

Longitudinal individual predictions from irregular repeated measurements data

Intensive longitudinal data can be used to explore important associations and patterns between various types of inputs and outcomes. Nonlinear relations and irregular measurement occasions can pose problems to develop an accurate model for these kinds of data. This paper focuses on the development, fitting and evaluation of a prediction model with irregular intensive longitudinal data. A three-step process for developing a prediction tool for (daily) monitoring and prediction is outlined and illustrated for intensive weight measurements in piglets. Step 1 addresses a nonlinear relation in the data by developing and applying a normalizing transformation. Step 2 addresses the intermittent nature of the time points by aligning the measurement times to a common time grid with a broken-stick model. Step 3 addresses the prediction problem by selecting and evaluating inputs and covariates in the model to obtain accurate predictions. The final model predicts future outcomes accurately, while allowing for nonlinearities between input and output and for different measurement histories of individuals. The methodology described can be used to develop a tool to deal with intensive irregular longitudinal data that uses the available information in an optimal way. The resulting tool demonstrated to perform well for piglet weight prediction and can be adapted to many different applications.

The creation of the Global Scales for Early Development (GSED) for children aged 0-3 years: combining subject matter expert judgements with big data

INTRODUCTION

With the ratification of the Sustainable Development Goals, there is an increased emphasis on early childhood development (ECD) and well-being. The WHO led Global Scales for Early Development (GSED) project aims to provide population and programmatic level measures of ECD for 0-3 years that are valid, reliable and have psychometrically stable performance across geographical, cultural and language contexts. This paper reports on the creation of two measures: (1) the GSED Short Form (GSED-SF)-a caregiver reported measure for population-evaluation-self-administered with no training required and (2) the GSED Long Form (GSED-LF)-a directly administered/observed measure for programmatic evaluation-administered by a trained professional.

METHODS

We selected 807 psychometrically best-performing items using a Rasch measurement model from an ECD measurement databank which comprised 66 075 children assessed on 2211 items from 18 ECD measures in 32 countries. From 766 of these items, in-depth subject matter expert judgements were gathered to inform final item selection. Specifically collected were data on (1) conceptual matches between pairs of items originating from different measures, (2) developmental domain(s) measured by each item and (3) perceptions of feasibility of administration of each item in diverse contexts. Prototypes were finalised through a combination of psychometric performance evaluation and expert consensus to optimally identify items.

RESULTS

We created the GSED-SF (139 items) and GSED-LF (157 items) for tablet-based and paper-based assessments, with an optimal set of items that fit the Rasch model, met subject matter expert criteria, avoided conceptual overlap, covered multiple domains of child development and were feasible to implement across diverse settings.

CONCLUSIONS

State-of-the-art quantitative and qualitative procedures were used to select of theoretically relevant and globally feasible items representing child development for children aged 0-3 years. GSED-SF and GSED-LF will be piloted and validated in children across diverse cultural, demographic, social and language contexts for global use.

Educational inequalities in self-rated health and emotional exhaustion among workers during the COVID-19 pandemic: a longitudinal study

OBJECTIVE

This study aimed to investigate trends in educational inequalities in poor health and emotional exhaustion during the pandemic among workers, and differences in trends between men and women.

METHODS

Five waves (2019-2021) from the longitudinal study 'the Netherlands Working Conditions Survey COVID-19 study' were used (response rates: 32-38%). Generalized logistic mixed models were used to estimate the changes in absolute and relative educational inequalities in poor health and emotional exhaustion for all workers (n = 12,479) and for men and women, separately.

RESULTS

Low and intermediate educated workers reported more often poor health (OR 2.54; 95% CI 1.71-3.77 and OR 2.09; 95% CI 1.68-2.61, respectively) than high educated workers. Intermediate educated women (OR 0.49; 95% CI 0.37-0.64) reported less emotional exhaustion than high educated women, but no differences were observed among men. The prevalence of poor health first decreased across all educational levels until March 2021, and bounced back in November 2021. A similar pattern was found for emotional exhaustion, but for low and intermediate educated workers only. Relative educational inequalities in poor health reduced among men during the pandemic, and absolute differences decreased among men and women by 2.4-2.6%. Relative educational inequalities in emotional exhaustion widened among men only. Absolute differences in emotional exhaustion first increased among both men and women, but narrowed between the last two waves.

DISCUSSION

Socioeconomic inequalities for poor self-rated health remained but narrowed in relative and absolute terms during the pandemic. With regard to emotional exhaustion, socioeconomic inequalities returned to pre-COVID-19 levels at the end of 2021.

Child development with the D-score: tuning instruments to unity

The chapter familiarises the reader with an intuitive yet powerful methodology to tune instruments to a common unit, presenting a fresh approach that expresses measurements made by different instruments on the D-score scale. As a result, the reader may compare D-scores between ages, children or cohorts. It shows how to exploit common developmental milestones to bridge instruments and cohorts; presents an analysis to obtain D-scores from 16 cohorts and 14 instruments; compares D-score age-distribution across populations from four continents; suggests an indicator for the United Nations Sustainable Development Goals; and defines developmentally-on-track.

Child development with the D-score: turning milestones into measurement

The chapter equips the reader with a basic understanding of robust psychometric methods that are needed to turn developmental milestones into measurements, introducing the fundamental issues in defining a unit for child development and demonstrates the relevant quantitative methodology. It reviews quantitative approaches to measuring child development;introduces the Rasch model in a non-technical way;shows how to estimate model parameters from real data;puts forth a set of principles for model evaluation and assessment of scale quality;analyses the relation between early D-scores and later intelligence;and compares the D-scores from three studies that all use the same instrument.

Sample size recommendations for studies on reliability and measurement error: an online application based on simulation studies

Simulation studies were performed to investigate for which conditions of sample size of patients (n) and number of repeated measurements (k) (e.g., raters) the optimal (i.e., balance between precise and efficient) estimations of intraclass correlation coefficients (ICCs) and standard error of measurements (SEMs) can be achieved. Subsequently, we developed an online application that shows the implications for decisions about sample sizes in reliability studies. We simulated scores for repeated measurements of patients, based on different conditions of n, k, the correlation between scores on repeated measurements (r), the variance between patients’ test scores (v), and the presence of systematic differences within k. The performance of the reliability parameters (based on one-way and two-way effects models) was determined by the calculation of bias, mean squared error (MSE), and coverage and width of the confidence intervals (CI). We showed that the gain in precision (i.e., largest change in MSE) of the ICC and SEM parameters diminishes at larger values of n or k. Next, we showed that the correlation and the presence of systematic differences have most influence on the MSE values, the coverage and the CI width. This influence differed between the models. As measurements can be expensive and burdensome for patients and professionals, we recommend to use an efficient design, in terms of the sample size and number of repeated measurements to come to precise ICC and SEM estimates. Utilizing the results, a user-friendly online application is developed to decide upon the optimal design, as ‘one size fits all’ doesn’t hold.

The Strengths and Difficulties Questionnaire Parent Form: Dutch norms and validity

Objective

This study provides Dutch national norms for the parent-reported Strengths and Difficulties Questionnaire (SDQ) for children aged 3-14 years, and assesses the test performance of the SDQ Total Difficulties Scale (TDS) and impairment Scale. We further compared Dutch SDQ norms with those of the United Kingdom (UK), to determine potential variation in country-specific norms.

Study design

We analyzed data of 3384 children aged 3 to 14 years. The data were obtained in schools, and in the context of Preventive Child Healthcare. Parents completed the SDQ parent form and the Child Behavior Checklist (CBCL). We determined clinical (10% elevated scores) and borderline (20% elevated scores) SDQ TDS norms. We assessed the test performance (validity) of the SDQ TDS and Impairment Score using the CBCL as criterion.

Results

The clinical SDQ TDS norms varied between > 10 and > 14 depending on the age group. The SDQ TDS discriminated between children with and without problems, as measured by the CBCL, for all age groups (AUCs varied from 0.92 to 0.96). The SDQ Impairment Score had added value (beyond the SDQ TDS) only for the age group 12-14 years. For the Netherlands we found lower clinical SDQ TDS norms than those previously reported for the UK (i.e. > 16).

Conclusion

The clinical SDQ TDS norms varied between > 10 and > 14 depending on the age groups. We found good test performance at these proposed norms. Dutch norms differed somewhat from UK norms. In the Netherlands, the SDQ performed better with Dutch-specific norms than with UK-specific norms.

Costs and effects of conventional vision screening and photoscreening in the Dutch preventive child health care system

Background

Little is known about costs and effects of vision screening strategies to detect amblyopia. Aim of this study was to compare costs and effects of conventional (optotype) vision screening, photoscreening or a combination in children aged 3–6 years.

Methods

Population-based, cross-sectional study in preventive child health care in The Hague. Children aged 3 years (3y), 3 years and 9 months (3y9m) or 5–6 years (5/6y) received the conventional chart vision screening and a test with a photoscreener (Plusoptix S12C). Costs were based on test duration and additional costs for devices and diagnostic work-up.

Results

Two thousand, one hundred and forty-four children were included. The estimated costs per child screened were €17.44, €20.37 and €6.90 for conventional vision screening at 3y, 3y9m and 5/6y, respectively. For photoscreening, these estimates were €6.61, €7.52 and €9.40 and for photoscreening followed by vision screening if the result was unclear (combination) €9.32 (3y) and €9.33 (3y9m). The number of children detected with amblyopia by age were 9, 14 and 5 (conventional screening), 6, 13 and 3 (photoscreening) and 10 (3y) and 15 (3y9m) (combination), respectively. The estimated costs per child diagnosed with amblyopia were €1500, €1050 and €860 for conventional vision screening, €860, €420 and €1940 for photoscreening and €730 (3y) and €450 (3y9m) for the combination.

Conclusions

Combining photoscreening with vision screening seems promising to detect amblyopia in children aged 3y/3y9m, whereas conventional screening seems preferable at 5/6y. As the number of study children with amblyopia is small, further research on the effects of these screening alternatives in detecting children with amblyopia is recommended.

Child development with the D-score: tuning instruments to unity

The chapter familiarises the reader with an intuitive yet powerful methodology to tune instruments to a common unit, presenting a fresh approach that expresses measurements made by different instruments on the D-score scale. As a result, the reader may compare D-scores between ages, children or cohorts. It shows how to exploit common developmental milestones to bridge instruments and cohorts; presents an analysis to obtain D-scores from 16 cohorts and 14 instruments; compares D-score age-distribution across populations from four continents; suggests an indicator for the United Nations Sustainable Development Goals; and defines developmentally-on-track.

Child development with the D-score: turning milestones into measurement

The chapter equips the reader with a basic understanding of robust psychometric methods that are needed to turn developmental milestones into measurements, introducing the fundamental issues in defining a unit for child development and demonstrates the relevant quantitative methodology. It reviews quantitative approaches to measuring child development;introduces the Rasch model in a non-technical way;shows how to estimate model parameters from real data;puts forth a set of principles for model evaluation and assessment of scale quality;analyses the relation between early D-scores and later intelligence;and compares the D-scores from three studies that all use the same instrument.

Missing item responses in latent growth analysis: Item response theory versus classical test theory

In medical research, repeated questionnaire data is often used to measure and model latent variables across time. Through a novel imputation method, a direct comparison is made between latent growth analysis under classical test theory and item response theory, while also including effects of missing item responses. For classical test theory and item response theory, by means of a simulation study the effects of item missingness on latent growth parameter estimates are examined given longitudinal item response data. Several missing data mechanisms and conditions are evaluated in the simulation study. The additional effects of missingness on differences in classical test theory- and item response theory-based latent growth analysis are directly assessed by rescaling the multiple imputations. The multiple imputation method is used to generate latent variable and item scores from the posterior predictive distributions to account for missing item responses in observed multilevel binary response data. It is shown that a multivariate probit model, as a novel imputation model, improves the latent growth analysis, when dealing with missing at random (MAR) in classical test theory. The study also shows that the parameter estimates for the latent growth model using item response theory show less bias and have smaller MSE’s compared to the estimates using classical test theory.

The D-score: a metric for interpreting the early development of infants and toddlers across global settings

INTRODUCTION

Early childhood development can be described by an underlying latent construct. Global comparisons of children's development are hindered by the lack of a validated metric that is comparable across cultures and contexts, especially for children under age 3 years. We constructed and validated a new metric, the Developmental Score (D-score), using existing data from 16 longitudinal studies.

METHODS

Studies had item-level developmental assessment data for children 0-48 months and longitudinal outcomes at ages >4-18 years, including measures of IQ and receptive vocabulary. Existing data from 11 low-income, middle-income and high-income countries were merged for >36 000 children. Item mapping produced 95 'equate groups' of same-skill items across 12 different assessment instruments. A statistical model was built using the Rasch model with item difficulties constrained to be equal in a subset of equate groups, linking instruments to a common scale, the D-score, a continuous metric with interval-scale properties. D-score-for-age z-scores (DAZ) were evaluated for discriminant, concurrent and predictive validity to outcomes in middle childhood to adolescence.

RESULTS

Concurrent validity of DAZ with original instruments was strong (average r=0.71), with few exceptions. In approximately 70% of data rounds collected across studies, DAZ discriminated between children above/below cut-points for low birth weight (<2500 g) and stunting (-2 SD below median height-for-age). DAZ increased significantly with maternal education in 55% of data rounds. Predictive correlations of DAZ with outcomes obtained 2-16 years later were generally between 0.20 and 0.40. Correlations equalled or exceeded those obtained with original instruments despite using an average of 55% fewer items to estimate the D-score.

CONCLUSION

The D-score metric enables quantitative comparisons of early childhood development across ages and sets the stage for creating simple, low-cost, global-use instruments to facilitate valid cross-national comparisons of early childhood development.

Self-reported adverse childhood experiences and quality of life among children in the two last grades of Dutch elementary education

BACKGROUND

Adverse Childhood Experiences (ACEs) may have a life-long impact on mental health and are related to physical disease, such as diabetes and cardiovascular diseases in adulthood. Research on ACEs suffers from recall bias when performed with adults.

OBJECTIVE

To estimate the prevalence of ACEs and the interrelationships between ACEs as reported by children, and to determine the impact on their self-reported quality of life (QoL). Children's opinions on the ACE-Questionnaire were also obtained.

METHOD

A cross-sectional study was conducted with a child version of the ACE-Questionnaire. This questionnaire assesses parental separation or divorce, physical and emotional child abuse and neglect, sexual violence, domestic violence, household substance abuse, psychological issues or suicide, and incarceration of a household member. QoL was measured with the Kidscreen-10.

PARTICIPANTS AND SETTING

The questionnaire was completed by 644 children at a mean age of 11 years (range 9-13 years), in the two last grades of regular elementary schools, recruited throughout the Netherlands.

RESULTS

Data were weighted by ethnicity to obtain a representative sample of children in Dutch elementary education. Of all children, 45.3% had one or more out of ten ACEs. Child maltreatment was experienced by 26.4%. ACEs often co-occurred. A higher number of ACEs correlated with a lower mean level of QoL (p <  0.001). Mean QoL was 8.5 points lower (Cohen's d = 0.8) in children who experienced child maltreatment. Children's opinions on the questionnaire were positive in 82.4%.

CONCLUSION

Prevention of ACEs, professional training and trauma-focus in schools are urgently needed.

Development of Prediction Models for Sick Leave Due to Musculoskeletal Disorders

Purpose The aim of this study was to develop prediction models to determine the risk of sick leave due to musculoskeletal disorders (MSD) in non-sick listed employees and to compare models for short-term (i.e., 3 months) and long-term (i.e., 12 months) predictions. Methods Cohort study including 49,158 Dutch employees who participated in occupational health checks between 2009 and 2015 and sick leave data recorded during 12 months follow-up. Prediction models for MSD sick leave within 3 and 12 months after the health check were developed with logistic regression analysis using routinely assessed health check variables. The performance of the prediction models was evaluated with explained variance (Nagelkerke's R-square), calibration (Hosmer-Lemeshow test) and discrimination (area under the receiver operating characteristic curve, AUC) measures. Results A total of 376 (0.8%) and 1193 (2.4%) employees had MSD sick leave within 3 and 12 months after the health check. The prediction models included similar predictor variables (educational level, musculoskeletal complaints, distress, supervisor social support, work-home interference, intrinsic motivation, development opportunities, and work pace). The explained variances were 7.6% and 8.8% for the model with 3 and 12 months follow-up, respectively. Both prediction models showed adequate calibration and discriminated between employees with and without MSD sick leave 3 months (AUC = 0.761; Interquartile range [IQR] 0.759-0.763) and 12 months (AUC = 0.740; IQR 0.738-0.741) after the health check. Conclusion The prediction models could be used to determine the risk of MSD sick leave in non-sick listed employees and invite them to preventive consultations with occupational health providers.

Educational inequalities in the impact of chronic diseases on exit from paid employment among older workers: a 7-year prospective study in the Netherlands

OBJECTIVES

The study aimed to investigate the relative and absolute risks of early exit from paid employment among older workers with a chronic disease, and to assess whether these risks differ across educational groups.

METHODS

Data on chronic diseases and demographics from 9160 Dutch workers aged 45-64 years were enriched with monthly information on employment status from Statistics Netherlands. Subdistribution hazard ratios (SHR) and 7-year probabilities among workers with a chronic disease of exit from paid employment through disability benefits, unemployment benefits, early retirement benefits or economic inactivity were estimated using competing risks regression analyses based on Fine and Gray's models.

RESULTS

Workers with one chronic disease had a higher risk to exit paid employment through disability benefits (SHR 4.48 (95%CI 3.22 to 6.25)) compared with workers without chronic disease, and this risk further increased for multiple chronic diseases (SHR 8.91 (95%CI 6.33 to 12.55)). As occurrence of chronic diseases was highest among low educated workers, the 7-year probabilities to exit paid employment through disability benefits were highest among this group. Cardiovascular, musculoskeletal, psychological and respiratory diseases were associated with disability benefits (SHRs ranging from 2.11 (95%CI 1.45 to 3.07) to 3.26 (95%CI 2.08 to 5.12)), whereas psychological diseases were also related to unemployment (SHR 1.78 (95%CI 1.33 to 2.38)).

CONCLUSIONS

Older workers with a chronic disease have a higher risk to exit paid employment through disability benefits. As multimorbidity has an additive effect, addressing multimorbidity as a risk factor for sustainable employment is needed.

Comparison of logistic-regression based methods for simple mediation analysis with a dichotomous outcome variable

Background

Logistic regression is often used for mediation analysis with a dichotomous outcome. However, previous studies showed that the indirect effect and proportion mediated are often affected by a change of scales in logistic regression models. To circumvent this, standardization has been proposed. The aim of this study was to show the relative performance of the unstandardized and standardized estimates of the indirect effect and proportion mediated based on multiple regression, structural equation modeling, and the potential outcomes framework for mediation models with a dichotomous outcome.

Methods

We compared the performance of the effect estimates yielded by the three methods using a simulation study and two real-life data examples from an observational cohort study (n = 360).

Results

Lowest bias and highest efficiency were observed for the estimates from the potential outcomes framework and for the crude indirect effect ab and the proportion mediated ab/(ab + c’) based on multiple regression and SEM.

Conclusions

We advise the use of either the potential outcomes framework estimates or the ab estimate of the indirect effect and the ab/(ab + c’) estimate of the proportion mediated based on multiple regression and SEM when mediation analysis is based on logistic regression. Standardization of the coefficients prior to estimating the indirect effect and the proportion mediated may not increase the performance of these estimates.

Childhood trauma and the role of self-blame on psychological well-being after deployment in male veterans

Background: Childhood trauma and combat-related trauma are both associated with decreased psychosocial functioning. Coping strategies play an important role in the adjustment to traumatic events. Objective: The present study examined childhood trauma and the mediating role of coping strategies in adult psychological symptoms in a non-clinical military population after deployment to Afghanistan. Additionally, the moderating role of coping strategies in vulnerability to combat events was explored. Method: Participants (N = 932) were drawn from a prospective study assessing psychological complaints (SCL-90), early trauma (ETISR-SF), combat-related events and coping strategies (Brief COPE). Mediation analyses via joint significance testing and moderation analyses were performed. Results: Childhood trauma is related to adult symptoms of general anxiety, depression and problems concerning interpersonal sensitivity through the mediation of self-blame as a coping strategy. Some evidence was found that self-blame moderated vulnerability to combat-related events resulting in psychological complaints, specifically symptoms of anxiety and depression. Conclusions: Military personnel should be made aware of self-criticizing maladaptive belief systems when dealing with aversive events. Negative beliefs about oneself and distorted trauma-related cognitions may have a basis in childhood events. Self-blame cognitions may be a potential mechanism of change in empirically supported trauma interventions such as cognitive processing therapy.

Development of the Arm Function in Multiple Sclerosis Questionnaire-Short Form (AMSQ-SF): A static 10-item version

BACKGROUND

Assessing arm and hand function of multiple sclerosis (MS) patients is important as impaired functioning may impact daily activities and reduce quality of life.

OBJECTIVE

A short-form of the Arm Function in Multiple Sclerosis Questionnaire (AMSQ), a recently developed patient-reported outcome measure containing 31 items, is developed to allow non-adaptive application.

METHODS

Complete data from 690 patients with MS, recruited via outpatient clinics, a residential center or via a Dutch website aimed at MS patients, were included in the analyses. A graded response model was fit to these data to estimate item response theory (IRT) parameters, which were used to perform post hoc computerized adaptive test (CAT) simulations with a cutoff standard error of measurement (SEM) of 0.32. The optimal test length was determined by the correlation between the static short-form and full-length theta, the mean SEM, and the amount of patients reaching a satisfactory SEM in CAT simulations.

RESULTS AND CONCLUSION

Based on five selection criteria (i.e. discrimination parameters, total information, times selected in CAT simulations, raw item means, and item content), 10 items were selected for inclusion in the short-form. The score on the final 10-item short-form correlated strongly with the full-length AMSQ and provided reliable ability estimations, indicating its usefulness instrument in research and clinical settings.

Gestational weight gain charts for different body mass index groups for women in Europe, North America, and Oceania

BACKGROUND

Gestational weight gain differs according to pre-pregnancy body mass index and is related to the risks of adverse maternal and child health outcomes. Gestational weight gain charts for women in different pre-pregnancy body mass index groups enable identification of women and offspring at risk for adverse health outcomes. We aimed to construct gestational weight gain reference charts for underweight, normal weight, overweight, and grades 1, 2 and 3 obese women and to compare these charts with those obtained in women with uncomplicated term pregnancies.

METHODS

We used individual participant data from 218,216 pregnant women participating in 33 cohorts from Europe, North America, and Oceania. Of these women, 9065 (4.2%), 148,697 (68.1%), 42,678 (19.6%), 13,084 (6.0%), 3597 (1.6%), and 1095 (0.5%) were underweight, normal weight, overweight, and grades 1, 2, and 3 obese women, respectively. A total of 138, 517 women from 26 cohorts had pregnancies with no hypertensive or diabetic disorders and with term deliveries of appropriate for gestational age at birth infants. Gestational weight gain charts for underweight, normal weight, overweight, and grade 1, 2, and 3 obese women were derived by the Box-Cox t method using the generalized additive model for location, scale, and shape.

RESULTS

We observed that gestational weight gain strongly differed per maternal pre-pregnancy body mass index group. The median (interquartile range) gestational weight gain at 40 weeks was 14.2 kg (11.4-17.4) for underweight women, 14.5 kg (11.5-17.7) for normal weight women, 13.9 kg (10.1-17.9) for overweight women, and 11.2 kg (7.0-15.7), 8.7 kg (4.3-13.4) and 6.3 kg (1.9-11.1) for grades 1, 2, and 3 obese women, respectively. The rate of weight gain was lower in the first half than in the second half of pregnancy. No differences in the patterns of weight gain were observed between cohorts or countries. Similar weight gain patterns were observed in mothers without pregnancy complications.

CONCLUSIONS

Gestational weight gain patterns are strongly related to pre-pregnancy body mass index. The derived charts can be used to assess gestational weight gain in etiological research and as a monitoring tool for weight gain during pregnancy in clinical practice.

MR Imaging Features of Retinoblastoma: Association with Gene Expression Profiles

Purpose To identify associations between magnetic resonance (MR) imaging features and gene expression in retinoblastoma. Materials and Methods A retinoblastoma MR imaging atlas was validated by using anonymized MR images from referral centers in Essen, Germany, and Paris, France. Images were from 39 patients with retinoblastoma (16 male and 18 female patients [the sex in five patients was unknown]; age range, 5-90 months; inclusion criterion: pretreatment MR imaging). This atlas was used to compare MR imaging features with genome-wide messenger RNA (mRNA) expression data from 60 consecutive patients obtained from 1995 to 2012 (35 male patients [58%]; age range, 2-69 months; inclusion criteria: pretreatment MR imaging, genome-wide mRNA expression data available). Imaging pathway associations were analyzed by means of gene enrichment. In addition, imaging features were compared with a predefined gene expression signature of photoreceptorness. Statistical analysis was performed with generalized linear modeling of radiology traits on normalized log2-transformed expression values. P values were corrected for multiple hypothesis testing. Results Radiogenomic analysis revealed 1336 differentially expressed genes for qualitative imaging features (threshold P = .05 after multiple hypothesis correction). Loss of photoreceptorness gene expression correlated with advanced stage imaging features, including multiple lesions (P = .03) and greater eye size (P < .001). The number of lesions on MR images was associated with expression of MYCN (P = .04). A newly defined radiophenotype of diffuse-growing, plaque-shaped, multifocal tumors displayed overexpression of SERTAD3 (P = .003, P = .049, and P = .06, respectively), a protein that stimulates cell growth by activating the E2F network. Conclusion Radiogenomic biomarkers can potentially help predict molecular features, such as photoreceptorness loss, that indicate tumor progression. Results imply a possible role for radiogenomics in future staging and treatment decision making in retinoblastoma.

Interrater and intrarater agreement of confocal microscopy imaging in diagnosing and subtyping basal cell carcinoma

BACKGROUND

Reflectance confocal microscopy (RCM) imaging can be used to diagnose and subtype basal cell carcinoma (BCC) but relies on individual morphologic pattern recognition that might vary among users.

OBJECTIVES

We assessed the inter-rater and intrarater agreement of RCM in correctly diagnosing and subtyping BCC.

METHODS

In this prospective study, we evaluated the inter-rater and intrarater agreement of RCM on BCC presence and subtype among three raters with varying experience who independently assessed static images of 48 RCM cases twice with four-week interval (T1 and T2). Histopathologic confirmation of presence and subtype of BCC from surgical excision specimen was defined as the reference standard.

RESULTS

The inter-rater agreement of RCM for BCC presence showed an agreement of 82% at T1 and 84% at T2. The agreements for subtyping BCC were lower (52% for T1 and 47% for T2). The intrarater agreement of RCM for BCC presence showed an observed agreement that varied from 79% to 92%. The observed agreements for subtyping varied from 56% to 71%.

CONCLUSIONS

In conclusion, our results show that RCM is reliable in correctly diagnosing BCC based on the assessment of static RCM images. RCM could potentially play an important role in BCC management if accurate subtyping will be achieved. Therefore, future clinical studies on reliability and specific RCM features for BCC subtypes are required.

Effectiveness of a Web-Based Tailored Intervention With Virtual Assistants Promoting the Acceptability of HPV Vaccination Among Mothers of Invited Girls: Randomized Controlled Trial

BACKGROUND

In 2010, the human papillomavirus (HPV) vaccination was introduced in the Dutch National Immunization Program for 12-year-old girls, aiming to reduce the incidence of cervical cancer in women. HPV vaccination uptake turned out to be lower than expected: 61% versus 70%, respectively. Mothers were shown to play the most important role in the immunization decision about this vaccination. They had also expressed their need for interactive personal information about the HPV vaccination over and above the existing universal general information. To improve the effectiveness of the existing education about the HPV vaccination, we systematically developed a Web-based tailored intervention with virtual assistants providing mothers of girls to be invited with tailored feedback on their decision making about the HPV vaccination.

OBJECTIVE

The aim of this study was to evaluate the effectiveness of the Web-based tailored intervention for promoting HPV vaccination acceptance by means of a randomized controlled trial (RCT).

METHODS

Mothers were recruited via the Dutch vaccination register (Praeventis) (n=36,000) and three Web-based panels (n=2483). Those who gave informed consent (N=8062) were randomly assigned to the control (n=4067) or intervention condition (n=3995). HPV vaccination uptake, as registered by Praeventis once the HPV vaccination round was completed, was used as the primary outcome. Secondary outcomes were differential scores across conditions between baseline (before the provided access to the new tailored intervention) and follow-up (just before the first vaccination) regarding the mothers' degree of informed decision making (IDM), decisional conflict, and critical determinants of HPV vaccination uptake among which are intention, attitude, risk perception, and outcome beliefs.

RESULTS

Intention-to-treat analysis (N=8062) showed a significant positive effect of the intervention on IDM, decisional conflict, and nearly all determinants of HPV vaccination uptake (P<.001). No effect was found on uptake (P=.60). This may be attributed to the overall high uptake rates in both conditions. Mothers evaluated the intervention as highly positive, including the website as well as the virtual assistants that were used to deliver the tailored feedback.

CONCLUSIONS

This computer-tailored intervention has the potential to improve HPV vaccination acceptability and IDM and to decrease decisional conflict among mothers of invited girls. Implications for future research are discussed.

TRIAL REGISTRATION

Trialregister.nl NTR4935; http://www.trialregister.nl/trialreg/admin/rctview.asp?TC=4935 (Archived by WebCite at http://www.webcitation.org/6srT7l9EM).

Methods for significance testing of categorical covariates in logistic regression models after multiple imputation: power and applicability analysis

Background

Multiple imputation is a recommended method to handle missing data. For significance testing after multiple imputation, Rubin’s Rules (RR) are easily applied to pool parameter estimates. In a logistic regression model, to consider whether a categorical covariate with more than two levels significantly contributes to the model, different methods are available. For example pooling chi-square tests with multiple degrees of freedom, pooling likelihood ratio test statistics, and pooling based on the covariance matrix of the regression model. These methods are more complex than RR and are not available in all mainstream statistical software packages. In addition, they do not always obtain optimal power levels. We argue that the median of the p-values from the overall significance tests from the analyses on the imputed datasets can be used as an alternative pooling rule for categorical variables. The aim of the current study is to compare different methods to test a categorical variable for significance after multiple imputation on applicability and power.

Methods

In a large simulation study, we demonstrated the control of the type I error and power levels of different pooling methods for categorical variables.

Results

This simulation study showed that for non-significant categorical covariates the type I error is controlled and the statistical power of the median pooling rule was at least equal to current multiple parameter tests. An empirical data example showed similar results.

Conclusions

It can therefore be concluded that using the median of the p-values from the imputed data analyses is an attractive and easy to use alternative method for significance testing of categorical variables.

Electronic supplementary material

The online version of this article (doi:10.1186/s12874-017-0404-7) contains supplementary material, which is available to authorized users.

Motivational and contextual determinants of HPV-vaccination uptake: A longitudinal study among mothers of girls invited for the HPV-vaccination

BACKGROUND

In the Netherlands, HPV-vaccination uptake among 12-year-old girls remains to be lower (61% in 2016) than expected. The present study is about 1) replicating the extent to which social-psychological determinants found in earlier cross-sectional studies explain HPV-vaccination intention, and 2) testing whether HPV-vaccination intention, as well as other social-psychological determinants, are good predictors of future HPV-vaccination uptake in a longitudinal design.

METHODS

A random sample of mothers of girls invited for the vaccination in 2015 was drawn from the Dutch vaccination register (Praeventis) (N=36,000) and from three online panels (N=2483). Two months prior to the vaccination of girls, their mothers were requested to complete a web-based questionnaire by letter (Praeventis sample) or by e-mail (panel samples). HPV-vaccination uptake was derived from Praeventis. Backward linear and logistic regression analyses were conducted to examine most dominant predictors of HPV-vaccination intention and uptake, respectively. The total sample used for data analyses consisted of 8062 mothers. Response rates were 18% for the Praeventis sample and 47% for the panel samples.

RESULTS

HPV-vaccination intention was best explained by attitude, beliefs, subjective norms, habit, and perceived relative effectiveness of the vaccination; they explained 83% of the variance in HPV-vaccination intention. Intention appeared to be the only stable predictor of HPV-vaccination uptake and explained 43% of the variance in HPV-vaccination uptake.

CONCLUSIONS

These results confirm what was found by earlier cross-sectional studies, and provide strong leads for selecting relevant targets in the planning of future communication strategies aiming to improve HPV-vaccination uptake.

Proteomics to predict the response to tumour necrosis factor-α inhibitors in rheumatoid arthritis using a supervised cluster-analysis based protein score

OBJECTIVE

In rheumatoid arthritis (RA), it is of major importance to identify non-responders to tumour necrosis factor-α inhibitors (TNFi) before starting treatment, to prevent a delay in effective treatment. We developed a protein score for the response to TNFi treatment in RA and investigated its predictive value.

METHOD

In RA patients eligible for biological treatment included in the BiOCURA registry, 53 inflammatory proteins were measured using xMAP® technology. A supervised cluster analysis method, partial least squares (PLS), was used to select the best combination of proteins. Using logistic regression, a predictive model containing readily available clinical parameters was developed and the potential of this model with and without the protein score to predict European League Against Rheumatism (EULAR) response was assessed using the area under the receiving operating characteristics curve (AUC-ROC) and the net reclassification index (NRI).

RESULTS

For the development step (n = 65 patient), PLS revealed 12 important proteins: CCL3 (macrophage inflammatory protein, MIP1a), CCL17 (thymus and activation-regulated chemokine), CCL19 (MIP3b), CCL22 (macrophage-derived chemokine), interleukin-4 (IL-4), IL-6, IL-7, IL-15, soluble cluster of differentiation 14 (sCD14), sCD74 (macrophage migration inhibitory factor), soluble IL-1 receptor I, and soluble tumour necrosis factor receptor II. The protein score scarcely improved the AUC-ROC (0.72 to 0.77) and the ability to improve classification and reclassification (NRI = 0.05). In validation (n = 185), the model including protein score did not improve the AUC-ROC (0.71 to 0.67) or the reclassification (NRI = -0.11).

CONCLUSION

No proteomic predictors were identified that were more suitable than clinical parameters in distinguishing TNFi non-responders from responders before the start of treatment. As the results of previous studies and this study are disparate, we currently have no proteomic predictors for the response to TNFi.

The association of self-regulation with weight loss maintenance after an intensive combined lifestyle intervention for children and adolescents with severe obesity

BACKGROUND

Knowledge is limited on the role the ability to self-regulate plays in the long-term outcome of obesity treatment in children and adolescents with severe obesity. The purpose of this study was to determine whether the ability to self-regulate after an one year intensive, partly inpatient, combined lifestyle intervention is associated with weight loss maintenance in children and adolescents with severe obesity.

METHODS

One hundred twenty participants (8-19 years) with an average SDS-BMI of 3.41 and their parents/caregivers were included in an intervention study. As primary determinant of weight loss maintenance, general self-regulation ability was evaluated using two behavioral computer tasks assessing inhibitory control and sensitivity to reward.

RESULTS

There was no association between inhibitory control at T12 and ∆SDS-BMI between T12 and T24 (β = 0.0002; CI 95% = -0.0010-0.0014; P = 0.761). There was also no relation between sensitivity to reward at T12 and ∆SDS-BMI between T12 and T24 (β = -0.0028; CI 95% = -0.0075-0.0019; P = 0.244). None of the psychosocial factors that were examined as moderators, showed a statistically significant interaction, except for parental feeding style (P = 0.023).

CONCLUSIONS

The ability to self-regulate after an intensive, partly inpatient, multidisciplinary one year intervention for severe obesity in children and adolescents was not associated with the ability to maintain the achieved weight loss during the following year. Factors that explain the large range of long term outcomes need to be elucidated.

TRIAL REGISTRATION

Netherlands Trial Register (NTR1678, registered 20-Feb-2009).

The anchor-based minimal important change, based on receiver operating characteristic analysis or predictive modeling, may need to be adjusted for the proportion of improved patients

OBJECTIVES

Patients have their individual minimal important changes (iMICs) as their personal benchmarks to determine whether a perceived health-related quality of life (HRQOL) change constitutes a (minimally) important change for them. We denote the mean iMIC in a group of patients as the "genuine MIC" (gMIC). The aims of this paper are (1) to examine the relationship between the gMIC and the anchor-based minimal important change (MIC), determined by receiver operating characteristic analysis or by predictive modeling; (2) to examine the impact of the proportion of improved patients on these MICs; and (3) to explore the possibility to adjust the MIC for the influence of the proportion of improved patients.

STUDY DESIGN AND SETTING

Multiple simulations of patient samples involved in anchor-based MIC studies with different characteristics of HRQOL (change) scores and distributions of iMICs. In addition, a real data set is analyzed for illustration.

RESULTS

The receiver operating characteristic-based and predictive modeling MICs equal the gMIC when the proportion of improved patients equals 0.5. The MIC is estimated higher than the gMIC when the proportion improved is greater than 0.5, and the MIC is estimated lower than the gMIC when the proportion improved is less than 0.5. Using an equation including the predictive modeling MIC, the log-odds of improvement, the standard deviation of the HRQOL change score, and the correlation between the HRQOL change score and the anchor results in an adjusted MIC reflecting the gMIC irrespective of the proportion of improved patients.

CONCLUSION

Adjusting the predictive modeling MIC for the proportion of improved patients assures that the adjusted MIC reflects the gMIC.

LIMITATIONS

We assumed normal distributions and global perceived change scores that were independent on the follow-up score. Additionally, floor and ceiling effects were not taken into account.

Multiple imputation strategies for zero-inflated cost data in economic evaluations: which method works best?

Cost and effect data often have missing data because economic evaluations are frequently added onto clinical studies where cost data are rarely the primary outcome. The objective of this article was to investigate which multiple imputation strategy is most appropriate to use for missing cost-effectiveness data in a randomized controlled trial. Three incomplete data sets were generated from a complete reference data set with 17, 35 and 50 % missing data in effects and costs. The strategies evaluated included complete case analysis (CCA), multiple imputation with predictive mean matching (MI-PMM), MI-PMM on log-transformed costs (log MI-PMM), and a two-step MI. Mean cost and effect estimates, standard errors and incremental net benefits were compared with the results of the analyses on the complete reference data set. The CCA, MI-PMM, and the two-step MI strategy diverged from the results for the reference data set when the amount of missing data increased. In contrast, the estimates of the Log MI-PMM strategy remained stable irrespective of the amount of missing data. MI provided better estimates than CCA in all scenarios. With low amounts of missing data the MI strategies appeared equivalent but we recommend using the log MI-PMM with missing data greater than 35 %.

The Cost-Effectiveness of Two Forms of Case Management Compared to a Control Group for Persons with Dementia and Their Informal Caregivers from a Societal Perspective

OBJECTIVES

The objective of this article was to compare the costs and cost-effectiveness of the two most prominent types of case management in the Netherlands (intensive case management and linkage models) against no access to case management (control group) for people with already diagnosed dementia and their informal caregivers.

METHODS

The economic evaluation was conducted from a societal perspective embedded within a two year prospective, observational, controlled, cohort study with 521 informal caregivers and community-dwelling persons with dementia. Case management provided within one care organization (intensive case management model, ICMM), case management where care was provided by different care organizations within one region (Linkage model, LM), and a group with no access to case management (control) were compared. The economic evaluation related incremental costs to incremental effects regarding neuropsychiatric symptoms (NPI), psychological health of the informal caregiver (GHQ-12), and quality adjusted life years (QALY) of the person with dementia and informal caregiver.

RESULTS

Inverse-propensity-score-weighted models showed no significant differences in clinical or total cost outcomes between the three groups. Informal care costs were significantly lower in the ICMM group compared to both other groups. Day center costs were significantly lower in the ICMM group compared to the control group. For all outcomes, the probability that the ICMM was cost-effective in comparison with LM and the control group was larger than 0.97 at a threshold ratio of 0 €/incremental unit of effect.

CONCLUSION

This study provides preliminary evidence that the ICMM is cost-effective compared to the control group and the LM. However, the findings should be interpreted with caution since this study was not a randomized controlled trial.

Passive imputation and parcel summaries are both valid to handle missing items in studies with many multi-item scales

Previous studies showed that missing data in multi-item scales can best be handled by multiple imputation of item scores. However, when many scales are used, the number of items will become too large for the imputation model to reliably estimate imputations. A solution is to use passive imputation or a parcel summary score that combine and consequently reduce the number of variables in the imputation model. The performance of these methods was evaluated in a simulation study and illustrated in an example. Passive imputation, which updated scale scores from imputed items, and parcel summary scores that use the average over available item scores were compared to using all items simultaneously, imputing total scores of scales and complete-case analysis. Scale scores and coefficient estimates from linear regression were compared to “true” parameters on bias and precision. Passive imputation and using parcel summaries showed smaller bias and more precision than imputing total scores and complete-case analyses. Passive imputation or using parcel summary scores are valid missing data solutions in studies that include many multi-item scales.

Post-traumatic stress symptoms 5 years after military deployment to Afghanistan: an observational cohort study

BACKGROUND

Deployment can put soldiers at risk of developing post-traumatic stress symptoms. Despite several longitudinal studies, little is known about the timing of an increase in post-traumatic stress symptoms relative to pre-deployment. Longitudinal studies starting pre-deployment, in which participants are repeatedly measured over time, are warranted to assess the timing of an increase in symptoms to ultimately assess the timing of an increase in treatment demand after deployment.

METHODS

In this large observational cohort study, Dutch military personnel who were deployed to Afghanistan as part of the International Security Assistance Forces between March, 2005, and September, 2008, were assessed for post-traumatic stress symptoms with the Self-Rating Inventory for Post-traumatic Stress Disorder (SRIP) questionnaire. Participants were assessed 1 month before deployment and followed up at 1 month, 6 months, 12 months, 2 years, and 5 years after deployment, with changes in SRIP scores compared with pre-deployment using a mixed model analysis. The primary outcome was the total score of post-traumatic stress symptoms measured with SRIP at pre-deployment and the five follow-up assessments, with a score of 38 used as the cutoff to indicate substantial post-traumatic stress symptoms.

FINDINGS

Between March, 2005, and September, 2008, 1007 participants were recruited to this study. The results show two important effects of deployment on post-traumatic stress symptoms. A short-term symptom increase within the first 6 months after deployment (symptom increase coefficient for SRIP score vs pre-deployment [β] 0·99, 95% CI 0·50-1·48); and a long-term symptom increase at 5 years after deployment (β 1·67, 1·14-2·20).

INTERPRETATION

This study underlines the importance of long-term monitoring of the psychological health of soldiers after deployment because early detection of symptoms is essential to early treatment, which is related to improved psychological health.

FUNDING

Dutch Ministry of Defense.

Minimal important change (MIC) based on a predictive modeling approach was more precise than MIC based on ROC analysis

OBJECTIVES

To present a new method to estimate a "minimal important change" (MIC) of health-related quality of life (HRQOL) scales, based on predictive modeling, and to compare its performance with the MIC based on receiver operating characteristic (ROC) analysis. To illustrate how the new method deals with variables that modify the MIC across subgroups.

STUDY DESIGN AND SETTING

The new method uses logistic regression analysis and identifies the change score associated with a likelihood ratio of 1 as the MIC. Simulation studies were conducted to investigate under which distributional circumstances both methods produce concordant or discordant results and whether the methods differ in accuracy and precision.

RESULTS

The "predictive MIC" and the ROC-based MIC were identical when the variances of the change scores in the improved and not-improved groups were equal and the distributions were normal or oppositely skewed. The predictive MIC turned out to be more precise than the ROC-based MIC. The predictive MIC allowed for the testing and estimation of modifying factors such as baseline severity.

CONCLUSION

In many situations, the newly described MIC based on predictive modeling yields the same value as the ROC-based MIC but with significantly greater precision. This advantage translates to increased statistical power in MIC studies.

Including auxiliary item information in longitudinal data analyses improved handling missing questionnaire outcome data

OBJECTIVES

Previous studies show that missing values in multi-item questionnaires can best be handled at item score level. The aim of this study was to demonstrate two novel methods for dealing with incomplete item scores in outcome variables in longitudinal studies. The performance of these methods was previously examined in a simulation study. The two methods incorporate item information at the background when simultaneously the study outcomes are estimated.

STUDY DESIGN AND SETTING

The investigated methods include the item scores or a summary of a parcel of available item scores as auxiliary variables while using the total score of the multi-item questionnaire as the main focus of the analysis in a latent growth model. That way the items help estimating the incomplete information of the total scores. The methods are demonstrated in two empirical data sets.

RESULTS

Including the item information results in more precise outcomes in terms of regression coefficient estimates and standard errors, compared with not including item information in the analysis.

CONCLUSION

The inclusion of a parcel summary is an efficient method that does not overcomplicate longitudinal growth estimates. Therefore, it is recommended in situations where multi-item questionnaires are used as outcome measure in longitudinal clinical studies with incomplete scores because of missing item scores.

Missing data in a multi-item instrument were best handled by multiple imputation at the item score level

OBJECTIVES

Regardless of the proportion of missing values, complete-case analysis is most frequently applied, although advanced techniques such as multiple imputation (MI) are available. The objective of this study was to explore the performance of simple and more advanced methods for handling missing data in cases when some, many, or all item scores are missing in a multi-item instrument.

STUDY DESIGN AND SETTING

Real-life missing data situations were simulated in a multi-item variable used as a covariate in a linear regression model. Various missing data mechanisms were simulated with an increasing percentage of missing data. Subsequently, several techniques to handle missing data were applied to decide on the most optimal technique for each scenario. Fitted regression coefficients were compared using the bias and coverage as performance parameters.

RESULTS

Mean imputation caused biased estimates in every missing data scenario when data are missing for more than 10% of the subjects. Furthermore, when a large percentage of subjects had missing items (>25%), MI methods applied to the items outperformed methods applied to the total score.

CONCLUSION

We recommend applying MI to the item scores to get the most accurate regression model estimates. Moreover, we advise not to use any form of mean imputation to handle missing data.