Account Missing Data Research Articles

The effects and process of the intervention “Individual Shantala Infant Massage” in preventive child healthcare to improve parent–child interaction: study protocol for a quasi-experimental study

BackgroundIndividual Shantala Infant Massage is an intervention that is offered by several Dutch Preventive Child Healthcare (PCH) organizations as optional preventive support, in addition to basic care as offered to all children. It targets vulnerable families and aims to enhance sensitive parenting and to reduce (effects of) parental stress. The intervention is carried out by a certified nurse. It consists of three structured home visits. Parents learn to massage their infant and receive parenting support. This study aims to investigate the effectiveness and the process of the intervention. The main hypothesis is that Individual Shantala Infant Massage leads to increased parental sensitive responsiveness, lower perceived and physiological parental stress, and improved child growth and development in the intervention group, compared to a control group where this intervention is not offered by PCH. Secondary research questions address effects on parenting confidence and parental concerns regarding the infant, the influence of background characteristics and the intervention process.MethodsThe study is a quasi-experimental non-randomized trial. The aim is to include 150 infant-parent dyads in both the intervention and the control group. This takes into account possible attrition and missing data as 105 dyads with complete data per group are sufficient for analysis. All participants complete questionnaires at T0 (pre-test, child age between six-sixteen weeks), T1 (post-intervention, or ± four weeks after T0), and T2 (follow-up at five months). At T2, a hair tuft is cut from the parents’ head to measure hair cortisol levels. Data on infant growth and development is obtained from PCH files. In the intervention group, additional data is collected to evaluate the intervention process: parents complete an evaluation questionnaire at T1, nurses keep semi-structured logbooks of intervention sessions and interviews are conducted with parents and professionals.DiscussionStudy results can contribute to the evidence base of infant massage as applied in Dutch PCH, and can inform parents, PCH practitioners, policy makers and researchers both inside and outside the Netherlands on feasibility and effectiveness of the infant massage intervention as applied in this format and setting.Trial registrationISRCTN registry: ISRCTN16929184. Date (retrospectively) registered: 29/03/2022.

A Bayesian predictive analytics model for improving long range epidemic forecasting during an infection wave

Following the outbreak of the coronavirus epidemic in early 2020, municipalities, regional governments and policymakers worldwide had to plan their Non-Pharmaceutical Interventions (NPIs) amidst a scenario of great uncertainty. At this early stage of an epidemic, where no vaccine or medical treatment is in sight, algorithmic prediction can become a powerful tool to inform local policymaking. However, when we replicated one prominent epidemiological model to inform health authorities in a region in the south of Brazil, we found that this model relied too heavily on manually predetermined covariates and was too reactive to changes in data trends. Our four proposed models access data of both daily reported deaths and infections as well as take into account missing data (e.g., the under-reporting of cases) more explicitly, with two of the proposed versions also attempting to model the delay in test reporting. We simulated weekly forecasting of deaths from the period from 31/05/2020 until 31/01/2021, with first week data being used as a cold-start to the algorithm, after which we use a lighter variant of the model for faster forecasting. Because our models are significantly more proactive in identifying trend changes, this has improved forecasting, especially in long-range predictions and after the peak of an infection wave, as they were quicker to adapt to scenarios after these peaks in reported deaths. Assuming reported cases were under-reported greatly benefited the model in its stability, and modelling retroactively-added data (due to the “hot” nature of the data used) had a negligible impact on performance.

SMIM: A unified framework of survival sensitivity analysis using multiple imputation and martingale.

Censored survival data are common in clinical trial studies. We propose a unified framework for sensitivity analysis to censoring at random in survival data using multiple imputation and martingale, called SMIM. The proposed framework adopts the δ-adjusted and control-based models, indexed by the sensitivity parameter, entailing censoring at random and a wide collection of censoring not at random assumptions. Also, it targets a broad class of treatment effect estimands defined as functionals of treatment-specific survival functions, taking into account missing data due to censoring. Multiple imputation facilitates the use of simple full-sample estimation; however, the standard Rubin's combining rule may overestimate the variance for inference in the sensitivity analysis framework. We decompose the multiple imputation estimator into a martingale series based on the sequential construction of the estimator and propose the wild bootstrap inference by resampling the martingale series. The new bootstrap inference has a theoretical guarantee for consistency and is computationally efficient compared to the nonparametric bootstrap counterpart. We evaluate the finite-sample performance of the proposed SMIM through simulation and an application on an HIV clinicaltrial.

Factors associated with non-nutritive sucking habits at 2 years of age among very preterm children: EPIPAGE-2 cohort study.

The association between prolonged non-nutritive sucking habits (NNSHs, ie, sucking pacifiers or fingers) and maxillofacial growth anomalies in the general population has been widely described. Because maturation of sucking abilities is not fully achieved in very preterm infants (<32 weeks' gestation), neonatal services worldwide rely on the use of pacifiers to promote the development of adequate sucking reflexes, possibly prolonging NNSHs during infancy. We aimed to describe the frequency and to identify factors associated with NNSHs at age 2 years in very preterm children. The study was based on data from EPIPAGE-2, a French national prospective cohort study of preterm births during 2011 that included 2593 children born between 24 and 31 weeks' gestation. The primary outcome was NNSHs at 2 years. Multivariable log-linear regression models with generalized estimation equations were used to study the association between the characteristics studied and NNSHs. Multiple imputations were used to take into account missing data. The frequency of NNSHs was 69% in the overall sample but higher among girls (adjusted risk ratio [RR] 1.12, 95% confidence interval [CI] 1.05, 1.17), children born from multiple pregnancies (eg, twins/triplets) (RR 1.07, 95% CI 1.00, 1.11), children who were fed by nasogastric tube (RR 1.07, 95% CI 1.01, 1.13), or those who benefitted from developmental care programmes (RR 1.10, 95% CI 1.02, 1.19). The NNSHs frequency was lower if mothers were not born in France (RR 0.70, 95% CI 0.64, 0.77), children had 2 or more older siblings (RR 0.88, 95% CI 0.82, 0.96), or children were breast-fed at discharge (RR 0.90, 95% CI 0.85, 0.95). NNSHs at 2 years seemed associated with cultural background, development care programmes, and breast feeding. Whether NNSHs at 2 years among very preterm children are associated with future maxillofacial growth anomalies deserves further attention.

Analysis of 7Be and 210Pb concentration and 7Be/210Pb activity ratio in ground level air in Palermo (Italy)

ABSTRACTA series of measurements of particulate samples collected on paper filters in Palermo (Italy) during 1995–2005 has been performed to detect Pb air activity concentration and, along with previous similar data on Be, to evaluate their activity ratio. Average values for the daily activity concentrations of Be and Pb are found to be 4.7 and 0.6 mBq m, respectively, 9.8 being the average activity ratio. Data confirm a seasonal behavior for both radionuclides. A correlation between Be and Pb air activity concentrations seems to exist, despite their different origins. The time behaviors of both Be and Pb air concentrations and Be/Pb activity ratio have been compared with some meteorological parameter time series. The analysis of the data shows a weak correlation with rainfall, while no correlation can be established with other parameters. Monthly and weekly average values have been reconstructed taking into account missing data, in order to carry out homogeneous comparisons with available data from other geographical areas. Data series correlations and the frequency analysis show the importance of atmospheric transport processes and the influence of meteorological and climatic parameters (in particular rainfall) in determining the concentrations in air of both radionuclides.

Biogas production from submerged macrophytes\u2014a case study of regional biomass potentials in Germany

BackgroundUtilization of energy crops for biogas production has been controversially discussed in Germany due to negative environmental effects and the “food vs. fuel” debate. This led to a search for alternative substrates focusing on material from landscape management measures. Aquatic biomass is harvested during water body management, yet it has not been considered for energy generation.MethodsThe information where and which amount of biomass is collected by aquatic de-weeding operations in rivers and lakes was gathered via a nationwide survey. In addition to that, the amount of aquatic plant biomass potentially available in water bodies was estimated exemplarily for the flowing waters of Baden-Württemberg—by using data from the European Water Framework Directive surveys.ResultsThe survey revealed 172 locations of de-weeding operations in flowing waters and 93 in standing waters. These locations are concentrated in lowland rivers of the North German Plain as well as the Upper Rhine Plain. Standing water de-weeding operations were reported mainly from the foothills of the Alps. The overall amount of biomass harvested per year is 36,244 t of fresh biomass. Taking into account missing data, a maximum of 100,000 t of fresh biomass per year can be estimated for Germany. The case study on plant biomass de-weeded from waters in Germany revealed that only a small part of the total aquatic plant biomass is actually harvested.ConclusionsThe amount of biomass harvested and removed from water bodies in Germany is considerably lower than the harvest of other substrates from landscape management measures such as mowing of meadows or trimming of trees and hedges. However, larger amounts are accumulating locally, concentrated in some regions or at specific water bodies, e.g., reservoirs, for which regional value chains could be established. In order to make the exploitation of these local potentials economically viable, changes regarding the economic and technological framework are required.

Combining RNA-seq and ddRAD-seq to SNP discovery of a highly migratory and apex predator

Combining RNA-seq and ddRAD-seq to SNP discovery of a highly migratory and apex predator

Electroacupuncture plus standard of care for managing refractory functional dyspepsia: protocol of a pragmatic trial with economic evaluation

IntroductionThis trial proposes to compare the effectiveness and cost-effectiveness of electroacupuncture (EA) plus on-demand gastrocaine with waiting list for EA plus on-demand gastrocaine in providing symptom relief and quality-of-life improvement...

Model-based co-clustering for ordinal data

A model-based co-clustering algorithm for ordinal data is presented. This algorithm relies on the latent block model embedding a probability distribution specific to ordinal data (the so-called BOS or Binary Ordinal Search distribution). Model inference relies on a Stochastic EM algorithm coupled with a Gibbs sampler, and the ICL-BIC criterion is used for selecting the number of co-clusters (or blocks). The main advantage of this ordinal dedicated co-clustering model is its parsimony, the interpretability of the co-cluster parameters (mode, precision) and the possibility to take into account missing data. Numerical experiments on simulated data show the efficiency of the inference strategy, and real data analyses illustrate the interest of the proposed procedure.

‘Quest for Quality’: an evaluation of the impact on elderly care

The Quest for Quality scheme was introduced with the aim of reducing hospital admission rates from care homes through interventions conducted by a multidisciplinary team (MDT). The objective of this was to study the types of interventions being made and establish their impact. The data was collected directly by the MDT prior to the research and therefore the analysis was retrospective. Exclusion criteria were set for each part of the analysis, taking into account missing data. A flow chart based on situations was developed allowing interventions to be classified into British National Formulary (BNF) categories and then assigned a priority. The categories found to have the most interventions were related to the gastrointestinal system (140) and to nutrition and blood (128), findings that were anticipated due to the medical issues arising in the elderly. One of the major problems identified was in the follow-up of interventions. Overall, 35% of intervention outcomes were unknown and only 20% were actioned. Reasons given for this were issues with patient record systems and the interventions not being received by the prescriber; a system allowing interventions to be tracked was therefore suggested. The causation between increasing reviews and decreasing hospitalisations could not be proved (Pearson correlation=−0.983, CI 95%) due to insufficient data. It was, however, shown that the MDT were making appropriate and valuable interventions, contributing to a reduction in polypharmacy and better quality of care when taking into account the specific case studies identified. With Quest for Quality now recommissioned, it will continue to make improvements to the benefit of patients.

0133 : Identifying familial hypercholesterolemia from registries of patients with acute myocardial infarction: an algorithm-based approach

Familial hypercholesterolemia (FH) is at very high risk of early myocardial infarction (MI). The prevalence of FH, which is estimated to be at least 1:500 in the general population, remains unclear in patients with acute MI. From databases of 3 French regional and nationwide registries of acute MI (RICO and FAST-MI 2005 and 2010, respectively), we aimed to determine FH prevalence by developing a specific algorithm. Consecutive patients with AMI ≤48 hours of onset included 1) in FAST-MI : during a one-month period in 223 institutions at the end of 2005 and 213 institutions at the end of 2010, and 2) in RICO :from January 2001 – December 2013 (≈ 13 y), were considered in the 3 databases. The algorithm was adapted from Dutch lipid clinic network criteria and was build upon 4 variables (i.e. LDL level and previous use of lipid lowering medications, premature and family history) to identify FH probability. The LDL level was adjusted on each type of lipid lowering medications and the probability of FH was defined taking into account missing data rate. Among the 7484 patients included in the RICO registry, 29.1% had premature vascular disease, 29.7% had familial history, 19.9% were under lipid lowering medications and 9.7% had LDL ≥5 mmol/L. FH prevalence was calculated as unlikely (72.6%), possible (24.6%) and probable /definite (2.8%). From the 1957 patients from FAST-MI 2005 with all data available, 29.7% had premature CV disease, 23% had a family history, 26.6% were on LLDs, and 5.4% had LDL ≥5 mmol/l. FH prevalence was calculated as unlikely (77.9%), possible (19.4%) and probable /definite (2.7%). In the 2223 patients from FASTMI 2010, 32.2% had premature CV disease, 24.9% had a family history, 28.1% were on LLDs, and 5.0% had LDL ≥5 mmol/l. FH prevalence was calculated as unlikely (75.7%), possible (21.5%) and probable /definite (2.7%). Our 4-variable algorithm yielded concordant results to determine FH probability in 3 different cohorts of AMI patients. In this large population reflecting routine clinical practice in acute MI, a high prevalence of FH was found, suggesting the opportunity for prevention strategies. The author hereby declares no conflict of interest

Conditional mode estimation for functional stationary ergodic data with responses missing at random

In this paper, we investigate the asymptotic properties of a non-parametric conditional mode estimation given a functional explanatory variable, when functional stationary ergodic data and missing at random responses are observed. First of all, we establish asymptotic properties for a conditional density estimator from which we derive almost sure convergence (with rate) and asymptotic normality of a conditional mode estimator. This new estimate take into account missing data, and a simulation study is performed to illustrate how this fact allows to get higher predictive performances than those obtained with standard estimates.

0132: Identifying familial hypercholesterolemia from registries of patients with acute myocardial infarction: an algorithm-based approach

Familial hypercholesterolemia (FH) is at very high risk of early myocardial infarction (MI). The prevalence of FH, which is estimated to be at least 1:500 in the general population, remains unclear in patients with acute MI. From databases of 2 French regional and nationwide registries of acute MI (RICO and FAST-MI, respectively), we aimed to determine FH prevalence by developing a specific algorithm. Consecutive patients with AMI ≤48 hours of onset included 1) in FAST-MI: during a one-month period in 213 institutions at the end of 2005 and 2) in RICO: from January 2001 December 2013 (≈13y), were considered in the 2 databases. The algorithm was adapted from Dutch lipid clinic network criteria and was build upon 4 variables (i.e. LDL level and previous use of lipid lowering medications, premature and family history) to identify FH probability. The LDL level was adjusted on each type of lipid lowering medications and the probability of FH was defined taking into account missing data rate. Among the 7484 patients included in the RICO registry, 29.1% had premature vascular disease, 29.7% had familial history, 19.9% were under lipid lowering medications and 9.7% had LDL ≥5mmol/L. FH prevalence was calculated as unlikely (72.6%), possible (24.6%) and probable /definite (2.8%). Our 4-variables algorithm is relevant to determine FH probability in databases from MI registries. In this large population reflecting routine clinical practice in acute MI, a high prevalence of FH was found, suggesting the opportunity for prevention strategies.

Longitudinal mixed-effects models for latent cognitive function

A mixed-effects regression model with a bent-cable change-point predictor is formulated to describe potential decline of cognitive function over time in the older population. For the individual trajectories, cognitive function is considered to be a latent variable measured through an item response theory model given longitudinal test data. Individual-specific parameters are defined for both cognitive function and the rate of change over time, using the change-point predictor for non-linear trends. Bayesian inference is used, where the Deviance Information Criterion and the L-criterion are investigated for model comparison. Special attention is given to the identifiability of the item response parameters. Item response theory makes it possible to use dichotomous and polytomous test items, and to take into account missing data and survey-design change during follow-up. This will be illustrated in an application where data stem from the Cambridge City over-75s Cohort Study.

Horvitz-Thompson estimation with randomized response and nonresponse

We consider the nonstandard situation where the randomized response theory is applied in a nonresponse set-up. The customary Horvitz-Thompson estimator for the population total of a sensitive variable is adjusted to take into account missing data, true response and randomized response, and the expression of its variance and the corresponding estimator provided. Finally, a numerical study is carried out to investigate how the estimation process can be affected by scrambled responses and how possible improvements in respondent cooperation can compensate for certain levels of nonresponse.

Linking Proteomic and Transcriptional Data through the Interactome and Epigenome Reveals a Map of Oncogene-induced Signaling

Cellular signal transduction generally involves cascades of post-translational protein modifications that rapidly catalyze changes in protein-DNA interactions and gene expression. High-throughput measurements are improving our ability to study each of these stages individually, but do not capture the connections between them. Here we present an approach for building a network of physical links among these data that can be used to prioritize targets for pharmacological intervention. Our method recovers the critical missing links between proteomic and transcriptional data by relating changes in chromatin accessibility to changes in expression and then uses these links to connect proteomic and transcriptome data. We applied our approach to integrate epigenomic, phosphoproteomic and transcriptome changes induced by the variant III mutation of the epidermal growth factor receptor (EGFRvIII) in a cell line model of glioblastoma multiforme (GBM). To test the relevance of the network, we used small molecules to target highly connected nodes implicated by the network model that were not detected by the experimental data in isolation and we found that a large fraction of these agents alter cell viability. Among these are two compounds, ICG-001, targeting CREB binding protein (CREBBP), and PKF118–310, targeting β-catenin (CTNNB1), which have not been tested previously for effectiveness against GBM. At the level of transcriptional regulation, we used chromatin immunoprecipitation sequencing (ChIP-Seq) to experimentally determine the genome-wide binding locations of p300, a transcriptional co-regulator highly connected in the network. Analysis of p300 target genes suggested its role in tumorigenesis. We propose that this general method, in which experimental measurements are used as constraints for building regulatory networks from the interactome while taking into account noise and missing data, should be applicable to a wide range of high-throughput datasets.

Power and sample size calculations for evaluating mediation effects in longitudinal studies

Current methods of power and sample size calculations for the design of longitudinal studies to evaluate mediation effects are mostly based on simulation studies and do not provide closed-form formulae. A further challenge due to the longitudinal study design is the consideration of missing data, which almost always occur in longitudinal studies due to staggered entry or drop out. In this article, we consider the product of coefficients as a measure for the longitudinal mediation effect and evaluate three methods for testing the hypothesis on the longitudinal mediation effect: the joint significant test, the normal approximation and the test of b methods. Formulae for power and sample size calculations are provided under each method while taking into account missing data. Performance of the three methods under limited sample size are examined using simulation studies. An example from the Einstein aging study is provided to illustrate the methods.

Sample size calculation for comparing time-averaged responses in [formula omitted]-group repeated-measurement studies

Many clinical trials compare the efficacy of K (≥3) treatments in repeated measurement studies. However, the design of such trials has received relatively less attention from researchers. Zhang and Ahn (2012) derived a closed-form sample size formula for two-sample comparisons of time-averaged responses using the generalized estimating equation (GEE) approach, which takes into account different correlation structures and missing data patterns. In this paper, we extend the sample size formula to scenarios where K (≥3) treatments are compared simultaneously to detect time-averaged differences in treatment effect. A closed-form sample size formula based on the noncentral χ2 test statistic is derived. We conduct simulation studies to assess the performance of the proposed sample size formula under various correlation structures from a damped exponential family, random and monotone missing patterns, and different observation probabilities. Simulation studies show that empirical powers and type I errors are close to their nominal levels. The proposed sample size formula is illustrated using a real clinical trial example.

Compliance Mixture Modelling with a Zero‐Effect Complier Class and Missing Data

Randomized experiments are the gold standard for evaluating proposed treatments. The intent to treat estimand measures the effect of treatment assignment, but not the effect of treatment if subjects take treatments to which they are not assigned. The desire to estimate the efficacy of the treatment in this case has been the impetus for a substantial literature on compliance over the last 15 years. In papers dealing with this issue, it is typically assumed there are different types of subjects, for example, those who will follow treatment assignment (compliers), and those who will always take a particular treatment irrespective of treatment assignment. The estimands of primary interest are the complier proportion and the complier average treatment effect (CACE). To estimate CACE, researchers have used various methods, for example, instrumental variables and parametric mixture models, treating compliers as a single class. However, it is often unreasonable to believe all compliers will be affected. This article therefore treats compliers as a mixture of two types, those belonging to a zero-effect class, others to an effect class. Second, in most experiments, some subjects drop out or simply do not report the value of the outcome variable, and the failure to take into account missing data can lead to biased estimates of treatment effects. Recent work on compliance in randomized experiments has addressed this issue by assuming missing data are missing at random or latently ignorable. We extend this work to the case where compliers are a mixture of types and also examine alternative types of nonignorable missing data assumptions.

P63 Treatment of pelvic inflammatory disease using short-course azithromycin

BackgroundIn our centre, a novel 5-day regimen of azithromycin (1 g stat then 500 mg daily for 4 days with or without metronidazole and/or ceftriaxone) is used and has been...