Comparison of methods for imputing limited-range variables: a simulation study

Flower production decreases with warmer and more humid atmospheric conditions in a western Amazonian forest

Climate models predict that everwet western Amazonian forests will face warmer and wetter atmospheric conditions, and increased cloud cover. It remains unclear how these changes will impact plant reproductive performance, such as flowering, which plays a central role in sustaining food webs and forest regeneration. Warmer and wetter nights may cause reduced flower production, via increased dark respiration rates or alteration in the reliability of flowering cue-based processes. Additionally, more persistent cloud cover should reduce the amounts of solar irradiance, which could limit flower production. We tested whether interannual variation in flower production has changed in response to fluctuations in irradiance, rainfall, temperature, and relative humidity over 18 yrs in an everwet forest in Ecuador. Analyses of 184 plant species showed that flower production declined as nighttime temperature and relative humidity increased, suggesting that warmer nights and greater atmospheric water saturation negatively impacted reproduction. Species varied in their flowering responses to climatic variables but this variation was not explained by life form or phylogeny. Our results shed light on how plant communities will respond to climatic changes in this everwet region, in which the impacts of these changes have been poorly studied compared with more seasonal Neotropical areas.

Statistical analysis plan for a cluster randomised controlled trial to compare screening, feedback and intervention for child anxiety problems to usual school practice: identifying Child Anxiety Through Schools-identification to intervention (iCATS-i2i)

BACKGROUND

The Identifying Child Anxiety Through Schools-identification to intervention (iCATS-i2i) trial is being conducted to establish whether 'screening and intervention', consisting of usual school practice plus a pathway comprising screening, feedback and a brief parent-led online intervention (OSI: Online Support and Intervention for child anxiety), bring clinical and health economic benefits compared to usual school practice and assessment only - 'usual school practice', for children aged 8-9 years in the following: (1) the 'target population', who initially screen positive for anxiety problems according to a two-item parent-report child anxiety questionnaire - iCATS-2, and (2) the 'total population', comprising all children in participating classes. This article describes the detailed statistical analysis plan for the trial.

METHODS AND DESIGN

iCATS-i2i is a definitive, superiority, pragmatic, school-based cluster randomised controlled trial (with internal pilot), with two parallel groups. Schools are randomised 1:1 to receive either screening and intervention or usual school practice. This article describes the following: trial objectives and outcomes; statistical analysis principles, including detailed estimand information necessary for aligning trial objectives, conduct, analyses and interpretation when there are different analysis populations and outcome measures to be considered; and planned main analyses, sensitivity and additional analyses.

TRIAL REGISTRATION

ClinicalTrials.gov ISRCTN76119074. Registered on 4 January 2022.

Comment on Oberman & Vink: Should we fix or simulate the complete data in simulation studies evaluating missing data methods?

For simulation studies that evaluate methods of handling missing data, we argue that generating partially observed data by fixing the complete data and repeatedly simulating the missingness indicators is a superficially attractive idea but only rarely appropriate to use.

Mental Pain Correlates with Mind Wandering, Self-Reflection, and Insight in Individuals with Psychotic Disorders: A Cross-Sectional Study

Understanding the cognitive processes that contribute to mental pain in individuals with psychotic disorders is important for refining therapeutic strategies and improving patient outcomes. This study investigated the potential relationship between mental pain, mind wandering, and self-reflection and insight in individuals diagnosed with psychotic disorders. We included individuals diagnosed with a 'schizophrenia spectrum disorder' according to DSM-5 criteria. Patients in the study were between 18 and 65 years old, clinically stable, and able to provide informed consent. A total of 34 participants, comprising 25 males and 9 females with an average age of 41.5 years (SD 11.5) were evaluated. The Psychache Scale (PAS), the Mind Wandering Deliberate and Spontaneous Scale (MWDS), and the Self-Reflection and Insight Scale (SRIS) were administered. Statistical analyses involved Spearman's rho correlations, controlled for potential confounders with partial correlations, and mediation and moderation analyses to understand the indirect effects of MWDS and SRIS on PAS and their potential interplay. Key findings revealed direct correlations between PAS and MWDS and inverse correlations between PAS and SRIS. The mediation effects on the relationship between the predictors and PAS ranged from 9.22% to 49.8%. The largest statistically significant mediation effect was observed with the SRIS-I subscale, suggesting that the self-reflection and insight component may play a role in the impact of mind wandering on mental pain. No evidence was found to suggest that any of the variables could function as relationship moderators for PAS. The results underscore the likely benefits of interventions aimed at reducing mind wandering and enhancing self-reflection in psychotic patients (e.g., metacognitive therapy, mindfulness). Further research will be essential to elucidate the underlying mechanisms.

Moderators of the cost-effectiveness of transdiagnostic CBT for anxiety disorders over an 8-month time horizon using a net-benefit regression framework

BACKGROUND

Access to evidence-based psychological treatment is a concern in many parts of the globe due to government-level financial constraints and patient-level barriers. Transdiagnostic cognitive behavioural therapy (tCBT) is an effective treatment approach that uses a single protocol for anxiety disorders which could enhance the dissemination of evidence-based psychotherapy. In a context of limited resources, the study of treatment moderators can allow to identify subgroups for which the cost-effectiveness of an intervention differs, information that could impact decision-making. So far, there has been no economic evaluation of tCBT for different subpopulations. The objectives of this study, using the net-benefit regression framework, were to explore clinical and sociodemographic factors as potential moderators of the cost-effectiveness of tCBT compared to treatment-as-usual (TAU).

METHODS

This is a secondary data analysis of a pragmatic randomized controlled trial opposing tCBT added to TAU (n = 117) to TAU only (n = 114). Data on costs from the health system and the limited societal perspectives, as well as anxiety-free days, an effectiveness measure based on the Beck Anxiety Inventory, were collected over an 8-month time horizon and used to derive individual net-benefits. The net-benefit regression framework was used to assess moderators of the cost-effectiveness of tCBT + TAU as opposed to TAU alone. Variables of sociodemographic and clinical nature were assessed.

RESULTS

Results showed that the number of comorbid anxiety disorders significantly moderated the cost-effectiveness of tCBT + TAU compared to TAU from the limited societal perspective.

CONCLUSIONS

The number of comorbid anxiety disorders was identified as a moderator affecting the cost-effectiveness of tCBT + TAU compared to TAU from the limited societal perspective. More research is needed to strengthen the case of tCBT from an economic standpoint for large-scale dissemination.

TRIAL REGISTRATION

ClinicalTrials.gov: NCT02811458, 23/06/2016.

Impact of comorbid allergic diseases on health-related quality of life of Hong Kong schoolchildren

BACKGROUND

Studies on the relationship between childhood allergic disease and health-related quality of life (HRQOL) have mostly been confined to a single allergic condition. Therefore, a composite allergic score (CAS) was derived to assess the accumulated effect of eczema, asthma, and allergic rhinitis on HRQOL in Hong Kong schoolchildren.

METHODS

Parents of grade one/two or grade eight/nine schoolchildren completed a questionnaire assessing the prevalence and severity of eczema (POEM), asthma (C-ACT/ACT), and allergic rhinitis (VAS) and schoolchildren's HRQOL (PedsQL). Three rounds of recruitment were conducted. A total of 19 primary and 25 secondary schools agreed to participate.

RESULTS

Data from 1140 caregivers of grade one/two schoolchildren and 1048 grade eight/nine schoolchildren were imputed and analyzed. The proportion of female respondents were lower in grade one/two (37.7%) but higher in grade eight/nine (57.3%). 63.8% of grade one/two and 58.1% of grade eight/nine schoolchildren reported having at least one allergic disease. In general, greater disease severity was significantly associated with lower HRQOL. After controlling for age, gender, and allergic comorbidity in hierarchical regressions, CAS significantly predicted all HRQOL outcomes in both grade one/two and grade eight/nine schoolchildren. Female grade eight/nine schoolchildren reported lower HRQOL outcomes.

CONCLUSION

Composite allergic score may be a practical tool to evaluate allergic comorbidity and the effectiveness of treatments targeting common pathological mechanisms of allergic diseases. Non-pharmaceutical approaches should be considered, especially for patients with more than one allergic disease and greater severity.

Deaf and hard-of-hearing children and adolescents' mental health, Quality of Life and communication

Mental health problems and lower Quality of Life (QoL) are more common in deaf and hard-of-hearing - (D)HH - children than in typically hearing (TH) children. Communication has been repeatedly linked to both mental health and QoL. The aims of this study were to compare mental health and QoL between signing deaf and hard-of-hearing (DHH), hard-of-hearing (HH) and TH children and to study associations between mental health/QoL and severity of hearing loss and communication. 106 children and adolescents (mean age 11;8; SD = 3.42), 59 of them DHH and 47 HH, and their parents reported child mental health and QoL outcomes. Parents also provided information about their children's communication, hearing loss and education while their children's cognitive ability was assessed. Although (D)HH and their parents rated their mental health similar to their TH peers, about twice as many (D)HH children rated themselves in the clinical range. However, (D)HH children rated their QoL as similar to their TH peers, while their parents rated it significantly lower. Associations between communicative competence, parent-reported mental health and QoL were found, whereas severity of hearing loss based on parent-report had no significant association with either mental health or QoL. These results are in line with other studies and emphasise the need to follow up on (D)HH children's mental health, QoL and communication.

Risk factors predictive of adverse drug events and drug-related falls in aged care residents: secondary analysis from the ReMInDAR trial

BACKGROUND

Residents of aged-care facilities have high rates of adverse drug events. This study aimed to identify risk factors for adverse drug events in aged-care residents.

METHOD

This was a secondary study using data from a multicentre randomised controlled trial. Data from 224 residents for whom there was 6 months of baseline information were analysed. We assessed the risk of adverse drug events and falls (post hoc) in the subsequent 6 months. Adverse events were identified via a key word search of the resident care record and adjudicated by a multidisciplinary panel using a modified version of the Naranjo criteria. Covariates identified through univariable logistic regression, including age, sex, medicines, physical activity, cognition (Montreal Cognitive Assessment), previous adverse events and health service use were included in multivariable models.

RESULTS

Overall, 224 residents were included, with a mean age of 86 years; 70% were female. 107 (48%) residents had an adverse drug event during the 6-month follow-up. Falls and bleeding were experienced by 73 (33%) and 28 (13%) residents, respectively. Age (odds ratio [OR] 1.05, 95% confidence interval [CI] 1.01-1.10), weight (OR 1.02, 95% CI 1.002-1.04), previous fall (OR 2.58, 95% CI 1.34-4.98) and sedative or hypnotic medicine use (OR 1.98, 95% CI 1.52-2.60) were associated with increased risk of adverse drug events. Increased cognition (OR 0.89, 95% CI 0.83-0.95) was protective. Risk factors for falls were previous fall (OR 3.27, 95% CI 1.68-6.35) and sedative or hypnotic medicines (OR 3.05, 95% CI 1.14-8.16). Increased cognition (OR 0.88, 95% CI 0.83-0.95) was protective.

CONCLUSION

Our results suggest residents with a previous fall, reduced cognition, and prescription of sedative or hypnotic medicines were at higher risk of adverse drug events and should be considered for proactive prevention.

Minimising Young Children's Anxiety through Schools (MY-CATS): statistical analysis plan for a cluster randomised controlled trial to evaluate the effectiveness and cost-effectiveness of an online parent-led intervention compared with usual school practice for young children identified as at risk for anxiety disorders

BACKGROUND

The Minimising Young Children's Anxiety through Schools (MY-CATS) trial is being conducted to determine whether an online evidence-based parent-guided cognitive behavioural therapy intervention in addition to usual school practice is effective and cost-effective compared with usual school practice in reducing anxiety disorders in children aged 4-7 deemed 'at risk' of anxiety disorders. This update article describes the detailed statistical analysis plan for the MY-CATS trial and reports a review of the underpinning sample size assumptions.

METHODS AND DESIGN

The MY-CATS study is a two-arm, definitive superiority pragmatic parallel group cluster randomised controlled trial in which schools will be randomised 1:1 to receive either the intervention (in addition to usual school practice) or the usual school practice only. This update to the (published) protocol provides a detailed description of the study methods, the statistical principles, the trial population and the planned statistical analyses, including additional analyses comprising instrumental variable regression and mediation analysis.

TRIAL REGISTRATION

ISRCTN82398107 . Prospectively registered on 14 January 2021.

Impute the missing data using retrieved dropouts

Background

In the past few decades various methods have been proposed to handle missing data of clinical studies, so as to assess the robustness of primary results. Some of the methods are based on the assumption of missing at random (MAR) which assumes subjects who discontinue the treatment will maintain the treatment effect after discontinuation. The agency, however, has expressed concern over methods based on this overly optimistic assumption, because it hardly holds for subjects discontinuing the investigational drug. Although in recent years a good number of sensitivity analyses based on missing not at random (MNAR) assumptions have been proposed, some use very conservative assumption on which it might be hard for sponsors and regulators to reach common ground.

Methods

Here we propose a multiple imputation method targeting at “treatment policy” estimand based on the MNAR assumption. This method can be used as the primary analysis, in addition to serving as a sensitivity analysis. It imputes missing data using information from retrieved dropouts defined as subjects who remain in the study despite occurrence of intercurrent events. Then imputed data long with completers and retrieved dropouts are analyzed altogether and finally multiple results are summarized into a single estimate. According to definition in ICH E9 (R1), this proposed approach fully aligns with the treatment policy estimand but its assumption is much more realistic and reasonable.

Results

Our approach has well controlled type I error rate with no loss of power. As expected, the effect size estimates take into account any dilution effect contributed by retrieved dropouts, conforming to the MNAR assumption.

Conclusions

Although multiple imputation approaches are always used as sensitivity analyses, this multiple imputation approach can be used as primary analysis for trials with sufficient retrieved dropouts or trials designed to collect retrieved dropouts.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-022-01509-9.

Reactive attachment disorder and disinhibited social engagement disorder in adolescence: co-occurring psychopathology and psychosocial problems

Insufficient care is associated with most psychiatric disorders and psychosocial problems, and is part of the etiology of reactive attachment disorder (RAD) and disinhibited social engagement disorder (DSED). To minimize the risk of misdiagnosis, and aid treatment and care, clinicians need to know to which degree RAD and DSED co-occur with other psychopathology and psychosocial problems, a topic little researched in adolescence. In a national study of all adolescents (N = 381; 67% consent; 12-20 years old; 58% girls) in Norwegian residential youth care, the Child and Adolescent Psychiatric Assessment interview yielded information about psychiatric diagnoses and psychosocial problems categorized as present/absent, and the Child Behavior Check List questionnaire was applied for dimensional measures of psychopathology. Most adolescents with a RAD or DSED diagnosis had several cooccurring psychiatric disorders and psychosocial problems. Prevalence rates of both emotional and behavioral disorders were high in adolescent RAD and DSED, as were rates of suicidality, self-harm, victimization from bullying, contact with police, risky sexual behavior and alcohol or drug misuse. Although categorical measures of co-occurring disorders and psychosocial problems revealed few and weak associations with RAD and DSED, dimensional measures uncovered associations between both emotional and behavioral problems and RAD/DSED symptom loads, as well as DSED diagnosis. Given the high degree of comorbidity, adolescents with RAD or DSED-or symptoms thereof-should be assessed for co-occurring psychopathology and related psychosocial problems. Treatment plans should be adjusted accordingly.

Multiple imputation of semi-continuous exposure variables that are categorized for analysis

Semi-continuous variables are characterized by a point mass at one value and a continuous range of values for remaining observations. An example is alcohol consumption quantity, with a spike of zeros representing non-drinkers and positive values for drinkers. If multiple imputation is used to handle missing values for semi-continuous variables, it is unclear how this should be implemented within the standard approaches of fully conditional specification (FCS) and multivariate normal imputation (MVNI). This question is brought into focus by the use of categorized versions of semi-continuous exposure variables in analyses (eg, no drinking, drinking below binge level, binge drinking, heavy binge drinking), raising the question of how best to achieve congeniality between imputation and analysis models. We performed a simulation study comparing nine approaches for imputing semi-continuous exposures requiring categorization for analysis. Three methods imputed the categories directly: ordinal logistic regression, and imputation of binary indicator variables representing the categories using MVNI (with two variants). Six methods (predictive mean matching, zero-inflated binomial imputation, and two-part imputation methods with variants in FCS and MVNI) imputed the semi-continuous variable, with categories derived after imputation. The ordinal and zero-inflated binomial methods had good performance across most scenarios, while MVNI methods requiring rounding after imputation did not perform well. There were mixed results for predictive mean matching and the two-part methods, depending on whether the estimands were proportions or regression coefficients. The results highlight the need to consider the parameter of interest when selecting an imputation procedure.

CRISTAL (a cluster-randomised, crossover, non-inferiority trial of aspirin compared to low molecular weight heparin for venous thromboembolism prophylaxis in hip or knee arthroplasty, a registry nested study): statistical analysis plan

BACKGROUND

This a priori statistical analysis plan describes the analysis for CRISTAL.

METHODS

CRISTAL (cluster-randomised, crossover, non-inferiority trial of aspirin compared to low molecular weight heparin for venous thromboembolism prophylaxis in hip or knee arthroplasty, a registry nested study) aims to determine whether aspirin is non-inferior to low molecular weight heparin (LMWH) in preventing symptomatic venous thromboembolism (VTE) following hip arthroplasty (HA) or knee arthroplasty (KA). The study is nested within the Australian Orthopaedic Association National Joint Replacement Registry. The trial was commenced in April 2019 and after an unplanned interim analysis, recruitment was stopped (December 2020), as the stopping rule was met for the primary outcome. The clusters comprised hospitals performing > 250 HA and/or KA procedures per annum, whereby all adults (> 18 years) undergoing HA or KA were recruited. Each hospital was randomised to commence with aspirin, orally, 85-150 mg daily or LMWH (enoxaparin), 40 mg, subcutaneously, daily within 24 h postoperatively, for 35 days after HA and 14 days after KA. Crossover was planned once the registration target was met for the first arm. The primary end point is symptomatic VTE within 90 days. Secondary outcomes include readmission, reoperation, major bleeding and death within 90 days, and reoperation and patient-reported pain, function and health status at 6 months. The main analyses will focus on the primary and secondary outcomes for patients undergoing elective primary total HA and KA for osteoarthritis. The analysis will use an intention-to-treat approach with cluster summary methods to compare treatment arms. As the trial stopped early, analyses will account for incomplete cluster crossover and unequal cluster sizes.

CONCLUSIONS

This paper provides a detailed statistical analysis plan for CRISTAL.

TRIAL REGISTRATION

Australian and New Zealand Clinical Trials Registry ACTRN12618001879257 . Registered on 19/11/2018.

A comparison of multiple imputation strategies for handling missing data in multi-item scales: Guidance for longitudinal studies

Medical research often involves using multi-item scales to assess individual characteristics, disease severity, and other health-related outcomes. It is common to observe missing data in the scale scores, due to missing data in one or more items that make up that score. Multiple imputation (MI) is a popular method for handling missing data. However, it is not clear how best to use MI in the context of scale scores, particularly when they are assessed at multiple waves of data collection resulting in large numbers of items. The aim of this article is to provide practical advice on how to impute missing values in a repeatedly measured multi-item scale using MI when inference on the scale score is of interest. We evaluated the performance of five MI strategies for imputing missing data at either the item or scale level using simulated data and a case study based on four waves of the Longitudinal Study of Australian Children (LSAC). MI was implemented using both multivariate normal imputation and fully conditional specification, with two rules for calculating the scale score. A complete case analysis was also performed for comparison. Based on our results, we caution against the use of a MI strategy that does not include the scale score in the imputation model(s) when the scale score is required for analysis.

DIMA: Data-Driven Selection of an Imputation Algorithm

Imputation is a prominent strategy when dealing with missing values (MVs) in proteomics data analysis pipelines. However, it is difficult to assess the performance of different imputation methods and varies strongly depending on data characteristics. To overcome this issue, we present the concept of a data-driven selection of an imputation algorithm (DIMA). The performance and broad applicability of DIMA are demonstrated on 142 quantitative proteomics data sets from the PRoteomics IDEntifications (PRIDE) database and on simulated data consisting of 5-50% MVs with different proportions of missing not at random and missing completely at random values. DIMA reliably suggests a high-performing imputation algorithm, which is always among the three best algorithms and results in a root mean square error difference (ΔRMSE) ≤ 10% in 80% of the cases. DIMA implementation is available in MATLAB at github.com/kreutz-lab/OmicsData and in R at github.com/kreutz-lab/DIMAR.

Gestational Weight Gain and Long-term Maternal Obesity Risk: A Multiple-Bias Analysis

BACKGROUND

Lifecourse research provides an important framework for chronic disease epidemiology. However, data collection to observe health characteristics over long periods is vulnerable to systematic error and statistical bias. We present a multiple-bias analysis using real-world data to estimate associations between excessive gestational weight gain and mid-life obesity, accounting for confounding, selection, and misclassification biases.

METHODS

Participants were from the multiethnic Study of Women's Health Across the Nation. Obesity was defined by waist circumference measured in 1996-1997 when women were age 42-53. Gestational weight gain was measured retrospectively by self-recall and was missing for over 40% of participants. We estimated relative risk (RR) and 95% confidence intervals (CI) of obesity at mid-life for presence versus absence of excessive gestational weight gain in any pregnancy. We imputed missing data via multiple imputation and used weighted regression to account for misclassification.

RESULTS

Among the 2,339 women in this analysis, 937 (40%) experienced obesity in mid-life. In complete case analysis, women with excessive gestational weight gain had an estimated 39% greater risk of obesity (RR = 1.4, CI = 1.1, 1.7), covariate-adjusted. Imputing data, then weighting estimates at the guidepost values of sensitivity = 80% and specificity = 75%, increased the RR (95% CI) for obesity to 2.3 (2.0, 2.6). Only models assuming a 20-point difference in specificity between those with and without obesity decreased the RR.

CONCLUSIONS

The inference of a positive association between excessive gestational weight gain and mid-life obesity is robust to methods accounting for selection and misclassification bias.

Utility of risk prediction models to detect atrial fibrillation in screened participants

AIMS

Atrial fibrillation (AF) is associated with higher risk of stroke. While the prevalence of AF is low in the general population, risk prediction models might identify individuals for selective screening of AF. We aimed to systematically identify and compare the utility of established models to predict prevalent AF.

METHODS AND RESULTS

Systematic search of PubMed and EMBASE for risk prediction models for AF. We adapted established risk prediction models and assessed their predictive performance using data from 2.5M individuals who attended vascular screening clinics in the USA and the UK and in the subset of 1.2M individuals with CHA2DS2-VASc ≥2. We assessed discrimination using area under the receiver operating characteristic (AUROC) curves and agreement between observed and predicted cases using calibration plots. After screening 6959 studies, 14 risk prediction models were identified. In our cohort, 10 464 (0.41%) participants had AF. For discrimination, six prediction model had AUROC curves of 0.70 or above in all individuals and those with CHA2DS2-VASc ≥2. In these models, calibration plots showed very good concordance between predicted and observed risks of AF. The two models with the highest observed prevalence in the highest decile of predicted risk, CHARGE-AF and MHS, showed an observed prevalence of AF of 1.6% with a number needed to screen of 63. Selective screening of the 10% highest risk identified 39% of cases with AF.

CONCLUSION

Prediction models can reliably identify individuals at high risk of AF. The best performing models showed an almost fourfold higher prevalence of AF by selective screening of individuals in the highest decile of risk compared with systematic screening of all cases.

REGISTRATION

This systematic review was registered (PROSPERO CRD42019123847).

Multiple imputation strategies for a bounded outcome variable in a competing risks analysis

In patient follow-up studies, events of interest may take place between periodic clinical assessments and so the exact time of onset is not observed. Such events are known as "bounded" or "interval-censored." Methods for handling such events can be categorized as either (i) applying multiple imputation (MI) strategies or (ii) taking a full likelihood-based (LB) approach. We focused on MI strategies, rather than LB methods, because of their flexibility. We evaluated MI strategies for bounded event times in a competing risks analysis, examining the extent to which interval boundaries, features of the data distribution and substantive analysis model are accounted for in the imputation model. Candidate imputation models were predictive mean matching (PMM); log-normal regression with postimputation back-transformation; normal regression with and without restrictions on the imputed values and Delord and Genin's method based on sampling from the cumulative incidence function. We used a simulation study to compare MI methods and one LB method when data were missing at random and missing not at random, also varying the proportion of missing data, and then applied the methods to a hematopoietic stem cell transplantation dataset. We found that cumulative incidence and median event time estimation were sensitive to model misspecification. In a competing risks analysis, we found that it is more important to account for features of the data distribution than to restrict imputed values based on interval boundaries or to ensure compatibility with the substantive analysis by sampling from the cumulative incidence function. We recommend MI by type 1 PMM.

Infant temperament, early-childhood parenting, and early-adolescent development: Testing alternative models of Parenting × Temperament interaction

Here we evaluate whether infant difficult temperament (6 months) functions as a vulnerability or more general plasticity factor when investigating effects of early-childhood parenting (8-42 months) on both positive and negative early-adolescent socioemotional development (age 8-11 years). Using data from the Avon Longitudinal Study of Parents and Children (ALSPAC, N = 14,541) and a re-parameterized model-testing approach to distinguish alternative person × environment conceptual models, results indicated that temperament × parenting interacted in predicting externalizing (i.e., hyperactivity, conduct problems), but not other behavior (i.e., emotional symptoms, peer problems), in a (weak) differential susceptibility manner. While more and less supportive parenting predicted, respectively, fewer and more behavior problems, it did so more strongly for children who were more difficult as infants.

Comprehensive Benefit-Risk Assessment of Noninferior Treatments Using Multicriteria Decision Analysis

OBJECTIVES

To develop a simple approach for evaluating the overall benefit-risk of a new noninferiority treatment compared with a standard of care.

METHODS

We propose using multicriteria decision analysis that accounts for uncertainty associated with both clinical outcomes and patient preference data. Because patients' preferences are likely to be influenced by their baseline characteristics, we suggest carrying out a preference study at the beginning of a trial. To reduce the burden of an additional study questionnaire, preference elicitation could be done on a small sample of trial participants. To restore preferences for all trial participants, we propose using multiple imputation (MI). Using simulations, we examine whether 3 different MI procedures lead to the same benefit-risk assessment conclusion, as if all trial participant preferences were obtained. We also compare MI results to complete case analysis, where only preferences of the small sample of trial participants are considered.

RESULTS

We show that the MI procedure successfully restores patients' preferences for the trial participants using different outcome criteria and preferences. For example, using 3 outcome criteria with only 10% of the trial participants providing their preferences, complete case analysis demonstrated a new noninferior treatment as favorable only 5.1% of the time, whereas MI procedures did so between 16.2% and 17.9% of the time. Given that 17.6% correspond to the fully observed weights, the MI methods demonstrate favorable results.

CONCLUSIONS

The MI procedure can help facilitate a simple comprehensive benefit-risk assessment for new noninferior treatments.

Age of First Walking and Associations with Symptom Severity in Children with Suspected or Diagnosed Autism Spectrum Disorder

Age of first walking (AOW) is reported to be later in autism spectrum disorder (ASD) compared with typical development. However, the relationship between AOW and variations in ASD symptoms across different neurodevelopmental disorders is largely unknown. This study investigated AOW and its association with autism symptom severity in a large sample of children (N = 490, 23% females) clinically evaluated for suspected ASD, differentiated into ASD (n = 376) and non-ASD (n = 114) diagnoses. Children with ASD achieved independent walking significantly later than children with non-ASD diagnoses. AOW was significantly associated with ASD symptom severity, and females had a non-significant later AOW. The current findings suggest that in cases with delayed AOW, ASD should be considered as an actual differential diagnosis, perhaps particularly in girls.

Disparities in Health Effects and Access to Health Care Among Houston Area Residents After Hurricane Harvey

OBJECTIVES

Although research shows that public health is substantially affected during and after disasters, few studies have examined the health effects of Hurricane Harvey, which made landfall on the Texas coast in August 2017. We assessed disparities in physical health, mental health, and health care access after Hurricane Harvey among residents of the Houston-The Woodlands-Sugar Land, Texas, metropolitan statistical area (ie, Houston MSA).

METHODS

We used structured survey data collected through telephone and online surveys from a population-based random sample of Houston MSA residents (n = 403) collected from November 29, 2017, through January 6, 2018. We used descriptive statistics to describe the prevalence of physical health/mental health and health care access outcomes and multivariable generalized linear models to assess disparities (eg, based on race/ethnicity, socioeconomic status, disability) in health outcomes.

RESULTS

Physical health problems disproportionately affected persons who did not evacuate (odds ratio [OR] = 0.41; 95% confidence interval [CI], 0.19-0.87). Non-Hispanic black persons were more likely than non-Hispanic white persons to have posttraumatic stress (OR = 5.03; 95% CI, 1.90-13.10), as were persons in households that experienced job loss post-Harvey (vs did not experience job loss post-Harvey; OR = 2.89; 95% CI, 1.14-7.32) and older persons (OR = 1.04; 95% CI, 1.01-1.06). Health care access was constrained for persons whose households lost jobs post-Harvey (vs did not lose jobs post-Harvey; OR = 2.73; 95% CI, 1.29-5.78) and for persons with disabilities (vs without disabilities; OR = 3.19; 95% CI, 1.37-7.45).

CONCLUSIONS

Our findings underscore the need to plan for and ameliorate public health disparities resulting from climate change-related disasters, which are expected to occur with increased frequency and magnitude.

Validation of Risk Prediction Models to Detect Asymptomatic Carotid Stenosis

Background Significant asymptomatic carotid stenosis (ACS) is associated with higher risk of strokes. While the prevalence of moderate and severe ACS is low in the general population, prediction models may allow identification of individuals at increased risk, thereby enabling targeted screening. We identified established prediction models for ACS and externally validated them in a large screening population. Methods and Results Prediction models for prevalent cases with ≥50% ACS were identified in a systematic review (975 studies reviewed and 6 prediction models identified [3 for moderate and 3 for severe ACS]) and then validated using data from 596 469 individuals who attended commercial vascular screening clinics in the United States and United Kingdom. We assessed discrimination and calibration. In the validation cohort, 11 178 (1.87%) participants had ≥50% ACS and 2033 (0.34%) had ≥70% ACS. The best model included age, sex, smoking, hypertension, hypercholesterolemia, diabetes mellitus, vascular and cerebrovascular disease, measured blood pressure, and blood lipids. The area under the receiver operating characteristic curve for this model was 0.75 (95% CI, 0.74-0.75) for ≥50% ACS and 0.78 (95% CI, 0.77-0.79) for ≥70% ACS. The prevalence of ≥50% ACS in the highest decile of risk was 6.51%, and 1.42% for ≥70% ACS. Targeted screening of the 10% highest risk identified 35% of cases with ≥50% ACS and 42% of cases with ≥70% ACS. Conclusions Individuals at high risk of significant ACS can be selected reliably using a prediction model. The best-performing prediction models identified over one third of all cases by targeted screening of individuals in the highest decile of risk only.

Do organization-level quality management systems influence department-level quality? A cross-sectional study across 32 large hospitals in Australia

Objective

Little is known about the influence that hospital quality systems have on quality at department level, in Australia and elsewhere. This study assessed the relationships between organizational-level quality management systems, and the extent to which hospital-level quality management systems and department-level quality management strategies are related.

Design

A multi-level, cross-sectional, mixed-method study.

Setting and participants

As part of the Deepening our Understanding of Quality in Australia (DUQuA) project, we invited all large hospitals in Australia (~200 or more beds) which provided acute myocardial infarction (AMI), hip fracture and stroke care. The quality managers of these hospitals were the respondents for one of seven measures of hospital quality management systems and strategies. Data across the six remaining measures were collected through site visits by external surveyors assessing the participating hospitals.

Main outcome measures

Relationships were assessed between three organization-level quality management system measures: a self-report measure assessing organization-level quality activities (quality management systems index, QMSI); externally assessed organization-level compliance to procedures used to plan, monitor and improve quality of care (quality management compliance index, QMCI); and externally assessed implementation of quality systems (clinical quality implementation index, CQII). Associations were also assessed between organization-level quality management systems and department-level quality management strategies: how clinical responsibilities are assigned for a particular condition; whether department organization processes are organized to facilitate evidence-based care recommendations; compliance with selected recommendations of international agencies; and whether clinical reviews are performed systematically.

Results

Of 78 invited hospitals, 32 participated in the study. QMSI was positively associated with QMCI and CQII, but after controlling for QMSI, no relationship was found between QMCI and CQII. There appears to be a cluster of relationships between QMSI and department-level measures, but this was not consistent across all departments.

Conclusion

This is the first national study undertaken in Australia to assess relationships within and between organization-level and department-level quality management systems. These quality management system tools align with many components of accreditation standards and may be useful for hospitals in continuously monitoring and driving improvement.

Dealing with indeterminate outcomes in antimalarial drug efficacy trials: a comparison between complete case analysis, multiple imputation and inverse probability weighting

BACKGROUND

Antimalarial clinical efficacy studies for uncomplicated Plasmodium falciparum malaria frequently encounter situations in which molecular genotyping is unable to discriminate between parasitic recurrence, either new infection or recrudescence. The current WHO guideline recommends excluding these individuals with indeterminate outcomes in a complete case (CC) analysis. Data from the four artemisinin-based combination (4ABC) trial was used to compare the performance of multiple imputation (MI) and inverse probability weighting (IPW) against the standard CC analysis for dealing with indeterminate recurrences.

METHODS

3369 study participants from the multicentre study (4ABC trial) with molecularly defined parasitic recurrence treated with three artemisinin-based combination therapies were used to represent a complete dataset. A set proportion of recurrent infections (10, 30 and 45%) were reclassified as missing using two mechanisms: a completely random selection (mechanism 1); missingness weakly dependent (mechanism 2a) and strongly dependent (mechanism 2b) on treatment and transmission intensity. The performance of MI, IPW and CC approaches in estimating the Kaplan-Meier (K-M) probability of parasitic recrudescence at day 28 was then compared. In addition, the maximum likelihood estimate of the cured proportion was presented for further comparison (analytical solution). Performance measures (bias, relative bias, standard error and coverage) were reported as an average from 1000 simulation runs.

RESULTS

The CC analyses resulted in absolute underestimation of K-M probability of day 28 recrudescence by up to 1.7% and were associated with reduced precision and poor coverage across all the scenarios studied. Both MI and IPW method performed better (greater consistency and greater efficiency) compared to CC analysis. In the absence of censoring, the analytical solution provided the most consistent and accurate estimate of cured proportion compared to the CC analyses.

CONCLUSIONS

The widely used CC approach underestimates antimalarial failure; IPW and MI procedures provided efficient and consistent estimates and should be considered when reporting the results of antimalarial clinical trials, especially in areas of high transmission, where the proportion of indeterminate outcomes could be large. The analytical solution estimating the cured proportion could provide an alternative approach, in scenarios with minimal censoring due to loss to follow-up or new infections.

Imputing missing data of function and disease activity in rheumatoid arthritis registers: what is the best technique?

Objective

To compare several methods of missing data imputation for function (Health Assessment Questionnaire) and for disease activity (Disease Activity Score-28 and Clinical Disease Activity Index) in rheumatoid arthritis (RA) patients.

Methods

One thousand RA patients from observational cohort studies with complete data for function and disease activity at baseline, 6, 12 and 24 months were selected to conduct a simulation study. Values were deleted at random or following a predicted attrition bias. Three types of imputation were performed: (1) methods imputing forward in time (last observation carried forward; linear forward extrapolation); (2) methods considering data both forward and backward in time (nearest available observation—NAO; linear extrapolation; polynomial extrapolation); and (3) methods using multi-individual models (linear mixed effects cubic regression—LME3; multiple imputation by chained equation—MICE). The performance of each estimation method was assessed using the difference between the mean outcome value, the remission and low disease activity rates after imputation of the missing values and the true value.

Results

When imputing missing baseline values, all methods underestimated equally the true value, but LME3 and MICE correctly estimated remission and low disease activity rates. When imputing missing follow-up values at 6, 12, or 24 months, NAO provided the least biassed estimate of the mean disease activity and corresponding remission rate. These results were not affected by the presence of attrition bias.

Conclusion

When imputing function and disease activity in large registers of active RA patients, researchers can consider the use of a simple method such as NAO for missing follow-up data, and the use of mixed-effects regression or multiple imputation for baseline data.

Analytical approaches and estimands to take account of missing patient-reported data in longitudinal studies

Patient-reported outcomes, such as quality of life, functioning, and symptoms, are used widely in therapeutic and behavioral trials and are increasingly used in drug development to represent the patient voice. Missing patient reported data is common and can undermine the validity of results reporting by reducing power, biasing estimates, and ultimately reducing confidence in the results. In this paper, we review statistically principled approaches for handling missing patient-reported outcome data and introduce the idea of estimands in the context of behavioral trials. Specifically, we outline a plan that considers missing data at each stage of research: design, data collection, analysis, and reporting. The design stage includes processes to prevent missing data, define the estimand, and specify primary and sensitivity analyses. The analytic strategy considering missing data depends on the estimand. Reviewed approaches include maximum likelihood-based models, multiple imputation, generalized estimating equations, and responder analysis. We outline sensitivity analyses to assess the robustness of the primary analysis results when data are missing. We also describe ad-hoc methods, including approaches to avoid. Last, we demonstrate methods using data from a behavioral intervention, where the primary outcome was self-reported cognition.

Multiple imputation methods for handling missing values in a longitudinal categorical variable with restrictions on transitions over time: a simulation study

Background

Longitudinal categorical variables are sometimes restricted in terms of how individuals transition between categories over time. For example, with a time-dependent measure of smoking categorised as never-smoker, ex-smoker, and current-smoker, current-smokers or ex-smokers cannot transition to a never-smoker at a subsequent wave. These longitudinal variables often contain missing values, however, there is little guidance on whether these restrictions need to be accommodated when using multiple imputation methods. Multiply imputing such missing values, ignoring the restrictions, could lead to implausible transitions.

Methods

We designed a simulation study based on the Longitudinal Study of Australian Children, where the target analysis was the association between (incomplete) maternal smoking and childhood obesity. We set varying proportions of data on maternal smoking to missing completely at random or missing at random. We compared the performance of fully conditional specification with multinomial and ordinal logistic imputation, and predictive mean matching, two-fold fully conditional specification, indicator based imputation under multivariate normal imputation with projected distance-based rounding, and continuous imputation under multivariate normal imputation with calibration, where each of these multiple imputation methods were applied, accounting for the restrictions using a semi-deterministic imputation procedure.

Results

Overall, we observed reduced bias when applying multiple imputation methods with restrictions, and fully conditional specification with predictive mean matching performed the best. Applying fully conditional specification and two-fold fully conditional specification for imputing nominal variables based on multinomial logistic regression had severe convergence issues. Both imputation methods under multivariate normal imputation produced biased estimates when restrictions were not accommodated, however, we observed substantial reductions in bias when restrictions were applied with continuous imputation under multivariate normal imputation with calibration.

Conclusion

In a similar longitudinal setting we recommend the use of fully conditional specification with predictive mean matching, with restrictions applied during the imputation stage.

Electronic supplementary material

The online version of this article (10.1186/s12874-018-0653-0) contains supplementary material, which is available to authorized users.

Multiple Imputation for Bounded Variables

Missing data are a common issue in statistical analyses. Multiple imputation is a technique that has been applied in countless research studies and has a strong theoretical basis. Most of the statistical literature on multiple imputation has focused on unbounded continuous variables, with mostly ad hoc remedies for variables with bounded support. These approaches can be unsatisfactory when applied to bounded variables as they can produce misleading inferences. In this paper, we propose a flexible quantile-based imputation model suitable for distributions defined over singly or doubly bounded intervals. Proper support of the imputed values is ensured by applying a family of transformations with singly or doubly bounded range. Simulation studies demonstrate that our method is able to deal with skewness, bimodality, and heteroscedasticity and has superior properties as compared to competing approaches, such as log-normal imputation and predictive mean matching. We demonstrate the application of the proposed imputation procedure by analysing data on mathematical development scores in children from the Millennium Cohort Study, UK. We also show a specific advantage of our methods using a small psychiatric dataset. Our methods are relevant in a number of fields, including education and psychology.

Treating loss-to-follow-up as a missing data problem: a case study using a longitudinal cohort of HIV-infected patients in Haiti

Background

HIV programs are often assessed by the proportion of patients who are alive and retained in care; however some patients are categorized as lost to follow-up (LTF) and have unknown vital status. LTF is not an outcome but a mixed category of patients who have undocumented death, transfer and disengagement from care. Estimating vital status (dead versus alive) among this category is critical for survival analyses and program evaluation.

Methods

We used three methods to estimate survival in the cohort and to ascertain factors associated with death among the first cohort of HIV positive patients to receive antiretroviral therapy in Haiti: complete case (CC) (drops missing), Inverse Probability Weights (IPW) (uses tracking data) and Multiple Imputation with Chained Equations (MICE) (imputes missing data). Logistic regression was used to calculate odds ratios and 95% confidence intervals for adjusted models for death at 10 years. The logistic regression models controlled for sex, age, severe poverty (living on <$1 USD per day), Port-au-Prince residence and baseline clinical characteristics of weight, CD4, WHO stage and tuberculosis diagnosis.

Results

Age, severe poverty, baseline weight and WHO stage were statistically significant predictors of AIDS related mortality across all models. Gender was only statistically significant in the MICE model but had at least a 10% difference in odds ratios across all models.

Conclusion

Each of these methods had different assumptions and differed in the number of observations included due to how missing values were addressed. We found MICE to be most robust in predicting survival status as it allowed us to impute missing data so that we had the maximum number of observations to perform regression analyses. MICE also provides a complementary alternative for estimating survival among patients with unassigned vital status. Additionally, the results were easier to interpret, less likely to be biased and provided an alternative to a problem that is often commented upon in the extant literature.

Electronic supplementary material

The online version of this article (10.1186/s12889-018-6115-0) contains supplementary material, which is available to authorized users.

Drinking or Smoking While Breastfeeding and Later Cognition in Children

BACKGROUND AND OBJECTIVES

Although prenatal alcohol and nicotine exposure are associated with reduced cognition in children, associations between consumption of alcohol during lactation and cognition have not been examined. We aimed to examine whether drinking or smoking while breastfeeding lowers children's cognitive scores. We hypothesized that increased drinking or smoking would be associated with dose-dependent cognitive reductions.

METHODS

Data were sourced from Growing Up in Australia: The Longitudinal Study of Australian Children. Participants were 5107 Australian infants recruited in 2004 and assessed every 2 years. Multivariable linear regression analyses assessed relationships between drinking and smoking habits of breastfeeding mothers and children's Matrix Reasoning, Peabody Picture Vocabulary Test-Third Edition and Who Am I? scores at later waves.

RESULTS

Increased or riskier wave 1 maternal alcohol consumption was associated with reductions in Matrix Reasoning scores at age 6 to 7 years in children who had been breastfed (B = -0.11; SE = 0.03; 95% confidence interval: -0.18 to -0.04; P = .01). This relationship was not evident in infants who had never breastfed (B = -0.02; SE = 0.10; 95% confidence interval = -0.20 to 0.17; P = .87). Smoking during lactation was not associated with any outcome variable.

CONCLUSIONS

Exposing infants to alcohol through breastmilk may cause dose-dependent reductions in their cognitive abilities. This reduction was observed at age 6 to 7 years but was not sustained at age 10 to 11 years. Although the relationship is small, it may be clinically significant when mothers consume alcohol regularly or binge drink. Further analyses will assess relationships between alcohol consumption or tobacco smoking during lactation and academic, developmental, physical, and behavioral outcomes in children.

Race/Ethnicity, Obesity, and the Risk of Being Verbally Bullied: a National Multilevel Study

OBJECTIVE

To examine the effects of obese/overweight status and race/ethnicity on the risk for being verbally bullied among second grade children, and to investigate if the relationship between weight status and verbal bullying varies based on race/ethnicity.

DESIGN

Data on second graders from the Early Childhood Longitudinal Study, Kindergarten Class of 2010-11 (Children = 18,130; Schools = 2419) were analyzed. Hierarchical generalized logistic modeling was used to address the objectives.

RESULTS

Independent of the child's sex, age, academic performance, family socioeconomic status, and school characteristics, obese/overweight children (relative to non-obese/overweight children) and Black children (relative to White children) were more likely to be verbally bullied. Hispanic and Asian children were less likely to be verbally bullied relative to White children. Hispanic obese/overweight children experienced less verbal bullying than White obese/overweight children.

CONCLUSIONS

This study documented disproportionate risks of being verbally bullied for obese/overweight US second graders. The risk of being verbally bullied was significantly greater for obese/overweight White children vs. obese/overweight Hispanic children.

IMPLICATIONS

Findings can inform the development of strategies to reduce verbal bullying of obese/overweight children in US elementary schools.

What Predicts a Good Adolescent to Adult Transition in ADHD? The Role of Self-Reported Resilience

OBJECTIVE

ADHD is a disorder associated with impairment and comorbid psychiatric problems in young adulthood; therefore, factors that may imply a more favorable outcome among adolescents with ADHD are of interest.

METHOD

This study used a longitudinal design to assess whether adolescent personal resilience characteristics during adolescence protected against psychosocial impairment, depression, and anxiety 3 years later. Self-reported protective factors were used as baseline measures in the assessment of 190 clinically referred adolescents with ADHD. A semi-structured diagnostic interview was performed at the follow-up.

RESULTS

In a group of youth with ADHD, personal resilience characteristics were associated with better psychosocial functioning in young adulthood, and less depression and anxiety.

CONCLUSION

Although further research is needed, these results indicate that personal resilience characteristics may be protective factors in the transitional period from adolescence to early adulthood.

Multiple Imputation for Incomplete Data in Epidemiologic Studies

Epidemiologic studies are frequently susceptible to missing information. Omitting observations with missing variables remains a common strategy in epidemiologic studies, yet this simple approach can often severely bias parameter estimates of interest if the values are not missing completely at random. Even when missingness is completely random, complete-case analysis can reduce the efficiency of estimated parameters, because large amounts of available data are simply tossed out with the incomplete observations. Alternative methods for mitigating the influence of missing information, such as multiple imputation, are becoming an increasing popular strategy in order to retain all available information, reduce potential bias, and improve efficiency in parameter estimation. In this paper, we describe the theoretical underpinnings of multiple imputation, and we illustrate application of this method as part of a collaborative challenge to assess the performance of various techniques for dealing with missing data (Am J Epidemiol. 2018;187(3):568-575). We detail the steps necessary to perform multiple imputation on a subset of data from the Collaborative Perinatal Project (1959-1974), where the goal is to estimate the odds of spontaneous abortion associated with smoking during pregnancy.

The conundrum of social class: Disparities in publishing among STEM students in undergraduate research programs at a Hispanic majority institution

Research on the science, technology, engineering, and math (STEM) student development pipeline has largely ignored social class and instead examined inequalities based on gender and race. We investigate the role of social class in undergraduate student research publications. Data come from a sample of 213 undergraduate research participants majoring in STEM at a Hispanic-majority institution. Based on generalized estimating equations that adjust for student demographics, research confidence, mentoring experiences, duration/number of research experiences, and clustering by major, we find that higher income students and continuing-generation students (vs. first-generation students) were significantly more likely to publish. Continuing-generation students had an even greater likelihood of publishing than first-generation students as students accrued more research confidence, spent more hours/week with faculty mentors, and conducted research for more months. Results suggest that undergraduate research programs designed to enhance diversity may help close some gaps (e.g., gender) but inadvertently reproduce class inequalities.

Limitations in Using Multiple Imputation to Harmonize Individual Participant Data for Meta-Analysis

Individual participant data (IPD) meta-analysis is a meta-analysis in which the individual-level data for each study are obtained and used for synthesis. A common challenge in IPD meta-analysis is when variables of interest are measured differently in different studies. The term harmonization has been coined to describe the procedure of placing variables on the same scale in order to permit pooling of data from a large number of studies. Using data from an IPD meta-analysis of 19 adolescent depression trials, we describe a multiple imputation approach for harmonizing 10 depression measures across the 19 trials by treating those depression measures that were not used in a study as missing data. We then apply diagnostics to address the fit of our imputation model. Even after reducing the scale of our application, we were still unable to produce accurate imputations of the missing values. We describe those features of the data that made it difficult to harmonize the depression measures and provide some guidelines for using multiple imputation for harmonization in IPD meta-analysis.

Sustained Posttransplantation Diabetes Is Associated With Long-Term Major Cardiovascular Events Following Liver Transplantation

Cardiovascular disease is a leading cause of death among liver transplant (LT) recipients. With a rising burden of posttransplantation metabolic disease, increases in cardiovascular-related morbidity and mortality may reduce life expectancy after LT. It is unknown if the risk of long-term major cardiovascular events (MCEs) differs among LT recipients with varying diabetic states. We performed a retrospective cohort study of LT recipients from 2003 through 2013 to compare the incidence of MCEs among patients (1) without diabetes, (2) with pretransplantation diabetes, (3) with de novo transient posttransplantation diabetes, and (4) with de novo sustained posttransplantation diabetes. We analyzed 994 eligible patients (39% without diabetes, 24% with pretransplantation diabetes, 16% with transient posttransplantation diabetes, and 20% with sustained posttransplantation diabetes). Median follow-up was 54.7 months. Overall, 12% of patients experienced a MCE. After adjustment for demographic and clinical variables, sustained posttransplantation diabetes was the only state associated with a significantly increased risk of MCEs (subdistribution hazard ratio 1.95, 95% confidence interval 1.20-3.18). Patients with sustained posttransplantation diabetes mellitus had a 13% and 27% cumulative incidence of MCEs at 5 and 10 years, respectively. While pretransplantation diabetes has traditionally been associated with cardiovascular disease, the long-term risk of MCEs is greatest in LT recipients with sustained posttransplantation diabetes mellitus.

A Pilot, Multicentre Pragmatic Randomised Trial to Explore the Impact of Carer Skills Training on Carer and Patient Behaviours: Testing the Cognitive Interpersonal Model in Adolescent Anorexia Nervosa

AIM

The aim of the study is to establish the acceptability, feasibility and approximate size of the effect of adding a carer intervention [Experienced Caregivers Helping Others (ECHO)] to treatment as usual (TAU) for adolescents with anorexia nervosa.

METHODS

The study is a pilot randomised trial comparing TAU (n = 50) alone or TAU plus ECHO with (n = 50) or without (n = 49) telephone guidance. Effect sizes (ESs) were regression coefficients standardised by baseline standard deviations of measure.

RESULTS

Although engagement with ECHO was poor (only 36% of carers in the ECHO group read over 50% of the book), there were markers of intervention fidelity, in that caregivers in the ECHO group showed a moderate increase in carer skills (ES = 0.4) at 12 months and a reduction in accommodating and enabling behaviour at 6 months (ES = 0.17). In terms of efficacy, in the ECHO group, carers spent less time care giving (ES = 0.40, p = 0.04) at 1 year, and patients had a minor advantage in body mass index (ES = 0.17), fewer admissions, decreased peer problems (ES = -0.36) and more pro-social behaviours (ES = 0.53). The addition of telephone guidance to ECHO produced little additional benefit.

CONCLUSIONS

The provision of self-management materials for carers to standard treatment for adolescent anorexia nervosa shows benefits for both carers and patients. This could be integrated as a form of early intervention in primary care. Copyright © 2017 John Wiley & Sons, Ltd and Eating Disorders Association.

Model checking in multiple imputation: an overview and case study

Background

Multiple imputation has become very popular as a general-purpose method for handling missing data. The validity of multiple-imputation-based analyses relies on the use of an appropriate model to impute the missing values. Despite the widespread use of multiple imputation, there are few guidelines available for checking imputation models.

Analysis

In this paper, we provide an overview of currently available methods for checking imputation models. These include graphical checks and numerical summaries, as well as simulation-based methods such as posterior predictive checking. These model checking techniques are illustrated using an analysis affected by missing data from the Longitudinal Study of Australian Children.

Conclusions

As multiple imputation becomes further established as a standard approach for handling missing data, it will become increasingly important that researchers employ appropriate model checking approaches to ensure that reliable results are obtained when using this method.

Electronic supplementary material

The online version of this article (doi:10.1186/s12982-017-0062-6) contains supplementary material, which is available to authorized users.

Faculty Motivation to Mentor Students Through Undergraduate Research Programs: A Study of Enabling and Constraining Factors

Undergraduate research experiences are a "high impact" educational practice that confer benefits to students. However, little attention has been paid to understanding faculty motivation to mentor undergraduate students through research training programs, even as the number of programs has grown, requiring increasing numbers of faculty mentors. To address this, we introduce a conceptual model for understanding faculty motivation to mentor and test it by using empirical data to identify factors that enable and constrain faculty engagement in an undergraduate research program. Using cross-sectional survey data collected in 2013, we employed generalized linear modeling to analyze data from 536 faculty across 13 research institutions to examine how expected costs/benefits, dispositional factors, situational factors, previous experience, and demographic factors predicted faculty motivation to mentor. Results show that faculty who placed greater value on the opportunity to increase diversity in the academy through mentorship of underrepresented minorities were more likely to be interested in serving as mentors. Faculty who agreed more strongly that mentoring undergraduate students was time consuming and their institution's reward structures were at odds with mentoring, or who had more constrained access to undergraduate students were less likely to be interested in serving as mentors. Mid-career faculty were more likely than late-career faculty to be interested in serving as mentors. Findings have implications for improving undergraduate research experiences, since the success of training programs hinges on engaging highly motivated faculty members as mentors.

Metabolic risk factors in schizophrenia and bipolar disorder: The effect of comedication with selective serotonin reuptake inhibitors and antipsychotics

BACKGROUND

The aim of this observational study was to investigate the relationship between metabolic factors and use of selective serotonin reuptake inhibitors (SSRIs) combined with olanzapine, quetiapine or risperidone.

METHODS

Data from the Norwegian Thematically Organized Psychosis study, a cross-sectional study on 1301 patients with schizophrenia (n=868) or bipolar disorder (n=433), were analyzed. As exposure variables in the linear regression model were included the dose or serum concentration of SSRIs (n=280) and of olanzapine (n=398), quetiapine (n=234) or risperidone (n=128). The main outcome variables were levels of total cholesterol, low and high density lipoprotein (LDL and HDL) cholesterol, triglycerides and glucose.

RESULTS

One defined daily dose (DDD) per day of an SSRI in addition to olanzapine was associated with an increase in total cholesterol of 0.16 (CI 0.01 to 0.32) mmol/L (P=0.042) and an increase in LDL-cholesterol of 0.17 (CI 0.02 to 0.31) mmol/L (P=0.022). An SSRI serum concentration in the middle of the reference interval in addition to quetiapine was associated with an increase in total cholesterol of 0.39 (CI 0.10 to 0.68) mmol/L (P=0.011) and an increase in LDL-cholesterol of 0.29 (0.02 to 0.56) mmol/L (P=0.037). There were no such effects when combined with risperidone.

CONCLUSIONS

The findings indicate only minor deteriorations of metabolic variables associated with treatment with an SSRI in addition to olanzapine and quetiapine, and none when combined with risperidone. These results suggest that SSRIs can be used in combination with antipsychotics, and that the possible increase in cardiovascular risk is negligible.

Factors Influencing Student Gains from Undergraduate Research Experiences at a Hispanic-Serving Institution

Undergraduate research experiences (UREs) confer many benefits to students, including improved self-confidence, better communication skills, and an increased likelihood of pursuing science careers. Additionally, UREs may be particularly important for racial/ethnic minority students who are underrepresented in the science workforce. We examined factors hypothetically relevant to underrepresented minority student gains from UREs at a Hispanic-serving institution, such as mentoring quality, family income, being Latino/a, and caring for dependents. Data came from a 2013 survey of University of Texas at El Paso students engaged in 10 URE programs (n = 227). Using generalized linear models (GzLMs) and adjusting for known covariates, we found that students who reported receiving higher-quality mentorship, spending more hours caring for dependents, and receiving more programmatic resources experienced significantly greater gains from their URE in all three areas we examined (i.e., thinking and working like a scientist, personal gains, and gains in skills). In two of three areas, duration of the URE was positive and significant. Being Latino/a was positive and significant only in the model predicting personal gains. Across the three models, quality of mentorship was the most important correlate of gains. This suggests that providing training to faculty mentors involved in UREs may improve student outcomes and increase program efficacy.

Increasing Research Productivity in Undergraduate Research Experiences: Exploring Predictors of Collaborative Faculty-Student Publications

Little attention has been paid to understanding faculty-student productivity via undergraduate research from the faculty member's perspective. This study examines predictors of faculty-student publications resulting from mentored undergraduate research, including measures of faculty-student collaboration, faculty commitment to undergraduate students, and faculty characteristics. Generalized estimating equations were used to analyze data from 468 faculty members across 13 research-intensive institutions, collected by a cross-sectional survey in 2013/2014. Results show that biomedical faculty mentors were more productive in publishing collaboratively with undergraduate students when they worked with students for more than 1 year on average, enjoyed teaching students about research, had mentored Black students, had received more funding from the National Institutes of Health, had a higher H-index scores, and had more years of experience working in higher education. This study suggests that college administrators and research program directors should strive to create incentives for faculty members to collaborate with undergraduate students and promote faculty awareness that undergraduates can contribute to their research.

Metabolic Abnormalities Related to Treatment With Selective Serotonin Reuptake Inhibitors in Patients With Schizophrenia or Bipolar Disorder

OBJECTIVE

The aim of the present study was to examine the effect of selective serotonin reuptake inhibitors (SSRIs) on cardiovascular risk factors in patients with schizophrenia or bipolar disorder.

METHOD

We used data from a cross-sectional study on 1301 patients with schizophrenia or bipolar disorder, of whom 280 were treated with SSRIs. The primary outcome variable was the serum concentration of total cholesterol. Secondary outcome variables were low-density lipoprotein (LDL) cholesterol, high-density lipoprotein cholesterol, triglyceride and glucose levels, body mass index, waist circumference, and systolic and diastolic blood pressure.

RESULTS

After adjusting for potential confounders, an SSRI serum concentration in the middle of the reference interval was associated with an increase of the total cholesterol level by 14.56 mg/dL (95% confidence interval (CI) 5.27-23.85 mg/dL, P = 0.002), the LDL cholesterol level by 8.50 mg/dL (CI 0.22-16.77 mg/dL, P = 0.044), the triglyceride level by 46.49 mg/dL (CI 26.53-66.46 mg/dL, P < 0.001) and the occurrence of the metabolic syndrome by a factor of 2.10 (CI 1.21-3.62, P = 0.008). There were also significant associations between the SSRI dose and total cholesterol and LDL cholesterol levels.

CONCLUSIONS

This study is the first to reveal significant associations between SSRI use and metabolic abnormalities in patients with schizophrenia or bipolar disorder. Although the effects were statistically significant, alterations were small. Thus, the clinical impact of the findings is most likely limited.

Multiple imputation methods for bivariate outcomes in cluster randomised trials

Missing observations are common in cluster randomised trials. The problem is exacerbated when modelling bivariate outcomes jointly, as the proportion of complete cases is often considerably smaller than the proportion having either of the outcomes fully observed. Approaches taken to handling such missing data include the following: complete case analysis, single‐level multiple imputation that ignores the clustering, multiple imputation with a fixed effect for each cluster and multilevel multiple imputation.

We contrasted the alternative approaches to handling missing data in a cost‐effectiveness analysis that uses data from a cluster randomised trial to evaluate an exercise intervention for care home residents.

We then conducted a simulation study to assess the performance of these approaches on bivariate continuous outcomes, in terms of confidence interval coverage and empirical bias in the estimated treatment effects. Missing‐at‐random clustered data scenarios were simulated following a full‐factorial design.

Across all the missing data mechanisms considered, the multiple imputation methods provided estimators with negligible bias, while complete case analysis resulted in biased treatment effect estimates in scenarios where the randomised treatment arm was associated with missingness. Confidence interval coverage was generally in excess of nominal levels (up to 99.8%) following fixed‐effects multiple imputation and too low following single‐level multiple imputation. Multilevel multiple imputation led to coverage levels of approximately 95% throughout. © 2016 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd.

Influences on Faculty Willingness to Mentor Undergraduate Students from Another University as Part of an Interinstitutional Research Training Program

In 2014, the National Institutes of Health invested $31 million in 10 primary institutions across the United States through the Building Undergraduate Infrastructure Leading to Diversity (BUILD) program; one requirement of BUILD is sending undergraduate trainees from those primary institutions to partner institutions for research experiences. Mechanisms like BUILD are designed to broaden research opportunities for students, especially those from underrepresented backgrounds. However, to our knowledge, no studies have examined faculty willingness to mentor undergraduates from other institutions through structured training programs. Survey data from 536 faculty members at 13 institutions were collected in Fall 2013 and analyzed using multiple statistical techniques. Results show that faculty who valued the opportunity to increase diversity in the academy and those who believed that mentoring undergraduates benefited their own research expressed greater willingness to serve as research mentors to visiting undergraduates, and faculty who perceived that they did not have the ability to accommodate additional students expressed less willingness to do so. Most respondents viewed student and faculty incentives as motivating factors in their willingness to mentor, but their perspectives on different types of incentives varied based on faculty career stage, discipline, and research funding status. Results have important implications for designing multi-institutional undergraduate research training programs.

Multiple imputation for harmonizing longitudinal non-commensurate measures in individual participant data meta-analysis

There are many advantages to individual participant data meta-analysis for combining data from multiple studies. These advantages include greater power to detect effects, increased sample heterogeneity, and the ability to perform more sophisticated analyses than meta-analyses that rely on published results. However, a fundamental challenge is that it is unlikely that variables of interest are measured the same way in all of the studies to be combined. We propose that this situation can be viewed as a missing data problem in which some outcomes are entirely missing within some trials and use multiple imputation to fill in missing measurements. We apply our method to five longitudinal adolescent depression trials where four studies used one depression measure and the fifth study used a different depression measure. None of the five studies contained both depression measures. We describe a multiple imputation approach for filling in missing depression measures that makes use of external calibration studies in which both depression measures were used. We discuss some practical issues in developing the imputation model including taking into account treatment group and study. We present diagnostics for checking the fit of the imputation model and investigate whether external information is appropriately incorporated into the imputed values.