Age-specific norms and validation of the German SDQ parent version based on a nationally representative sample (KiGGS).

The Strengths and Difficulties Questionnaire (SDQ) is the most widely used mental health screening instrument for children and adolescents. It is a short questionnaire including 25 items that can be answered by parents, teachers or children. There are two studies which report norms for the German SDQ parent version. They do not include children younger than 6 years. Moreover, whether the German SDQ parent version is measurement invariant across age has not yet been investigated. The absence of measurement invariance across age would support the use of age-specific norms that are not yet available for the German SDQ parent version. We used data of the German Health Interview and Examination Survey for Children and Adolescents (KiGGS), a nationally representative survey including 14,835 children aged 3-17 years, to assess measurement invariance of the German SDQ parent version across the full age range. Multi-group confirmatory factor analysis revealed that the hyperactivity and emotional symptoms subscales are not comparable between children of different ages. This supports the use of age-specific norms for these two subscales and for the total SDQ. We used methods of centile estimation to smoothly model the centiles of the SDQ total score and the subscale scores in dependence on age. These age-specific centiles reflect the developmental course of SDQ problems in children (including preschoolers) and adolescents living in Germany. They can be used to identify children and adolescents with abnormal behaviour, while accounting for the developmental course of emotional and behaviour problems.

Making complex prediction rules applicable for readers: Current practice in random forest literature and recommendations.

Ideally, prediction rules should be published in such a way that readers may apply them, for example, to make predictions for their own data. While this is straightforward for simple prediction rules, such as those based on the logistic regression model, this is much more difficult for complex prediction rules derived by machine learning tools. We conducted a survey of articles reporting prediction rules that were constructed using the random forest algorithm and published in PLOS ONE in 2014-2015 in the field "medical and health sciences", with the aim of identifying issues related to their applicability. Making a prediction rule reproducible is a possible way to ensure that it is applicable; thus reproducibility is also examined in our survey. The presented prediction rules were applicable in only 2 of 30 identified papers, while for further eight prediction rules it was possible to obtain the necessary information by contacting the authors. Various problems, such as nonresponse of the authors, hampered the applicability of prediction rules in the other cases. Based on our experiences from this illustrative survey, we formulate a set of recommendations for authors who aim to make complex prediction rules applicable for readers. All data including the description of the considered studies and analysis codes are available as supplementary materials.

Time trends in depression prevalence and health-related correlates: results from population-based surveys in Germany 1997-1999 vs. 2009-2012.

BACKGROUND: Although an "epidemic" of depression is frequently claimed, empirical evidence is inconsistent, depending on country, study design and depression assessment. Little is known about changes in depression over time in Germany, although health insurance companies report frequency increases. Here we examined time trends in depression prevalence, severity and health-related correlates in the general population. METHODS: Data were obtained from the mental health module of the "German Health Interview and Examination Survey for Adults" (2009-2012, n = 3265) and the mental health supplement of the "German National Health Interview and Examination Survey 1998" (1997-1999, n = 4176), excluding respondents older than 65. 12-month major depressive disorder (MDD), severity and symptoms were assessed based on the WHO Composite International Diagnostic Interview. Health-related quality of life (SF-36), self-reported sick days or days with limitations in normal daily life activities were examined, too. Calculations were carried out population-weighted. Additional age-standardized analyses were conducted to account for demographic changes. RESULTS: Overall, MDD 12-month prevalence remained stable at 7.4%. Women showed a shifted age distribution with increased prevalence at younger ages, and increasing MDD severity. Time trends in health-related correlates occurred both in participants with and without MDD. Mental health disability increased over time, particularly among men with MDD, reflected by the mental component score of the SF-36 and days with activity limitation due to mental health problems. Demographic changes had a marginal impact on the time trends. CONCLUSIONS: In contrast to the ongoing international debate regarding increased depression rates in western countries, we found no increase in overall MDD prevalence in Germany over a long period. In conclusion, increased depression frequencies in national health insurance data and growing health care costs associated with depression are not attributable to overall prevalence changes at a population level. However, shifted age distribution and increased severity among women may reflect a rising depression risk within this specific subgroup, and changes in health-related correlates indicate a growing mental health care need for depression, particularly among men.

Quantification of Adipose Tissue and Muscle Mass Based on Computed Tomography Scans: Comparison of Eight Planimetric and Diametric Techniques Including a Step-By-Step Guide.

BACKGROUND: Recent scientific work proved that knowledge about body composition beyond the body mass index is essential. Both adipose tissue and muscular status are determining risk factors of morbidity and mortality. Analysis of single cross-sectional computed tomography (CT) images, acquired during routine care only to prevent additional radiation exposure, provide a detailed insight into the body composition of chronically and critically ill patients. METHODS: This retrospective study included 490 trauma patients of whom a whole-body multiple detector CT scan was acquired at admission. From a single cross-sectional CT, we compared eight diametric and planimetric techniques for the assessment of core muscle mass as well as visceral and subcutaneous adipose tissue. Furthermore, we derived formulas for converting the measurement results of various techniques into each other. RESULTS: For intra- and interobserver reliability, we obtained intraclass correlation coefficients (ICCs) ranging from 0.947 to 0.997 (intraobserver reliability) and from 0.850 to 0.998 (interobserver reliability) for planimetric measurements. Diametric techniques conferred lower ICCs with 0.851-0.995 and 0.833-0.971, respectively. Overall, area-based measurements of abdominal adipose tissue yielded highly correlated results with diametric measures of obesity. For example, the Pearson correlation of visceral adipose tissue and sagittal abdominal diameter was 0.87 for male and 0.82 for female patients. Planimetric and diametric muscle measurements correlated best for lean psoas area and bilateral diametric measurement of the psoas with a Pearson correlation of 0.90 and 0.93 for male and female patients, respectively. CONCLUSION: Planimetric measurements should remain the gold standard to describe fat and muscle compartments. Diametric measurements could however serve as a surrogate if planimetric techniques are not readily available or feasible as for example in large registries.

Lung volumes predict survival in patients with chronic lung allograft dysfunction.

Identification of disease phenotypes might improve the understanding of patients with chronic lung allograft dysfunction (CLAD). The aim of the study was to assess the impact of pulmonary restriction and air trapping by lung volume measurements at the onset of CLAD.A total of 396 bilateral lung transplant recipients were analysed. At onset, CLAD was further categorised based on plethysmography. A restrictive CLAD (R-CLAD) was defined as a loss of total lung capacity from baseline. CLAD with air trapping (AT-CLAD) was defined as an increased ratio of residual volume to total lung capacity. Outcome was survival after CLAD onset. Patients with insufficient clinical information were excluded (n=95).Of 301 lung transplant recipients, 94 (31.2%) developed CLAD. Patients with R-CLAD (n=20) and AT-CLAD (n=21), respectively, had a significantly worse survival (p<0.001) than patients with non-R/AT-CLAD. Both R-CLAD and AT-CLAD were associated with increased mortality when controlling for multiple confounding variables (hazard ratio (HR) 3.57, 95% CI 1.39-9.18; p=0.008; and HR 2.65, 95% CI 1.05-6.68; p=0.039). Furthermore, measurement of lung volumes was useful to identify patients with combined phenotypes.Measurement of lung volumes in the long-term follow-up of lung transplant recipients allows the identification of patients who are at risk for worse outcome and warrant special consideration.

Letter to the Editor: On the term 'interaction' and related phrases in the literature on Random Forests.

In an interesting and quite exhaustive review on Random Forests (RF) methodology in bioinformatics Touw et al. address--among other topics--the problem of the detection of interactions between variables based on RF methodology. We feel that some important statistical concepts, such as 'interaction', 'conditional dependence' or 'correlation', are sometimes employed inconsistently in the bioinformatics literature in general and in the literature on RF in particular. In this letter to the Editor, we aim to clarify some of the central statistical concepts and point out some confusing interpretations concerning RF given by Touw et al. and other authors.

Subsampling versus bootstrapping in resampling-based model selection for multivariable regression.

In recent years, increasing attention has been devoted to the problem of the stability of multivariable regression models, understood as the resistance of the model to small changes in the data on which it has been fitted. Resampling techniques, mainly based on the bootstrap, have been developed to address this issue. In particular, the approaches based on the idea of "inclusion frequency" consider the repeated implementation of a variable selection procedure, for example backward elimination, on several bootstrap samples. The analysis of the variables selected in each iteration provides useful information on the model stability and on the variables' importance. Recent findings, nevertheless, show possible pitfalls in the use of the bootstrap, and alternatives such as subsampling have begun to be taken into consideration in the literature. Using model selection frequencies and variable inclusion frequencies, we empirically compare these two different resampling techniques, investigating the effect of their use in selected classical model selection procedures for multivariable regression. We conduct our investigations by analyzing two real data examples and by performing a simulation study. Our results reveal some advantages in using a subsampling technique rather than the bootstrap in this context.

Categorical variables with many categories are preferentially selected in bootstrap-based model selection procedures for multivariable regression models.

Automated variable selection procedures, such as backward elimination, are commonly employed to perform model selection in the context of multivariable regression. The stability of such procedures can be investigated using a bootstrap-based approach. The idea is to apply the variable selection procedure on a large number of bootstrap samples successively and to examine the obtained models, for instance, in terms of the inclusion of specific predictor variables. In this paper, we aim to investigate a particular important problem affecting this method in the case of categorical predictor variables with different numbers of categories and to give recommendations on how to avoid it. For this purpose, we systematically assess the behavior of automated variable selection based on the likelihood ratio test using either bootstrap samples drawn with replacement or subsamples drawn without replacement from the original dataset. Our study consists of extensive simulations and a real data example from the NHANES study. Our main result is that if automated variable selection is conducted on bootstrap samples, variables with more categories are substantially favored over variables with fewer categories and over metric variables even if none of them have any effect. Importantly, variables with no effect and many categories may be (wrongly) preferred to variables with an effect but few categories. We suggest the use of subsamples instead of bootstrap samples to bypass these drawbacks.

Pitfalls of hypothesis tests and model selection on bootstrap samples: Causes and consequences in biometrical applications.

The bootstrap method has become a widely used tool applied in diverse areas where results based on asymptotic theory are scarce. It can be applied, for example, for assessing the variance of a statistic, a quantile of interest or for significance testing by resampling from the null hypothesis. Recently, some approaches have been proposed in the biometrical field where hypothesis testing or model selection is performed on a bootstrap sample as if it were the original sample. P-values computed from bootstrap samples have been used, for example, in the statistics and bioinformatics literature for ranking genes with respect to their differential expression, for estimating the variability of p-values and for model stability investigations. Procedures which make use of bootstrapped information criteria are often applied in model stability investigations and model averaging approaches as well as when estimating the error of model selection procedures which involve tuning parameters. From the literature, however, there is evidence that p-values and model selection criteria evaluated on bootstrap data sets do not represent what would be obtained on the original data or new data drawn from the overall population. We explain the reasons for this and, through the use of a real data set and simulations, we assess the practical impact on procedures relevant to biometrical applications in cases where it has not yet been studied. Moreover, we investigate the behavior of subsampling (i.e., drawing from a data set without replacement) as a potential alternative solution to the bootstrap for these procedures.

An AUC-based permutation variable importance measure for random forests.

BACKGROUND: The random forest (RF) method is a commonly used tool for classification with high dimensional data as well as for ranking candidate predictors based on the so-called random forest variable importance measures (VIMs). However the classification performance of RF is known to be suboptimal in case of strongly unbalanced data, i.e. data where response class sizes differ considerably. Suggestions were made to obtain better classification performance based either on sampling procedures or on cost sensitivity analyses. However to our knowledge the performance of the VIMs has not yet been examined in the case of unbalanced response classes. In this paper we explore the performance of the permutation VIM for unbalanced data settings and introduce an alternative permutation VIM based on the area under the curve (AUC) that is expected to be more robust towards class imbalance. RESULTS: We investigated the performance of the standard permutation VIM and of our novel AUC-based permutation VIM for different class imbalance levels using simulated data and real data. The results suggest that the new AUC-based permutation VIM outperforms the standard permutation VIM for unbalanced data settings while both permutation VIMs have equal performance for balanced data settings. CONCLUSIONS: The standard permutation VIM loses its ability to discriminate between associated predictors and predictors not associated with the response for increasing class imbalance. It is outperformed by our new AUC-based permutation VIM for unbalanced data settings, while the performance of both VIMs is very similar in the case of balanced classes. The new AUC-based VIM is implemented in the R package party for the unbiased RF variant based on conditional inference trees. The codes implementing our study are available from the companion website: http://www.ibe.med.uni-muenchen.de/organisation/mitarbeiter/070_drittmittel/janitza/index.html.

Determination of breath isoprene allows the identification of the expiratory fraction of the propofol breath signal during real-time propofol breath monitoring.

Real-time measurement of propofol in the breath may be used for routine clinical monitoring. However, this requires unequivocal identification of the expiratory phase of the respiratory propofol signal as only expiratory propofol reflects propofol blood concentrations. Determination of CO2 breath concentrations is the current gold standard for the identification of expiratory gas but usually requires additional equipment. Human breath also contains isoprene, a volatile organic compound with low inspiratory breath concentration and an expiratory concentration plateau. We investigated whether breath isoprene could be used similarly to CO2 to identify the expiratory fraction of the propofol breath signal. We investigated real-time breath data obtained from 40 study subjects during routine anesthesia. Propofol, isoprene, and CO2 breath concentrations were determined by a combined ion molecule reaction/electron impact mass spectrometry system. The expiratory propofol signal was identified according to breath CO2 and isoprene concentrations and presented as median of intervals of 30 s duration. Bland-Altman analysis was applied to detect differences (bias) in the expiratory propofol signal extracted by the two identification methods. We investigated propofol signals in a total of 3,590 observation intervals of 30 s duration in the 40 study subjects. In 51.4 % of the intervals (1,844/3,590) both methods extracted the same results for expiratory propofol signal. Overall bias between the two data extraction methods was -0.12 ppb. The lower and the upper limits of the 95 % CI were -0.69 and 0.45 ppb. Determination of isoprene breath concentrations allows the identification of the expiratory propofol signal during real-time breath monitoring.

elaborator: A Novel App for Insights into Laboratory Data of Clinical Trials.

In clinical studies there are huge numbers of laboratory parameters available that are measured at several visits for several treatment groups. The status quo for presenting laboratory data in clinical trials consists in generating large numbers of tables and data listings. Such tables and listings are required for submissions to health authorities. However, reviewing laboratory data presented in the form of tables and listings is a lengthy and tedious process. Thus, to enable efficient exploration of laboratory data we developed elaborator, a comprehensive and easy-to-use interactive browser-based application. The elaborator app comprises three analyses types for addressing different questions, for example about changes in laboratory values that frequently occur, treatment-related changes and changes beyond the normal ranges. In this way, the app can be used by study teams for identifying safety signals in a clinical trial as well as for generating hypotheses that are further inspected with detailed analyses and possibly data from other sources. The elaborator app is implemented in the statistical software R. The R package elaborator can be obtained from https://cran.r-project.org/package=elaborator . Patients' laboratory data need to be extracted from the clinical database and pre-processed locally for feeding into the app. For exploring data by means of the elaborator, the user needs some familiarity with R but no programming knowledge is required.

On the overestimation of random forest's out-of-bag error.

The ensemble method random forests has become a popular classification tool in bioinformatics and related fields. The out-of-bag error is an error estimation technique often used to evaluate the accuracy of a random forest and to select appropriate values for tuning parameters, such as the number of candidate predictors that are randomly drawn for a split, referred to as mtry. However, for binary classification problems with metric predictors it has been shown that the out-of-bag error can overestimate the true prediction error depending on the choices of random forests parameters. Based on simulated and real data this paper aims to identify settings for which this overestimation is likely. It is, moreover, questionable whether the out-of-bag error can be used in classification tasks for selecting tuning parameters like mtry, because the overestimation is seen to depend on the parameter mtry. The simulation-based and real-data based studies with metric predictor variables performed in this paper show that the overestimation is largest in balanced settings and in settings with few observations, a large number of predictor variables, small correlations between predictors and weak effects. There was hardly any impact of the overestimation on tuning parameter selection. However, although the prediction performance of random forests was not substantially affected when using the out-of-bag error for tuning parameter selection in the present studies, one cannot be sure that this applies to all future data. For settings with metric predictor variables it is therefore strongly recommended to use stratified subsampling with sampling fractions that are proportional to the class sizes for both tuning parameter selection and error estimation in random forests. This yielded less biased estimates of the true prediction error. In unbalanced settings, in which there is a strong interest in predicting observations from the smaller classes well, sampling the same number of observations from each class is a promising alternative.

Gram-negative and -positive bacteria differentiation in blood culture samples by headspace volatile compound analysis.

BACKGROUND: Identification of microorganisms in positive blood cultures still relies on standard techniques such as Gram staining followed by culturing with definite microorganism identification. Alternatively, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry or the analysis of headspace volatile compound (VC) composition produced by cultures can help to differentiate between microorganisms under experimental conditions. This study assessed the efficacy of volatile compound based microorganism differentiation into Gram-negatives and -positives in unselected positive blood culture samples from patients. METHODS: Headspace gas samples of positive blood culture samples were transferred to sterilized, sealed, and evacuated 20 ml glass vials and stored at -30 °C until batch analysis. Headspace gas VC content analysis was carried out via an auto sampler connected to an ion-molecule reaction mass spectrometer (IMR-MS). Measurements covered a mass range from 16 to 135 u including CO2, H2, N2, and O2. Prediction rules for microorganism identification based on VC composition were derived using a training data set and evaluated using a validation data set within a random split validation procedure. RESULTS: One-hundred-fifty-two aerobic samples growing 27 Gram-negatives, 106 Gram-positives, and 19 fungi and 130 anaerobic samples growing 37 Gram-negatives, 91 Gram-positives, and two fungi were analysed. In anaerobic samples, ten discriminators were identified by the random forest method allowing for bacteria differentiation into Gram-negative and -positive (error rate: 16.7 % in validation data set). For aerobic samples the error rate was not better than random. CONCLUSIONS: In anaerobic blood culture samples of patients IMR-MS based headspace VC composition analysis facilitates bacteria differentiation into Gram-negative and -positive.

Gender-Specific Differences in Low-Dose Haloperidol Response for Prevention of Postoperative Nausea and Vomiting: A Register-Based Cohort Study.

BACKGROUND: Postoperative nausea and vomiting (PONV) is one of the most common and distressing complications after general anesthesia and surgery, with young non-smoking females receiving postoperative opioids being high-risk patients. This register-based study aims to evaluate the effect of low-dose haloperidol (0.5 mg intravenously) directly after induction of general anesthesia to reduce the incidence of PONV in the postoperative anesthesiological care unit (PACU). METHODS: Multivariable regression models were used to investigate the association between low-dose haloperidol and the occurrence of PONV using a patient registry containing 2,617 surgical procedures carried out at an university hospital. RESULTS: Haloperidol 0.5 mg is associated with a reduced risk of PONV in the total collective (adjusted odds ratio = 0.75, 95% confidence interval: [0.56, 0.99], p = 0.05). The results indicate that there is a reduced risk in male patients (adjusted odds ratio = 0.45, 95% confidence interval: [0.28, 0.73], p = 0.001) if a dose of 0.5 mg haloperidol was administered while there seems to be no effect in females (adjusted odds ratio = 1.02, 95% confidence interval: [0.71, 1.46], p = 0.93). Currently known risk factors for PONV such as female gender, duration of anesthesia and the use of opioids were confirmed in our analysis. CONCLUSION: This study suggests that low-dose haloperidol has an antiemetic effect in male patients but has no effect in female patients. A confirmation of the gender-specific effects we have observed in this register-based cohort study might have major implications on clinical daily routine.

Core Muscle Size Predicts Postoperative Outcome in Lung Transplant Candidates.

BACKGROUND: Careful patient selection is the prerequisite to raise transplant benefit. In lung transplant (LT) candidates, the effect of body mass index (BMI) on postoperative outcome remains controversial, possibly due to the inaccuracy of BMI in discriminating between fat and muscle mass. We therefore hypothesized that assessment of body composition by muscle mass measures is more accurate than by BMI regarding postoperative outcome. METHODS: All LT recipients from 2011 to 2014 were included and retrospectively analyzed. Lean psoas area (LPA) was assessed from pretransplant computed tomography scans, and associations with postoperative outcomes were investigated. RESULTS: Included were 103 consecutive LT recipients with a mean pre-LT BMI of 22.0 ± 4.0 kg/m(2) and a mean LPA of 22.3 ± 8.3 cm(2). LPA was inversely associated with length of mechanical ventilation (p = 0.03), requirement of tracheostomy (p = 0.035), and length of stay in the intensive care unit (p = 0.02), while controlling for underlying disease, BMI, sex, age, and procedure; in contrast, BMI was not (p = 0.25, p = 0.54, and p = 0.42, respectively.). Multiple regression analysis revealed that the 6-minute walk distance at the end of pulmonary rehabilitation was significantly associated with LPA (p = 0.02). CONCLUSIONS: LPA can easily be assessed in LT candidates as part of pretransplant evaluation and was significantly associated with short-term outcome, whereas BMI was not. Assessment of LPA may provide additional information on body composition beyond BMI. However, the clinical utility has to be further evaluated.

Influence of abdominal obesity on multiorgan dysfunction and mortality in acute respiratory distress syndrome patients treated with prone positioning.

PURPOSE: Obesity is a worldwide pandemic, and obese patients face an increased risk of developing acute respiratory distress syndrome (ARDS). Prone positioning (PP) is a frequently used intervention in the treatment of ARDS. There are no data describing the impact of PP on morbidity and mortality in abdominally obese patients. We report our observations in abdominally obese ARDS patients treated with PP. MATERIALS AND METHODS: Patients with ARDS (n=82) were retrospectively divided into 2 groups characterized by presence (n=41) or absence (n=41) of abdominal obesity as defined by a sagittal abdominal diameter of 26 cm or more. RESULTS: There was no difference in cumulative time abdominally obese patients were placed in prone position from admission to day 7 (41.0 hours [interquartile range, 50.5 hours] vs 39.5 hours [interquartile range, 61.5 hours]; P=.65) or in overall intensive care unit mortality (34% vs 34%; P=1). However, abdominally obese patients developed renal failure (83% vs 35%; P<.001) and hypoxic hepatitis (22% vs 2%; P=.015) more frequently. A significant interaction effect between abdominal obesity and prone position with respect to mortality risk (likelihood ratio, P=.0004) was seen if abdominally obese patients were treated with prolonged cumulative PP. CONCLUSION: A cautious approach to PP should be considered in abdominally obese patients.