Search

Background Hamiltonian Monte Carlo is one of the algorithms of the Markov chain Monte Carlo method that uses Hamiltonian dynamics to propose samples that follow a target distribution. The method can avoid the random walk behavior to achieve a more effective and consistent exploration of the probability space and sensitivity to correlated parameters, which are shortcomings that plague many Markov chain Monte Carlo methods. However, the performance of Hamiltonian Monte Carlo is highly sensitive to two hyperparameters. The No-U-Turn Sampler, an extension of Hamiltonian Monte Carlo, was recently introduced to automate the tuning of these hyperparameters. Thus, this study compared the performances of Gibbs sampling, Hamiltonian Monte Carlo, and the No-U-Turn Sampler for estimating genetic parameters and breeding values as well as sampling qualities in both simulated and real pig data. For all datasets, we used a pedigree-based univariate linear mixed model. Results For all datasets, the No-U-Turn Sampler and Gibbs sampling performed comparably regarding the estimation of heritabilities and accuracies of breeding values. Compared with Gibbs sampling, the estimates of effective sample sizes for simulated and pig data with the No-U-Turn Sampler were 3.2 to 22.6 and 3.5 to 5.9 times larger, respectively. Autocorrelations decreased more quickly with the No-U-Turn Sampler than with Gibbs sampling. When true heritability was low in the simulated data, the skewness of the marginal posterior distributions with the No-U-Turn Sampler was smaller than that with Gibbs sampling. The performance of Hamiltonian Monte Carlo for sampling quality was inferior to that of No-U-Turn Sampler in the simulated data. Moreover, Hamiltonian Monte Carlo could not estimate genetic parameters because of difficulties with the hyperparameter settings with pig data. Conclusions The No-U-Turn Sampler is a promising sampling method for animal breeding because of its good sampling qualities: large effective sample sizes, low autocorrelations, and low skewness of marginal posterior distributions, particularly when heritability is low. Meanwhile, Hamiltonian Monte Carlo failed to converge with a simple univariate model for pig data. Thus, it might be difficult to use Hamiltonian Monte Carlo for usual complex models in animal breeding.

Especially in the early phase, it is difficult to obtain reliable figures on the spread of a pandemic. The effects of the COVID-19 pandemic and the associated comprehensive but incomplete data monitoring provide a strong reason to estimate the number of unreported cases.The aim of this paper is to present a simple mathematical model that allows early estimation of the number of unregistered cases (underreporting).Prevalences of reported infections in different age groups are combined with additional assumptions on relative contact rates. From this, a corrected prevalence is derived for each age group, which can then be used to estimate the number of unreported cases.Our model derives for Germany in mid-April 2020 about 2.8 times more total infections than registered cases. For Italy, the model results in a factor of 8.3. The case mortalities derived from this are 0.98% for Germany and 1.51% for Italy, which are much closer together than the case mortalities of 2.7% and 12.6% derived purely from the number of reports available at that time.The number of unreported SARS-CoV-2-infected cases derived from the model can largely explain the difference in observations in case mortalities and of conditions in the early phase of the COVID-19 pandemic in Germany and Italy. The model is simple, fast, and robust to implement, and can respond well when the reporting numbers are not representative of the population in terms of age structure. We suggest considering this model for efficient and early estimations of unreported case numbers in future epidemics and pandemics.HINTERGRUND: Gerade in der Frühphase einer Pandemie ist es schwierig, verlässliche Zahlen über deren Ausbreitung zu erhalten. Die derzeitige COVID-19-Pandemie und das damit verbundene umfassende, aber nicht vollständige Datenmonitoring bieten die Möglichkeit, die Dunkelziffer der nicht erfassten Fälle zu schätzen. ZIEL: Vorstellung eines einfachen mathematischen Modells, welches eine frühzeitige Abschätzung der Zahl nichtregistrierter Fälle (Dunkelziffer) ermöglicht.Es werden die Prävalenzen der gemeldeten Infektionen in verschiedenen Altersgruppen mit Kennzahlen der altersabhängigen Kontaktzahlen kombiniert. Daraus wird für jede Altersgruppe eine korrigierte Prävalenz abgeleitet, mit der dann die Dunkelziffer geschätzt werden kann.Unser Modell berechnet für Mitte April 2020 in Deutschland insgesamt 2,8-mal so viele Infektionen wie die Zahl der registrierten Infektionen (Fälle). Für Italien ergibt sich Mitte April 2020 ein Faktor von 8,3. Die daraus abgeleiteten Fallsterblichkeiten betragen 0,98 % für Deutschland und 1,51 % für Italien, welche deutlich näher zusammenliegen als die rein aus den zu dem Zeitpunkt vorhandenen Meldezahlen abgeleiteten Fallsterblichkeiten von 2,7 % und 12,6 %.Die aus dem Modell abgeleitete Dunkelziffer kann die unterschiedlichen Beobachtungen in den Fallsterblichkeiten und der Zustände in der Frühphase der COVID-19-Pandemie in Deutschland und Italien zu einem großen Teil erklären. Das Modell ist einfach, schnell und robust implementierbar und kann gut darauf reagieren, wenn die Meldezahlen hinsichtlich der Altersstruktur nicht repräsentativ für die Bevölkerung sind. Wir empfehlen, dieses Modell für eine effiziente und frühzeitige Schätzung nichtgemeldeter Fallzahlen bei zukünftigen Epidemien und Pandemien in Betracht zu ziehen.

After the global outbreak of the COVID-19 pandemic, an infection dynamic of immense extent developed. Since then, numerous measures have been taken to bring the infection under control. This was very successful in the spring of 2020, while the number of infections rose sharply the following autumn. To predict the occurrence of infections, epidemiological models are used. These are in principle a very valuable tool in pandemic management. However, they still partly need to be based on assumptions regarding the transmission routes and possible drivers of the infection dynamics. Despite numerous individual approaches, systematic epidemiological data are still lacking with which, for example, the effectiveness of individual measures could be quantified. Such information generated in studies is needed to enable reliable predictions regarding the further course of the pandemic. Thereby, the complexity of the models could develop hand in hand with the complexity of the available data. In this article, after delineating two basic classes of models, the contribution of epidemiological models to the assessment of various central aspects of the pandemic, such as the reproduction rate, the number of unreported cases, infection fatality rate, and the consideration of regionality, is shown. Subsequently, the use of the models to quantify the impact of measures and the effects of the "test-trace-isolate" strategy is described. In the concluding discussion, the limitations of such modelling approaches are juxtaposed with their advantages.Nach dem globalen Ausbruch der COVID-19-Pandemie entwickelte sich eine Infektionsdynamik von immensen Ausmaßen. Seitdem wird versucht, das Infektionsgeschehen mit zahlreichen Maßnahmen unter Kontrolle zu bringen. Das gelang im Frühjahr 2020 sehr gut, während im darauffolgenden Herbst die Anzahl der Infektionen stark anstieg. Zur Vorhersage des Infektionsgeschehens werden epidemiologische Modelle eingesetzt, die grundsätzlich ein sehr wertvolles Werkzeug im Pandemiemanagement sind. Allerdings beruhen sie teils immer noch auf Vermutungen bzgl. der Übertragungswege und möglicher Treiber der Infektionsdynamik. Trotz zahlreicher einzelner Ansätze fehlen auch noch heute in vielen Bereichen systematische epidemiologische Daten, mit denen z. B. die Wirksamkeit einzelner Maßnahmen nachgewiesen werden könnte. In Studien generierte Daten werden aber benötigt, um möglichst belastbare Vorhersagen bzgl. des weiteren Verlaufs der Pandemie treffen zu können. Dabei entwickelt sich die Komplexität der Modelle Hand in Hand mit der Komplexität der zur Verfügung stehenden Daten. In diesem Artikel wird nach einer Abgrenzung zweier grundsätzlicher Modellklassen der Beitrag epidemiologischer Modelle zur Beurteilung verschiedener zentraler Aspekte des Pandemieverlaufs, wie z. B. Reproduktionszahl, Dunkelziffer, Infektionssterblichkeit, sowie zur Berücksichtigung der Regionalität aufgezeigt. Anschließend wird der Einsatz der Modelle zur Quantifizierung der Wirkung von Maßnahmen und der Effekte der Strategie des Testens, Nachverfolgens und Isolierens („test-trace-isolate strategy“) beschrieben. In der abschließenden Diskussion werden die Limitationen solcher Modellierungsansätze ihren Vorteilen gegenübergestellt.

In rehabilitation science dichotomous data formats are frequently used to indicate or analyze patient characteristics. Although underlying information may be more fine graded, dichotomous or dichotomized data formats like 'yes vs. no', 'salient vs. not salient' or 'capable to work vs. not capable to work' are often adequate information formats for communication or decision making purposes. Definitions and properties of risk, odds and logits - typically applied in rehabilitation research - are outlined as frequency measures of occurrence. The association of two variables can be characterized by relative risk, risk reduction, logit reduction, odds ratio and ln(odds ratio). As odds ratio is independent from base rates it reflects association strength unambiguously. Hence, important correlational measures are defined as functions of odds ratio. It will be illustrated that statistical aspects as well as aspects regarding comprehensibility of the represented information are decisive to select appropriate measures and present data information appropriately.In den Rehabilitationswissenschaften sind Merkmale häufig dichotom ausgeprägt oder werden dichotom erfasst. Für die Kommunikation von Merkmalsausprägungen oder als Entscheidungsgrundlage erweisen sich 2-gestufte Formate wie ‚ja vs. nein‘‚ ‚unauffällig vs. auffällig‘ oder ‚arbeitsfähig vs. nicht arbeitsfähig‘ oft als günstig, auch wenn die unterliegende Information als feiner gestuft angenommen werden kann. Für die statistische Darstellung der Auftretenshäufigkeit eines Merkmals werden die Definitionen und Eigenschaften der in der Rehabilitationsforschung typischerweise angewendeten Maße Risiko, Odds und Logits behandelt. Relatives Risiko, Risiko-Reduktion, Logit-Reduktion, Odds Ratio und ln(Odds Ratio) bilden den Zusammenhang zweier Merkmale ab. Odds Ratio und ln(Odds Ratio) bleiben konstant, unabhängig davon, wie häufig die Einzelmerkmale auftreten (z. B. seltene vs. häufige Erkrankung). Diese Maße bilden die unterliegende Zusammenhangsstärke unabhängig von anderen Verteilungseigenschaften ab. Aufgrund dieser sehr günstigen Eigenschaft basieren wichtige korrelationsstatistische Maßzahlen auf Odds Ratio. Es wird gezeigt, dass sowohl statistische Eigenschaften als auch Aspekte der verständlichen und fehlerfreien Interpretierbarkeit der Analysebefunde bei der Auswahl und Darstellung der Maßzahlen berücksichtigt werden müssen.

Systematic errors and related phenomena represent an intrinsic challenge to the quality of clinical research. As a consequence even otherwise methodologically demanding studies may produce results that systematically differ from the true values. Systematic errors relating to investigative medicine are divided into six groups according to their affiliation with the consecutive chronological sections of the study. Bias can occur in preliminary literature research in the field, specifying the study design and selecting the study sample, measuring exposure and outcome, analyzing the data, interpreting the analyses and publishing the results. The most important systematic errors that concern diagnostic and interventional studies are created by access to the data of previous tests, calculated study design, preselection of the participants, comparison with non-contemporaneous controls, antedating the time of diagnosis and overdiagnosis of slowly progressive forms of diseases examined. Checking the measured values often leads to a mosaic of several biases with one being more or less dominant. Even by exercising due care in the preparation and performance of the study, the majority of distortions cannot be eliminated but only diminished. It is essential to consider each detected bias as a potential full or partial argument in support of an observed correlation. The control of systematic errors and related phenomena is both a significant element of the discussion of the study report and a key element for assessment of its scientific value.

The use and the understanding of statistics are very important for biomedical research and for the clinical practice. This is particularly true for estimation of the possibilities for different diagnostic and therapy options in the field of glaucoma. The apparent complexity and contraintuitiveness of statistics along with a cautious acceptance by many physicians, might be the cause of conscious and unconscious manipulation with data representation and interpretation.Comprehendable clarification of some typical errors in the handling of medical statistical data.Using two hypothetical examples from glaucoma diagnostics the presentation of the effect of a hypotensive drug and interpretation of the results of a diagnostic test and typical statistical applications and sources of error are analyzed in detail and discussed.Mechanisms of data manipulation and incorrect data interpretation are elucidated. Typical sources of error in the statistical analysis and data presentation are explained.The practical examples analyzed demonstrate the need to understand the basics of statistics and to be able to apply them correctly. The lack of basic knowledge or half-knowledge in medical statistics can lead to misunderstandings, confusion and wrong decisions in medical research and also in clinical practice.

Borderline personality disorders (BPD) are severe mental diseases which place high pressure on the psychiatric healthcare system. Nowadays, well-tested, disorder-specific treatment concepts are available also for inpatient treatment in Germany. These show very good and long-term improvements in the psychopathology as well as posttreatment social participation; however, prerequisites for the implementation of these evidence-based inpatient psychotherapy programs are well-trained treatment teams and appropriate financing of resource expenditure.The aim was to formulate a definition of normative needs for treatment duration and intensity for a guideline-conform, empirically proven and effective inpatient treatment of borderline personality disorder as well as the derived personnel requirements in comparison to the currently available resources within the framework of the Psychiatry Personnel Act (Psych-PV).The resource requirements were established based on evaluated hospital ward models, the recommendations of the S2 guidelines and the criteria of specialist societies and compared with the personnel stipulations according to the Psych-PV.The results for a normatively established treatment program showed a pronounced deficit in the financing of the evaluated resource requirements, even when the stipulations laid down in the Psych-PV were implemented to 100 %.Disorder-specific inpatient treatment programs for borderline personality disorder have been scientifically proven to be highly effective; however, resource analyses show that the personnel requirements necessary for effective implementation of these programs are much higher than those allocated by the funding according to the Pysch-PV. The current underfunding leads to inadequate treatment outcomes with high readmission rates and as a result high direct and indirect costs of illness.

Although bronchoalveolar lavage (BAL) is often used in the diagnosis of interstitial lung diseases (ILDs), its importance in investigating, in particular, idiopathic pulmonary fibrosis (IPF) is controversial. The cell distributions in the BAL are taken into account in the clinical routine, non-cellular characteristics of the BAL play no role.Using mathematical modeling of data, the present work investigated the extent to which BAL features enable drawing conclusions about the underlying ILK or help exclude IPF. Included in the calculation are cellular findings of the BAL, in addition the protein and albumin content of the BAL, the nicotine history (pack years), and spirometry (FEV1, IVC).Using linear discriminant analysis and creating classification trees, the relevance of the characteristics of 806 patients with ILK was examined (183 IPF, 191 cryptogenic organizing pneumonia, 147 lung involvement in autoimmune disease, 97 respiratory bronchiolitis interstitial lung disease, 118 extrinsic allergic alveolitis, 41 lymphocytic interstitial pneumonia (LIP), 23 non-specific interstitial pneumonia (NSIP), 88 controls).There was a close positive relationship between protein levels and lymphocytes in the group as a whole. No such correlations were seen in IPF and NSIP. Albumin was closely correlated with the protein content in all groups.The lymphocytes are best suited to distinguish between different ILDs. Yet, a reliable calculation of the ILD is not possible on the basis of the investigated factors, the classification error ranged from 23.5 % (IPF) to 100 % (LIP, NSIP).Constellations that likely ( 99 %) speak against an IPF are lymphocytosis 34 % or protein content 347 mg/l. The same applies to the constellation: lymphocytes 25 % together with protein 250 mg/l.In ILD, BAL findings can narrow the diagnosis, but they are seldom diagnostic. BAL can make an important contribution to excluding of IPF.

Guideline-oriented inpatient psychiatric and psychotherapeutic treatment of patients with obsessive-compulsive disorder (OCD) is an important part of the care available for these patients. It may not be adequately reflected in the current personnel resources available according to the German psychiatry personnel regulation (Psych-PV).The goal of this work was to assess the personnel resources necessary for a guideline-oriented inpatient psychiatric and psychotherapeutic treatment of patients with OCD and compare the necessary resources with the resources available according to Psych-PV.Based on the German national guidelines for OCD and a meta-analysis on treatment intensity, we formulated a normative weekly treatment plan. Based on this plan we calculated the necessary personnel resources and compared these with the resources available according to Psych-PV category A1 (standard care).The weekly treatment time for a guideline-oriented inpatient psychiatric and psychotherapeutic treatment of patients with OCD is 23.5 h per week. This corresponds to a weekly personnel requirement of 20.9 h. This requirement is only partly reflected in the Psych-PV (17.3 h, 82.8 %). The coverage of personnel resources by Psych-PV is even lower for psychotherapy provided by psychiatrist and psychologists (38.3 %, i. e. 183 min in the normative plan versus 70 min in Psych-PV).The current paper shows that the personal resources required for a guideline-oriented inpatient psychiatric and psychotherapeutic treatment of patients with OCD is not adequately reflected in the German psychiatry personnel regulation (Psych-PV). The actual shortage may be underestimated in our paper.

Multilevel Analysis (MLA) are still rarely used in Health Services Research in Germany, though hierarchical data, e. g. from patients clustered in hospitals, is often present. MLA provide the valuable opportunity to study the health care context in health care organizations and the associations between context and health care outcomes. This article's aims are to introduce this particular method of data analysis, to discuss its' benefits and its' applicability particularly for Health Services Research focusing on organizational characteristics and to provide a concise guideline for performing the analysis. First, the benefits and the necessity for MLA compared to ordinary correlation analyses in the case of hierarchical data are discussed. Furthermore, the statistical requirements and key decisions for the performance of MLA are illustrated.

Estimation of the number of deaths as a consequence of the influenza pandemics in the twentieth and twenty-first centuries (i.e. 1918-1919, 1957-1958, 1968-1970 and 2009) is a challenge worldwide and also in Germany. After conducting a systematic literature search complemented by our own calculations, values and estimates for all four pandemics were collated and evaluated.A systematic literature search including the terms death, mortality, pandemic, epidemic, Germany, 1918, 1957, 1968, 2009 was performed. Hits were reviewed by title and abstract and selected for possible relevance. We derived our own estimates using excess mortality calculations, which estimate the mortality exceeding that to be expected. All identified values were evaluated by methodology and quality of the database. Numbers of pandemic deaths were used to calculate case fatality rates and were compared with global values provided by the World Health Organization.For the pandemic 1918-1919 we identified 5 relevant publications, 3 for the pandemics 1957-1958 and 1968-1970 and 3 for 2009. For all four pandemics the most plausible estimations were based on time series analyses, taken either from the literature or from our own calculations based on monthly or weekly all cause death statistics. For the four pandemics these estimates were in chronological order 426,600 (1918-1919), 29,100 (1957-1958), 46,900 (1968-1970) and 350 (2009) excess pandemic-related deaths. This translates to an excess mortality ranging between 691 per 100,000 (0.69 % in 1918-1919) and 0.43 per 100,000 (0.00043 % in 2009). Case fatality rates showed good agreement with global estimates.We have proposed plausible estimates of pandemic-related excess number of deaths for the last four pandemics as well as excess mortality in Germany. The heterogeneity among pandemics is large with a variation factor of more than 1000. Possible explanations include characteristics of the virus or host (immunity), social conditions, status of the healthcare system and medical advances.

The past and future development of occupancy rates in prisons is of crucial importance for the judicial administration of every country. Basic factors for planning the required penal facilities are seasonal fluctuations, minimum, maximum and average occupancy as well as the present situation and potential development of certain imprisonment categories. As the prisoner number of a country is determined by a complex set of interdependent conditions, it has turned out to be difficult to provide any theoretical explanations. The idea accepted in criminology for a long time that prisoner numbers are interdependent with criminal policy must be regarded as having failed. Statistical and time series analyses may help, however, to identify the factors having influenced the development of prisoner numbers in the past. The analyses presented here, first describe such influencing factors from a criminological perspective and then deal with their statistical identification and modelling. Using the development of prisoner numbers in Hesse as an example, it has been found that modelling methods in which the independent variables predict the dependent variable with a time lag are particularly helpful. A potential complication is, however, that for predicting the number of prisoners the different dynamics in German and foreign prisoners require the development of further models.

Introduction: For survival data the coefficient of determination cannot be used to describe how good a model fits to the data. Therefore, several measures of explained variation for survival data have been proposed in recent years. Methods: We analyse an existing measure of explained variation with regard to minimisation aspects and demonstrate that these are not fulfilled for the measure. Results: In analogy to the least squares method from linear regression analysis we develop a novel measure for categorical covariates which is based only on the Kaplan-Meier estimator. Hence, the novel measure is a completely nonparametric measure with an easy graphical interpretation. For the novel measure different weighting possibilities are available and a statistical test of significance can be performed. Eventually, we apply the novel measure and further measures of explained variation to a dataset comprising persons with a histopathological papillary thyroid carcinoma. Conclusion: We propose a novel measure of explained variation with a comprehensible derivation as well as a graphical interpretation, which may be used in further analyses with survival data., Einleitung: Das Bestimmtheitsmaß kann bei Überlebenszeitdaten nicht verwendet werden um mithilfe einer Maßzahl anzugeben, wie gut ein Modell zu den vorliegenden Daten passt. Daher wurden in den letzten Jahren mehrere Maße der Erklärten Variation für Überlebenszeitdaten vorgeschlagen. Methoden: Wir analysieren eines dieser Maße der Erklärten Variation bezüglich gewisser Minimierungseigenschaften und zeigen, dass diese für das Maß nicht erfüllt sind. Ergebnisse: In Analogie zu der Kleinste-Quadrate-Methode aus der linearen Regression entwickeln wir ein neues Maß für kategorielle Kovariaten, welches nur auf dem Kaplan-Meier-Schätzer basiert. Dadurch ist das neue Maß komplett nichtparametrisch und besitzt eine einfache grafische Interpretation. Für das neue Maß stehen verschiedene Gewichtungsmöglichkeiten zur Verfügung und ein statistischer Signifikanztest kann angewendet werden. Abschließend bestimmen wir das neue Maß sowie weitere Maße der Erklärten Variation für die Personen eines Datensatzes mit einem histopathologisch gesicherten papillären Schilddrüsenkarzinom. Schlussfolgerung: Wir schlagen ein neues Maß der Erklärten Variation mit einer eingängigen Herleitung sowie einer grafischen Interpretation vor, welches bei künftigen Auswertungen von Überlebenszeitdaten verwendet werden könnte., GMS German Medical Science; 13:Doc18

In the last article, you learned about the interpretation of survival curves. This article is about the comparison of "survival" - or the time to any defined event. Two survival curves can be summarized with the hazard ratio, which is discussed in more detail in this article. In order to compare the "survival"of two treatment groups - particularly in non-randomised studies - additional variables (potential confounders) must be taken into account. This is done using a regression model, the so-called Cox-regression. After reading this article, hopefully these terms are more familiar to you.

This paper summarizes statistical methods to describe susceptibility data. A frequent data basis in resistance studies are minimum inhibitory concentrations (MICs), measured for different antimicrobial agents. In the statistical context these (semi) quantitative MIC values are ordinal scaled. Therefore, they should be analysed with statistical tools appropriate for ordinal data. The resistance situation for each antimicrobial agent is often described using frequency distributions of MIC values. Resistance patterns can be described by frequencies of resistance profiles. More detailed insights into appearance and changes of simultaneous resistance against different agents are provided by a systematic analysis of dependency structure in susceptibility data. Furthermore, the calculation of differences between resistance profiles using appropriate distance measures enables the application of common methods of multivariate statistic for description and more complex analysis of susceptibility data. To improve the comparability of study results, it is desirable to present as much information as possible in a uniform way.

Within the next 8 years about 2 000 general practitioners (GPs) will be seeking a successor in the federal state of Baden-Württemberg, Germany. Both the small number of newly qualified GPs and the wish to work as an employee or in a group practice will lead to a situation in which about 500 practices will likely not find a successor. Using a single administrative district, the aim of this analysis was to develop a projection of the demand for GP health care at the community level.Using the administrative district of Rottweil with its 21 communities, a community-based demographic forecast on the basis of current birth and death probabilities was performed. From the projected population structure, the demand for GP care in the year 2023 was derived under the assumption of unchanged age- and gender-specific numbers of GP visits. The anticipated deficit or, respectively, overrun of GPs at the community level was calculated as the difference between expected demand and number of GPs not retiring for age-related reasons.Until the year 2023 the demographic change will cause a shrinking population. However, with unchanged age- and gender-specific numbers of GP visits, a slightly higher demand of 0.6 GPs will occur as a result of population-aging. The expected age-related retirement of physicians will have a stronger impact on primary care demand than demography. Up to 32 (37%) GPs might need a successor. In addition to 4 communities today, this would result in another 5 communities not having a GP in 10 years.Communities that are at higher risk of GP shortage based on demographic changes and age of practicing GPs, can be identified by the approach described and applied here in order to implement targeted comprehensive community models of care.

Monte Carlo simulations are regarded as the most accurate method of solving complex problems in the field of dosimetry and radiation transport. In (external) radiation therapy they are increasingly used for the calculation of dose distributions during treatment planning. In comparison to other algorithms for the calculation of dose distributions, Monte Carlo methods have the capability of improving the accuracy of dose calculations - especially under complex circumstances (e.g. consideration of inhomogeneities). However, there is a lack of knowledge of how accurate the results of Monte Carlo calculations are on an absolute basis. A practical verification of the calculations can be performed by direct comparison with the results of a benchmark experiment. This work presents such a benchmark experiment and compares its results (with detailed consideration of measurement uncertainty) with the results of Monte Carlo calculations using the well-established Monte Carlo code EGSnrc. The experiment was designed to have parallels to external beam radiation therapy with respect to the type and energy of the radiation, the materials used and the kind of dose measurement. Because the properties of the beam have to be well known in order to compare the results of the experiment and the simulation on an absolute basis, the benchmark experiment was performed using the research electron accelerator of the Physikalisch-Technische Bundesanstalt (PTB), whose beam was accurately characterized in advance. The benchmark experiment and the corresponding Monte Carlo simulations were carried out for two different types of ionization chambers and the results were compared. Considering the uncertainty, which is about 0.7 % for the experimental values and about 1.0 % for the Monte Carlo simulation, the results of the simulation and the experiment coincide.

is to test individual level and regional labour market hypotheses about return to work (RTW) after vocational retraining derived from matching theory.In multilevel analyses individual data of graduates in 2006 (n=3620) by the association of German "Berufsförderungswerke" and for contextual level the regional unemployment rates in 2007 (n=159 Federal Employment Agency of Germany districts) are used.Probability of RTW rises with decreasing regional unemployment. There's an age effect only in context of high unemployment. In context of low unemployment partnership promotes RTW - but in context of high unemployment only for men and not for women. It's compatible with matching theory because family obligations lead to high individual reservation wages and entry wages are to low.Direct and indirect effects of labour market should be taken into account in research on effectiveness and in comparative evaluation on intervention quality.

A novel multiple regression method (RM) is developed to predict identity-by-descent probabilities at a locus L (IBD L ), among individuals without pedigree, given information on surrounding markers and population history. These IBD L probabilities are a function of the increase in linkage disequilibrium (LD) generated by drift in a homogeneous population over generations. Three parameters are sufficient to describe population history: effective population size (Ne), number of generations since foundation (T), and marker allele frequencies among founders (p). IBD L are used in a simulation study to map a quantitative trait locus (QTL) via variance component estimation. RM is compared to a coalescent method (CM) in terms of power and robustness of QTL detection. Differences between RM and CM are small but significant. For example, RM is more powerful than CM in dioecious populations, but not in monoecious populations. Moreover, RM is more robust than CM when marker phases are unknown or when there is complete LD among founders or Ne is wrong, and less robust when p is wrong. CM utilises all marker haplotype information, whereas RM utilises information contained in each individual marker and all possible marker pairs but not in higher order interactions. RM consists of a family of models encompassing four different population structures, and two ways of using marker information, which contrasts with the single model that must cater for all possible evolutionary scenarios in CM.

The influence of the demographic change and the impact on the entire health care system have been discussed by different theories (Expansion of Morbidity, Compression of Morbidity, Bimodality of Morbidity). This paper aims to project for the outpatient health care of statutory health insurance, the influence of the demographic development, the size and structure of statutory health insured and the change in morbidity based utilization on the outpatient volume of services for 2020.Considered impact factors and their trend specifications are separately modelled and modular combined as components of scenarios.The compound annual growth rate of the outpatient volume of services varies between almost 0 and 2.2%.For 2020 a moderate increase of outpatient volume of services for statutory health insurance is expected, if costs of dying are taken into account, even though cross-sectional data for 2011 show a decrease in expenditure profiles with increased age. Projections with the scenario technique can support strategic decision making by estimating likely trends and effects of measures given the causal process in health care systems.

In Germany a considerable increase in the number of urological cancers is expected due to demographic change. Small-scale analyses are important for directed planning of uro-oncological health care due to significant regional variability in the demographic development. In this study the number of new urological cancer cases was extrapolated on the county level for Lower Saxony.The incidence rates for penile (C60), prostate (C61), testis (C62), kidney (C64), renal pelvis/ureter (C65-66) and bladder cancer (C67, D09.0, D41.4) were extrapolated for counties and urban communes from 2010 to 2020 and to 2030 based on the regional population forecast of the State Office for Statistics of Lower Saxony (2009-2031) and gender- and 5-year age-specific incidence rates for Lower Saxony (averaged for 2006-2010).From 2010 (n=12.668) to 2020 and 2030, increases of 15% (n = 14.519; men: 15%, women: 10%) and 28% (n=16.201; men: 29%, women: 20%) are expected in urological cancers for Lower Saxony. The greatest rise is predicted for prostate cancer (2030: 31%, n = 9.732; C67 + D09.0 + D41.4: 30%; C60: 28%; C65-66: 27%; C64: 19%). Only testicular carcinomas are expected to decrease (-13%). The increase varies considerably between regions. In the counties the rates range from 7% (2030; C61: 10%) in Osterode am Harz to 63% in Vechta (C61: 70%). In the urban communes the greatest increase is predicted for Oldenburg (total: 40%; C61: 45%) and the lowest increase for Wolfsburg (total: 3%; C61: 3%).Demographic change is expected to lead to a sharp increase in urological cancers. In health care planning (e.g. specialist care) regionally very heterogeneous developments and in particular high growth and close to home care of more and more older and less mobile cancer sufferers, respectively, must be considered for rural areas.

A common indicator of the measurement of relative poverty is the disposable income of a household. Current research introduces the living standard approach as an alternative concept for describing and measuring relative poverty. This study compares both approaches with regard to subjective health status of the German population, and provides theoretical implications for the utilisation of the income and living standard approach in health research.Analyses are based on the German Socio-Economic Panel (GSOEP) from the year 2011 that includes 12 290 private households and 21106 survey members. Self-rated health was based on a subjective assessment of general health status. Income poverty is based on the equalised disposable income and is applied to a threshold of 60% of the median-based average income. A person will be denoted as deprived (inadequate living standard) if 3 or more out of 11 living standard items are lacking due to financial reasons. To calculate the discriminate power of both poverty indicators, descriptive analyses and stepwise logistic regression models were applied separately for men and women adjusted for age, residence, nationality, educational level, occupational status and marital status.The results of the stepwise regression revealed a stronger poverty-health relationship for the living standard indicator. After adjusting for all control variables and the respective poverty indicator, income poverty was statistically not significantly associated with a poor subjective health status among men (OR Men: 1.33; 95% CI: 1.00-1.77) and women (OR Women: 0.98; 95% CI: 0.78-1.22). In contrast, the association between deprivation and subjective health status was statistically significant for men (OR Men: 2.00; 95% CI: 1.57-2.52) and women (OR Women: 2.11; 95% CI: 1.76-2.64).The results of the present study indicate that the income and standard of living approach measure different dimensions of poverty. In comparison to the income approach, the living standard approach measures stronger shortages of wealth and is relatively robust towards gender differences. This study expands the current debate about complementary research on the association between poverty and health.

The patient reported outcome in interventional studies is often measured with questionnaires at baseline and after the intervention. A person whose difference in outcome exceeds a critical threshold (the minimal important difference, MID) is classified as a responder, otherwise as a non-responder. The generally low reliability of differences causes misclassifications. False positives and false negatives usually do not cancel out: an MID above the average difference results in an overestimated proportion of responders, while an MID below the average difference results in an underestimated proportion of responders. Such misclassifications can be substantial. We introduce a new and simple method for estimating the true proportion of responders which is based on the assumptions of classical test theory. The consequences of the method are demonstrated with empirical data. It is recommended to report the estimates of true responders. This applies to settings with one study group as well as to settings with an additional control group.

Multiple studies successfully applied multivariate analysis to neuroimaging data demonstrating the potential utility of neuroimaging for clinical diagnostic and prognostic purposes.Summary of the current state of research regarding the application of neuroimaging in the field of psychiatry.Literature review of current studies.Results of current studies indicate the potential application of neuroimaging data across various diagnoses, such as depression, schizophrenia, bipolar disorder and dementia. Potential applications include disease classification, differential diagnosis and prediction of disease course.The results of the studies are heterogeneous although some studies report promising findings. Further multicentre studies are needed with clearly specified patient populations to systematically investigate the potential utility of neuroimaging for the clinical routine.

Comparative quality measurements between organisations or institutions are common. Quality measures need to be standardised and risk adjusted. Random error must also be taken adequately into account. Rankings without consideration of the precision lead to flawed interpretations and enhances "gaming". Application of confidence intervals is one possibility to take chance variation into account. Funnel plots are modified control charts based on Statistical Process Control (SPC) theory. The quality measures are plotted against their sample size. Warning and control limits that are 2 or 3 standard deviations from the center line are added. With increasing group size the precision increases and so the control limits are forming a funnel. Data points within the control limits are considered to show common cause variation; data points outside special cause variation without the focus of spurious rankings. Funnel plots offer data based information about how to evaluate institutional performance within quality management contexts.Kennzahlenbasierte Qualitätsvergleiche von Einrichtungen sind verbreitet. Neben der Standardisierung und Risikoadjustierung muss die zufallsbedingte Variabilität in den Daten adäquat berücksichtigt werden. Rankings ohne Berücksichtigung der Präzision führen zu Fehlinterpretationen und fördern Datenmanipulationen. Konfidenzintervalle um die Kennzahlen der Einzeleinrichtungen sind eine Möglichkeit, die Präzision zu berücksichtigen. Funnel Plots als Spezialform von Regelkarten basieren auf der Theorie der Statistischen Prozesskontrolle (SPC). Darin werden die Kennzahlen in Beziehung zur Gruppengröße gesetzt. Warn- und Kontrollgrenzen, die sich 2 bzw. 3 Standardabweichungen vom Mittelwert aller Kennzahlen befinden, laufen mit steigender Fallzahl aufeinander zu und bilden einen Trichter. Kennzahlen innerhalb der Kontrollgrenzen weisen eine natürlicherweise zu erwartende Variabilität (common cause variation), Kennzahlen außerhalb der Kontrollgrenzen spezielle Variabilität (special cause variation) auf. Der Rankinggedanke ist aufgehoben. Funnel Plots bieten datenbasierte Kriterien, die Leistung von Einrichtungen im Qualitätsmanagement-Kontext zu bewerten.

This study examined the validity of the classification of intraoperative difficulties and its usefulness in surgical practice.Data on general surgical patients were collected in four German hospitals within a multicentre validation study. Before and immediately after surgery, the operating surgeon rated the relative difficulty of the operation using a score of 1 (easy), 2 (not easy), 3 (difficult) and 4 (very difficult). Data on the duration of surgery and on the occurrence of intraoperative and postoperative complications were collected. Multivariate regression models were constructed to examine whether different clinical variables and the surgeon's preoperative assessment of surgical difficulty increased the power of the prognostic model. The R(2) statistics, which describe explained variance (EV) as a percentage was used to compare regression models.From July 2010 to August 2011 overall 500 patients were analyzed. Most patients were classified as being ideal (30 %) or relatively ideal (49 %) candidates for surgery. Preoperative and postoperative classification results were identical in 64 % of patients and were partly determined by classical risk factors (ASA score, number of previous surgeries, type of surgery, body mass index and gender). The addition of the surgeon's risk estimation to the multivariate models improved the prediction of duration of surgery (from 41.4% to 45.5 % EV), complications (from 22.5% to 24.5 % EV) and length of stay (from 32.6% to 34.5 % EV).The classification of intraoperative difficulty can be applicable in surgical daily practice in terms of surgical decision-making in difficult intraoperative situations as well as in operating room management. It could also be useful for other surgical disciplines.

Accurate modeling of spatial dependencies between observations is a significant challenge in research on regional health-care services. This article provides insight into current methods of modeling relationships in regional health-care service research, with consideration of spatial dependencies. Spatial dependencies may be triggered by spillover effects between neighboring regions and spatially distributed differences in - e.g., morbidity - which are not observable. If not considered in the model, the results of the analyses may be biased. Spatial dependencies can be added to the regression model as a spatial lag or a spatial error term. Using an example study, we illustrate that failing to consider spatial autocorrelation may lead to biased coefficients and/or standard errors. Research on regional health-care services should, therefore, if possible, test for spatial autocorrelation in the data and adjust the model accordingly.

Purpose: The purpose of study was to compare the diagnostic value of a thick slap MIP projection in patients with mesenteric ischemia. Materials and methods: 36 individuals suspect of having mesenteric ischemia were included in the study. Standard biphasic (arterial and venous phase) computed tomography was performed in all patients. Two readers read the datasets twice in a randomized order. Reader 1 was a medical student, reader 2 an experienced radiologist. In the first read, the readers were instructed to read the datasets with the thick slap MIP projection, in the second read without MIP projection. Results: 26 out of the 36 patients turned out to have mesenteric ischemia confirmed by surgery and histopathology. Of the 26 patients having mesenteric ischemia, two turned out to have a non-occlusive disease. Sensitivity in detecting mesenteric ischemia was 76 % using the MIP projection and also without MIP projection and for reader 1. Specificity was 100 % for both reading methods. For reader 2 sensitivity without MIP projection was 76 % but 92 % without MIP projection. However, reader 2 increased specificity form 83 % up to 90 % using the MIP projection. Conclusion: A thick slap MIP projection does not increase readers sensitivity in detecting mesenteric ischemia. However, the experienced reader was able to increase specificity.

This study examines the quantitative changes in the somatic classification according to birth weight and duration of pregnancy of German neonates when maternal height is considered (5 maternal height groups). Our calculations were performed using data of 319 884 girls born in 2010. Overall, about 6% (18 792 girls) are classified differently (more appropriately) when group-specific norm values were used.

Being confronted with missing data, the most important step is the knowledge of the underlying mechanism. In consequence of this, modern missing data techniques in contrast to past techniques are presented and compared, which can deal with more common situations. Finally, concrete instructions for the correct treatment of missing data are given.

Background Spurious associations between single nucleotide polymorphisms and phenotypes are a major issue in genome-wide association studies and have led to underestimation of type 1 error rate and overestimation of the number of quantitative trait loci found. Many authors have investigated the influence of population structure on the robustness of methods by simulation. This paper is aimed at developing further the algebraic formalization of power and type 1 error rate for some of the classical statistical methods used: simple regression, two approximate methods of mixed models involving the effect of a single nucleotide polymorphism (SNP) and a random polygenic effect (GRAMMAR and FASTA) and the transmission/disequilibrium test for quantitative traits and nuclear families. Analytical formulae were derived using matrix algebra for the first and second moments of the statistical tests, assuming a true mixed model with a polygenic effect and SNP effects. Results The expectation and variance of the test statistics and their marginal expectations and variances according to the distribution of genotypes and estimators of variance components are given as a function of the relationship matrix and of the heritability of the polygenic effect. These formulae were used to compute type 1 error rate and power for any kind of relationship matrix between phenotyped and genotyped individuals for any level of heritability. For the regression method, type 1 error rate increased with the variability of relationships and with heritability, but decreased with the GRAMMAR method and was not affected with the FASTA and quantitative transmission/disequilibrium test methods. Conclusions The formulae can be easily used to provide the correct threshold of type 1 error rate and to calculate the power when designing experiments or data collection protocols. The results concerning the efficacy of each method agree with simulation results in the literature but were generalized in this work. The power of the GRAMMAR method was equal to the power of the FASTA method at the same type 1 error rate. The power of the quantitative transmission/disequilibrium test was low. In conclusion, the FASTA method, which is very close to the full mixed model, is recommended in association mapping studies.

There is a linear relationship between maternal height and birth weight. For each 1 cm increase in maternal height, birth weight increases by 16.7 g. Birth weight percentiles should be calculated by taking maternal height into account. We present birth weight percentile values for girls and boys born after 23-43 completed weeks of gestation for 5 maternal height groups. With these percentiles "genetically" small and "genetically" large, but healthy, neonates can be classified more adequately. The calculations are based on data of about 2.2 million singleton pregnancies from the German Perinatal Survey of 1995-2000.

Background Infectious bovine keratoconjunctivitis (IBK) or ‘pinkeye’ is an economically important ocular disease that significantly impacts animal performance. Genetic parameters for IBK infection and its genetic and phenotypic correlations with cattle tick counts, number of helminth (unspecified species) eggs per gram of faeces and growth traits in Australian tropically adapted Bos taurus cattle were estimated. Methods Animals were clinically examined for the presence of IBK infection before and after weaning when the calves were 3 to 6 months and 15 to 18 months old, respectively and were also recorded for tick counts, helminth eggs counts as an indicator of intestinal parasites and live weights at several ages including 18 months. Results Negative genetic correlations were estimated between IBK incidence and weight traits for animals in pre-weaning and post-weaning datasets. Genetic correlations among weight measurements were positive, with moderate to high values. Genetic correlations of IBK incidence with tick counts were positive for the pre-weaning and negative for the post-weaning datasets but negative with helminth eggs counts for the pre-weaning dataset and slightly positive for the post-weaning dataset. Genetic correlations between tick and helminth eggs counts were moderate and positive for both datasets. Phenotypic correlations of IBK incidence with helminth eggs per gram of faeces were moderate and positive for both datasets, but were close to zero for both datasets with tick counts. Conclusions Our results suggest that genetic selection against IBK incidence in tropical cattle is feasible and that calves genetically prone to acquire IBK infection could also be genetically prone to have a slower growth. The positive genetic correlations among weight traits and between tick and helminth eggs counts suggest that they are controlled by common genes (with pleiotropic effects). Genetic correlations between IBK incidence and tick and helminth egg counts were moderate and opposite between pre-weaning and post-weaning datasets, suggesting that the environmental and (or) maternal effects differ between these two growth phases. This preliminary study provides estimated genetic parameters for IBK incidence, which could be used to design selection and breeding programs for tropical adaptation in beef cattle.

In contrast to usual ingredients in processed packaged foodstuffs, there are no suitable and binding regulations for the labeling of unintentional allergen traces in these foods as yet. This situation is unsatisfactory in regard to the fact that even traces of undeclared "hidden" allergens can constitute a considerable health risk for food allergic consumers. Furthermore, the unintentional cross-contact (cross-contamination) of allergens is also an issue in regard to food manufacturer product liability and due diligence. Therefore, stakeholders consider imperative need for the scientific determination of maximum tolerable levels of allergen traces in order to establish thresholds for legally binding food labeling. In addition to conventional toxicological risk assessments, the risk assessment of allergen traces that would be necessary in this context nowadays also incorporates modern approaches such as benchmark procedures and probabilistic modeling and methods. The scientific debate concerning the establishment of safe threshold levels continues, and a consensus must still be reached.

The representative health surveys conducted by the Department of Epidemiology and Health Reporting weren't used before to provide estimates for the spatial distribution of health outcomes. We are discussing the possibilities of providing these outcomes using methods for 'Small-Area-Estimation'. In the study we are using data of the "German Health Update 2009" (GEDA) to analyze regional inequalities for self-assessed health status, smoking and obesity on the district level in Germany. The small area estimates are provided by multilevel logistic regression models using additional regional statistical data from the official INKAR 2009 database of regional indicators for Germany. We are mapping the results of our analysis for the district level (NUTS-3) using simple thematic maps. Afterwards we compared the results of our small area models with conventional estimates that were based on the official German small scale census. The results showed that our estimates are in line with the prevalences of the census. Overall the results suggest that Small-Area-Estimation methods have a big potential to provide regionalized health indicators for the health reporting in Germany.

Scoring systems are used in all diagnostic areas of medicine. Several parameters are evaluated and rated with points according to their value in order to simplify a complex clinical situation with a score. The application ranges from the classification of disease severity through determining the number of staff for the intensive care unit (ICU) to the evaluation of new therapies under study conditions. Since the introduction of scoring systems in the 1980's a variety of different score models has been developed. The scoring systems that are employed in intensive care and are discussed in this article can be categorized into prognostic scores, expenses scores and disease-specific scores. Since the introduction of compulsory recording of two scoring systems for accounting in the German diagnosis-related groups (DRG) system, these tools have gained more importance for all intensive care physicians. Problems remain in the valid calculation of scores and interpretation of the results.

Spinal anesthesia causes sympathetic blockade which leads to changes in the local temperature of the skin surface due to hyperemia.These changes in skin temperature were used in a newly developed method for estimating the level of analgesia. A total of 11 patients who were scheduled for surgical procedures of the lower extremities with symmetrical spinal anesthesia were included in the clinical study. By means of an electronic digital multi-channel body temperature measurement device with eight high precision temperature sensors placed on defined dermatomes, patient skin temperature was continuously measured at 2 s intervals and documented before, during and for 45 min after spinal anesthesia. Simultaneously, a neurological pin-prick test was carried on at regular intervals every 2 min on the defined dermatomes to calculate the correlation between the effects of analgesia and corresponding changes in skin temperature.The analyzed correlations showed that there is a minimum of 1.05°C temperature difference before and after spinal anesthesia especially on the lower extremities (foot, knee, inguinal) of patient dermatomes. The collected data of varying temperature differences were systematically evaluated using statistical software which led to a deeper understanding of the interdependency between temperature differences at different dermatomes. These interdependencies of temperature differences were used to develop a systematic analgesia level measurement algorithm. The algorithm calculates the skin temperature differences at specified dermatomes to find the accurate level of analgesia and also to find the forward and reverse progresses of analgesia. The developed mathematical method shows that it is possible to predict the level of analgesia up to an accuracy of 95% after spinal anesthesia.Therefore, it can be concluded that systematic processing of skin temperature data, collected at defined dermatomes can be used as a promising parameter for predicting surgical tolerance. The objective is to improve this experimental method with an extended patient population study.