Incidence trend of type 2 diabetes from 2012 to 2021 in Germany: an analysis of health claims data of 11 million statutorily insured people

AIMS/HYPOTHESIS

The aim of the study is to describe the time trend of type 2 diabetes incidence in the largest state of Germany, Bavaria, from 2012 to 2021, and to compare the incidence rates during the pandemic period (2020-2021) to the pre-pandemic period (2012-2019).

METHODS

This secondary data analysis uses health claims data provided by the Bavarian Association of Statutory Health Insurance Physicians (KVB), covering approximately 11 million insurees, accounting for 85% of the total population of Bavaria, Germany. Newly diagnosed type 2 diabetes cases in adults (≥20 years) coded as E11 (Diabetes mellitus, Type 2) or E14 (Unspecified diabetes mellitus) under ICD-10, German modification (ICD-10-GM) for the study period 2012 to 2021 were included. Annual and quarterly age-standardised incidence rates (ASIR) stratified by sex, age and region were calculated using the European standard population. Sex-specific crude incidence rates (CIR) were calculated using 10-year age groups. Regression analyses adjusted for time trends, seasonal effects, and pandemic effects were used to analyse the incidence trend and to assess the effect of the pandemic.

RESULTS

Overall, 745,861 new cases of type 2 diabetes were diagnosed between 2012 and 2021: 50.4% (376,193 cases) in women. The male/female ratio remained stable over the observation period, while the median age at diagnosis decreased from 61 to 58 years in men and from 66 years to 61 years in women. ASIR were consistently higher for men compared with women, with the yearly difference remaining stable over time (2012: 18%; 2021: 20%). An overall decreasing trend in ASIR was observed during the study period, with a strong decrease from 2012 to 2017, followed by a less pronounced decline from 2018 to 2021 for both sexes. For men, ASIR decreased from 1514 per 100,000 person-years in 2012 to 995 per 100,000 person-years in 2021 (4.6% average annual reduction), and for women from 1238 per 100,000 person-years in 2012 to 796 per 100,000 person-years in 2021 (4.8% average annual reduction). This downward trend was also observed for age groups above 50 years. Regression analyses showed no significant change in incidence rates during the pandemic period (2020 and 2021) compared with the pre-pandemic period.

CONCLUSIONS/INTERPRETATION

For the first time, a 10-year incidence trend of type 2 diabetes is reported for Germany, showing a strong decline from 2012 to 2017, followed by a less pronounced decline from 2018 to 2021. The incidence trend of type 2 diabetes appears not to have been affected by the first 2 years of the COVID-19 pandemic. Despite an overall increasing prevalence, the incidence is decreasing, potentially resulting from robust screening by family physicians, reducing the median age at diagnosis by 3 to 5 years. However, further investigation is needed to fully identify the reasons for the declining incidence trend. Continued incidence monitoring is necessary to identify the long-term trend and the potential effect of the pandemic on diagnoses of type 2 diabetes.

An audit of 1632 routinely collected cervical cancer screening smears from 398 women in Germany: Results from the TeQaZ Study

OBJECTIVE

There is evidence in Germany that half of the cervical cancer (CC) cases had undergone screening frequently in the decade preceding their diagnosis, signaling cytology quality issues. This study investigates routine smear assessment accuracy in Germany.

METHODS

Within a population-based case-control study in 9 German states, we recruited cases (women with a histologically confirmed diagnosis of CC) and population controls (women with no history of CC or hysterectomy). Two independent expert cytologists audited Pap smears taken within the 10 years preceding CC diagnosis (cases)/study entry (controls). We report the prevalence of positive results, as well as routine assessment's accuracy, as sensitivity, specificity, false-positive and false-negative rates along with 95% confidence intervals (95% CI). We also examined cases' smear history, to investigate possible false-positive recurrence.

RESULTS

We audited 1632 smears of 392 women (18.9% cases, 81.1% controls). In the routine assessment, the overall prevalence of positive results was 4.5% (29.0% among cases). According to the expert audit, the overall prevalence of positive results was 7.7% (40.8% among cases). When restricting analyses to the 3 years preceding diagnosis/study entry, this prevalence increased to 11.9% overall (61.4% among cases). The overall sensitivity of the routine assessment was 54.9% (66.8% for cases).

CONCLUSION

As cytology remains an important part of CC screening, quality issues must be urgently addressed in Germany. Shifting to objective methods such as primary high-risk HPV screening followed by triaging may help CC elimination in Germany.

Clusters of longitudinal risk profile trajectories are associated with cardiometabolic diseases: Results from the population-based KORA cohort

BACKGROUND

Multiple risk factors contribute jointly to the development and progression of cardiometabolic diseases. Therefore, joint longitudinal trajectories of multiple risk factors might represent different degrees of cardiometabolic risk.

METHODS

We analyzed population-based data comprising three examinations (Exam 1: 1999-2001, Exam 2: 2006-2008, Exam 3: 2013-2014) of 976 male and 1004 female participants of the KORA cohort (Southern Germany). Participants were followed up for cardiometabolic diseases, including cardiovascular mortality, myocardial infarction and stroke, or a diagnosis of type 2 diabetes, until 2016. Longitudinal multivariate k-means clustering identified sex-specific trajectory clusters based on nine cardiometabolic risk factors (age, systolic and diastolic blood pressure, body-mass-index, waist circumference, Hemoglobin-A1c, total cholesterol, high- and low-density lipoprotein cholesterol). Associations between clusters and cardiometabolic events were assessed by logistic regression models.

RESULTS

We identified three trajectory clusters for men and women, respectively. Trajectory clusters reflected a distinct distribution of cardiometabolic risk burden and were associated with prevalent cardiometabolic disease at Exam 3 (men: odds ratio (OR)ClusterII = 2.0, 95% confidence interval: (0.9-4.5); ORClusterIII = 10.5 (4.8-22.9); women: ORClusterII = 1.7 (0.6-4.7); ORClusterIII = 5.8 (2.6-12.9)). Trajectory clusters were furthermore associated with incident cardiometabolic cases after Exam 3 (men: ORClusterII = 3.5 (1.1-15.6); ORClusterIII = 7.5 (2.4-32.7); women: ORClusterII = 5.0 (1.1-34.1); ORClusterIII = 8.0 (2.2-51.7)). Associations remained significant after adjusting for a single time point cardiovascular risk score (Framingham).

CONCLUSIONS

On a population-based level, distinct longitudinal risk profiles over a 14-year time period are differentially associated with cardiometabolic events. Our results suggest that longitudinal data may provide additional information beyond single time-point measures. Their inclusion in cardiometabolic risk assessment might improve early identification of individuals at risk.

A systematic review of clinical health conditions predicted by machine learning diagnostic and prognostic models trained or validated using real-world primary health care data

With the advances in technology and data science, machine learning (ML) is being rapidly adopted by the health care sector. However, there is a lack of literature addressing the health conditions targeted by the ML prediction models within primary health care (PHC) to date. To fill this gap in knowledge, we conducted a systematic review following the PRISMA guidelines to identify health conditions targeted by ML in PHC. We searched the Cochrane Library, Web of Science, PubMed, Elsevier, BioRxiv, Association of Computing Machinery (ACM), and IEEE Xplore databases for studies published from January 1990 to January 2022. We included primary studies addressing ML diagnostic or prognostic predictive models that were supplied completely or partially by real-world PHC data. Studies selection, data extraction, and risk of bias assessment using the prediction model study risk of bias assessment tool were performed by two investigators. Health conditions were categorized according to international classification of diseases (ICD-10). Extracted data were analyzed quantitatively. We identified 106 studies investigating 42 health conditions. These studies included 207 ML prediction models supplied by the PHC data of 24.2 million participants from 19 countries. We found that 92.4% of the studies were retrospective and 77.3% of the studies reported diagnostic predictive ML models. A majority (76.4%) of all the studies were for models' development without conducting external validation. Risk of bias assessment revealed that 90.8% of the studies were of high or unclear risk of bias. The most frequently reported health conditions were diabetes mellitus (19.8%) and Alzheimer's disease (11.3%). Our study provides a summary on the presently available ML prediction models within PHC. We draw the attention of digital health policy makers, ML models developer, and health care professionals for more future interdisciplinary research collaboration in this regard.

A tree-based modeling approach for matched case-control studies

Conditional logistic regression (CLR) is the indisputable standard method for the analysis of matched case-control studies. However, CLR is strongly restricted with respect to the inclusion of non-linear effects and interactions of confounding variables. A novel tree-based modeling method is proposed which accounts for this issue and provides a flexible framework allowing for a more complex confounding structure. The proposed machine learning model is fitted within the framework of CLR and, therefore, allows to account for the matched strata in the data. A simulation study demonstrates the efficacy of the method. Furthermore, for illustration the method is applied to a matched case-control study on cervical cancer.

Colposcopy non-attendance following an abnormal cervical cancer screening result: a prospective population-based cohort study

Background

A considerable proportion of cervical cancer diagnoses in high-income countries are due to lack of timely follow-up of an abnormal screening result. We estimated colposcopy non-attendance, examined the potential factors associated and described non-attendance reasons in a population-based screening study.

Methods

Data from the MARZY prospective cohort study were analysed. Co-test screen-positive women (atypical squamous cells of undetermined significance or worse [ASC-US+] or high-risk human papillomavirus [hrHPV] positive) aged 30 to 65 years were referred to colposcopy within two screening rounds (3-year interval). Women were surveyed for sociodemographic, HPV-related and other data, and interviewed for non-attendance reasons. Logistic regression was used to examine potential associations with colposcopy attendance.

Results

At baseline, 2,627 women were screened (screen-positive = 8.7%), and 2,093 again at follow-up (screen-positive = 5.1%; median 2.7 years later). All screen-positives were referred to colposcopy, however 28.9% did not attend despite active recall. Among co-test positives (ASC-US+ and hrHPV) and only hrHPV positives, 19.6% were non-attendees. Half of only ASC-US+ screenees attended colposcopy. Middle age (adjusted odds ratio [aOR] = 1.55, 95% CI 1.02, 4.96) and hrHPV positive result (aOR = 3.04, 95% CI 1.49, 7.22) were associated with attendance. Non-attendance was associated with having ≥ 3 children (aOR = 0.32, 95% CI 0.10, 0.86). Major reasons for non-attendance were lack of time, barriers such as travel time, need for childcare arrangements and the advice against colposcopy given by the gynaecologist who conducted screening.

Conclusions

Follow-up rates of abnormal screening results needs improvement. A systematic recall system integrating enhanced communication and addressing follow-up barriers may improve screening effectiveness.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12905-022-01851-6.

Association between Habitual Diet and the Postprandial Glucose Response-An Enable Study

SCOPE

It is inconclusive which factors influence inter-individual variations of postprandial glucose response (PPGR). This study investigates whether the habitual diet is associated with PPGR.

METHODS AND RESULTS

Data from healthy adults (young adults with 18-25 years, middle-aged adults with 40-65 years, and older adults with 75-85 years) is collected at baseline and during an oral glucose tolerance test (OGTT) collected. Habitual diet is assessed by a food frequency questionnaire and two 24-h food lists. Associations between habitual diet and glucose incremental area under the curve (iAUC_min ) are examined by regression models. The intake of cereals and cereal products is negatively associated with glucose iAUC_min (p = 0.002) in the total cohort (N = 459, 50% women, 55 ± 21 years, BMI 26 ± 5 kg m^{- 2} ). Up to 9% of the variance in the glycemic response is explained by the respective dietary parameters identified in the models of the specific age groups.

CONCLUSION

There are age-specific diet-related effects on PPGR. The usual intake of cereals and cereal products seems to play a greater role in PPGR in more than one age group. Further research is needed, to establish how diet can be optimized based on age and PPGR.

Effect of legroom proportions and individual factors for sitting with crossed legs: implications on the interior design of automated driving vehicles

Sitting with crossed legs is a commonly adopted sitting posture in everyday situations. Yet, little is known about suitable design criteria to facilitate such a position inside a vehicle. This study is aimed at determining how much space is necessary for crossing the legs while considering legroom restrictions, anthropometric measures, and individual flexibility. More specifically, 3 D-kinematics of an ankle-on-knee leg-crossing task and the easiness to move ratings of 30 participants were assessed with restrictions of the legroom (2 heights × 3 distances) as well as without restrictions. Functional regression models revealed adaptations to a legroom restriction in the execution of movement, which occurred mainly in the knee joint and increased with more restricted legroom proportions. Therefore, the present study suggests a distance of 120% of the buttock-knee length between the dashboard and the occupant, as it requires only moderate adaptations and does not affect the perceived easiness of move. Practitioner Summary: This research investigated how much space is needed to cross the legs while sitting in a vehicle, finding that the movement execution is affected by legroom proportions, as well as individual anthropometry and flexibility. The study further presents the use of predicted motion traces to determine spatial requirements of movements. Abbreviations: BKL: buttock-knee length; H-point: hip point.

976Screening for cervical cancer with Human Papillomavirus testing: stand-alone is preferable over co-testing with cytology

Background

Cervical cancer screening can be conducted with cytology and Human Papillomavirus (HPV) testing but few studies have compared the latter directly to concomitant testing (co-testing). We compared these strategies to determine appropriate screening.

Methods

Within a randomised population-based cohort study conducted around Mainz, Germany, eligible women (≥30 years) were screened via Pap smear, liquid-based cytology (LBC) and HPV testing (HC2) and HPV genotyped post hoc (PCR). These tests formed three strategies: cytology (Pap or LBC) and HPV (HC2 or PCR) stand-alone and co-testing. Screen positives and 5% negative women were invited to colposcopy. Absolute and relative sensitivity, specificity, false positive rates (FPR) and number needed to colposcopy to detect one lesion (NNC) were calculated. Estimates were crude and verification bias-adjusted using stratified sampling with bootstrapped confidence intervals.

Results

Of 2,627 screened women, cytology stand-alone demonstrated lowest sensitivities (47%) and highest specificities (97%-99%) while HPV stand-alone demonstrated higher sensitivities (79%-95%) but lower specificities (94%-95%). Co-testing increased sensitivity (84%-99%) but not specificity (92%-95%). Relative sensitivities were similar between crude and adjusted estimates, with greater detection via HPV-based strategies. Specificity of co-testing with LBC relative to HPV stand-alone was near unity (0.99, 95% CI 0.99-1.00) but significantly lower than unity with Pap co-testing. FPR and NNC were greatest under co-testing.

Conclusions

HPV stand-alone screening in women over 30 years appears appropriate over co-testing as a screening strategy.

Key messages

Co-testing for cervical cancer does not appear to add any benefit in detection and may introduce unnecessary harms compared to HPV stand-alone screening.

Impact of opportunistic screening on squamous cell and adenocarcinoma of the cervix in Germany: A population-based case-control study

Background

We investigated the uptake of opportunistic cervical cancer screening (CCS) and other risk factors and their association with cervical cancer in Germany in a case-control study.

Methods and findings

We recruited incident cases of cervical cancer (ICD-10 C53) diagnosed between 2012 and 2016 and matched with three population-based controls, based on age and region of residence. Cases and controls reported their CCS participation during the past ten years (frequent: every three years; no or infrequent: less than every three years) and other relevant variables. We fitted conditional logistic regression models, reporting odds ratios (OR) and 95% confidence intervals (95% CI). We report overall and stratified analyses by histologic group (squamous cell–SCC, and adenocarcinoma–AC), T category (T1 and T2+), and age (<50 and ≥50 years). We analysed 217 cases and 652 matched controls. 53.0% of cases and 85.7% of controls attended CCS frequently. In the overall adjusted model, no or infrequent participation in CCS (OR 5.63; 95% CI 3.51 to 9.04), having had more than one sexual partner (OR 2.86; 95%CI 1.50 to 5.45) and obesity (OR 1.69; 95% CI 1.01 to 2.83) were associated with cervical cancer. Twelve years of schooling (OR 0.37; 95% CI 0.23 to 0.60) and a net monthly income of €3000 or more (OR 0.50; 95% CI 0.30 to 0.82) were protective factors. In the stratified analyses, no or infrequent participation was associated with T1 (OR 4.37; 95% CI 2.48 to 7.71), T2+ (OR 10.67; 95% CI 3.83 to 29.74), SCC (OR 6.88; 95% CI 4.08 to 11.59) and AC (OR 3.95; 95% CI 1.47 to 10.63).

Conclusion

Although women who frequently attended CCS were less likely to develop cervical cancer, especially larger tumours, the high proportion of cases who had been frequently screened prior to diagnosis underscores the need to investigate the quality of cytology and treatment of precancerous lesions in Germany.

Multivariate ordinal random effects models including subject and group specific response style effects

Common random effects models for repeated measurements account for the heterogeneity in the population by including subject-specific intercepts or variable effects. They do not account for the heterogeneity in answering tendencies. For ordinal responses in particular, the tendency to choose extreme or middle responses can vary in the population. Extended models are proposed that account for this type of heterogeneity. Location effects as well as the tendency to extreme or middle responses are modelled as functions of explanatory variables. It is demonstrated that ignoring response styles may affect the accuracy of parameter estimates. An example demonstrates the applicability of the method.

Cervical Cancer Screening: Comparison of Conventional Pap Smear Test, Liquid-Based Cytology, and Human Papillomavirus Testing as Stand-alone or Cotesting Strategies

BACKGROUND

Some countries have implemented stand-alone human papillomavirus (HPV) testing while others consider cotesting for cervical cancer screening. We compared both strategies within a population-based study.

METHODS

The MARZY cohort study was conducted in Germany. Randomly selected women from population registries aged ≥30 years (n = 5,275) were invited to screening with Pap smear, liquid-based cytology (LBC, ThinPrep), and HPV testing (Hybrid Capture2, HC2). Screen-positive participants [ASC-US+ or high-risk HC2 (hrHC2)] and a random 5% sample of screen-negatives were referred to colposcopy. Post hoc HPV genotyping was conducted by GP5+/6+ PCR-EIA with reverse line blotting. Sensitivity, specificity (adjusted for verification bias), and potential harms, including number of colposcopies needed to detect 1 precancerous lesion (NNC), were calculated.

RESULTS

In 2,627 screened women, cytological sensitivities (Pap, LBC: 47%) were lower than HC2 (95%) and PCR (79%) for CIN2+. Cotesting demonstrated higher sensitivities (HC2 cotesting: 99%; PCR cotesting: 84%), but at the cost of lower specificities (92%-95%) compared with HPV stand-alone (HC2: 95%; PCR: 94%) and cytology (97% or 99%). Cotesting versus HPV stand-alone showed equivalent relative sensitivity [HC2: 1.06, 95% confidence interval (CI), 1.00-1.21; PCR: 1.07, 95% CI, 1.00-1.27]. Relative specificity of Pap cotesting with either HPV test was inferior to stand-alone HPV. LBC cotesting demonstrated equivalent specificity (both tests: 0.99, 95% CI, 0.99-1.00). NNC was highest for Pap cotesting.

CONCLUSIONS

Cotesting offers no benefit in detection over stand-alone HPV testing, resulting in more false positive results and colposcopy referrals.

IMPACT

HPV stand-alone screening offers a better balance of benefits and harms than cotesting.See related commentary by Wentzensen and Clarke, p. 432.

Uncertainty in Latent Trait Models

A model that extends the Rasch model and the Partial Credit Model to account for subject-specific uncertainty when responding to items is proposed. It is demonstrated that ignoring the subject-specific uncertainty may yield biased estimates of model parameters. In the extended version of the model, uncertainty and the underlying trait are linked to explanatory variables. The parameterization allows to identify subgroups that differ in uncertainty and the underlying trait. The modeling approach is illustrated using data on the confidence of citizens in public institutions.

Responsibility of Individuals and Stakeholders for Obesity and a Healthy Diet: Results From a German Survey

INTRODUCTION

Overweight and obesity are thought to be mainly caused by an energy-rich diet and a sedentary lifestyle. The opinions of those with and without obesity about an individual´s and stakeholder´s responsibility for overweight and obesity as well as a healthy diet is rather unclear. Therefore, a survey was conducted to assess the thoughts of persons with and without obesity about the responsibilities for a high body weight and healthy diet.

METHODS

This telephone-based survey was conducted in Germany. Landline and mobile phone users older than 17 years were quota sampled to represent the German population (n=1,003). Additionally, 354 adults with a body mass index (BMI) ≥ 30 kg/m2 were included in the survey population. Questions on weight management, eating and drinking and anthropometry were asked. Furthermore, the opinions of participants on the responsibility of individuals and stakeholders for obesity and a healthy diet were collected. Data was statistically weighted by age, gender, education, domicile, and BMI.

RESULTS

Data of 1,357 persons (51.1% female, age: 50.5 ± 18.5 years, 15.9% with BMI ≥ 30 kg/m2) were analyzed. Participants responded that the general causes of a high body weight were low physical activity (82.7%) and excessive caloric intake (80.5%) followed by a lack of will power (72.1%). Almost 90% of the survey population reported that each individual is responsible for his/her own healthy diet. More than 85% of the survey population agreed that a healthy diet in kindergarten and nutrition education at schools should be the preferred approaches when politics take care of a person´s healthy diet. Sub-analyses revealed that BMI, sex, age, and education are potential confounders.

CONCLUSION

This German survey showed that the majority of participants indicated that the responsibility for a healthy diet lies with the individual and high body weight is caused by self-controlled attitudes. These results suggest that the survey population underestimates societal and environmental factors that contribute to the development of obesity, which could lead to attitudes that facilitate weight-related stigmatization. Furthermore, survey participants indicated that they would support policy-driven measures that promote a healthy diet.

Patient-related and anesthesia-dependent determinants for postoperative delirium after oral and maxillofacial surgery. Results from a register-based case-control study

OBJECTIVE

To identify risk factors for postoperative delirium (POD) after general oral and maxillofacial surgery.

MATERIAL AND METHODS

2420 patients were screened postoperatively for POD using the Nursing Delirium Screening Scale (NuDESC) before discharge from the post anesthesia caring unit (PACU). Basic health data and risk factors were collected. For analysis the study group (n=41) was compared to a control group of 164 randomly selected patients (case-control-ratio=1:4). To identify risk factors for POD multivariable logistic regression models were used. To see whether estimations remain stable, regression analysis was repeated for the subgroup of patients not undergoing dentoalveolar surgery (n=105). To estimate the risk for dentoalveolar surgery a logistic regression model was performed.

RESULTS

Dementia was the only significant risk factor for POD (Odds ratio 41.5; 95% CI 5.48-314), also for patients undergoing other than dentoalveolar surgery (58.1; 1.70-1983). Patients undergoing dentoalveolar surgery were more often suffering from dementia (35.5; 2.85-441), other psychiatric and neurological disorders (3.15; 1.05-9.43), were of younger age (0.97; 0.94-1.00) and had higher anesthesiological risk (3.95; 1.04-14.9).

CONCLUSION

Patients with dementia are at higher risk to develop POD after oral and maxillofacial surgery. We found a strong interdependence between age, dementia, ASA-Score and dentoalveolar surgery.

Short-Term Effects of the Serious Game "Fit, Food, Fun" on Nutritional Knowledge: A Pilot Study among Children and Adolescents

“Serious games” are a novel and entertaining approach for nutritional education. The aim of this pilot study was to evaluate the short-term effectiveness of “Fit, Food, Fun” (FFF), a serious game to impart nutritional knowledge among children and adolescents. Data collection was conducted at two secondary schools in Bavaria, Germany. The gameplay intervention (gameplay group; GG) consisted of a 15-minute FFF gameplay session during each of three consecutive days. The teaching intervention (teaching group; TG) was performed in a classic lecture format. Nutritional knowledge was evaluated via questionnaires at baseline and post-intervention. Statistical analyses were performed using R (R Core Team, 2018). In total, baseline data were available for 39 participants in the GG and 44 participants in the TG. The mean age was 13.5 ± 0.7 years in the GG and 12.8 ± 0.9 years in the TG. There was a significant (p-value < 0.001) improvement in nutritional knowledge in both intervention groups. Moreover, a between-group difference with a significantly (p-value = 0.01) higher increase in nutritional knowledge was detected for the TG. This pilot study provides evidence for the short-term effectiveness of both educational interventions on the improvement in nutritional knowledge. Finally, the FFF game might be an adequate educational tool for the transfer of nutritional knowledge among children and adolescents.

Predicting matches in international football tournaments with random forests

Many approaches that analyse and predict results of international matches in football are based on statistical models incorporating several potentially influential covariates with respect to a national team's success, such as the bookmakers’ ratings or the FIFA ranking. Based on all matches from the four previous FIFA World Cups 2002–2014, we compare the most common regression models that are based on the teams’ covariate information with regard to their predictive performances with an alternative modelling class, the so-called random forests. Random forests can be seen as a mixture between machine learning and statistical modelling and are known for their high predictive power. Here, we consider two different types of random forests depending on the choice of response. One type of random forests predicts the precise numbers of goals, while the other type considers the three match outcomes—win, draw and loss—using special algorithms for ordinal responses. To account for the specific data structure of football matches, in particular at FIFA World Cups, the random forest methods are slightly altered compared to their standard versions and adapted to the specific needs of the application to FIFA World Cup data.

Response Styles in the Partial Credit Model

In the modeling of ordinal responses in psychological measurement and survey-based research, response styles that represent specific answering patterns of respondents are typically ignored. One consequence is that estimates of item parameters can be poor and considerably biased. The focus here is on the modeling of a tendency to extreme or middle categories. An extension of the partial credit model is proposed that explicitly accounts for this specific response style. In contrast to existing approaches, which are based on finite mixtures, explicit person-specific response style parameters are introduced. The resulting model can be estimated within the framework of generalized mixed linear models. It is shown that estimates can be seriously biased if the response style is ignored. In applications, it is demonstrated that a tendency to extreme or middle categories is not uncommon. A software tool is developed that makes the model easy to apply.

Subject-specific modelling of paired comparison data: A lasso-type penalty approach

In traditional paired comparison models heterogeneity in the population is simply ignored and it is assumed that all persons or subjects have the same preference structure. In the models considered here the preference of an object over another object is explicitly modelled as depending on subject-specific covariates, therefore allowing for heterogeneity in the population. Since by construction the models contain a large number of parameters we propose to use penalized estimation procedures to obtain estimates of the parameters. The used regularized estimation approach penalizes the differences between the parameters corresponding to single covariates. It enforces variable selection and allows to find clusters of objects with respect to covariates. We consider simple binary but also ordinal paired comparisons models. The method is applied to data from a pre-election study from Germany.

Detection of differential item functioning in Rasch models by boosting techniques

Methods for the identification of differential item functioning (DIF) in Rasch models are typically restricted to the case of two subgroups. A boosting algorithm is proposed that is able to handle the more general setting where DIF can be induced by several covariates at the same time. The covariates can be both continuous and (multi-)categorical, and interactions between covariates can also be considered. The method works for a general parametric model for DIF in Rasch models. Since the boosting algorithm selects variables automatically, it is able to detect the items which induce DIF. It is demonstrated that boosting competes well with traditional methods in the case of subgroups. The method is illustrated by an extensive simulation study and an application to real data.

A penalty approach to differential item functioning in Rasch models

A new diagnostic tool for the identification of differential item functioning (DIF) is proposed. Classical approaches to DIF allow to consider only few subpopulations like ethnic groups when investigating if the solution of items depends on the membership to a subpopulation. We propose an explicit model for differential item functioning that includes a set of variables, containing metric as well as categorical components, as potential candidates for inducing DIF. The ability to include a set of covariates entails that the model contains a large number of parameters. Regularized estimators, in particular penalized maximum likelihood estimators, are used to solve the estimation problem and to identify the items that induce DIF. It is shown that the method is able to detect items with DIF. Simulations and two applications demonstrate the applicability of the method.

Subject-specific Bradley–Terry–Luce models with implicit variable selection

The Bradley–Terry–Luce (BTL) model for paired comparison data is able to obtain a ranking of the objects that are compared pairwise by subjects. The task of each subject is to make preference decisions in favour of one of the objects. This decision is binary when subjects prefer either the first object or the second object, but can also be ordinal when subjects make their decisions on more than two preference categories. Since subject-specific covariates, which reflect characteristics of the subject, may affect the preference decision, it is essential to incorporate subject-specific covariates into the model. However, the inclusion of subject-specific covariates yields a model that contains many parameters and thus estimation becomes challenging. To overcome this problem, we propose a procedure that is able to select and estimate only relevant variables.

Site-dependent decrease of odour-related peak-to-mean factors with distance

Abstract. The peak-to-mean concept developed earlier by the authors to calculate odour-related separation distances is applied here to meteorological input for dispersion models provided by ultrasonic anemometers. In addition to conventional meteorological input parameters like wind direction, wind speed and stability classes, three-dimensional sonics provide also turbulence information via the Obukhov stability parameter and the variance of the wind speed, which can be used directly to determine peak-to-mean ratios depending on the distance from the source. The influence and importance of these site-specific peak-to-mean ratios on the resulting direction-dependent separation distances is investigated and discussed.

Novel approach for assessing performance of PCR cyclers used for diagnostic testing

As part of a large international project for validation and standardization of PCR, the influence of thermocyclers on PCR was tested. Six brand-new, Peltier technology-driven 96-well thermocyclers were subjected to a novel and stringent in-tube (not block) physical testing. The temperature was directly monitored in PCR tubes containing 50 microl of distilled water at 13 different block positions. The certified temperature accuracy of the measurement system was +/-0.3 degrees C. Finally, the results of the physical testing were compared to those of an amplification efficiency study running an in-house PCR assay. The cyclers did not perform within the manufacturer's specification. Premature timing, under- and overshooting, and spatial variation of heat transfer were found to be the critical factors. The physical testing allowed us to distinguish accurate from less-accurate (2/6) cyclers. The lack of thermal homogeneities became most evident at the denaturation level during the first 15 s. At the time point zero, the accurate cyclers showed temperature deviations of 0.5 to 1.5 degrees C, whereas less-accurate cyclers failed to reach the set temperature by 13 to 20 degrees C. Consequently, the two less-accurate cyclers could not gain positive PCR results by running an in-house PCR assay. However, by modifying the original temperature protocol by increasing the denaturation temperature and time, the amplification efficiency of these two cyclers could be improved significantly. The results have implication for laboratories using diagnostic PCR testing.

Ultraviolet protective performance of photoprotective lipsticks: change of spectral transmittance because of ultraviolet exposure

BACKGROUND

Photoinstability of sunscreens because of ultraviolet (UV) exposure is a well-known and common phenomenon. Recently, it was also shown that sunscreens with complex filter combinations are photoinactivated by UV exposures, which can easily be acquired by solar exposure over several hours.

OBJECTIVES

To assess the change of the spectral transmission after UV exposure (UV-challenged protective performance) of 27 commercially available photoprotective lipsticks.

METHODS

Quartz slides were covered with a lipstick layer (area density 1.0+/-0.1 mg/cm2) and irradiated with increasing doses of solar-simulated radiation. The spectral transmission (T) was measured spectrophotometrically before and after 5, 12.5, 25, and 50 standard erythema doses (SED) of exposure. We calculated the change in transmission (photoinstability) as the difference between the spectral transmission before and after a defined UV exposure, DeltaT, and the arithmetic mean, for both the UVA (DeltaTA) and UVB (DeltaTB) ranges. A product was labelled as photounstable if the mean photoinstability in the UVA, DeltaTA, or UVB range, DeltaTB, was higher than 5% for an UV exposure of 12.5 SED.

RESULTS

Eleven products showed a significant photoinstability in the UVA range (DeltaTA between 6% and 27%), only one product in the UVB range (DeltaTB = 13%), and one product in both the UVA (DeltaTA = 31%) and UVB (DeltaTB = 9%) range. In one product photoinstability became significant in the UVA range at higher UV exposures.

CONCLUSIONS

Out of 27 lipsticks only 13 products showed a photostable performance (DeltaTA < 5% and DeltaTB < 5% for 12.5 SED). We propose therefore that only products, which fulfil these UV photostability criteria should be marketed.

Does the weather influence the occurrence of abomasal displacement in dairy cows?

Aim of the study was to investigate the correlation between the occurrence of abomasal displacement and the weather situations. Data were collected for the years 2000 and 2001 in the area around Lisbon at 26 farms keeping about 6500 Holstein-Friesian milk cows; the analysis of 372 cases of abomasal displacement were included in the study. The weather situation was assessed on daily basis: atmospheric pressure, relative humidity, precipitation, daily insolation, maximum air temperature, minimum air temperature and average wind velocity. These original parameters were modified for the statistical analyses according to atmospheric pressure, water vapour pressure, relative humidity, temperature, temperature range, precipitation, insolation and wind speed. The change of the weather situation was described by the mean absolute deviation (deviation of the meteorological parameters from the running mean over a period of 5 and 10 days) and by a method which was based on the principal component analysis of the entire data set. A weak influence of water vapour pressure, relative humidity, temperature, temperature range, precipitation, and insolation on the occurrence of abomasal displacements was found by linear correlation- and regression analysis. A higher probability for abomasal displacement was verified for low water vapour pressure, high relative humidity, low air temperature, low temperature range, high precipitation and low insolation. No statistically significant relations were found between occurrence of abomasal displacement and either wind velocity or atmospheric pressure. A time series analysis exhibited a higher prevalence for abomasal displacement for periods with a change from sunny, warm and dry days to cool, overcast and humid days. From the present work it can be concluded that the meteorological situation has an influence on the occurrence of abomasal displacement. Therefore, the weather situation should be included among the predisposing causes of the occurrence of abomasal displacements.

Assessment of thickness of photoprotective lipsticks and frequency of reapplication: results from a laboratory test and a field experiment

BACKGROUND

The thickness of the sunscreen layer that is actually applied by consumers under usual conditions has been determined for photoprotective lotions and creams; however, this question is still unanswered for photoprotective lipsticks.

OBJECTIVES

To assess lipstick thickness (area density) and frequency of application per day for two commercially available photoprotective lipsticks with different consistency.

METHODS

The study consisted of a laboratory test and a field experiment. In the laboratory test the applied lipstick thickness was determined as area density in mg cm(-2) for a group of 28 panellists under standardized conditions. In a separate group of 18 subjects we assessed the area density and the frequency of application per day for two photoprotective lipsticks during a 6-day skiing course.

RESULTS

In the laboratory test the median and 95% confidence interval of the area density was 0.98 mg cm(-2) (0.66-1.65) and 0.86 mg cm(-2) (0.63-1.40) for products A and B, respectively. The respective values of the field experiment were 1.58 mg cm(-2) (0.79-2.23) (product A) and 1.76 mg cm(-2) (1.16-3.50) (product B). Only 11% of all applications of lipstick A and 6% of all applications of lipstick B reached the reference area density of 2.0 mg cm(-2). The difference between the median of the area density for lipstick A (firm consistency) and lipstick B (soft consistency) was not statistically significant. No statistically significant influence on the area density was found for age, sex, photobiological skin type or regular lipstick use. The median daily frequency of application was 2.2 times for lipstick A and 3.0 times for lipstick B.

CONCLUSION

Our investigation shows that photoprotective lipsticks are applied in a much thinner layer than recommended by international standards (2 mg cm(-2)). This results in a significant reduction of the photoprotective capacity. Furthermore, the frequency of application is too low for adequate protection. Therefore, we propose that the sun protection factor (SPF) should be assessed for an area density that reflects the actual usage patterns. As long as the test protocol is not adapted to the reduced area density, photoprotective lipsticks with high and ultrahigh SPF should be recommended, especially for individuals with increased risk for the development of lip malignancies.

Change of ultraviolet absorbance of sunscreens by exposure to solar-simulated radiation

Regarding the outdoor behavior of the Caucasian population, modern sunscreens should provide high and broad-spectrum ultraviolet protection in the ultraviolet B as well as in the ultraviolet A range and should be photochemically stable for ultraviolet doses, which can be expected in solar radiation. At present an assessment of the photostability of suncare products is not a general requirement before marketing. In order to evaluate the photostability of suncare products we conducted an in vitro test and measured the spectral absorbance of 16 sunscreens before, and after exposure to increasing biologically weighted standard erythema doses (5, 12.5, 25, 50) of solar-simulated radiation. Seven of 16 suncare products showed a significant dose- and wavelength-dependent decrease of the ultraviolet A protective capacity, whereas the ability to absorb ultraviolet B was not affected. In the ultraviolet A range, the decrease of absorbance (photoinactivation), respectively, the increase of transmission was 12-48% for an ultraviolet exposure of 25 standard erythema dose. Photoinactivation started in the wavelength range between 320 and 335 nm with a maximum above 350 nm. Furthermore, our analysis showed that the behavior of suncare products was not predictable from its individual ingredients. Neither complex combinations of organic filters nor addition of inorganic filters could absolutely prevent photoinactivation. The inclusion of a single photounstable filter did not mean photoinstability of the complete suncare product. Photoinactivation of sunscreens appears to be an underestimated hazard to the skin, first, by formation of free radicals, second, by increased ultraviolet A transmission.

Predicting odour impact using the Austrian odour dispersion model (AODM)

Using a dispersion model to calculate ambient odour concentrations, the separation distance between livestock buildings and residential areas is defined by the odour impact criteria using a combination of a pre-selected odour threshold and an exceeding probability. The dynamic Austrian Odour Dispersion Model (AODM), a Gaussian model, is used to calculate the direction-dependent separation distances for several combinations of these two values, which represent the protection level of various land use categories. The calculated direction-dependent separation distances are a function of the prevailing wind velocity and atmospheric stability conditions. At a site in the Austrian North-alpine foreland, the direction-dependent separation distance (calculated on the basis of a two year time series of meteorological data) for pure residential areas (3% exceeding probability over the year for an odour threshold of 1 OU/m3) lies between 99 m (for northerly winds with a probability of less than 3%) and 362 m (for westerly winds with a probability of 34%). For west and east the main wind directions, odour sensation can be expected more often for higher wind velocities and a neutral or stable atmosphere around sunset. Northerly and southerly winds show the typical diurnal variation of a local valley wind system with predominantly northerly daytime up-valley and southerly night-time down-valley winds.

Steady-state balance model to calculate the indoor climate of livestock buildings, demonstrated for finishing pigs

The indoor climate of livestock buildings is of importance for the well-being and health of animals and their production performance (daily weight gain, milk yield etc). By using a steady-state model for the sensible and latent heat fluxes and the CO2 and odour mass flows, the indoor climate of mechanically ventilated livestock buildings can be calculated. These equations depend on the livestock (number of animals and how they are kept), the insulation of the building and the characteristics of the ventilation system (ventilation rate). Since the model can only be applied to animal houses where the ventilation systems are mechanically controlled (this is the case for a majority of finishing pig units), the calculations were done for an example of a finishing pig unit with 1,000 animal places. The model presented used 30 min values of the outdoor parameters temperature and humidity, collected over a 2-year period, as input. The projected environment inside the livestock building was compared with recommended values. The duration of condensation on the inside surfaces was also calculated.

Comparison of models used for UV index calculations

Eighteen radiative transfer models in use for calculation of UV index are compared with respect to their results for more that 100 cloud-free atmospheres, which describe present, possible future and extreme conditions. The comparison includes six multiple-scattering spectral models, eight fast spectral models and four empirical models. Averages of the results of the six participating multiple-scattering spectral models are taken as a basis for assessment. The agreement among the multiple-scattering models is within +/- 0.5 UV index values for more than 80% of chosen atmospheric parameters. The fast spectral models have very different agreement, between +/- 1 and up to 12 UV index values. The results of the empirical models agree reasonably well with the reference models but only for the atmospheres for which they have been developed. The data to describe the atmospheric conditions, which are used for the comparison, together with the individual results of all participating models and model descriptions are available on the Internet: http://www.meteo.physik.uni-muenchen.de/++ +strahlung/cost/.

[Human behavior in the solar radiation field with reference to ultraviolet exposure]

There is a causal relation between solar ultraviolet radiation and skin cancer. For epidemiological investigations, quantification of the UV exposure is essential. To set up a risk assessment for the whole population, a representative survey was performed in Austria. The questionnaire refers to three sectors of everyday life: work, recreation and holidays; in addition the use of solaria is asked about for a further investigation. The UV exposure caused by humans' behaviour in the field of solar radiation was analysed from various demographic aspects. For some subpopulations the UV exposure sustained during work, recreation and holidays was compared. Groups with high occupational UV exposure show a weaker tendency to stay outdoors during leisure time and holidays than groups characterized by high UV exposure in their leisure time, who also prefer sun-intensive activities during holidays.

Trend analysis of solar ultraviolet exposure of the Austrian population caused by holiday patterns since 1969

Over the last few decades, a steady increase in the incidence of skin cancer has been observed. Solar ultraviolet (UV) exposure is undoubtedly one of the primary causes, and of these, holiday patterns are an intrinsic component. To assess solar UV exposure for this trend analysis, changes in holiday patterns were investigated. For Austrians, a preponderance of holidays are spent at the Mediterranean Sea. These holidays represent not only a potentially high UV exposure because of location but also because sun-intensive activities are preferred. The analysis shows an increase in solar UV exposure from all holidays of 1.6% per year. About 22% of the population spends holidays at the Mediterranean Sea, and UV exposure of the total population from these holidays has increased annually by about 3.3%. Because of the latency period, it is not possible to draw conclusions from the increase in UV exposure and the temporal trends in skin cancer. A longer time series would be necessary to do this. A direct comparison of data from other countries reporting increases in skin cancer does not seem appropriate, as levels of UV exposure vary greatly and such factors as genetic susceptibility are involved.

Model for the global irradiance of the solar biologically-effective ultraviolet-radiation on inclined surfaces

The body surface area of man is the relevant receiving surface for solar UV radiation. To consider this body surface geometry, the biologically-effective UV radiation of the solar global radiation was measured. This was done at 26 differently aligned measuring points whose orientation was determined by the angle of inclination (vertical) and the azimuth (horizontal). Approximately eight hundred sets of measurement series were carried out at 33 different sites. A simple model, developed from the data obtained, made it possible to calculate relative irradiance as a function of the angle of inclination and the ground reflection (UV albedo). Thus relative risk of solar UV exposure to different regions of the body can be assessed. In addition to this, if the irradiance on a horizontal plane (measured or calculated by a corresponding model) is taken into consideration, the absolute values for UV irradiance on tilted planes can be determined.

Effect of topical solvents on ultraviolet B erythema

The skin of subjects was pretreated with 3 solvents, normal saline, 95% ethanol and diethylether. No significant difference in erythemal sensitivity to UVB was demonstrated for any of the solvents.

[UV-inactivation of microorganisms in water]

UV-Inactivation of Escherichia coli, Bacillus subtilis spores, Staphylococcus-Phage A 994, Poliovirus type Mahoney and Rotavirus SA 11 was tested under controlled physical conditions. B. subtilis-spores were found to be the most resistant of these microorganisms, followed by Rotavirus, Bacteriophage and Poliovirus. E. coli required the lowest irradiation dose for inactivation. Causes and meaning of these dose-survival-reactions are discussed.

[Computer simulation of the stable climate on the PC in veterinary practice]

The wide distribution of personal computers allows the application of programs which enable veterinary surgeons to assist in the management of live stock. A simulation model is made for the field of indoor climates of stables which makes a full-year calculation of temperature and humidity of the indoor climate and the quality of the air in stables. The calculation is based on a steady-state balance model of energy (heat and humidity) and matter. Using the example of a pig fattening sty, all the necessary system parameters are stated that are needed to totally present the model. The results of the simulation are then interpreted. Furthermore the assertions and conclusions will be shown which result thereby.