Treatment specific survival curves are an important tool to illustrate the treatment effect in studies with time‐to‐event outcomes. In non‐randomized studies, unadjusted estimates can lead to biased depictions due to confounding. Multiple methods to adjust survival curves for confounders exist. However, it is currently unclear which method is the most appropriate in which situation. Our goal is to compare forms of inverse probability of treatment weighting, the G‐Formula, propensity score matching, empirical likelihood estimation and augmented estimators as well as their pseudo‐values based counterparts in different scenarios with a focus on their bias and goodness‐of‐fit. We provide a short review of all methods and illustrate their usage by contrasting the survival of smokers and non‐smokers, using data from the German Epidemiological Trial on Ankle‐Brachial‐Index. Subsequently, we compare the methods using a Monte‐Carlo simulation. We consider scenarios in which correctly or incorrectly specified models for describing the treatment assignment and the time‐to‐event outcome are used with varying sample sizes. The bias and goodness‐of‐fit is determined by taking the entire survival curve into account. When used properly, all methods showed no systematic bias in medium to large samples. Cox regression based methods, however, showed systematic bias in small samples. The goodness‐of‐fit varied greatly between different methods and scenarios. Methods utilizing an outcome model were more efficient than other techniques, while augmented estimators using an additional treatment assignment model were unbiased when either model was correct with a goodness‐of‐fit comparable to other methods. These “doubly‐robust” methods have important advantages in every considered scenario.

Visualization is a key aspect of communicating the results of any study aiming to estimate causal effects. In studies with time-to-event outcomes, the most popular visualization approach is depicting survival curves stratified by the variable of interest. This approach cannot be used when the variable of interest is continuous. Simple workarounds, such as categorizing the continuous covariate and plotting survival curves for each category, can result in misleading depictions of the main effects. Instead, we propose a new graphic, the survival area plot, to directly depict the survival probability over time and as a function of a continuous covariate simultaneously. This plot utilizes g-computation based on a suitable time-to-event model to obtain the relevant estimates. Through the use of g-computation, those estimates can be adjusted for confounding without additional effort, allowing a causal interpretation under the standard causal identifiability assumptions. If those assumptions are not met, the proposed plot may still be used to depict noncausal associations. We illustrate and compare the proposed graphics to simpler alternatives using data from a large German observational study investigating the effect of the Ankle-Brachial Index on survival. To facilitate the usage of these plots, we additionally developed the contsurvplot R-package, which includes all methods discussed in this paper.

In recent times, large language models (LLMs) have made significant strides in generating computer code, blurring the lines between code created by humans and code produced by artificial intelligence (AI). As these technologies evolve rapidly, it is crucial to explore how they influence code generation, especially given the risk of misuse in areas such as higher education. The present paper explores this issue by using advanced classification techniques to differentiate between code written by humans and code generated by ChatGPT, a type of LLM. We employ a new approach that combines powerful embedding features (black-box) with supervised learning algorithms including Deep Neural Networks, Random Forests, and Extreme Gradient Boosting to achieve this differentiation with an impressive accuracy of $98\%$. For the successful combinations, we also examine their model calibration, showing that some of the models are extremely well calibrated. Additionally, we present white-box features and an interpretable Bayes classifier to elucidate critical differences between the code sources, enhancing the explainability and transparency of our approach. Both approaches work well, but provide at most 85–88% accuracy. Tests on a small sample of untrained humans suggest that humans do not solve the task much better than random guessing. This study is crucial in understanding and mitigating the potential risks associated with using AI in code generation, particularly in the context of higher education, software development, and competitive programming.

Aim: This article presents the study protocol of a retrospective cohort study designed to compare the effectiveness of herpes zoster, and influenza vaccines in individuals with multiple sclerosis (MS), chronic inflammatory bowel diseases (IBD), or chronic inflammatory rheumatic diseases (CIRD) to individuals without these diseases, using claims data of one of the largest population based health insurances in Germany.Background: Individuals with autoimmune diseases such as MS, IBD, and CIRD are more susceptible to vaccine preventable infectious diseases such as influenza and herpes zoster, due to the autoimmune disease itself, the presence of comorbidities and immunosuppressive therapies. Vaccines are the primary means to prevent such diseases. The efficacy of these vaccines is usually estimated using large randomized controlled trials, from which patients with MS, IBD, and CIRD are often excluded. It is therefore unclear whether these vaccines are also effective for these patients.Design: A target trial emulation based on observational claims data of a statutory health insurance company is proposed.Methods: This study will aim to emulate multiple target trials to compare the effectiveness of herpes zoster and influenza vaccines in patients with and without MS, IBD and CIRD using data from a large German statutory health insurance provider (BARMER). The primary outcome for each vaccine effectiveness analysis is the disease itself. The analysis will be carried out using both time-dependent matching and a multivariable Cox proportional hazards model in conjunction with g-computation. Additionally, the moderating effect of immunosuppressive therapies on the vaccine effectiveness will be estimated using a stratified secondary analysis.Discussion: This study will estimate and compare the effectiveness of influenza and herpes zoster vaccines in individuals with and without MS, IBD, and CIRD. Because of the large amount of data, this study will also be able to investigate the role of the immunosuppressive medication on vaccine effectiveness, which may provide guidance for vaccine administration guidelines.

In 2022, over 1,000,000 people from Ukraine fled to Germany, mostly women and children. The aim of this study was to determine health status and socio-demographic background of Ukrainian refugee minors in Germany and to compare that to German minors. In this study, Ukrainian refugees of all ages, who voluntarily made contact with the study centres via flyers or refugee shelters, were interviewed with the help of trained interpreters from 09–12/2022. We assessed data from minors and their families on socio-economic status, general health perception, current and pre-existing infectious and non-contagious diseases, and vaccination status. The data for the minors were collected by interviewing the accompanying adults. In addition, a physical examination and serological screening for infectious and vaccine-preventable diseases were performed. Overall, 392 minors and accompanying adults (358) were included. Minors’ general and mental health were reported as “at least good" in 83% (316) and 80% (304), respectively. In 16% (63), infectious or other underlying diseases were recorded. Tuberculosis infections (7, 1.8%) and hepatitis C (2, 0.5%) were found through the serological tests. HIV could only be detected in one previously known case. Self-reported vaccinations against measles and varicella, for example, did not match serological findings (protection against measles 304; 84%; varicella 230; 64%). Conclusion: Accompanying adults often rate the general health and well-being of Ukrainian refugee minors as rather poor. Frequently, self-reported vaccinations and preexisting diseases did not match serology results; therefore, Ukrainian refugee minors without formal vaccination certificate should be offered all recommended vaccinations and be generously screened for hepatitis B, C and HIV in addition to tuberculosis. What is Known:• Infectious diseases (TB, hepatitis B/C, HIV) are more common among Ukrainian minors and vaccine hesitancy is more widespread than in many host countries and require appropriate medical care.• Displacement and war worsen medical care and have negative impact on mental health.. What is New:• Although intrafamilial spread of infectious diseases was rare (in this cohort), perceived general health of Ukrainian minors is rather poor.• Semi-structured interviews—even with interpreters—are often insufficient to collect valid medical information.

Misfolding and aggregation of alpha-synuclein (αSyn) plays a key role in the pathophysiology of Parkinson’s disease (PD). It induces cellular and axonal damage already in the early stages of the disease. Despite considerable advances in PD diagnostics by αSyn seed-amplification assays (SAAs), an early and differential diagnosis of PD still represents a major challenge. Here, we extended the immuno-infrared sensor (iRS) platform technology from Alzheimer’s disease (AD), in which β-amyloid misfolding was monitored as a fluid biomarker towards αSyn misfolding in PD. Using the iRS platform technology, we analyzed cerebrospinal fluid (CSF) from two independent cohorts, a discovery and a validation cohort comprising clinically diagnosed PD (n=57), atypical Parkinsonian disorders with αSyn pathology (multiple system atrophy (MSA), n= 5) or Tau pathology (corticobasal degeneration (CBD), n=5, progressive supranuclear palsy (PSP) n=9), and further disease controls (frontotemporal dementia (FTD) n=7 and other, n=51). In the discovery cohort, an AUC of 0.90, 95 %-CL 0.85 – 0.96 is obtained for the differentiation of PD/MSA vs. all controls, and in the validation cohort, an AUC of 0.86, 95 %-CL 0.80 - 0.93, respectively. In the combined dataset, the αSyn misfolding classifies PD/MSA from controls with an AUC of 0.90 (n=134, 95 %-CL 0.85 - 0.96). Using two threshold values instead of one identified people in the continuum between clearly unaffected (low misfolding group) and affected by PD/MSA (high misfolding group) with an intermediate area in between. The controls versus PD/MSA in the low vs. high misfolding group were classified with 97% sensitivity and 92% specificity. The spectral data showed misfolding in CSF from an α-helical/random-coil secondary structure of αSyn in controls to β-sheet enriched secondary structures in PD and MSA patients. In subgroups, the iRS platform implied a potential for stratifying patients with overlapping clinical symptoms. With high accuracy, the iRS αSyn misfolding platform provides a novel diagnostic tool using body fluids for differential, biological classification of the αSyn associated disorders PD and MSA. The iRS platform, indicating directly and fast all αSyn conformers, is complementary to the αSyn SAAs, which also uses αSyn misfolding as a fluid biomarker. However, monomers amplify competent misfolded conformers in SAA over time. The iRS platform opens new avenues for the stratification of PD by a body-fluid analysis and follow-ups in the continuum of healthy to clinically impaired individuals.