Using Negative Control Populations to Assess Unmeasured Confounding and Direct Effects.

Sometimes treatment effects are absent in a subgroup of the population. For example, penicillin has no effect on severe symptoms in individuals infected by resistant Staphylococcus aureus , and codeine has no effect on pain in individuals with certain polymorphisms in the CYP2D6 enzyme. Subgroups where a treatment is ineffective are often called negative control populations or placebo groups. They are leveraged to detect bias in different disciplines. Here we present formal criteria that justify the use of negative control populations to rule out unmeasured confounding and mechanistic (direct) causal effects. We further argue that negative control populations, satisfying our formal conditions, are available in many settings, spanning from clinical studies of infectious diseases to epidemiologic studies of public health interventions. Negative control populations can also be used to rule out placebo effects in unblinded randomized experiments. As a case study, we evaluate the effect of mobile stroke unit dispatches on functional outcomes at discharge in individuals with suspected stroke, using data from a large trial. Our analysis supports the hypothesis that mobile stroke units improve functional outcomes in these individuals.

Economic Evaluation of a Mobile Stroke Unit Service in Germany.

BACKGROUND: Lower global disability and higher quality of life among ischemic stroke patients was found to be associated with the dispatch of mobile stroke units (MSUs) among patients eligible for recanalizing treatments in the Berlin_Prehospital Or Usual Delivery of stroke care (B_PROUD) study. The current study assessed the cost-utility and cost-effectiveness of additional MSU dispatch using data from this prospective, controlled, intervention study. METHODS: Outcomes considered in the economic evaluation included quality-adjusted life years (QALYs) derived from the 3-level version of EQ-5D (EQ-5D-3L) and modified Rankin Scale (mRS) scores for functional outcomes 3-months after stroke. Costs were prospectively collected during the study by the MSU provider (Berlin Fire Brigade) and the B_PROUD research team. We focus our results on the societal perspective. As we aimed to determine the economic consequences of the intervention beyond the study's follow-up period, both care costs and QALYs were extrapolated over 5 years. RESULTS: The additional MSU dispatch resulted in an incremental €40,984 per QALY. The best-case scenario and the worst-case scenario yielded additional costs of, respectively, €24,470.76 and €61,690.88 per QALY. In the cost-effectiveness analysis, MSU dispatch resulted in incremental costs of €81,491 per survival without disability. The best-case scenario and the worst-case scenario yielded additional costs of, respectively, €44,455.30 and €116,491.15 per survival without disability. INTERPRETATION: Among patients eligible for recanalizing treatments in ischemic stroke, MSU dispatch was associated with both higher QALYs and higher costs and is cost-effective when considering internationally accepted thresholds ranging from an additional €40,000 to €80,000 per QALY. ANN NEUROL 2023;93:942-951.

Effect of Mobile Stroke Unit Dispatch in all Patients with Acute Stroke or TIA.

OBJECTIVE: To determine the effect of additional mobile stroke unit (MSU) dispatch on functional outcomes among the full spectrum of stroke patients, regardless of subtype or potential contraindications to reperfusion therapies. METHODS: We used data from the nonrandomized Berlin-based B_PROUD study (02/2017 to 05/2019), in which MSUs were dispatched based solely on availability, and the linked B-SPATIAL stroke registry. All patients with final stroke or transient ischemic attack (TIA) diagnoses were eligible. The intervention under study was the additional dispatch of an MSU, an emergency physician-staffed ambulance equipped to provide prehospital imaging and thrombolytic treatment, compared to conventional ambulance alone. The primary outcome was the 3-month modified Rankin Scale (mRS) score, and the co-primary outcome was a 3-tiered disability scale. We identified confounders using directed acyclic graphs and obtained adjusted effect estimates using inverse probability of treatment weighting. RESULTS: MSUs were dispatched to 1,125 patients (mean age: 74 years, 46.5% female), while for 1,141 patients only conventional ambulances were dispatched (75 years, 49.9% female). After confounding adjustment, MSU dispatch was associated with more favorable 3-month mRS scores (common odds ratio [cOR] = 0.82; 95% confidence interval [CI]: 0.71-0.94). No statistically significant association was found with the co-primary outcome (cOR = 0.86; 9% CI: 0.72-1.01) or 7-day mortality (OR = 0.94; 95% CI: 0.59-1.48). INTERPRETATION: When considering the entire population of stroke/TIA patients, MSU dispatch improved 3-month functional outcomes without evidence of compromised safety. Our results are relevant for decision-makers since stroke subtype and treatment eligibility are unknown at time of dispatch. ANN NEUROL 2023;93:50-63.

Understanding random resampling techniques for class imbalance correction and their consequences on calibration and discrimination of clinical risk prediction models.

OBJECTIVE: Class imbalance is sometimes considered a problem when developing clinical prediction models and assessing their performance. To address it, correction strategies involving manipulations of the training dataset, such as random undersampling or oversampling, are frequently used. The aim of this article is to illustrate the consequences of these class imbalance correction strategies on clinical prediction models' internal validity in terms of calibration and discrimination performances. METHODS: We used both heuristic intuition and formal mathematical reasoning to characterize the relations between conditional probabilities of interest and probabilities targeted when using random undersampling or oversampling. We propose a plug-in estimator that represents a natural correction for predictions obtained from models that have been trained on artificially balanced datasets ("naïve" models). We conducted a Monte Carlo simulation with two different data generation processes and present a real-world example using data from the International Stroke Trial database to empirically demonstrate the consequences of applying random resampling techniques for class imbalance correction on calibration and discrimination (in terms of Area Under the ROC, AUC) for logistic regression and tree-based prediction models. RESULTS: Across our simulations and in the real-world example, calibration of the naïve models was very poor. The models using the plug-in estimator generally outperformed the models relying on class imbalance correction in terms of calibration while achieving the same discrimination performance. CONCLUSION: Random resampling techniques for class imbalance correction do not generally improve discrimination performance (i.e., AUC), and their use is hard to justify when aiming at providing calibrated predictions. Improper use of such class imbalance correction techniques can lead to suboptimal data usage and less valid risk prediction models.

Consequences of age and education correction of cognitive screening tests - A simulation study of the MoCA test in Italy.

BACKGROUND: Cognitive screening tools are widely used in clinical practice to screen for age-related cognitive impairment and dementia. These tools' test scores are known to be influenced by age and education, leading to routine correction of raw scores for these factors. Despite these corrections being common practice, there is evidence suggesting that corrected scores may perform worse in terms of discrimination than raw scores. OBJECTIVE: To address the ongoing debate in the field of dementia research, we assessed the impact of the corrections on discrimination, specificity, and sensitivity of the Montreal Cognitive Assessment test in Italy, both for the overall population and across age and education strata. METHODOLOGY: We created a realistic model of the resident population in Italy in terms of age, education, cognitive impairment and test scores, and performed a simulation study. RESULTS: We confirmed that the discrimination performance was higher for raw scores than for corrected scores in discriminating patients with cognitive impairment from individuals without (areas under the curve of 0.947 and 0.923 respectively). With thresholds determined on the overall population, raw scores showed higher sensitivities for higher-risk age-education groups and higher specificities for lower-risk groups. Conversely, corrected scores showed uniform sensitivity and specificity across demographic strata, and thus better performance for certain age-education groups. CONCLUSION: Raw and corrected scores show different performances due to the underlying causal relationships between the variables. Each approach has advantages and disadvantages, the optimal choice between raw and corrected scores depends on the aims and preferences of practitioners and policymakers.

Should Cognitive Screening Tests Be Corrected for Age and Education? Insights From a Causal Perspective.

Cognitive screening tests such as the Mini-Mental State Examination are widely used in clinical routine to predict cognitive impairment. The raw test scores are often corrected for age and education, although documented poorer discrimination performance of corrected scores has challenged this practice. Nonetheless, test correction persists, perhaps due to the seemingly counterintuitive nature of the underlying problem. We used a causal framework to inform the long-standing debate from a more intuitive angle. We illustrate and quantify the consequences of applying the age-education correction of cognitive tests on discrimination performance. In an effort to bridge theory and practical implementation, we computed differences in discrimination performance under plausible causal scenarios using Open Access Series of Imaging Studies (OASIS)-1 data. We show that when age and education are causal risk factors for cognitive impairment and independently also affect the test score, correcting test scores for age and education removes meaningful information, thereby diminishing discrimination performance.

Using LASSO Regression to Estimate the Population-Level Impact of Pneumococcal Conjugate Vaccines.

Pneumococcal conjugate vaccines (PCVs) protect against diseases caused by Streptococcus pneumoniae, such as meningitis, bacteremia, and pneumonia. It is challenging to estimate their population-level impact due to the lack of a perfect control population and the subtleness of signals when the endpoint-such as all-cause pneumonia-is nonspecific. Here we present a new approach for estimating the impact of PCVs: using least absolute shrinkage and selection operator (LASSO) regression to select variables in a synthetic control model to predict the counterfactual outcome for vaccine impact inference. We first used a simulation study based on hospitalization data from Mexico (2000-2013) to test the performance of LASSO and established methods, including the synthetic control model with Bayesian variable selection (SC). We found that LASSO achieved accurate and precise estimation, even in complex simulation scenarios where the association between the outcome and all control variables was noncausal. We then applied LASSO to real-world hospitalization data from Chile (2001-2012), Ecuador (2001-2012), Mexico (2000-2013), and the United States (1996-2005), and found that it yielded estimates of vaccine impact similar to SC. The LASSO method is accurate and easily implementable and can be applied to study the impact of PCVs and other vaccines.

The Effect of Mobile Stroke Unit Care on Functional Outcomes: An Application of the Front-door Formula.

BACKGROUND: The Berlin-based B_PROUD study was designed to assess the effect of mobile stroke unit (MSU) dispatch among ischemic stroke and transient ischemic attack (TIA) patients without contraindications to reperfusion treatments. However, a large proportion of patients for whom the MSU was dispatched did not ultimately receive MSU care. We estimated the causal effect of additional MSU care on 3-month functional outcomes among B_PROUD patients for whom an MSU was dispatched. METHODS: We used data from the B_PROUD study (1 February 2017-8 May 2019). Given the presence of exposure-outcome unmeasured confounding, we used the front-door formula to identify the distribution of modified Rankin scale (mRS) outcomes under two hypothetical interventions: (1) receiving additional MSU care and (2) only receiving conventional care. We considered the time from dispatch to thrombolysis as the full mediator and adjusted for exposure-mediator and mediator-outcome confounding. We used a parametric estimator to estimate the common odds ratio (cOR) and 95% bootstrapped confidence intervals (CI). RESULTS: We included in total 768 ischemic stroke/TIA patients with MSU dispatch. The MSU was canceled for 180 (23%) patients, whereas 588 (77%) received MSU care. The unadjusted association between the care group and mRS favored conventional care (cOR = 1.7; 95% CI = 1.2, 2.3); however, after applying the front-door formula, the mRS distribution favored MSU care (cOR = 0.88; 95% CI = 0.81, 0.96). CONCLUSIONS: Receiving MSU care was associated with better functional outcomes than conventional care only, compatible with the hypothesized beneficial effect of MSU care on poststroke outcomes, among stroke and TIA patients without contraindications to reperfusion treatments.

Assessing the transportability of clinical prediction models for cognitive impairment using causal models.

BACKGROUND: Machine learning models promise to support diagnostic predictions, but may not perform well in new settings. Selecting the best model for a new setting without available data is challenging. We aimed to investigate the transportability by calibration and discrimination of prediction models for cognitive impairment in simulated external settings with different distributions of demographic and clinical characteristics. METHODS: We mapped and quantified relationships between variables associated with cognitive impairment using causal graphs, structural equation models, and data from the ADNI study. These estimates were then used to generate datasets and evaluate prediction models with different sets of predictors. We measured transportability to external settings under guided interventions on age, APOE ε4, and tau-protein, using performance differences between internal and external settings measured by calibration metrics and area under the receiver operating curve (AUC). RESULTS: Calibration differences indicated that models predicting with causes of the outcome were more transportable than those predicting with consequences. AUC differences indicated inconsistent trends of transportability between the different external settings. Models predicting with consequences tended to show higher AUC in the external settings compared to internal settings, while models predicting with parents or all variables showed similar AUC. CONCLUSIONS: We demonstrated with a practical prediction task example that predicting with causes of the outcome results in better transportability compared to anti-causal predictions when considering calibration differences. We conclude that calibration performance is crucial when assessing model transportability to external settings.

Access, referral, service provision and management of individuals with primary progressive aphasia: A survey of speech-language therapists in Italy.

BACKGROUND AND OBJECTIVES: In Italy, approximately 650 individuals receive a diagnosis of primary progressive aphasia (PPA) every year. Unfortunately, the frequency with which patients are referred to speech-language services is suboptimal, likely due to skepticism regarding the value of speech-language therapy in the context of neurodegeneration. MATERIALS AND METHODS: We conducted a virtual survey of speech and language therapists (SLTs) across Italy, to collect information about the assessment, intervention and management of patients with PPA. To ensure that as many SLTs as possible received the survey, the Italian Federation of SLTs (Federazione Logopedisti Italiani, FLI) aided in disseminating the survey. RESULTS: In total, 336 respondents participated in the online survey, 140 of whom had previous experience with PPA patients. Respondents indicated having seen a total of 428 PPA patients in the previous 24 months (three patients on average, range: 0-40). SLTs who reported never working with PPA identified underdiagnoses, low referral rates and the rarity of the clinical syndrome as major reasons for their lack of experience with PPA. SLTs with experience working with PPA indicated that patients may not have accessed services because of service dysfunction and geographical barriers. Respondents reported using informal interviews during assessments and tests developed for post-stroke aphasia, while impairment-based/restitutive interventions were utilised most often. CONCLUSION: Findings may serve to inform health policy organisations regarding the current shortcomings and needed recommendations for improving the care of individuals with PPA in Italy. Improving awareness of the utility of rehabilitation among SLTs and other clinical service providers may serve to facilitate access to intervention, which in turn will serve to better support individuals living with PPA. WHAT THIS PAPER ADDS: What is already known on the subject Speech and language therapists (SLTs) play a crucial role in the assessment, diagnosis and treatment of people with primary progressive aphasia (PPA). However, the frequency with which individuals with PPA are referred for speech and language services is suboptimal due to skepticism regarding the value of speech and language therapy in the context of neurodegeneration, the scarcity of SLTs with expertise in the treatment of PPA and the lack of awareness of the SLT role amongst referrers. What this paper adds to existing knowledge In recognition of the lack of published information on the provision of speech and language therapy services and clinicians' approaches to the assessment and treatment of individuals with PPA in Italy, we conducted an online survey to evaluate the current referral patterns for speech and language therapy services and to examine the current barriers to access these services for individuals with PPA in Italy. What are the potential or actual clinical implications of this work? The data presented here support that SLTs view treatment as useful for individuals with PPA and other professional figures and may serve to improve access to intervention, which in turn will serve to better support individuals living with PPA. The results highlight the need to inform health policy organisations about current gaps and aid in developing recommendations for improving the care of individuals with PPA, in order to understand how SLTs can best support individuals with PPA and their families.

Estimation of migraine prevalence considering active and inactive states across different age groups.

BACKGROUND: Migraine is a very common headache disorder on the population level, characterized by symptomatic attacks (activity). For many people with migraine, the migraine symptoms intermittently or permanently cease during their lifetime (inactive migraine). The current diagnostic classification of migraine considers two states: active migraine (having migraine symptoms within the last year) and not having active migraine (including both individuals with inactive migraine and those who never had migraine). Defining a state of inactive migraine that has gone into remission may better capture the trajectories of migraine across the lifespan and contribute to a better understanding of its biological processes. We aimed to quantify the prevalence of never, active, and inactive migraine separately, using modern prevalence and incidence estimation methodology to better describe the complexity of migraine trajectories at the population level. METHODS: Using a multistate modeling approach, data from the Global Burden of Disease (GBD) study, and results from a population-based study, we estimated the transition rates by which individuals moved between migraine disease states and estimated prevalences of never, active and inactive migraine. We used data from the GBD project and a hypothetical cohort of 100,000 people with a starting age of 30 and 30 years of follow-up, both in Germany and globally, stratified by sex. RESULTS: In Germany, the estimated rate of transition from active to inactive migraine (remission rate) increased after the age of 22.5 in women and 27.5 in men. The pattern for men in Germany was similar to the one observed on the global level. The prevalence of inactive migraine among women reaches 25.7% in Germany and 16.5% globally at age 60. For men, the inactive migraine prevalence estimates at the same age were 10.4% in Germany and 7.1% globally. CONCLUSIONS: Considering an inactive migraine state explicitly reflects a different epidemiological picture of migraine across the lifecourse. We have demonstrated that many women of older ages may be in an inactive migraine state. Many pressing research questions can only be answered if population-based cohort studies collect information not only on active migraine but also on inactive migraine states.

COVID-19 Vaccine-Associated Cerebral Venous Thrombosis in Germany.

OBJECTIVE: We aimed to estimate the incidence of cerebral sinus and venous thrombosis (CVT) within 1 month from first dose administration and the frequency of vaccine-induced immune thrombotic thrombocytopenia (VITT) as the underlying mechanism after vaccination with BNT162b2, ChAdOx1, and mRNA-1273, in Germany. METHODS: A web-based questionnaire was e-mailed to all departments of neurology. We requested a report of cases of CVT occurring within 1 month of a COVID-19 vaccination. Other cerebral events could also be reported. Incidence rates of CVT were calculated by using official statistics of 9 German states. RESULTS: A total of 45 CVT cases were reported. In addition, 9 primary ischemic strokes, 4 primary intracerebral hemorrhages, and 4 other neurological events were recorded. Of the CVT patients, 35 (77.8%) were female, and 36 (80.0%) were younger than 60 years. Fifty-three events were observed after vaccination with ChAdOx1 (85.5%), 9 after BNT162b2 (14.5%) vaccination, and none after mRNA-1273 vaccination. After 7,126,434 first vaccine doses, the incidence rate of CVT within 1 month from first dose administration was 0.55 (95% confidence interval [CI] = 0.38-0.78) per 100,000 person-months (which corresponds to a risk of CVT within the first 31 days of 0.55 per 100,000 individuals) for all vaccines and 1.52 (95% CI = 1.00-2.21) for ChAdOx1 (after 2,320,535 ChAdOx1 first doses). The adjusted incidence rate ratio was 9.68 (95% CI = 3.46-34.98) for ChAdOx1 compared to mRNA-based vaccines and 3.14 (95% CI = 1.22-10.65) for females compared to non-females. In 26 of 45 patients with CVT (57.8%), VITT was graded highly probable. INTERPRETATION: Given an incidence of 0.02 to 0.15 per 100,000 person-months for CVT in the general population, these findings point toward a higher risk for CVT after ChAdOx1 vaccination, especially for women. ANN NEUROL 2021;90:627-639.

FRONTotemporal dementia Incidence European Research Study-FRONTIERS: Rationale and design.

INTRODUCTION: The incidence of Frontotemporal Lobar Degeneration (FTLD)-related disorders and their characteristics are not well known. The "FRONTotemporal dementia Incidence European Research Study" (FRONTIERS) is designed to fill this gap. METHODS: FRONTIERS is a European prospective, observational population study based on multinational registries. FRONTIERS comprises 11 tertiary referral centers across Europe with long-lasting experience in FTLD-related disorders and comprehensive regional referral networks, enabling incidence estimation over well-defined geographical areas. ENDPOINTS: The primary endpoints are (1) the incidence of FTLD-related disorders across Europe; (2) geographic trends of FTLD-related disorders; (3) the distribution of FTLD phenotypes in different populations and ethnicities in Europe; (4) inheritance of FTLD-related disorders, including the frequencies of monogenic FTLD as compared to overall disease burden; and (5) implementation of data banking for clinical and biological material. EXPECTED IMPACTS: FRONTIERS will improve the understanding of FTLD-related disorders and their epidemiology, promoting appropriate public health service policies and treatment strategies.

The Reconstructed Cohort Design: A Method to Study Rare Neurodegenerative Diseases in Population-Based Settings.

Rare neurodegenerative diseases are characterized by high heterogeneity and high clinical complexity, as well as low incidence and prevalence, thus making tracking small numbers of incident cases in the general population very challenging. Since it is not possible to use classical cohort studies to estimate the incidence of these rare diseases, we can "reconstruct" a theoretical cohort using case information from a well-defined geographic region collected through a surveillance system. The incidence rate is estimated as the ratio between the number of individuals at risk who were diagnosed with the disease of interest during the study period and the estimated overall amount of time individuals in the reference population spent at risk during the study period. If a series of assumptions are met, the approximate incidence proportion of a closed theoretical cohort without competing events and with the same follow-up duration can be calculated by multiplying the incidence rate with the length of the study time. This rationale relies on the presence of an effective referral system, which links all levels of the healthcare system together in the region, from general practitioners to specialized clinical centers. The reconstructed cohort design is a valid and cost-effective method to collect data on the incidence of rare neurodegenerative diseases and represents the theoretical framework for building up population-based registries.

Directed acyclic graphs and causal thinking in clinical risk prediction modeling.

BACKGROUND: In epidemiology, causal inference and prediction modeling methodologies have been historically distinct. Directed Acyclic Graphs (DAGs) are used to model a priori causal assumptions and inform variable selection strategies for causal questions. Although tools originally designed for prediction are finding applications in causal inference, the counterpart has remained largely unexplored. The aim of this theoretical and simulation-based study is to assess the potential benefit of using DAGs in clinical risk prediction modeling. METHODS: We explore how incorporating knowledge about the underlying causal structure can provide insights about the transportability of diagnostic clinical risk prediction models to different settings. We further probe whether causal knowledge can be used to improve predictor selection in clinical risk prediction models. RESULTS: A single-predictor model in the causal direction is likely to have better transportability than one in the anticausal direction in some scenarios. We empirically show that the Markov Blanket, the set of variables including the parents, children, and parents of the children of the outcome node in a DAG, is the optimal set of predictors for that outcome. CONCLUSIONS: Our findings provide a theoretical basis for the intuition that a diagnostic clinical risk prediction model including causes as predictors is likely to be more transportable. Furthermore, using DAGs to identify Markov Blanket variables may be a useful, efficient strategy to select predictors in clinical risk prediction models if strong knowledge of the underlying causal structure exists or can be learned.

Amyotrophic Lateral Sclerosis Descriptive Epidemiology: The Origin of Geographic Difference.

Amyotrophic lateral sclerosis (ALS) epidemiology has rapidly developed in the last 30 years alongside the evolving changes in concepts in the field of clinical ALS and also due to the recent proposals of new classification system for motor neuron diseases. Many of these changes in the clinical scenario have been determined through the results of ALS population-based studies conducted in the last 20 years primarily in Europe. All the evidences converge to show that ALS risk is different across continents and ethnicities. In a European registry consortium named EURALS, ALS incident cases were drawn from a source population comprising almost 24 million people across Europe (ALS cases: 1028) and the estimated incidence was 2.2 per 100,000 person-years (py) for the general population. In contrast, other population-based studies have measured the lowest incidence in East Asia to be 0.89 per 100,000 py and in South Asia to be 0.79 per 100,000 py. A large part of Africa, Latin America and Asia does not have any population-based studies. The origin of geographic difference in ALS incidence is a matter of debate. Probably, this is partly due to genes (C9ORF72) and partly due to environmental risk factors. The rapid disappearance of ALS Foci in Guam, Kii, and West Guinea underline the importance of changes in lifestyle and environmental factors. The Global Burden of Disease, a project aiming to describe the burden of all diseases and injuries across all the countries of the world with a standardized protocol, has collected heterogeneous sources of data to estimate the burden of motor neuron diseases. The demographic changes related to increased expectation of life and the growth of the world population indicate that the load of motor neuron disease is rapidly moving toward 400 thousand prevalent cases. The burden is expected to shift toward Asia and Africa in the next decades for the rapid increase of expectation of life of countries with high demographic impact.

Screening for Aphasia in NeuroDegeneration for the Diagnosis of Patients with Primary Progressive Aphasia: Clinical Validity and Psychometric Properties.

BACKGROUND: We evaluated the psychometric proprieties of the Screening for Aphasia in NeuroDegeneration (SAND) battery in Italian primary progressive aphasia (PPA) and movement disorder (MD) patients. METHODS: The sample included 30 consecutive PPA and 45 MD patients who completed the SAND battery together with a clinical interview and a neurological/neuropsychological examination and 130 healthy controls (HC). RESULTS: The SAND battery showed good internal consistency and good convergent and divergent validity. receiver operating characteristic analysis revealed an area under the curve of 0.978 for PPA versus HC and of 0.786 for PPA versus MD. A cutoff ≥3 gave a sensitivity of 0.933% and a specificity of 0.946% for discriminating PPA versus HC, whereas a cutoff ≥5 gave a sensitivity of 0.767% and a specificity of 0.667% for discriminating PPA versus MD. CONCLUSION: These results indicate that the SAND battery is an adequate, reliable, and valid diagnostic tool for PPA.

Rethinking animal attrition in preclinical research: Expressing causal mechanisms of selection bias using directed acyclic graphs.

Animal attrition in preclinical experiments can introduce bias in the estimation of causal treatment effects, as the treatment-outcome association in surviving animals may not represent the causal effect of interest. This can compromise the internal validity of the study despite randomization at the outset. Directed Acyclic Graphs (DAGs) are useful tools to transparently visualize assumptions about the causal structure underlying observed data. By illustrating relationships between relevant variables, DAGs enable the detection of even less intuitive biases, and can thereby inform strategies for their mitigation. In this study, we present an illustrative causal model for preclinical stroke research, in which animal attrition induces a specific type of selection bias (i.e., collider stratification bias) due to the interplay of animal welfare, initial disease severity and negative side effects of treatment. Even when the treatment had no causal effect, our simulations revealed substantial bias across different scenarios. We show how researchers can detect and potentially mitigate this bias in the analysis phase, even when only data from surviving animals are available, if knowledge of the underlying causal process that gave rise to the data is available. Collider stratification bias should be a concern in preclinical animal studies with severe side effects and high post-randomization attrition.

Effect of elevated depressive symptoms during adolescence on health-related quality of life in young adulthood-a six-year cohort study with repeated exposure measurements.

Objectives: Depression is a major contributor of young people's burden of disease. In this study we aim to estimate the effect of elevated depressive symptoms on physical health-related quality of life. Design: We used self-reported information from the prospective BELLA cohort study, which included adolescents selected from the general population in Germany. The baseline assessment (2003-2006) and the 1-, 2-, and 6-year follow-up waves provide the data basis. Participants: The baseline study population consisted of 1,460 adolescents between the ages of 12 and 17 who, according to their caregivers, did not suffer from depression. Variables: The primary outcome, as measured by the physical component score (PCS) of the SF-36 at a 6-year follow-up (range: 0-100), is physical health-related quality of life. The exposure of interest is depressive symptoms, as measured by the Center for Epidemiological Studies Depression Scale for Children (CES-DC) at baseline, 1-year follow-up and 2-year follow-ups (range: 0-60). We dichotomized the exposure into subthreshold (≤15) and elevated depressive symptoms (>15). For the main analyses we considered a cumulative index for elevated depressive symptoms across the three time points (range: 0-3). Considered confounders are sex, age, socioeconomic status, migrant background, social support, anxiety symptoms, physical activity, chronic diseases, and sleeping problems. Statistical methods: We used multiple imputation to account for missing values. Within each imputed dataset, we applied inverse probability weighting (IPW) to estimate the effect of the cumulative index for elevated depressive symptoms at baseline, 1- and 2-year follow-up on physical health-related quality of life at 6-year follow-up. We derived 95% confidence intervals by bootstrapping. Results: After adjusting with IPW, the effect of the cumulative index per one unit increase of elevated depressive symptoms on the physical component score was -1.71 (95% CI: -3.51 to -0.04). The adjusted effect estimates of single exposure of elevated depressive symptoms on physical health-related quality of life were -0.83 (95% CI: -3.69 to 1.87) at baseline, -2.96 (95% CI: -4.94 to -0.52) at 1-year follow-up and -1.32 (95% CI: -3.85 to 1.15) at 2-year follow-up. Conclusion: Findings suggest that elevated depressive symptoms during adolescence decrease physical health-related quality of life in young adulthood.

Sex and gender bias in chronic coronary syndromes research: analysis of studies used to inform the 2019 European Society of Cardiology guidelines.

Background: Sex and gender inequalities in ischemic heart diseases persist. Although ischemic heart disease is less common in women, they experience worse clinical outcomes and are less likely to receive guideline-recommended treatments. The primary scientific literature from which clinical guideline recommendations are derived may not have considered potential sex- and gender biases. This study aims to determine whether the literature cited in recent cardiovascular guidelines' clinical recommendations contain sex and gender biases. Methods: We analysed publications cited in the 2019 European Society of Cardiology (ESC) guideline recommendations on chronic coronary syndromes, using a checklist to guide data extraction and evaluate the individual studies for sex- and gender-related aspects, such as inclusion/exclusion criteria, outcome measures, and demographic data reporting. To assess representation over time, the proportion of women participants in each study was computed and analysed using a beta regression model. We also examined the associations between women's representation, journal impact factor and author gender. Findings: Among the 20 ESC recommendations on chronic coronary syndromes, four contained sex-related statements; we did not identify any gender-specific suggestions. The referenced literature upon which these recommendations were based consisted of 108 articles published between 1991 and 2019, encompassing more than 1.6 million study participants (26.8%; 432,284 women). Only three studies incorporated sex-sensitive designs; none were gender-specific. The term "gender" did not occur in 84% (n = 91/108) of the publications; when used, it was exclusively to denote biological sex. The proportion of women (assumed by investigators) among study participants fluctuated over time. Having a woman as first (odds ratio (OR) = 1.68, 95% CI: 1.19-2.39) or last author (OR = 2.28, 95% CI: 1.31-3.97), was significantly associated with having more women participants in the study. Interpretation: The data underlying ESC guideline recommendations largely lack reporting of possible sex- and gender-specific aspects, and women are distinctly underrepresented. To what extent these recommendations apply to members of specific population groups who are not well-represented in the underlying evidence base remains unknown. Funding: This study is part of the Gender and Health Inequalities (GENDHI) project, ERC-2019-SyG. This project has received funding from the European Research Council (ERC).