Search

The presence of one or more covariates that perfectly or almost perfectly predict the outcome of interest (which is referred to as complete or quasi-complete separation, the latter denoting the case when such perfect prediction occurs only for a subset of observations in the data) has been extensively studied in the last four decades. Since 1984, when Albert and Anderson (1984) differentiated between complete and quasi-complete separation, several authors have studied this phenomenon and tried to provide answers or ways of identifying the problem (Lesaffre and Albert, 1989; Firth, 1993; Christmann and Rousseeuw, 2001; Rousseeuw and Christmann, 2003; Allison, 2004; Zorn, 2005; Heinze, 2006). From an estimation perspective, separation leads to infinite coefficients and standard errors, which makes the algorithm collapse or give inappropriate results. As a practical matter, separation forces the analyst to choose from a number of problematic alternatives for dealing with the problem, and in the past the elimination of such problematic variables were common practice to deal with such situations. In the last decade, solutions using penalized likelihood have been proposed, but always dealing with independent binary data. Here we will propose a Bayesian solution to the problem when we deal with clustered binary data using informative priors that are supported by the data and compare it with an alternative procedure proposed by Gelman et al. (2008). [ABSTRACT FROM AUTHOR]

The linear mixed-effects model (Verbeke and Molenberghs, 2000) has become a standard tool for the analysis of continuous hierarchical data such as, for example, repeated measures or data from meta-analyses. However, in certain situations the model does pose insurmountable computational problems. Precisely this has been the experience of Buyse et al. (2000a) who proposed an estimation- and prediction-based approach for evaluating surrogate endpoints. Their approach requires fitting linear mixed models to data from several clinical trials. In doing so, these authors built on the earlier, single-trial based, work by Prentice (1989), Freedman et al. (1992), and Buyse and Molenberghs (1998). While Buyse et al. (2000a) claim their approach has a number of advantages over the classical single-trial methods, a solution needs to be found for the computational complexity of the corresponding linear mixed model. In this paper, we propose and study a number of possible simplifications. This is done by means of a simulation study and by applying the various strategies to data from three clinical studies: Pharmacological Therapy for Macular Degeneration Study Group (1977), Ovarian Cancer Meta-analysis Project (1991) and Corfu-A Study Group (1995). [ABSTRACT FROM AUTHOR]

In 2023, 14 Member States were affected by African swine fever (ASF), including Croatia and Sweden where ASF emerged (wild boar outbreaks only) and Greece where ASF re‐emerged after being free since 2021. The number of ASF outbreaks among domestic pigs in the EU was five times higher than in 2022, reaching a similar magnitude to that in 2019. This was predominantly driven by the introduction and subsequent spread of ASF in Croatia and its resurgence in Romania, representing 96% of the EU outbreaks. ASF outbreaks in domestic pigs were clearly seasonal in all countries, with 88% of outbreaks reported between July and October. Most of the ASF outbreaks among domestic pigs were detected through clinical suspicion (94%), followed by tracing from affected establishments (3%), and the weekly testing of at least two dead pigs in establishments (3%). In wild boar, a 10% increase in the number of notified outbreaks was observed in the EU in comparison with 2022, with considerable variations between countries. A winter peak was observed only in Poland, Slovakia and Hungary. The epidemiological situation in wild boar improved in Germany and Hungary, as suggested by the decrease in the number of outbreaks and in the proportions of PCR‐positive samples from dead wild boar. Overall, 31% of wild boar carcasses found during passive surveillance tested positive by PCR, representing 69% of the ASF outbreaks in wild boar in the EU. In contrast, 0.4% of hunted wild boar tested positive, representing 31% of the outbreaks. Despite the introduction of ASF into new countries and the increase in the number of outbreaks, the size of restricted zones in the EU remained stable, due to the highly clustered outbreaks in Croatia, and the reduction of restricted zones in Poland, Slovakia and Bulgaria (in domestic pigs), and Hungary (in wild boar). [ABSTRACT FROM AUTHOR]

This technical report describes the protocol for the first Term of Reference (TOR 1) of the European Commission request for the scientific and technical assistance on African swine fever (ASF). This TOR 1 requests annual descriptive epidemiological analysis of the spread and impact of ASF in the domestic pig and wild boar populations in the affected EU Member States and neighbouring countries. This report presents the interpretation of the TOR, the questions and elements of the mandate, and the data and methods chosen to address those questions. The ASF epidemiological reports to be produced from 2023 to 2028 will include data from six different sources to address: i) the surveillance activities performed for ASF and their performance; iii) the spatial and temporal dynamics of the disease in the last year, and iii) the impact of the disease in domestic pigs and wild boar populations. This publication is linked to the following EFSA Journal article: http://onlinelibrary.wiley.com/doi/10.2903/j.efsa.2023.8016/full [ABSTRACT FROM AUTHOR]

This report presents the epidemiological analysis of African swine fever (ASF) during 2022 based on the surveillance and pig population data submitted by the European Union (EU) affected countries and one neighbouring country. Coinciding with regulatory changes and an important decrease in ASF outbreaks in 2022 in the EU, the number of domestic pig samples tested as part of active surveillance decreased by 80%, while the number of samples from passive surveillance almost doubled compared with 2021. Most outbreaks among domestic pigs in the EU were detected by testing clinical suspicions (93% of outbreaks), followed by tracing activities (5%) and weekly testing of the first two dead pigs per establishment (2%). Although most of the wild boar samples came from hunted animals, the probability of detecting PCR‐positive animals was much higher in wild boar found dead. The ASF outbreaks among domestic pigs in the EU decreased by 79% while a decrease of 40% in the wild boar cases was observed in comparison with 2021. This was strongly marked in Romania, Poland and Bulgaria, with a reduction of 50–80% compared with 2021. In many countries, an important decrease in the number of pig establishments was observed, especially of small establishments with fewer than 100 pigs. The regional between farm incidence and proportion of pigs lost due to ASF in the EU was in general very low (average of 1%) apart from some regions in Romania. The impact of ASF on wild boar populations was variable, with a decline in wild boar abundance observed in certain countries versus a stable or even increased population after ASF introduction. This supports the negative relationship observed in this report between the proportion of the country with restricted zones due to ASF in wild boar and wild boar hunting bags. This publication is linked to the following EFSA Supporting Publications article: http://onlinelibrary.wiley.com/doi/10.2903/sp.efsa.2023.EN-8029/full [ABSTRACT FROM AUTHOR]

The European Commission requested scientific and technical assistance in the preparation of a EU‐wide baseline survey protocol for a European Union (EU) coordinated monitoring programme on the prevalence of methicillin‐resistant Staphylococcus Aureus (MRSA) in pigs. The objective of the survey is to estimate the MRSA prevalence in batches of fattening pigs at slaughter at both European and national levels, with a 95% level of confidence and a level of precision of 10% considering an expected prevalence of 50%. The survey protocol defines the target population, the sample size for the survey, sample collection requirements, the analytical methods (for isolation, identification, phenotypic susceptibility testing and further genotypic testing of MRSA isolates), the data reporting requirements and the plan of analysis. The samples are to be analysed according to the laboratory protocols available on the European Union Reference Laboratory (EURL‐AR) website. Generalised linear models will be used to estimate proportion (with 95% confidence intervals) of batches of slaughter pigs tested positive to MRSA. The necessary data to be reported by the EU Member States to support this baseline survey are presented in three data models. The results of the survey should be reported using the EFSA data collection framework. [ABSTRACT FROM AUTHOR]

Following a request from the European Commission, the EFSA Panel on Nutrition, Novel Foods and Food Allergens (NDA) was asked to deliver a scientific opinion on the conversion of calcium‐l‐methylfolate and (6S)‐5‐methyltetrahydrofolic acid glucosamine salt (collectively called 5‐MTHF hereafter) into dietary folate equivalents (DFE). Following a systematic review, the conclusions of the opinion are based on one intervention study in adults for intakes < 400 μg/day and three intervention studies in adults for intakes ≥ 400 μg/day. At intakes below 400 μg/day, folic acid (FA) is assumed to be linearly related to responses of biomarkers of intake and status and is an appropriate comparator for deriving a DFE conversion factor for 5‐MTHF. It is proposed to use the same factor as for folic acid for conversion of 5‐MTHF into DFE for intakes < 400 μg/day. As such intake levels are unlikely to be exceeded through fortified food consumption, the conversion factor of 1.7 relative to natural food folate (NF) could be applied to 5‐MTHF added to foods and to food supplements providing < 400 μg/day. At 400 μg/day, 5‐MTHF was found to be more bioavailable than folic acid and a conversion factor of 2 is proposed for this intake level and for higher intakes. The derived DFE equations are DFE = NF + 1.7 × FA + 1.7 × 5‐MTHF for fortified foods and food supplements providing intakes < 400 μg/day; and DFE = NF + 1.7 × FA + 2.0 × 5‐MTHF for food supplements providing intakes ≥ 400 μg/day. Although this assessment applies to calcium‐L‐methylfolate and 5‐MTHF glucosamine salt, it is considered that the influence of the cation on bioavailability is likely to be within the margin of error of the proposed DFE equations. Therefore, the proposed equations can also be applied to 5‐MTHF associated with other cations. [ABSTRACT FROM AUTHOR]

At the request of the European Commission, EFSA has prepared these specific guidelines to guide the surveyor through the design of statistically sound and risk‐based surveys for Agrilus planipennis, integrating the key biological information. Based on examples, three different survey designs are simulated: (i) detection surveys to substantiate pest freedom, (ii) delimiting surveys to determine the boundaries of an infested zone, and (iii) buffer zone surveys to monitor a zone that ensures pest detection at low prevalence. The first step of the survey design consists of defining the objectives of the survey, characterising the host plant population and the methods used to detect the pest.The corresponding survey parameters are set while taking into account the importance of the underlying assumptions. The robustness of the survey conclusions will depend on the quality of the information used in this first step. The second step of the survey design consists of the sample size (i.e. number of traps) estimation using the survey parameters as inputs tothe proposed statistical tool (EFSA RiBESS+). The final step of the survey design is the allocation of the sample size in the survey area which depends on the available information about the target population and risk factors. The proposed methodology allows surveys to be compared across time and space, thus contributing to the harmonisation ofA. planipennissurveillance across the EU. The flexible approach allows surveys to be tailored to the specific situation in a Member State, taking into account the ash tree distribution and available resources. Thesuccess of a good survey design relies on the survey preparation and on the involvement of the risk managers. [ABSTRACT FROM AUTHOR]

Globally, aquaculture products are expected to account for >60% of total seafood produce by 2030. In the European Union, the seafood sector is of considerable economic importance to member states with household spending on seafood produce totalling €54.8 billion in 2016. Within the EU, shellfish aquaculture supports livelihoods and employment in many rural communities throughout the region. Harmful algal blooms pose considerable risk to consumer safety and in turn, stability of the shellfish market. If contaminated produce was to make it to the market the health risk to the public could be considerable, but the damage to the sector through loss of trust in producers would also be significant. Mytilus edulis account for a considerable portion of the aquaculture sector in the Northeast Atlantic. At present, no scientific tool is available to industry regulators, to allow them to trace mussel produce to its source, uphold food safety standards and ensure consumer confidence. The present study uses chemical analysis of shells and soft tissues to classify individual M. edulis to their site of harvest. The use of random forest classification of trace element composition has revealed location specific elemental signatures for all examined sites. This led to the correct classification of 100% of individuals sampled to their respective harvesting locations, including two sites located just 6 km apart within the same bay. The protocol demonstrated here provides the basis for a scientifically driven traceability framework for shellfish produce. Unlabelled Image • Trace element concentrations of mussels varied spatially, between rearing sites. • 100% of individuals were correctly assigned to their respective harvesting location. • Distinct elemental fingerprinting of mussels between sites 6 km apart within a bay • A viable scientific traceability tool for farmed shellfish products is presented. [ABSTRACT FROM AUTHOR]

Objectives: The objective of this study was to explore the association between resistance and outpatient antibiotic use, expressed as either DDDs per 1000 inhabitants per day (DID) or packages per 1000 inhabitants per day (PID). Methods: IMS Health data on outpatient penicillin and cephalosporin (β-lactam) and tetracycline, macrolide, lincosamide and streptogramin (TMLS) use, aggregated at the level of the active substance (WHO version 2011) expressed as DID and PID (2000-07) were linked to European Antimicrobial Resistance Surveillance System (EARSS) data on proportions of penicillin-non-susceptible Streptococcus pneumoniae (PNSP) and erythromycin-non-susceptible S. pneumoniae (ENSP) (2000-09). Combined data for 27 European countries were analysed with a generalized linear mixed model. Model fit for use in DID, PID or both and 0, 1 or 2 year time lags between use and resistance was assessed and predictions of resistance were made for decreasing use expressed as DID, PID or both. Results: When exploring the association between β-lactam use and PNSP, the best model fit was obtained for use in PID without time lag. For the association between TMLS use and ENSP, the best model fit was obtained for use in both PID and DID with a 1 year time lag. PNSP and ENSP are predicted to decrease when use decreases in PID, but not when use decreases in DID. Conclusions: Associations between outpatient antibiotic use and resistance and predictions of resistance were inconsistent whether expressing antibiotic use as DID or PID. We recommend that data on antibiotic use be expressed as PID and that time lags between use and resistance be considered when exploring these associations. [ABSTRACT FROM AUTHOR]

In the summer of 2011 Schmallenberg virus (SBV), a Culicoides -borne orthobunyavirus, emerged in Germany and The Netherlands and subsequently spread across much of Europe. To draw inferences about the transmission of SBV we have developed two models to describe its spread within and between farms. The within-farm model was fitted to seroprevalence data for cattle and sheep farms in Belgium and The Netherlands, with parameters estimated using approximate Bayesian computation. Despite the short duration of viraemia in cattle and sheep (mean of 3–4 days) the within-farm seroprevalence can reach high levels (mean within-herd seroprevalence >80%), largely because the probability of transmission from host to vector is high (14%) and SBV is able to replicate quickly (0.03 per day-degree) and at relatively low temperatures (threshold for replication: 12.3 °C). Parameter estimates from the within-farm model were then used in a separate between-farm model to simulate the regional spread of SBV. This showed that the rapid spread of SBV at a regional level is primarily a consequence of the high probability of transmission from host to vector and the temperature requirements for virus replication. Our results, obtained for a region of the UK in a typical year with regard to animal movements, indicate that there is no need to invoke additional transmission mechanisms to explain the observed patterns of rapid spread of SBV in Europe. Moreover, the imposition of movement restrictions, even a total movement ban, has little effect on the spread of SBV at this scale. [ABSTRACT FROM AUTHOR]

Summary Following its emergence in northern Europe in 2011 Schmallenberg virus (SBV), a vector-borne disease transmitted by the bites of Culicoides midges, has spread across much of the continent. Here we develop simple models to describe the spread of SBV at a continental scale and, more specifically, within and between NUTS2 regions in Europe. The model for the transmission of SBV between regions suggests that vector dispersal is the principle mechanism for transmission, even at the continental scale. The within-region model indicates that there is substantial heterogeneity amongst regions in the force of infection for cattle and sheep farms. Moreover, there is considerable under-ascertainment of SBV-affected holdings, though the level of under-ascertainment varies between regions. We contrast the relatively simple approach adopted in this study with the more complex continental-scale micro-simulation models which have been developed for pandemic influenza and discuss the strengths, weaknesses and data requirements of both approaches. [ABSTRACT FROM AUTHOR]

Abstract: According to the European Food Safety Authority, salmonellosis is still one of the main causes of infectious foodborne gastroenteritis in humans. Broilers are an important source of salmonellosis after eggs and pork. Between 1987 and 1999 the trend of human salmonellosis incidence in Belgium increased constantly. However, from 2000 until 2005 a decrease in human cases was observed, probably following the sanitary measures implemented in the poultry breeder and laying sector. In order to decrease human infections it is essential to tackle the problem at the farm level to minimize cross-contamination from farm to fork. This paper seeks to answer two questions: (i) given the Salmonella status of the farm at a certain occasion (equal to the sampling time of the flock), what are the risk factors that the farm will be Salmonella positive at a following occasion? And (ii) what are the risk factors for a farm to be persistently positive for two consecutive flocks? We used surveillance data on 6824 broiler flocks studied for Salmonella infectivity from 2005 to 2006 in Belgium. The farms were tested regularly (3 weeks before slaughter of each broiler flock) for the presence of Salmonella based on multiple faecal samples per flock on a farm yielding clustered data. Generalized estimating equations, alternating logistic regression models, and random-intercept logistic regression models were employed to analyse these correlated binary data. Our results indicated that there are many factors that influence Salmonella risk in broiler flocks, and that they interact. Accounting for interactions between risk factors leads to an improved determination of those risk factors that increase infection with Salmonella. For the conditional analysis, the risk factors found to increase the risk of Salmonella infection on a farm at a current occasion given the previous Salmonella status included: Salmonella infection of day-old chicks (of the current flock); a previously infected flock even though the farm was equipped with a hygiene place to change clothes prior to entering the broiler house; having temporary workmen when there was a separation between birds of different species; and separating birds of different species in the Walloon region relative to the Flanders region. Sanitary measures such as a cleaning and disinfecting procedure conducted by an external cleaning firm, applying the all-in all-out procedure, and hand washing decreased the risk despite their interaction with other factors. From the joint analysis, the most important factors identified for increased risk for persistent Salmonella on a farm involved the interaction between having temporary workmen when there were poultry or farmers in contact with foreign poultry or persons, and the interaction between having temporary workmen when there were poultry or farmers in contact with external poultry or persons. [Copyright &y& Elsevier]

Abstract: Since the 1980s, the prevalence of Salmonella in Belgian poultry layers and broilers has greatly fluctuated with a rise observed in 2003 and a significant decrease in 2005. In order to alleviate the risk at egg consumer level, it is crucial to understand the factors which influence the contamination and the spread of Salmonella in laying hens. To study such determinants we explored the Belgian data from the 2005 baseline study on the prevalence of Salmonella in laying flocks of Gallus gallus in the European Union. The response variables corresponded to presence or absence of Salmonella from dust and faecal samples taken from the environment of a Belgian layer flock. The explanatory variables included: region of Belgium, sampling time (month the flock was sampled), production type (cage or barn and free range), Salmonella vaccination status, flock age and flock size. Analyses of these data were performed using a bivariate logistic regression model assuming independence between the two responses and bivariate generalized estimating equations model, which incorporates the correlation between the two responses on the same flock. The main risk factor that was identified was rearing flocks in cages compared to barns and free-range systems. The results also showed a significant higher risk for Salmonella for a 1 week increase in flocks’ age as well as with a unit increase in the size of the flock. [Copyright &y& Elsevier]