Complication Prediction after Esophagectomy with Machine Learning.

Esophageal cancer can be treated effectively with esophagectomy; however, the postoperative complication rate is high. In this paper, we study to what extent machine learning methods can predict anastomotic leakage and pneumonia up to two days in advance. We use a dataset with 417 patients who underwent esophagectomy between 2011 and 2021. The dataset contains multimodal temporal information, specifically, laboratory results, vital signs, thorax images, and preoperative patient characteristics. The best models scored mean test set AUROCs of 0.87 and 0.82 for leakage 1 and 2 days ahead, respectively. For pneumonia, this was 0.74 and 0.61 for 1 and 2 days ahead, respectively. We conclude that machine learning models can effectively predict anastomotic leakage and pneumonia after esophagectomy.

Time-dependent association between STOPP and START criteria and gastrointestinal bleeding in older patients using routinely collected primary care data.

PURPOSE: Only few studies have assessed the preventive effect of the STOPP/START criteria on adverse events. We aim to quantify 1) the association between nonadherence to STOPP/START criteria and gastrointestinal bleedings, and 2) the association between exposure to the potentially harmful START-medications and gastrointestinal bleedings. DESIGN: A retrospective cohort study using routinely collected data of patients aged ≥ 65 years from the electronic health records (EHR) of 49 general practitioners (GPs) in 6 GP practices, from 2007 to 2014. The database is maintained in the academic research network database (AHA) of Amsterdam UMC, the Netherlands. METHODS: Gastrointestinal bleedings were identified using ICPC codes and free text inspections. Three STOPP and six START criteria pertaining to gastrointestinal bleedings were selected. Cox proportional hazards regression with time-dependent covariate analysis was performed to assess the independent association between nonadherence to the STOPP/START criteria and gastrointestinal bleedings. The analysis was performed with all criteria as a composite outcome, as well as separately for the individual criteria. RESULTS: Out of 26,576 participants, we identified 19,070 Potential Inappropriate Medications (PIM)/Potential Prescribing Omission (PPO) instances for 3,193 participants and 146 gastrointestinal bleedings in 143 participants. The hazard ratio for gastrointestinal bleedings of STOPP/STARTs, taken as composite outcome, was 5.45 (95% CI 3.62-8.21). When analysed separately, two out of nine STOPP/STARTs showed significant associations. CONCLUSION: This study demonstrates a significant positive association between nonadherence to the STOPP/START criteria and gastrointestinal bleeding. We emphasize the importance of adherence to the relevant criteria for gastrointestinal bleeding, which may be endorsed by decision support systems.

Bayesian diagnostic test evaluation and true prevalence estimation of mycoplasma bovis in dairy herds.

The true prevalence of dairy cattle herds with M. bovis infections in the Netherlands is unknown. Previous attempts to estimate prevalences were hampered by the absence of a diagnostic serological test that was validated under field conditions. This study estimated sensitivity and specificity of two commercial serum ELISAs and the true M. bovis herd prevalence using different Bayesian latent class models. A total of 7305 serum samples from 415 randomly chosen dairy herds were collected in fall/winter 2019 and investigated for presence of antibodies against M. bovis using the BIO-K-260 ELISA from Bio-X. Serum samples from 100 of these herds were also tested with a second ELISA, from IDvet. A Bayesian latent class model using the paired test results estimated a sensitivity of 14.1% (95% Bayesian probability interval (BPI): 11.6-16.7%) for the Bio-X ELISA and a specificity of 97.2% (95% BPI: 95.9-98.4%). Sensitivity and specificity for the IDvet ELISA were estimated at 92.5% (95% BPI: 88.3-96.5%) and 99.3% (95% BPI: 98.7-99.8%), respectively. Also, Bio-X ELISA sensitivity was considerably higher with data from calves only and with data from a selection of herds with a clinical outbreak, whereas the IDvet ELISA sensitivity was fairly constant under these conditions. These differences in test sensitivity is expected to be related to an effect of time since infection. A second Bayesian model, applied on test results of all 415 herds, estimated a true herd prevalence of 69.9% (95% BPI: 62.7-77.6%), suggesting M. bovis in endemic amongst dairy cattle herds in the Netherlands. To what extent seropositive herds have experienced a clinical outbreak needs further investigation.

Added Value of Meat Inspection Data for Monitoring of Dairy Cattle Health in the Netherlands.

Meat inspection records of one large cattle slaughterhouse were analyzed to evaluate the added value of slaughterhouse data for cattle health surveillance in the Netherlands. Data were available from January 2015 to September 2018, consisting of 467,361 meat inspection records. Analyses included (1) an assessment of the representativeness of the cattle herds in the slaughterhouse data in relation to the cattle herd population in the Netherlands, and (2) multivariable analyses to quantify associations between meat inspection findings and farm of origin characteristics, and the trends in time of the findings in slaughtered cattle. Ninety percent of the meat inspection records originated from dairy cattle therefore this paper only presents the results of dairy herds (N = 422,194 cattle). The dairy herds in the slaughterhouse data seemed representative for the Dutch dairy population although their regional coverage differed from the distribution of dairy herds in the Netherlands. Non-dairy herds were underrepresented in the slaughterhouse data which stresses the importance of the inclusion of data from other slaughterhouses that may be more specialized in slaughtering beef cattle. Inspection records were categorized into 15 indicators related to ante-mortem and post-mortem findings. Following multivariable analyses, seven indicators were deemed of added value to existing cattle health surveillance components, as they provided either new information or information regarding specific health problems.

The effect of bovine viral diarrhea virus introduction on milk production of Dutch dairy herds.

Dairy cows are negatively affected by the introduction of bovine viral diarrhea virus (BVDV), and consequently, produce less milk. Existing literature on potential milk production losses is based on relatively outdated data and hardly evaluates milk production loss in relation to a new BVDV infection in a surveillance system. This study determined the annual and quarterly loss in milk production of BVDV introduction in 3,126 dairy herds participating in the Dutch BVDV-free program between 2007 and 2017. Among these herds, 640 were "breakdown-herds" that obtained and subsequently lost their BVDV-free status during the study period, and 2,486 herds obtained and retained their BVDV-free status during the study period. Milk yields before and after BVDV introduction were compared through annual and quarterly linear mixed models. The fixed variables for both models included herd type (breakdown-herd or free-herd), bovine viral diarrhea status (on an annual and quarterly basis), year, season, and a random herd effect. The dependent variable was the average daily milk yield on the test day. To define the possible BVDV-introduction dates, 4 scenarios were developed. In the default scenario, the date of breakdown (i.e., loss of the BVDV-free status) was assumed as the BVDV-introduction date. For the other 3 scenarios, the BVDV-introduction dates were set at 4, 6, and 9 mo before the date of breakdown, based on the estimated birth date of a persistently infected calf. In the default scenario, the loss in milk yield due to BVDV introduction occurred mainly in the first year after breakdown, with a reduction in yield of 0.08 kg/cow per day compared with the last year before breakdown. For the other 3 scenarios, the greatest yield reduction occurred in the second year after BVDV introduction, with a loss of 0.09, 0.09, and 0.1 kg/cow per day, respectively. For the first 4 quarters after BVDV introduction in the default scenario, milk yield loss was 0.14, 0.09, 0.02, and 0.08 kg/cow per day, respectively. These quarterly results indicated that milk yield loss was greatest in the first quarter after BVDV introduction. Overall, BVDV introduction had a negative, but on average a relatively small, effect on milk yield for herds participating in the BVDV-free program. This study will enable dairy farmers and policymakers to have a clearer understanding of the quantitative milk production effect of BVDV on dairy farms in a control program.

The Comparison of Three Statistical Models for Syndromic Surveillance in Cattle Using Milk Production Data.

Two vector-borne infections have emerged and spread throughout the north-western part of Europe in the last decade: Bluetongue virus serotype-8 (BTV-8) and the Schmallenberg virus (SBV). The objective of the current study was to compare three statistical methods when applied in a syndromic surveillance context for the early detection of emerging diseases in cattle in the Netherlands. Since BTV-8 and SBV both have a negative effect on milk production in dairy cattle, routinely collected bulk milk recordings were used to compare the three statistical methods in their potential to detect drops in milk production during a period of seven years in which BTV-8 and SBV emerged. A Cusum algorithm, Bayesian disease mapping model, and spatiotemporal cluster analysis using the space-time scan statistic were performed and their performance in terms of sensitivity and specificity was compared. Spatiotemporal cluster analysis performed best for early detection of SBV in cattle in the Netherlands with a relative sensitivity of 71% compared to clinical surveillance and 100% specificity in a year without major disease outbreaks. Sensitivity to detect BTV-8 was low for all methods. However, many alerts of reduced milk production were generated several weeks before the week in which first clinical suspicions were reported. It cannot be excluded that these alerts represent the actual first signs of BTV-8 infections in cattle in the Netherlands thus leading to an underestimation of the sensitivity of the syndromic surveillance methods relative to the clinical surveillance in place.

Comparison of international legislation and standards on veterinary drug residues in food of animal origin.

Current legislation governing monitoring of drug residues in foodstuff of animal origin is being revised at the European level. This study provides a qualitative comparison of the legislation, public and private standards in the European Union, the United States of America (USA) and the Eurasian Customs Union/Russia. We made a quantitative comparison of Maximum Residue Levels (MRLs) applied in each region for pork kidneys for tetracycline (with a focus on doxycycline), penicillin and chloramphenicol. The Customs Union generally applied lower levels than the other regions, with MRLs for tetracyclines in pig kidneys being 1200 times lower than those applied in the USA. Growing consumer interest and concern about chemicals in their food could be leveraged to support and enhance the implementation of new initiatives to improve veterinary public health. Farmers and veterinarians could help reduce findings of drug residues in meat through the judicious use of preventive actions when using veterinary medicine.

Changing surveillance objectives during the different phases of an emerging vector-borne disease outbreak: The Schmallenberg virus example.

In the late summer of 2011, a sudden rise in incidence of fever, drop in milk production and diarrhoea was observed in dairy cows in the eastern region of the Netherlands and in north-western Germany. In the autumn of 2011, a novel orthobunyavirus was identified by metagenomic analyses in samples from acutely diseased cows on a farm near the German city of Schmallenberg, and was thereafter named Schmallenberg virus (SBV). Due to the novelty of the virus, there was an immediate need for knowledge regarding the epidemiological characteristics of SBV-infections to inform surveillance and control strategies. A rapid assessment of the spread and impact of an emerging disease supports decision-makers on allocation of resources. This paper reviews the disease mitigation activities during and after the SBV epidemic in the Netherlands, to illustrate the phases in surveillance when a new (vector-borne) pathogen emerges in a country or region. Immediate and short-term disease mitigation activities that were initiated after SBV was identified are discussed in detail, as well as ways to enhance future surveillance (e.g. by syndromic surveillance) and preparedness for similar disease outbreaks. By doing so, lessons learnt from the SBV epidemic will also improve surveillance for other emerging diseases in cattle.

The Association Between the STOPP/START Criteria and Gastro-Intestinal Track Bleedings in Elderly Patients.

Potentially inappropriate prescribing is a common problem, especially in elderly care. To tackle this problem, Irish medical experts have developed a list of criteria when medication should be added or omitted based upon the patient's physical condition and medication use, known as the STOPP and START criteria. The STOPP and START criteria have been formulated to identify the prescribing of potentially inappropriate medicines (PIMs) and potential prescribing omissions (PPOs). One of the most common problems of inappropriate prescribing is gastro-intestinal track bleedings. For this purpose, nine of the 87 STOPP and START criteria are designed to prevent this. However, the prevalence of gastro-intestinal track bleedings has not been established when these nine STOPP and START criteria are violated. The database contained 182,000 patients belonging to 49 general practitioners in the region of Amsterdam, The Netherlands. We estimated both the incidence of PIMs and PPOs and whether harm, in this case a gastro-intestinal track bleeding, occurred. We found that although violation of the nine STOPP or START criteria were possibly associated with harm (OR = 1.30), this association was not statistically significant (p = 0.323). Searching for evidence for harm informs decision support design aimed at improving quality of medication prescription as it prioritizes the many suggested criteria based on their relevance.

Epidemiological performance and subsequent costs of different surveillance strategies to control bovine herpesvirus type 1 in dairy farms.

This study aimed at comparing the surveillance program of bovine herpesvirus type 1 (BHV1) as laid down by EU Decision 2004/558/EC and 2007/584/EC ('conventional design') with an alternative design. The alternative design was based on monthly bulk-milk testing, clinical surveillance and a risk-based component that involves testing of animals that are purchased from non-free cattle herds. Scenario-tree analyses were carried out to determine sensitivities of the surveillance system (and its components) and the monthly confidence of freedom on herd-level. Also, the expected costs per surveillance design and components thereof were calculated. Results showed that the conventional (EU) and alternative surveillance designs to obtain a BHV1-free status performed equally well in terms of sensitivity. However, total costs per cattle herd to obtain a free status were highest in the conventional design. In an endemic situation and with a within-herd design prevalence of 10%, the conventional design led to a varying probability of freedom ranging from 99.6% to 100% per month. With the alternative design, in this situation, a constant probability of freedom of >99.9% per month was found. In a disease-free situation, both designs performed equally well (probability of freedom >99.9% per month). The yearly costs per farm for monitoring the disease-free status decreased by approximately 25% in the alternative design. The alternative strategy based on monthly bulk-milk monitoring therefore was deemed most cost-effective. This study showed that the surveillance regime to attain and maintain a BHV1-free status as described by EU-legislation can be improved to reduce the monitoring costs without reduction of the system's sensitivity, given a within-herd design prevalence of 10%. The assessment of various surveillance designs could be highly useful to support decision-making towards a more risk-based approach of animal health surveillance.

Application of syndromic surveillance on routinely collected cattle reproduction and milk production data for the early detection of outbreaks of Bluetongue and Schmallenberg viruses.

This study aimed to evaluate the use of routinely collected reproductive and milk production data for the early detection of emerging vector-borne diseases in cattle in the Netherlands and the Flanders region of Belgium (i.e., the northern part of Belgium). Prospective space-time cluster analyses on residuals from a model on milk production were carried out to detect clusters of reduced milk yield. A CUSUM algorithm was used to detect temporal aberrations in model residuals of reproductive performance models on two indicators of gestation length. The Bluetongue serotype-8 (BTV-8) epidemics of 2006 and 2007 and the Schmallenberg virus (SBV) epidemic of 2011 were used as case studies to evaluate the sensitivity and timeliness of these methods. The methods investigated in this study did not result in a more timely detection of BTV-8 and SBV in the Netherlands and BTV-8 in Belgium given the surveillance systems in place when these viruses emerged. This could be due to (i) the large geographical units used in the analyses (country, region and province level), and (ii) the high level of sensitivity of the surveillance systems in place when these viruses emerged. Nevertheless, it might be worthwhile to use a syndromic surveillance system based on non-specific animal health data in real-time alongside regular surveillance, to increase the sense of urgency and to provide valuable quantitative information for decision makers in the initial phase of an emerging disease outbreak.

Evaluation of a hierarchical ascendant clustering process implemented in a veterinary syndromic surveillance system.

Syndromic surveillance is considered as one of the surveillance components for early warning of health-related events, as it allows detection of aberrations in health indicators before laboratory confirmation. "MoSS-Emergences 2" (MoSS-E2), a tool for veterinary syndromic surveillance, aggregates groups of similar clinical observations by hierarchical ascendant classification (HAC). In the present study, this HAC clustering process was evaluated using a reference set of data that, for the purpose of this evaluation, was a priori divided and defined as Bluetongue (BTV) positive cases (PC) on the one hand and BTV negative cases (NC) on the other hand. By comparing the clustering result of MoSS-E2 with the expected outcome, the sensitivity (the ability to cluster PC together) and specificity (the ability to exclude NC from PC) of the clustering process were determined for this set of data. The stability of the classes obtained with the clustering algorithm was evaluated by comparing the MoSS-E2 generated dendrogram (applying complete linkage) with dendrograms of STATA® software applying average and single linkage methods. To assess the systems' robustness, the parameters of the distance measure were adjusted according to different scenarios and obtained outcomes were compared to the expected outcome based on the a priori known labels. Rand indexes were calculated to measure similarity between clustering outcomes. The clustering algorithm in its default settings successfully segregated the reference BTV cases from the non-BTV cases, resulting in a sensitivity of 100.0% (95% CI: 89.0-100.0) and a specificity of 100.0% (95% CI: 80.0-100.0) for this set of data. The different linkage methods showed similar clustering results indicating stability of the classes (Rand indexes of respectively 0.77 for average and 0.75 for single linkage). The system proved to be robust when changing the parameters as the BTV cases remained together in meaningful clusters (Rand indexes between 0.72 and 1). The configurable MoSS-E2 system demonstrated its suitability to identify meaningful clusters of clinical syndromes.

Can routinely recorded reproductive events be used as indicators of disease emergence in dairy cattle? An evaluation of 5 indicators during the emergence of bluetongue virus in France in 2007 and 2008.

In response to increasing risks of emerging infectious diseases, syndromic surveillance can be a suitable approach to detect outbreaks of such diseases across a large territory in an early phase. To implement a syndromic surveillance system, the primary challenge is to find appropriate health-related data. The objective of this study was to evaluate whether routinely collected dates of reproductive events in dairy cattle could be used to build indicators of health anomalies for syndromic surveillance. The evaluation was performed on data collected in France between 2003 and 2009. First, a set of 5 indicators was proposed to assess several types of reproductive disorders. For each indicator, the demographic coverage over the total number of cattle at risk was analyzed in time and space. Second, the ability to detect an emerging disease in an early phase was retrospectively evaluated during epidemics of bluetongue serotypes 1 and 8 (BTV-1, BTV-8) in France in 2007 and 2008. Reproductive indicators were analyzed weekly during these epidemics for each indicator in each infected French district (16 in 2007 and 50 in 2008 out of 94 districts). The indicators were able to detect the BTV epidemics despite their low demographic coverage on a weekly basis relatively to total number of cattle (median=1.21%; range=0-11.7%). Four indicators related to abortions, late embryonic death, and short gestations were abnormally elevated during both BTV epidemics. Median times to abnormal elevations in these indicators were 20 to 71 d after the first notification of clinical signs of BTV by veterinarians. These results demonstrate that reproduction data can be used as indicators of disease emergences, whereas in the specific case of these BTV epidemics, detection via these indicators was later than clinical detection by veterinarians. The emergence of bluetongue in 2007 in France was associated with gestations that were a few days shorter than expected. A short gestation indicator underwent high elevations relative to prior random fluctuations and was the earliest (out of the 4 indicators) to show abnormal elevations, making it possible to detect this emergence.

Risk factors for malformations and impact on reproductive performance and mortality rates of Schmallenberg virus in sheep flocks in the Netherlands.

In Northwestern Europe, an epizootic outbreak of congenital malformations in newborn lambs due to infection with Schmallenberg virus (SBV) started at the end of 2011. The objectives of this study were to describe clinical symptoms of SBV infection, the effect of infection on mortality rates, and reproductive performance in sheep, as well as to identify and quantify flock level risk factors for SBV infections resulting in malformations in newborn lambs. A case-control study design was used, with 93 case flocks that had notified malformed lambs and 84 control flocks with no such lambs. Overall animal seroprevalence in case flocks was estimated at 82.0% (95% CI: 74.3-87.8), and was not significantly different from the prevalence in control flocks being 76.4% (95% CI: 67.2-83.6). The percentages of stillborn lambs or lambs that died before weaning, repeat breeders, and lambs with abnormal suckling behaviour were significantly higher in case flocks compared to control flocks. However, effect of SBV infection on mortality rates and reproductive performance seemed to be limited. Multivariable analysis showed that sheep flocks with an early start of the mating season, i.e. before August 2011 (OR = 33.1; 95% CI: 10.0-109.8) and in August 2011 (OR = 8.2; 95% CI: 2.7-24.6) had increased odds of malformations in newborn lambs caused by SBV compared to sheep flocks with a start of the mating season in October 2011. Other flock-level risk factors for malformations in newborn lambs were purchase of silage (OR 5.0; 95% CI: 1.7-15.0) and flocks with one or more dogs (OR = 3.3; 95% CI: 1.3-8.3). Delaying mating until October could be a potential preventive measure for naïve animals to reduce SBV induced losses. As duration of immunity after infection with SBV is expected to last for several years, future SBV induced congenital malformations are mainly expected in offspring of early mated seronegative animals.

The Schmallenberg virus epidemic in Europe-2011-2013.

During the Schmallenberg virus (SBV) epidemic, the European Food Safety Authority (EFSA) collected data on SBV occurrence across Europe in order to provide an assessment of spread and impact. By May 2013, twenty-nine countries were reporting to EFSA and twenty-two countries had reported cases of SBV. The total number of SBV herds reported was 13,846 and the number of SBV laboratory confirmed herds was 8730. The surveillance activities were based on the detection of SBV clinical cases (either adults or newborns). Malformation in newborns was the most commonly reported clinical sign of SBV-infection. All countries were able to provide the date when the first suspicion of SBV in the herd was reported and nineteen could report the location of the herd at a regional level. This allowed the spread of SBV in Europe to be measured both temporally and spatially. The number of SBV confirmed herds started to increase in December 2011 and two peaks were observed in 2012 (February and May). Confirmed herds continued to be reported in 2012 and into 2013. An increase during winter 2012 and spring 2013 was again observed, but the number of confirmed herds was lower than in the previous year. SBV spread rapidly throughout Europe from the initial area of detection. SBV was detected above the latitude of 60° North, which exceeds the northern expansion observed during the bluetongue virus serotype 8 epidemic in 2006-2009. The impact of SBV was calculated as ratio of the number of herds with at least one malformed SBV positive foetus and the total number of herds in this region. The 75th percentile of the malformations ratio in the various affected countries for the whole reporting period was below 1% and 3% for cattle and sheep herds, respectively. International data collection on emerging diseases represents a challenge as the nature of available data, data quality and the proportion of reported cases may vary widely between affected countries. Surveillance activities on emerging animal diseases are often structured only for case detection making the estimation of infection/diseases prevalence and the investigation of risk factors difficult. The impact of the disease must be determined to allow risk managers to take appropriate decisions. Simple within-herd impact indicators suitable for emerging disease outbreaks should be defined that could be measured as part of routine animal health surveillance programmes and allow for rapid and reliable impact assessment of emerging animal health diseases.

Serological and virological BVDV prevalence and risk factor analysis for herds to be BVDV seropositive in Belgian cattle herds.

Bovine viral diarrhoea virus (BVDV) is a worldwide spread virus that most commonly infects cattle and can cause considerable economic losses. To determine the prevalence of BVDV in Belgium, a cross-sectional study was performed between November 2009 and March 2010. Young stock aged between 6 and 12 months from 773 randomly selected Belgian cattle herds were tested for BVDV-specific antibodies and antigen. With a target and maximum of 10 animals per sampled herd, a total of 5246 animals were selected. Additionally a questionnaire including different herd management topics and questions about participation in animal health programmes, including BVDV, was sent to 1100 Belgian cattle herds, including the 773 herds for BVDV testing. This paper focuses on results regarding these 773 herds. The true prevalence of BVDV-specific antibodies and antigen at herd level was respectively 47.4% and 4.4%, while at animal level this was respectively 32.9% and 0.3%. In 44.4% of the herds where BVDV-specific antibodies were detected at least 60% of the sampled young stock was BVDV seropositive. Interestingly, 83.4% of these farmers stated not to have suffered from problems related to BVDV. Moreover, only 8.4% of all farmers who completed the questionnaire (n=895) reported problems possibly related to BVDV the past 3 years. This demonstrates that farmers are often unaware of the presence of BVDV in their herd. Risk factors for a herd to be BVDV seropositive were identified by means of a multivariable logistic regression model. Large herds were significantly more likely to be BVDV seropositive (OR=1.004, p<0.01). The interaction between "Antigen positive animal detected in this study" and "BVDV vaccination in 2009" was significant (p<0.01). In non-vaccinating herds, the detection of antigen positive animals was significantly associated with BVDV seropositive herds (OR=13.8, p<0.01). In herds with no antigen positive animals detected, vaccination resulted in a significant risk factor to be BVDV seropositive compared to non-vaccinating herds (OR=3.4, p<0.01). Herds reporting BVDV-related problems the past 3 years were more likely to be BVDV seropositive (OR=1.9, p<0.05). This relation became non-significant (OR=1.8, p=0.08) when only a subset of herds with no vaccination of animals <12 months was taken into account. The results of the current study suggest an active circulation of BVDV in a considerable number of Belgian cattle herds.