Bias--is it a problem, and what should we do?

An Evaluation of Temporal Distributions of High, Low, and Zero Cohort Morbidity of Cumulative First Treatment Bovine Respiratory Disease and Their Associations with Demographic, Health, and Performance Outcomes in US Feedlot Cattle

Timing and magnitude of bovine respiratory disease (BRD) can impact intervention and overall economics of cattle on feed. Furthermore, there is a need to better describe when cattle are being treated for BRD. The first objective was to perform a cluster analysis on the temporal distributions of cumulative first treatment BRD from HIGH (≥15% of cattle received treated for BRD) and LOW cohorts (>0 and <15% of cattle received treated for BRD) to assess cohort-level timing (days on feed) of BRD first treatments. The second objective was to determine associations among cluster groups (temporal patterns) and demographic risk factors, health outcomes, and performance. Cluster analysis determined that optimal number of clustering groups for the HIGH morbidity cohort was six clusters and LOW morbidity cohort was seven clusters. Cohorts with zero BRD treatment records were added for statistical comparisons. Total death loss, BRD morbidity, average daily gain (ADG), railing rate, days to 50% BRD, cattle received, shrink, arrival weight, and sex were associated with temporal groups (p < 0.05). These data could be used as a tool for earlier identification and potential interventions for cohorts based on the BRD temporal pattern.

Malignant testicular tumors in children: A single institution's 12-year experience

Testicular neoplasms are not commonly found in children and are a formidable threat if treated inappropriately. However, there is no consensus regarding its management. This study aimed to create a holistic picture of the interprofessional team in the management of malignant testicular tumors. Seventeen patients had mixed germ cell tumors, 15 had pure yolk sac tumors, 2 had immature teratomas, 2 had teratocarcinomas, and 1 had a sex cord stromal tumor. Five lesions were diagnosed as nongerm cell tumors: 2 embryonal rhabdomyosarcomas, 2 lymphomas, and 1 acute myeloid leukemia. At initial presentation, retroperitoneal (n = 2), bone marrow (n =1), and mediastinal (n = 1) metastases were identified in 4 (10%) patients. The operative interventions performed included radical inguinal orchiectomy (n = 5), scrotal orchiectomy (n = 31), and testicular biopsy or testis-sparing enucleation of the tumor (n = 6). Postoperatively, 18 patients received either adjuvant chemotherapy (n = 14) or chemoradiation (n = 5). Five patients with mixed germ cell tumors (n = 2), group IV paratesticular rhabdomyosarcoma (n = 2), and acute myeloid leukemia with myeloid sarcoma (n =1) died of disease progression. Thirty-six patients remained alive and disease-free at the last visit. Malignant testicular tumors in children deserve proper diagnostic support from a therapeutic perspective. Any concern or suspicion of a testicular tumor warrants an inguinal approach to avoid scrotal violation.

Veterinarian barriers to knowledge translation (KT) within the context of swine infectious disease research: an international survey of swine veterinarians

Background

Food animal veterinarians face commodity specific and urgent global challenges yet conditions preventing use of best available knowledge have been sparsely studied. The American Association of Swine Veterinarians (AASV) membership (N = 1289) was surveyed online to benchmark their information priorities and their motivations and sources for keeping current with infectious disease research, and to describe their reported time, skill, access, and process as barriers to knowledge translation (KT).

Results

Respondents (n = 80) were mostly from Canada (n = 40) and the U.S.A (n = 31) and demographics approximated the AASV’s. Colleagues are the first choice for information on difficult cases (49%, 95%CI: 38–61). Half of respondents (53%, 95%CI: 41–64) spend an hour or less per week keeping up with infectious disease research. The majority reported moderate or less than moderate efficiency (62%, 95%CI: 51–72), and moderate or greater stress (59%, 95%CI: 48–70) with their process for keeping up. Journal article methods sections are commonly not read, almost a third (32%, 95% CI: 22–43) reported either they do not evaluate statistical methods or that they had poor confidence to do so, and half (52, 95%CI: 41–63) could not explain ‘confounding bias’. Approximately half (55%, 95%CI: 41-69) with direct oversight of swine herds had full access to 2 or fewer academic journals. Approximately a third of respondents (34%, 95%CI: 24–46) selected only formats involving single research studies (either full text or summaries) as preferred reading materials for keeping current over expert summaries of the body of evidence.

Conclusion

KT barriers are considerable and a source of stress for many swine veterinarians. Sub-optimal efficiency with keeping up and low confidence to appraise aspects of research are concerns. Results are consistent with previous literature and illustrate need for improved KT infrastructure and for additional training in statistical methods and interpretation of primary research. Further evaluation is warranted of why approximately a third of veterinarians in this study, for the purpose of keeping up, preferentially choose to review individual research studies over choices that would include an expert summary of the body of evidence. Consideration of reasons for this preference will be important in the planning of KT infrastructure improvements.

Supplementary information

Supplementary information accompanies this paper at 10.1186/s12917-020-02617-8.

Reporting of Safety Events during Anti-VEGF Treatment: Pharmacovigilance in a Noninterventional Trial

Aim

The prospective, noninterventional OCEAN study assessed the safety of intravitreal ranibizumab injections for treatment of neovascular age-related macular degeneration, diabetic macular edema, and retinal vein occlusion under real-world conditions in Germany.

Methods

Adults receiving ≥1 ranibizumab (0.5 mg) injections were recruited by 369 ophthalmologists and followed for 24 months. Information on adverse events (AEs) was reported by the treating physician or detected by the data management team. Collected information included observed AE, AE start and end date, intensity, causal relationship, outcome, severity, suspected drug, and actions taken.

Results

2,687 AEs were reported for 1,176 of the 5,781 patients who had received a total of 32,621 injections: 27.4% nonserious AEs, 30.3% serious AEs, 27.3% nonserious adverse drug reactions (ADRs), and 15.0% serious ADRs. Most patients reported no AEs (79.7%) or only 1 AE (10.3%). AEs were most commonly reported in the Medical Dictionary for Regulatory Activities (MedDRA) System Organ Class (SOC) Eye disorders (9.4% of patients) and General disorders and administration site conditions (5.8%). The most frequent AEs by MedDRA preferred term (PT) were visual acuity reduced (3.5% of patients), intraocular pressure increased (2.5%), and drug ineffective (2.1%). AEs occurred most frequently after 3 or 4 injections (1,129 of 2,687 AEs). The proportion of AEs in the SOC Eye disorders decreased slightly with increasing number of injections, from 39.8% of events after 1 or 2 injections to 29.1% after 5 or more injections. Rates of the most frequently reported PTs did not show any consistent increase with increasing number of injections. A decrease in overall AE rates was observed over the study course.

Conclusions

The results did not raise any new safety concerns for ranibizumab. The findings allow conclusions to be drawn on how pharmacovigilance data can be collected even more effectively in real-world studies to facilitate discussion on benefit-risk ratio.

Variations in association of nasal microbiota with virulent and non-virulent strains of Glaesserella (Haemophilus) parasuis in weaning piglets

Glaesserella (formerly Haemophilus) parasuis causes Glässer’s disease, which results in high economic loss in the swine industry. To understand the polymicrobial interactions of G. parasuis and the nasal microbiota, the statistical association patterns of nasal colonizing bacteria with virulent and non-virulent strains of G. parasuis were studied accounting for the farm management practices as potential risk factors for the occurrence of Glässer’s disease. The nasal microbiota from 51 weaned-piglets from four farms with Glässer’s disease and three farms with no respiratory diseases was previously characterized and included in this study. The presence of virulent and/or non-virulent G. parasuis strains in the nasal cavities was determined in order to establish the potential association with other members of the nasal microbiota. Multivariate logistic and linear regression models were performed among the various members of nasal microbiota and G. parasuis. The multi-site production system and disease presence in the farm were both significantly associated with the presence of G. parasuis virulent strains in the nose of the piglets. Differential bacterial associations were observed with virulent or non-virulent G. parasuis. Chitinophagaceae, Corynebacteriaceae and Corynebacterium were positively associated with the virulent G. parasuis strains, while Enterobacteriaceae, Peptostreptococcaceae, Clostridium XI, and Escherichia/Shigella were negatively associated with virulent G. parasuis. On the other hand, Flavobacteriaceae, Planobacterium, and Phascolarctobacterium were positively associated with the non-virulent G. parasuis strains, while Rikenellaceae, Enterococcaceae, Odoribacter, and Corynebacterium were negatively associated with non-virulent G. parasuis. In conclusion, the nasal microbiota communities showed variations in the association with the G. parasuis strains type.

Bayesian estimation of qPCR and bacterial culture accuracy for detection of bovine coagulase-negative staphylococci from milk and teat apex at different test cut-off points

AIM

To primarily estimate the sensitivity (Se) and specificity (Sp) of the commercially available Mastit4 quantitative PCR (qPCR) assay and bacterial culture (BC) for diagnosis of intramammary infections (IMI) and teat apex colonization (TAC) with coagulase-negative staphylococci (CNS) at different cut-offs for qPCR cycle threshold values using Bayesian latent class analysis. A secondary objective was to evaluate two cut-offs of BC for diagnosis of IMI and TAC with CNS.

METHODS AND RESULTS

We randomly selected 13-20 cows with subclinical mastitis from eight dairy herds. Teat skin samples and aseptically collected foremilk samples were collected from the right hindquarters (n = 149) for BC and qPCR analysis. The Se of qPCR was always higher than BC_Se in diagnosis of IMI, however; the Sp of BC was higher than qPCR_Sp . BC_Se and BC_Sp showed no substantial difference between the tested BC cut-offs. In contrast to IMI, estimates of BC and qPCR in diagnosing TAC were different. BC_Se was higher than qPCR_Se at all tested cut-offs, however; qPCR_Sp was higher than BC_Sp .

CONCLUSION

The overall performance of qPCR is higher than BC in the diagnosis of IMI; however, the performance of BC is better than qPCR in diagnosis of TAC. The qPCR and BC are valid diagnostics for bovine IMI with CNS. However, for TAC, both techniques require further investigation to reduce the uncertainty of the true status of the quarter and teat skin.

SIGNIFICANCE AND IMPACT OF THE STUDY

We reported, for the first time, the diagnostic performance of new mastitis technology (Mastit4 PCR) and culture for detection of CNS in milk and nonmilk samples in dairy herds with automatic milking systems. Our findings will improve the interpretation of the test results of culture and qPCR assay and subsequently, will strengthen the control of IMI with CNS in dairy cows.

Complex System Approaches for Animal Health Surveillance

Many new and highly variable data are currently being produced by the many participants in farmed animal productions systems. These data hold the promise of new information with potential value for animal health surveillance. The current analytical paradigm for dealing with these new data is to implement syndromic surveillance systems, which focus mainly on univariate event detection methods applied to individual time series, with the goal of identifying epidemics in the population. This approach is relatively limited in the scope and not well-suited for extracting much of the additional information that is contained within these data. These approaches have value and should not be abandoned. However, an additional, new analytical paradigm will be needed if surveillance and disease control agencies wish to extract additional information from these data. We propose a more holistic analytical approach borrowed from complex system science that considers animal disease to be a product of the complex interactions between the many individuals, organizations and other factors that are involved in, or influence food production systems. We will discuss the characteristics of farmed animal food production systems that make them complex adaptive systems and propose practical applications of methods borrowed from complex system science to help animal health surveillance practitioners extract additional information from these new data.

Seroprevalence and participatory epidemiology of camelpox in Afar region of Ethiopia

Camelpox is endemic in most camel rearing regions of the world, causing significant economic losses. However, its epidemiology is not extensively investigated. We conducted a cross sectional seroprevalence study of camelpox in Amibara and Awash Fentale districts in Afar region of Ethiopia from November 2014 to May 2015. In addition, participatory epidemiology (PE) was conducted to identify seasonal occurrence of the disease in the study districts. Blood samples were collected from 384 dromedary camels from 31 herds distributed in five pastoral associations (PAs) in the two districts. Serum samples were separated from the blood samples and tested for the presence of viral antibodies using virus neutralization test. Seroprevalence data were analyzed using multilevel mixed effects logistic regression models accounting for the 4-level hierarchical data structure (camels nested in herds-herds in PA, and PA in district). For the participatory data, Kendall's coefficient of concordance was used to assess agreements between the informants in identifying seasonal occurrences of the top five camel diseases. Camelpox antibodies were detected in 19.3% of camels (n = 384), 81% of herds (n = 31), and in all five PAs from the two districts in the Gabi Rasu zone of Afar region, Ethiopia. The seroprevalence did not significantly vary between herds, PAs or districts suggesting the widespread occurrence of the disease. Estimated age stratified basic reproduction number (R₀) was 1.25 (95% CI: 0.62-2.19). Camelpox was identified as one of the top five common camel diseases in the area. The widespread occurrence of the disease can be attributed mainly to the commingling of camels from many herds during seasonal migration in search of feed and water, a practice very common under pastoral production systems. Although the PE informants indicated the clinical disease to be more common in young animals, seropositivity was higher in older animals. Camelpox commonly occurs during the minor and major rainy seasons. In conclusion, camelpox is found to be endemic in Afar pastoral region with sporadic outbreaks occurring during rainy seasons. Vaccination and improved camel management practices particularly during the high-risk period can be viable strategies to reduce the burden of the disease.

Rethinking bias and truth in evidence-based health care

In modern philosophy, the concept of truth has been problematized from different angles, yet in evidence-based health care (EBHC), it continues to operate hidden and almost undisputed through the linked concept of "bias." To prevent unwarranted relativism and make better inferences in clinical practice, clinicians may benefit from a closer analysis of existing assumptions about truth, validity, and reality. In this paper, we give a brief overview of several important theories of truth, notably the ideal limit theorem (which assumes an ultimate and absolute truth towards which scientific inquiry progresses), the dominant way truth is conceptualized in the discourse and practice of EBHC. We draw on Belgian philosopher Isabelle Stengers' work to demonstrate that bias means one thing if one assumes a world of hard facts "out there," waiting to be collected. It means something different if one takes a critical view of the knowledge-power complex in research trials. Bias appears to have both an unproductive aspect and a productive aspect as argued by Stengers and others: Facts are not absolute but result from an interest, or interesse: a bias towards a certain line of questioning that cannot be eliminated. The duality that Stengers' view invokes draws attention to and challenges the assumptions underlying the ideal limit theory of truth in several ways. Most importantly, it casts doubt on the ideal limit theory as it applies to the single case scenario of the clinical encounter, the cornerstone of EBHC. To the extent that the goal of EBHC is to support inferencing in the clinical encounter, then the ideal limit as the sole concept of truth appears to be conceptually insufficient. We contend that EBHC could usefully incorporate a more pluralist understanding of truth and bias and provide an example how this would work out in a clinical scenario.

Effect of incomplete milking during the first 5 days in milk on udder and reproductive tract health: Results from a randomized controlled trial

The aim of this study was to investigate the effect of an incomplete milking on risk of mastitis and reproductive tract disease. Multiparous dairy cows (n = 878) from 13 commercial herds were enrolled in a randomized controlled trial. Cows were randomly assigned to either a control (milked conventionally) or a treatment group, which consisted of an incomplete milking (10-14 L of milk collected/d) from 1 to 5 d in milk (DIM). Quarter milk samples were collected at approximately 11 and 18 DIM to measure somatic cell count (SCC). Quarters were considered negative for intramammary infection if SCC was <100,000 cells/mL and positive if SCC was ≥200,000 cells/mL. To calculate intramammary infection incidence, negative quarters of the initial samples collected were tested again 1 wk later. This was done to deter incidence of positive quarters. To calculate elimination rate, positive quarters were tested again 1 wk later to detect mastitis elimination. Farmers recorded clinical mastitis events. Cows were also examined at approximately 35 DIM with a Metricheck device (Simcro, Hamilton, New Zealand) for detection of purulent vaginal discharge (PVD) and with an endometrial cytobrush for presence of leukocytes [endometrial cytology for smear (ENDO) and for leukocyte esterase test (LE)]. A threshold ≥3 was used to define a positive PVD or LE test, whereas a polymorphonuclear cell count ≥6% was used to define a positive ENDO. Five generalized mixed models with cow or herd as random intercepts were used to determine the effects of incomplete milking on odds of new intramammary infection, odds of intramammary infection elimination, and odds of a positive PVD, LE, or ENDO status. To investigate time until first clinical mastitis event, a Cox model with a herd frailty term was used. The odds of new intramammary infection and intramammary infection elimination for incompletely milked cows were 0.90 [95% confidence interval (CI): 0.49, 1.7] and 2.9 (95% CI: 1.4, 6.0) times those of conventionally milked cows, respectively. The hazard of clinical mastitis in incompletely milked cows was 0.96 (95% CI: 0.59, 1.6) times that of conventionally milked cows. The odds of PVD, LE, and ENDO for incompletely milked cows were 1.4 (95% CI: 0.89, 2.1), 1.3 (95% CI: 0.88, 1.8), and 1.2 (95% CI: 0.81, 1.7) times those of conventionally milked cows. These results suggest that incomplete milking during the first 5 DIM increases the odds of a decrease in SCC from 11 to 18 DIM but does not affect odds of increase in SCC in the same period. The incomplete milking had no effect on clinical mastitis incidence in the first 90 DIM or on reproductive tract health at 35 DIM.

Mastitis on Rabbit Farms: Prevalence and Risk Factors

In this cross-sectional study, prevalence of clinical mastitis (PCM) and farm-specific risk factors were determined on 531 doe rabbit farms in Spain and Portugal, from January 2001 through March 2017. The information was obtained by carrying out 2367 visits and doing physical examinations of 144,455 lactating does, sorted in 2635 cohorts. Overall mean PCM was 4.05% (CI95% [3.87⁻4.22]), (minimum to maximum: 0⁻36.00% PCM). This study suggests that PCM was influenced by the variable number of batches (a batch was a group of does served the same day), per maternity barn (p < 0.0001). The duo system (does being moved to clean disinfected barns for parturition), was also an enabling risk factor for CM. The day of service or lactation stage also affected PCM. Lastly, there was an effect of the breeds or lines (p < 0.0001); mean PCM ranged between 1.29% and 7.09%. A subset of data obtained from 200 farms visited during January 2012⁻March 2017, was recorded to describe the use of antimicrobials against mastitis. Changes in host, husbandry, environment, and biosecurity practices, are highlighted to provide health and welfare benefits for breeding rabbits.

Validation of the register-based lifetime antimicrobial usage measurement for finisher batches based on comparison with recorded antimicrobial usage at farm level

Assessing the relationship between antimicrobial usage (AMU) and antimicrobial resistance (AMR) requires the accurate and precise utilisation of register data. Therefore, validation of register-based data is essential for evaluating the quality and, subsequently, the internal validity of studies based on the data. In this study, different smoothing methods for Veterinary Medicine Statistic Program database (VetStat)-records were validated by comparing these with farm-records. Comparison between measurements included accuracy as; completeness and correctness, and precision as; a relative difference of the error, correlation with Fisher's z transformation and reliability coefficient. The most valid methods of those examined were then used in re-analyses of the abundance of AMR genes in 10 finisher batches from a previous study. Improved accuracy was found when detailed smoothing methods were applied. Although the precision also increased, the effect was not as pronounced, as the usage estimate of all smoothing methods deviated moderately compared with the farm-registrations. Applying the most valid methods to the 10 finisher batches increased estimates of statistical model fit for aminoglycosides, lincosamides, tetracyclines and decreased estimates of statistical model fit for macrolides. The estimates of statistical model fit for sulfonamides and broad-spectrum penicillins remained the same. Through refined data transformation, VetStat-records can be used to calculate a daily amount of AMU per pig reflecting the true usage accurately and moderately precisely, which is the foundation for calculating lifetime AMU.

Palivizumab Exposure and the Risk of Autoimmune Disease: A Cross-National Cohort Study

BACKGROUND

Treatment with biologic pharmaceuticals may be associated with an increased risk of immune-mediated disease. Palivizumab is a humanized monoclonal antibody designed to provide passive immunity against respiratory syncytial virus infection. Palivizumab is primarily used in preterm children known to be immunologically immature. The long-term effect of palivizumab in terms of autoimmune diseases has not yet been investigated.

AIM

Our objective was to investigate whether exposure to palivizumab was associated with the development of autoimmune diseases in children.

METHODS

This was a population-based cohort study including data from 769,523 Danish children born between 1 January 1999 and 31 December 2010 and data from 581,742 Swedish children born between 1 July 2005 and 31 December 2010.

RESULTS

Of the 1,351,265 children included, 1192 (0.09 %) were exposed to palivizumab. Nine cases of autoimmune disease were diagnosed among palivizumab-exposed children during the period of observation. Among the children exposed to palivizumab, one child in Denmark developed inflammatory bowel disease; in Sweden, children developed juvenile arthritis (one child), diabetes mellitus (two children), celiac disease (four children), and inflammatory bowel disease (one child). The risk of autoimmune disease was not significantly increased after palivizumab exposure (hazard ratio adjusted for age and country: 1.54; 95 % confidence interval 0.80-2.95).

CONCLUSION

The risk of autoimmune disease was not increased after palivizumab exposure. Given the small number of incident cases of autoimmune disease observed, this finding should be interpreted with caution.

Explanation and Elaboration Document for the STROBE-Vet Statement: Strengthening the Reporting of Observational Studies in Epidemiology-Veterinary Extension

The STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement was first published in 2007 and again in 2014. The purpose of the original STROBE was to provide guidance for authors, reviewers, and editors to improve the comprehensiveness of reporting; however, STROBE has a unique focus on observational studies. Although much of the guidance provided by the original STROBE document is directly applicable, it was deemed useful to map those statements to veterinary concepts, provide veterinary examples, and highlight unique aspects of reporting in veterinary observational studies. Here, we present the examples and explanations for the checklist items included in the STROBE-Vet statement. Thus, this is a companion document to the STROBE-Vet statement methods and process document (JVIM_14575 "Methods and Processes of Developing the Strengthening the Reporting of Observational Studies in Epidemiology-Veterinary (STROBE-Vet) Statement" undergoing proofing), which describes the checklist and how it was developed.

Diagnostic performance of molecular and conventional methods for identification of dermatophyte species from clinically infected Arabian horses in Egypt

BACKGROUND

Rapid and accurate identification of dermatophytes is crucial for the effective control of disease outbreaks. Current methods based on culture and microscopic characteristics may require weeks before positive identification is made.

OBJECTIVES

To (i) identify the most common pathogenic dermatophytes affecting Arabian horses; (ii) compare the performance of direct microscopy (DM), culture, PCR using hair samples (PCRhair ) and PCR based on culture isolates (PCRculture ) for the diagnosis of dermatophytosis.

METHODS

Samples of hair and crusts of skin lesions from Arabian horses were collected on a monthly basis by scraping skin of affected horses. Samples were divided into three portions: the first portion was used for microscopic examination, the second for culture and the third portion for PCR amplification of intergenic spacer (ITS) regions.

RESULTS

Out of 200 horses examined, 70 (35%) showed cutaneous lesions characteristic of dermatophytosis. DM revealed that 70.4% were positive for fungal elements and 85.7% were culture positive. The identified species were Microsporum canis, Trichophyton verrucosum, T. mentagrophytes var. mentagrophytes and M. equinum. Among 25 selected samples, 64, 92, 91.3 and 52% were positive for dermatophytes, as determined by DM, culture, PCRculture and PCRhair , respectively.

CONCLUSIONS

The dermatophytes M. canis, T. verrucosum, T. mentagrophytes var. mentagrophytes and M. equinum were the most common cause of dermatophytosis in Arabian horses. Although the number of samples was small, the ITS-based PCR may be a useful diagnostic tool when combined with culture.

Associations between prior management of cattle and risk of bovine respiratory disease in feedlot cattle

Bovine respiratory disease (BRD) is the major cause of clinical disease and death in feedlot populations worldwide. A longitudinal study was conducted to assess associations between risk factors related to on-farm management prior to transport to the feedlot and risk of BRD in a population of feedlot beef cattle sourced from throughout the cattle producing regions of Australia. Exposure variables were derived from questionnaire data provided by farmers supplying cattle (N=10,721) that were a subset of the population included in a nationwide prospective study investigating numerous putative risk factors for BRD. Causal diagrams were used to inform model building to allow estimation of effects of interest. Multilevel mixed effects logistic regression models were fitted within the Bayesian framework. Animals that were yard weaned were at reduced risk (OR: 0.7, 95% credible interval: 0.5-1.0) of BRD at the feedlot compared to animals immediately returned to pasture after weaning. Animals that had previously been fed grain (OR: 0.6, 95% credible interval: 0.3-1.1) were probably at reduced risk of BRD at the feedlot compared to animals not previously fed grain. Animals that received prior vaccinations against Bovine viral diarrhoea virus 1 (OR: 0.8, 95% credible interval: 0.5-1.1) or Mannheimia haemolytica (OR: 0.8, 95% credible interval: 0.6-1.0) were also probably at reduced risk compared to non-vaccinated animals. The results of this study confirm that on-farm management before feedlot entry can alter risk of BRD after beef cattle enter feedlots.

Investigation of the association between the test day milk fat-protein ratio and clinical mastitis using a Poisson regression approach for analysis of time-to-event data

The objective of the present study was to investigate the association between the milk fat-protein ratio and the incidence rate of clinical mastitis including repeated cases of clinical mastitis to determine the usefulness of this association to monitor metabolic disorders as risk factors for udder health. Herd records from 10 dairy herds of Holstein cows in Saxony, Germany, from September 2005-2011 (36,827 lactations of 17,657 cows) were used for statistical analysis. A mixed Poisson regression model with the weekly incidence rate of clinical mastitis as outcome variable was fitted. The model included repeated events of the outcome, time-varying covariates and multilevel clustering. Because the recording of clinical mastitis might have been imperfect, a probabilistic bias analysis was conducted to assess the impact of the misclassification of clinical mastitis on the conventional results. The lactational incidence of clinical mastitis was 38.2%. In 36.2% and 34.9% of the lactations, there was at least one dairy herd test day with a fat-protein ratio of <1.0 or >1.5, respectively. Misclassification of clinical mastitis was assumed to have resulted in bias towards the null. A clinical mastitis case increased the incidence rate of following cases of the same cow. Fat-protein ratios of <1.0 and >1.5 were associated with higher incidence rates of clinical mastitis depending on week in milk. The effect of a fat-protein ratio >1.5 on the incidence rate of clinical mastitis increased considerably over the course of lactation, whereas the effect of a fat-protein ratio <1.0 decreased. Fat-protein ratios <1.0 or >1.5 on the precedent test days of all cows irrespective of their time in milk seemed to be better predictors for clinical mastitis than the first test day results per lactation.

Farm characteristics related to on-farm cow mortality in dairy herds: a questionnaire study

Dairy cow mortality is an important animal welfare issue that also causes financial losses. The objective of this study was to identify farm characteristics and herd management practices associated with high on-farm cow mortality in Swedish dairy herds. A postal questionnaire was sent to farmers that had either high or low mortality rates for 3 consecutive years. The questionnaire consisted of five sections: 'About the farm', 'Milking and housing', 'Feeding', 'Routines' and 'Lame and sick cows'. A total of 145 questionnaires were returned (response rate=33%). Ten of the 77 characteristics investigated met the inclusion criteria for multivariable analysis. The final logistic regression model included: herd size, breed, use of natural service bull, bedding improvement frequency and pasture system. Herds with Swedish Holstein as the predominant breed (odds ratio (OR) 22.9, 95% confidence interval (CI) 5.2 to 101.8) or with mixed breeds (OR 5.5, 95% CI 1.7 to 17.5) had a greater risk of being high mortality herds than herds that were predominantly Swedish Red (OR 1). Herds larger than 100 cows (OR 19.6, 95% CI 3.5 to 110.4) and herds with 50 to 99 cows (OR 13.8, 95% CI 3.2 to 60.6) had greater risk of mortality than herds numbering 35 to 50 cows (OR 1). Being a high mortality herd was also associated with having cows on exercise lots during the summer season (OR 3.6, 95% CI 1.3 to 9.9) compared with on pasture. A missing answer on the question of bedding improvement frequency was associated with high mortality herds. Overall, this study suggests that characteristics that are related to intensification of the dairy industry are also associated with high on-farm mortality of dairy cows.

Making valid causal inferences from observational data

The ability to make strong causal inferences, based on data derived from outside of the laboratory, is largely restricted to data arising from well-designed randomized control trials. Nonetheless, a number of methods have been developed to improve our ability to make valid causal inferences from data arising from observational studies. In this paper, I review concepts of causation as a background to counterfactual causal ideas; the latter ideas are central to much of current causal theory. Confounding greatly constrains causal inferences in all observational studies. Confounding is a biased measure of effect that results when one or more variables, that are both antecedent to the exposure and associated with the outcome, are differentially distributed between the exposed and non-exposed groups. Historically, the most common approach to control confounding has been multivariable modeling; however, the limitations of this approach are discussed. My suggestions for improving causal inferences include asking better questions (relates to counterfactual ideas and "thought" trials); improving study design through the use of forward projection; and using propensity scores to identify potential confounders and enhance exchangeability, prior to seeing the outcome data. If time-dependent confounders are present (as they are in many longitudinal studies), more-advanced methods such as marginal structural models need to be implemented. Tutorials and examples are cited where possible.