Haemorrhagic bowel syndrome (HBS) is one of the most common causes of death in fattening pigs worldwide. The objective of this descriptive study was to systematically assess predictors or causal components for the appearance of HBS using case farms (mortality rate caused by HBS ≥ 1.5%) in comparison with control farms (mortality rate caused by HBS ≤ 0.25%), focusing on feed ingredients, feed quality and size, and gastrointestinal pathogens. The inclusion of sugar beet as a feed component in liquid feeding systems was found to be associated (p = 0.03) with farms identified as HBS cases. Another predictive or causal factor found for liquid feeding systems, but only for those using meal, was particle size. A higher percentage of small particles (< 2 mm) in the meal was associated with a higher risk of being an HBS case farm (p = 0.02), while no relevant association was detected for the use of pellets. Sugar beet in the diet was also associated with the incidence of HBS. The microbial quality of the feed in dry feeding systems, specifically the number of total aerobes at the first and last outlet tubes, was associated with a higher incidence of HBS (p = 0.03). Faecal sample analysis showed a difference (p < 0.05) in the prevalence of B. pilosicoli in the herd category (case vs. control herd). In this descriptive study, five predictive or causal factors were identified for an HBS farm with a mortality rate due to HBS ≥ 1.5%. These included the number of aerobes in dry matter samples from the first and last feeders, the particle diameter of the meal used in liquid feeding systems and sugar beet as a component of liquid feeding rations, and the presence of B. pilosicoli as an infectious agent at animal level. Relevant associations reinforce the findings of the previously published Swiss study that HBS is a multifactorial syndrome involving different aspects of pig production and cannot be attributed to a single cause. Further studies are needed to develop evidenced based causal models for HBS in swine.

Blood tests are widely used in modern medicine to diagnose certain illnesses and evaluate the overall health of a patient. To enable testing in resource-limited areas, there has been increasing interest in point-of-care (PoC) testing devices. To process blood samples, liquid mixing with active pumps is usually required, making PoC blood testing expensive and bulky. We explored the possibility of processing approximately 2 μL of whole blood for image flow cytometry using capillary structures that allowed test times of a few minutes without active pumps. Capillary pump structures with five different pillar shapes were simulated using Ansys Fluent to determine which resulted in the fastest whole blood uptake. The simulation results showed a strong influence of the capillary pump pillar shape on the chip filling time. Long and thin structures with a high aspect ratio exhibited faster filling times. Microfluidic chips using the simulated pump design with the most efficient blood uptake were fabricated with polydimethylsiloxane (PDMS) and polyethylene oxide (PEO). The chip filling times were tested with 2 μL of both water and whole blood, resulting in uptake times of 24 s for water and 111 s for blood. The simulated blood plasma results deviated from the experimental filling times by about 35% without accounting for any cell-induced effects. By comparing the flow speed induced by different pump pillar geometries, this study offers insights for the design and optimization of passive microfluidic devices for inhomogenous liquids such as whole blood in sensing applications.