Prevention of perioperative thromboembolism in patients with atrial fibrillation.

BACKGROUND: The aim of this retrospective study was to review the safety and efficacy of two regimens for the prophylaxis of perioperative thromboembolism in patients with atrial fibrillation. METHODS: From a database of 14 801 procedures, atrial fibrillation occurred in 1.9 per cent of patients. Those not on oral anticoagulation (n = 146) received low molecular weight heparin (LMWH) before and after surgery (nadroparine 40 units per kg). Patients on oral anticoagulation before surgery (n = 136) received intravenous unfractionated heparin (UFH) after surgery at a dose adequate to maintain the thrombin time at a therapeutic level. RESULTS: The incidence of perioperative arterial or venous thromboembolism was independent of pre-existing risk factors and occurred in 4.6 per cent of patients, without significant difference between the two regimens (P = 0.780). Logistic regression revealed that thromboembolism was significantly associated with increased perioperative mortality (odds ratio 9.5, (95 per cent confidence interval 2.5 to 35.8); P = 0.001). The rate of postoperative bleeding was 4.8 per cent in patients who had LMWH and 17.6 per cent in those who had UFH (P < 0.001). CONCLUSION: Postoperative anticoagulation with therapeutic UFH in patients with atrial fibrillation was associated with an increased rate of bleeding without reducing the risk of thromboembolism.

Local vectors in desert ants: context-dependent landmark learning during outbound and homebound runs.

Desert ants, Cataglyphis fortis, associate nestward-directed vector memories (local vectors) with the sight of landmarks along a familiar route. This view-based navigational strategy works in parallel to the self-centred path integration system. In the present study we ask at what temporal stage during a foraging journey does the ant acquire nestward-directed local vector information from feeder-associated landmarks: during its outbound run to a feeding site or during its homebound run to the nest. Tests performed after two reversed-image training paradigms revealed that the ants associated such vectors exclusively with landmarks present during their homebound runs.

Egocentric information helps desert ants to navigate around familiar obstacles.

Homing ants have been shown to associate directional information with familiar landmarks. The sight of these local cues might either directly guide the path of the ant or it might activate a landmark-based vector that points towards the goal position. In either case, the ants define their courses within allocentric systems of reference. Here, we show that desert ants, Cataglyphis fortis, forced to run along a devious path can use egocentric information as well. The ants were trained to deviate from their straight homebound course by a wide inconspicuous barrier that was placed between the feeding and nesting sites. At a distant test area, the ants were confronted with an identical barrier rotated through 45 degrees. After passing the edge of the obstacle, the ants did not proceed in the trained direction, defined by the skylight compass, but rotated their courses to match the rotation of the barrier. Visual guidance could be excluded because, as soon as the ants turned around the end of the barrier, the visual cue it provided vanished from their field of view. Instead, the ants must have maintained a constant angle relative to their previous walking trajectory along the obstacle and, hence, must have determined their new vector course in an egocentric way.