Simultaneous placement of a PEG and a VP shunt should be avoided in the acute phase of a patient's hospital admission.

Excessive bruising is a symptom noted by parents of some children treated with the ketogenic diet for epilepsy control, although this side effect is not reported in the literature. We evaluated our cohort of current and past diet‐treated patients for symptoms of bruising or bleeding through chart review and prospective screening at clinic follow‐up visits. A significant increase in bruising or other minor bleeding was reported and/or observed in 16 of 51 patients (31.4%). There were no differences in sex distribution or number of anticonvulsants used between patients with bruising/bleeding and those without this symptom, although the group with bruising/bleeding was significantly younger. No specific anticonvulsant was associated with bruising/bleeding. Six patients with diet‐induced bruising/bleeding underwent an investigation for bleeding diathesis. Five of these patients had prolonged bleeding times and all had diminished responsiveness to various platelet aggregating agents, with no evidence of a release defect. The abnormalities all normalized in the 1 patient tested after ceasing the diet. No patients had serious hemorrhage. One patient had mild von Willebrand disease, which had been asymptomatic before diet initiation. Some patients were Stimate® responsive, suggesting a treatment for more severe bouts of symptoms. These data suggest that a ketogenic diet–related bleeding tendency occurs in about one third of treated patients owing to preexisting factors defining susceptibility in combination with diet‐induced depression of platelet responsiveness, possibly related to changes in platelet membrane lipid composition and/or concentration and resultant effects on function of membrane‐embedded proteins. Patients on the diet undergoing anticoagulation or surgery should be evaluated carefully for symptoms of bleeding tendency. Ann Neurol 2001;49:98–103

A broad-spectrum 5-day regimen is superior to 'single-shot' antibiotic prophylaxis in preventing infection-related wound complications. However, this study needs to be conducted in a larger number of patients to have statistical power.

The balloon water manometer system has long been employed in animals and in man as a means of recording graphically the motility of the alimentary tract (1, 2). As others have done, we have used this method to study the behavior of adjacent segments of the small intestine in man under control conditions and in response to drugs and placebos (3, 4). The purpose of this communication is to describe the recording procedure and the evaluation of the tracing patterns. The results of the drug and placebo studies are to be reported elsewhere.Certain drawbacks to the use of the balloon water manometer system have been presented by Quigley (5). He has pointed out that this method necessitates using a large volume/pressure coefficient which does not maintain a linear relationship. The pressure, moreover, registered by the balloon varies depending on its changing shape and on the degree of patency of the intestinal lumen. These objections did not primarily concern us as we were interested in qualitative and relative quantitative changes in motility rather than in absolute measurements of intraluminal pressures.Various workers have described graphic records of motility in terms of their non-propulsive and propulsive character (6-9). Adler, Atkinson and Ivy (6) observed that in the colon of patients with colostomies low, relatively rapid wave patterns were non-propulsive while high, sustained contractions were associated with the forward movement of the intestinal contents provided adequate coordination existed between bowel segments. Abbott and Pendergrass (7) felt that a