Diarrhea occurs frequently in the critically ill tube-fed population and may result from a multitude of causes. Despite the availability of antidiarrheal medications, diarrhea associated with enteral feedings remains a problem for clinicians and for the patients affected by it. We tested the hypothesis that administration of banana flakes would control diarrhea in critically ill patients receiving enteral feedings. Thirty-one patients with diarrhea and receiving enteral feedings were randomized to receive either banana flakes or medical treatment for diarrhea. Medical treatments included the use of pharmacological agents according to the discretion of the patient's physician or reducing feeding rates. Both banana flakes and medical treatments reduced the severity of diarrhea in critically ill tube-fed patients. Over the course of treatment, mean diarrhea scores were 21.64 +/- 7.81 for the banana flake group and 25.41 +/- 9.76 for the medical group. These differences were not statistically significant. Both groups achieved similar levels of nutrition support. The banana flake group had less diarrhea clinically, with 57% of the subjects diarrhea free on their last study day as opposed to 24% of the medically treated subjects. This occurred despite a threefold increase in the number of patients testing positive for Clostridium difficile toxin in the banana flake group. We conclude that banana flakes can be used as a safe, cost-effective treatment for diarrhea in critically ill tube-fed patients. Banana flakes can be given concurrently with a workup for C. difficile colitis, thereby expediting treatment of diarrhea.

To evaluate the ionic requirements of colonic base secretion, segments of rat distal colon were studied under short-circuited conditions. Net base flux was composed of an active secretory component and a diffusive component. Studied in the absence of a transepithelial HCO3- concentration gradient, active base secretion was dependent on the HCO3- concentration of the bathing solution but was not influenced by the CO2 tension or pH. Base secretion appeared to saturate with a Km of 33 +/- 9 mM and was inhibited by ouabain. The diffusive component was characterized by an apparent permeability coefficient to HCO3- of 8.9 +/- 0.9 x 10(-6) cm/s. In addition to requiring HCO3- on the serosal surface, net base secretion was inhibited by reducing the Na+ concentration in the serosal medium and the Cl- concentration in the mucosal medium. These data suggest that colonic base secretion involves HCO3- entry across the basolateral surface, energized by the Na+ gradient, and HCO3- exit across the apical surface in exchange for Cl-.

No abstract

Although plasma separation has been reported to have a relatively small complication rate in large series of healthy outpatients, little attention has been directed to the evaluation of the safety and effectiveness of the therapy in acutely ill, hospitalized patients. The experience of using standard hemodialysis equipment and membrane plasma separators with 281 plasma separation treatments in 49 patients over the last 7 yr is reported and analyzed. The data reveal a 1.4% incidence of hypotension and a 0.4% incidence of hematuria in the 281 treatments--rates similar to those reported in outpatients. In addition, analysis of the diseases and patients treated over the 7 yr reported demonstrates a marked shift from immunological and hematological disorders towards neurological disorders. The data suggest that plasma separation may be easily and safely performed by any institution capable of performing acute hemodialysis.

To evaluate changes in feto-placental markers with plasma exchange in pregnancy, two patients at varying stages of pregnancy referred to a tertiary care hospital and requiring plasma exchange for intercurrent problems were evaluated. Alpha-fetoprotein, human chorionic gonadotropin, and free estriol were sequentially measured in the patients' plasma and in the fluid removed, thus permitting calculations of permeability rates and clearances. Despite markedly different molecular weights, all three feto-placental markers had similar permeabilities and clearances. While in both patients maternal levels of alpha-fetoprotein and human chorionic gonadotropin decreased rapidly with plasma separation and rebounded rapidly to baseline, free estriol responded differently and did not appear to decrease with therapy. Maternal levels of feto-placental markers only transiently changed with plasma exchange during pregnancy and rapidly returned to baseline with no apparent consequences to the pregnancy.