Drug use evaluation of antibiotics in a pediatric teaching hospital.

OBJECTIVE: To determine the incidence and type of antibiotic use variances at our institution. DESIGN: Inpatient bacterial culture and susceptibility results were reviewed for 1 week per month. Medication administration records were evaluated to determine whether antibiotic selection was appropriate, given the susceptibility of the organism. Process indicators included use of the least costly antibiotic, as well as appropriate dose, interval, and route of administration. The complete medical record was reviewed for all patients if management did not appear to meet criteria. SETTING: A 225-bed, tertiary-care children's teaching hospital. RESULTS: Thirty-five (8.2%) of 428 patients reviewed over 12 months had a total of 49 variances: failure to treat (3), treatment of contaminant/colonizer (2), use of more costly agent (10), failure to revise therapy (8), inappropriate route (2), inappropriate empiric antibiotic (4), incorrect dose (3), unnecessary multiple antibiotics (6), inappropriate drug (8), and prolonged prophylaxis (3). CONCLUSIONS: Thirty-five patients with 10 types of variances were identified during the study. Follow-up monitoring will assess the impact of educational efforts on the incidence of variances. Specific problem antibiotics have been identified for further audits.

Effect of an infusion device on the integrity of whole blood and packed red blood cells.

The effects of the Gemini PC-2 linear peristaltic infusion device on the integrity of packed red blood cells (RBC) and whole blood products are reported. Thirty-eight units of blood products were infused at rates of 999, 100, 50, and 5 mL/hr under simulated clinical conditions. To evaluate the effect of hematocrit on cell survival, fresh and stored packed RBCs preserved with adenine-saline 3 (AS-3) and fresh and stored packed RBCs and fresh and stored whole blood preserved with citrate-phosphate-dextrose-adenine 1 (CPDA-1) were used. No two units tested came from the same donor. Plasma potassium and plasma free hemoglobin concentrations were determined before and after simulated infusion for 80 experimental runs. Preinfusion plasma potassium and free hemoglobin concentrations varied significantly among the blood products. Stored products were associated with higher plasma potassium and free hemoglobin levels than fresh units, both before and after infusion. Concentrations also differed significantly between AS-3-preserved and CPDA-1-preserved fresh and stored packed RBCs. Infusion did not change plasma potassium values appreciably under any conditions. Plasma free hemoglobin increased in the fresh products only. Donor-specific differences were significant for potassium but not for free hemoglobin. There was no significant effect of infusion rate on either biochemical marker. In all the experimental runs, less than 0.01% of cells were lysed. The Gemini PC-2 linear peristaltic infusion device delivered a variety of blood products at a wide range of infusion rates without inducing a substantial degree of hemolysis.

Administrating incompatible drugs by a retrograde intravenous infusion system.

Use of a retrograde infusion system for concurrent intravenous administration of two incompatible drugs separated by a barrier fluid was studied. Four different barrier fluids (5% dextrose injection, 10% dextrose injection, 0.9% sodium chloride injection, and sterile water for injection) were used to separate sodium bicarbonate and calcium chloride. The primary infusion was 5% dextrose injection. The delivery system was visually inspected for formation of precipitate (calcium carbonate) upon addition of the two drug solutions and 2 ml of the barrier fluid. If precipitation occurred, the procedure was repeated, increasing the volume of barrier fluid incrementally until no precipitate could be seen. If no precipitate was seen with the initial 2 ml of barrier fluid, the volume of barrier fluid was decreased until precipitation occurred. These procedures were repeated for each barrier fluid at flow rates of 5-20 ml/hr for 90 minutes. The minimum volume of barrier fluid needed to prevent precipitate formation was determined in triplicate at each flow rate. The minimum volume of barrier fluid that prevented precipitate formation was approximately 2.0 ml. This volume did not differ significantly by barrier solution type or by flow rate. Sodium bicarbonate and calcium chloride can be administered concurrently by the retrograde intravenous method without visual incompatibility when separated by greater than or equal to 2 ml of barrier fluid, and this method can probably be used for administration of other potentially incompatible drugs.