Systems for the preparation and administration of drugs are designed to ensure that the drug is not contaminated. They do not necessarily consider the work environment for the medical staff and new techniques are therefore desirable. The aim of this work is to compare a new closed system for the preparation and administration of drugs with the traditional technique with regard to airborne emission and surface spillage of drugs. Platinum, determined using adsorptive voltammetry, was used as the tracer for airborne emission. Air samples were collected during the preparation and administration, and the collected platinum on the filters was determined by adsorptive voltammetry. For determination of spills and leakage onto surfaces the radioisotope 99m-technetium was used as a tracer. The radiation from the isotope was determined on protective gloves and bench covers after preparation and administration. The mean airborne emission was 6 ng m(-3) with the closed system and 15 ng m(-3) with the traditional pump technique. The average surface spillage using the closed technique was 0.005 microL. This is significantly smaller than with the traditional technique, which resulted in an average spillage of 64 microL. Our results also show that the dominant part of the leakage is surface spillage. Inexperienced nurses could also adequately handle the closed system.

Of 5,125 mostly elderly out-patients, average age 78.2 years, 1,594 (31%) had at least one interacting drug combination according to the Swedish National Formulary. On average, each patient with drug-drug interactions (DDIs) had 1.6 DDIs. There was no difference in the sex or age between those having DDIs and those without. However, patients with one or more DDIs used a significantly larger number of drugs than those without DDIs, on average 8.1 versus 5.2. This significant difference was independent of age and sex. When classified for clinical significance according to the handbook "Drug Interactions and Updates" by Hansten and Horn, 155 (3%) patients had interactions of 'major clinical significance'. The most common were interactions between beta-adrenergic blockers and antidiabetics, followed by potassium-sparing diuretics and potassium, and carbamazepine and dextropropoxyphene. The methodology allows us to screen routinely for DDIs and to plan further studies with emphasis on clinical outcome.

This study has shown that the combination of CBZ and dextropropoxyphene is hazardous in elderly patients and should be used with caution.