[Potential and real drug interactions in critical care patients]

To detect potential drug interactions occurring during treatment of critically ill patients and to evaluate the real clinical and analytical repercussions of such interactions.

Seventy consecutive patients in a six-bed medical-surgical intensive care unit (ICU) were studied prospectively over a period of five months. The pharmacological department of the hospital provided the drugs through a computerized unit dosing system, using a program that gave warnings of possible drug interactions as each patient's treatment was entered. Interactions were cataloged according to clinical importance following the criteria of Hansten. The repercussions were evaluated by clinical and analytical monitoring.

On hundred two potential interactions were detected in 44.3% of patients (1.5 interactions per patient). In 6 patients (19.3%) analytical alterations were observed but only 2 (6.4%) of them presented clinical manifestations. Most of the interactions were cataloged as being of moderate clinical importance. Digoxin was the most frequently implicated drug. The number of drugs administered per patient was 14.7 and patients receiving the most drugs were the most likely to experience and interaction. We found a significant correlation (p < 0.05) between the number of interactions and age, APACHE III score, length of ICU stay and number of drugs administered.

The frequently of drug interactions in critically ill patients can be high, given the large number of drugs they receive. Although the real clinical repercussions of interactions are few, physicians should watch for them and the act appropriately.