Prospective surveillance of imipenem/cilastatin use and associated seizures using a hospital information system

To develop computerized methods to monitor and recommend dosage changes for patients treated with excessive dosages of imipenem/cilastatin (I/C) and to determine the incidence of I/C-associated seizures in our patient population.

Prospective observational and interventional study of all patients admitted to LDS Hospital and treated with I/C from May 1, 1987, through June 30, 1991.

LDS Hospital, Salt Lake City, UT, a 520-bed, tertiary care center associated with the University of Utah School of Medicine.

Using a hospital information system we developed computerized algorithms to identify and monitor patients receiving I/C. These algorithms screened the computer-stored medical records of all inpatient admissions for I/C prescription orders. Computer-decision support algorithms estimated the renal function of each I/C-treated patient and provided suggestions when dosages were determined to be excessive. Additional computer-generated alerts identified patients who were receiving anticonvulsants concomitantly with I/C or whose therapy reflected dosage changes in the previous 24 hours. A list of all I/C-treated patients with alerts was reviewed daily by a clinical pharmacist and prescribing physicians were contacted if the computer-generated suggestions were clinically relevant.

The number and characterization of I/C-associated seizures.

From May 1, 1987, through June 30, 1991, we prospectively monitored 107,600 patients of whom 1951 were treated with I/C. The following risk factors for I/C-associated seizures were observed in the I/C-treated population: CNS disease (6 percent), seizure disorders (0.6 percent), and abnormal renal function (70 percent). The observational and interventional methods employed in this study resulted in 79 percent of the patients receiving I/C dosages appropriate for their corresponding renal function. During the 50-month study period, we detected four seizures (0.20 percent) in the I/C-treated patients. All 4 patients were receiving I/C dosages that were excessive with respect to their renal function.

Our rate of seizure (0.2 percent) was lower than the 1-2 percent rate reported in the literature despite the fact that more than 70 percent of the patients who received I/C had risk factors for seizure. We believe that appropriate dosing of I/C results in a low rate of associated seizures. Computer-assisted monitoring of I/C dosages in relation to renal function resulted in a reduced incidence of seizures.