Longitudinal Assessment of Cocaine Toxicity in Emergency Department Patients

Abstract

Longitudinal data are necessary to identify changes in drug abuse and resultant toxicity. The development of COC-LOG, a longitudinal database used to assess cocaine-associated emergency department patient visits at Truman Medical Center, the University of Missouri-Kansas City, is described. COC-LOG currently contains 1,054 patient visits. The data demonstrate that cocaine toxicity has remained relatively constant in our emergency department (ED) over the last 4 years. The majority of cocaine use is through smoking crack, and major differences in toxicity are not seen between this route and IV cocaine use. The relatively low prevalence of severe cardiac toxicity and deaths has lead to a series of projects to evaluate cocaine toxicity's clinical course. Electrocardiographic abnormalities are common, but do not frequently indicate severe clinical toxicity. In the first 3 years after ED discharge, deaths in cocaine patients occur almost twice as frequently as in age-, race-, and sex-matched control ED patients. Urine drug testing of pregnant patients suggests that documentation of recent cocaine use by history alone underestimates the prevalence of cocaine use in ED patients approximately four-fold. COC-LOG is a useful method of following cocaine's impact on department patients. Study results can be applied to patient care and the education of ED faculty, residents, students, and personnel.