Multicenter assessment of the revisit risk for a further drug-related problem in the emergency department in cocaine users (MARRIED-cocaine study).

ABSTRACT

INTRODUCTION AND OBJECTIVES: Emergency departments (EDs) in Spanish hospitals daily attend a large number of patients for adverse reactions or clinical complications resulting from cocaine use. After discharge, some of these patients revisit the ED for the same reason within a year. The objective of the present study was to quantify the rate of such revisits and identify the factors associated with them. METHOD: We performed a retrospective, multicenter study with cohort follow-up and without a control group, conducted in the EDs of six Spanish hospitals during 12 months (January-December 2009). We included all ED patients attended for cocaine-related symptoms who reported recent cocaine use and those with cocaine-positive urine analysis by immunoassay without declared consumption. Twelve independent variables assessed for each hospital ED were collected sex, age, place of consumption, month, day, and time of consumption, mode of arrival at the ED, discharge diagnosis, psychiatric assessment on the ED episode, concomitant drugs, destination on discharge, and history of previous ED visits related with drug use and alcohol use. The dependent variable was a subsequent visit to the ED associated with drug use, identified using the computerized hospital admissions system. RESULTS: The study included 807 patients, of whom 6.7% revisited the ED within 30 days, 11.9% within 3 months and 18.9% within 1 year. The variables significantly associated with ED revisits were presence of clinical manifestations directly related to cocaine (p < 0.05), ED attendance on a working day (p < 0.05), history of ED visits related with the consumption of alcohol (p < 0.001) or drugs (p < 0.001), and the need for urgent consultation with a psychiatrist (p < 0.001), although only the last four were independent predictors in multivariate analysis. We derived a score based on these variables to predict risk of revisits (MARRIED-score, ranging from 0 to 400 points), which had a reasonably good predictive value for revisit (area under ROC of 0.75; 95% CI 0.71-0.79).