[Outcome of community-acquired bacteremia admitted to an Internal Medicine Department].
OBJECTIVE: To analyze mortality and associated factors in community-acquired bacteremia admitted to an Internal Medicine Department. PATIENT AND METHODS: Prospective study of bacteremia admitted in 1 year (May 1999-April 2000). We have collected demographic data, previous comorbid conditions, functional status, source of infection, complications, vital signs, laboratory values, APACHE II and SOFA scores, blood cultures, therapy and 28-day mortality. In bivariate analysis, we have used Chi-square, Student-t test and Mann-Whitney U as needed. Significant variables have been introduced in a stepwise backward logistic regression model with mortality as the dependent variable. RESULTS: We have observed 115 episodes of bacteremia in 114 patients. The source of bacteremia was urinary tract in 57.4% episodes and the most common isolate was Escherichia coli (54.4%). 28-day mortality was 15.3%. Factors independently associated with mortality were septic shock (OR 10.4), non-urinary source of bacteremia (OR 9.3), APACHE II score higher than 20 (OR 5.5), and previous dependent functional status (OR 4.8). CONCLUSIONS: Mortality risk factors were septic shock, non-urinary source of bacteremia, APACHE II score and dependent functional status.