Initial approach of patients admitted to third level hospitals with systemic inflamatory response syndrome

Abstract

Objective: To know the initial approach of patients with diagnosis of SIRS from infectious origin in emergency rooms.

Design: Analytic observational study of a concurrent cohort.

Setting: Emergency rooms of Hospital Universitario San Vicente de Paúl (HUSVP) and Hospital General de Medellín (HGM).

Patients: Admitted through emergency rooms with non-traumatic SIRS between August, 1998, and March, 1999, older than 14 and with suspicion of infection as one of the main diagnosis at admission.

Measurements: Description of the associated diseases frequency, the risk factors and findings on basic physical examination; also the usefulness of diagnostic tools, previous antibiotic use and empirical antibiotic therapy at the time of admission. We used chi2 or Fisher´s exact test to compare proportions.

Results: 502 patients were admitted. The main antecedents were chronic obstructive pulmonary disease (COPD) (21.5%) and trauma or previous surgery (18.7%); vital signs were determined as follows: heart rate in 100%, breath rate in 94.8%, blood pressure in 99.2%, temperature in 80.3%, Glasgow scale in 75.6% of the patients. Laboratory requests were done in the following proportions: white blood cell count in 98.4%, chest X rays in 71.1%, platelet count in 94.4% and creatinine determination in 89% of patients. In 26.5% of the patients cultures were not requested; blood cultures were done in 48.8% of the patients and results were positive in 19.2% of the specimens. In 22.3% of the patients antibiotics had previously been used but this fact did not show association with the growth of microorganisms or the prescription of empiric antibiotic therapy in the emergency room (p=0.65).

Conclusions: All of the signs that define SIRS are not determined in patients with suspected infection; neurological state, as determined by the Glasgow scale, and temperature, although being mandatory parameters in emergency room patients, are not determined in 25% of the patients. Laboratory requests and microbiological studies did not in every case correlate with the initial diagnosis and the empirical antibiotic therapy.