We aimed to evaluate the impact of physicians' educational programs in the reduction of inappropriate intravenous lines in internal medicine.

Fifty-six French internal medicine units were enrolled in a nationwide, prospective, blinded, randomized controlled trial. Forms describing the patients with an intravenous line and internal medicine department characteristics were filled out on 2 separate days in January and April 2007. Following the first visit, all units were randomly assigned to either a specific education program on the appropriate indications of an intravenous line, during February and March 2007, or no training (control group). The Investigators' Committee then blindly evaluated the clinical relevance of the intravenous line according to pre-established criteria. The primary outcome was the percentage of inappropriate intravenous lines.

During January 2007, intravenous lines were used in 475 (24.9%) of the 1910 hospitalized patients. Of these, 80 (16.8%) were considered inappropriate. In April 2007, 416 (22.8%) of the 1823 hospitalized patients received an intravenous line, which was considered in 10.2% (21/205) of patients managed by trained physicians, versus 16.6% (35/211) of patients in the control group (relative difference 39%; 95% confidence interval, −0.6-13.3; P = .05). Reduced intravenous administration of fluids, antibiotics, and analgesics accounted for the observed decrease.

The use of a simple education program reduced the rate of inappropriate intravenous lines by almost 40% in an internal medicine setting (NCT01633307).