Information intervention to decrease inappropriate antibiotic use

Inappropriate antibiotic use negatively impacts health care costs and resistance. Simple, non-intrusive, and effective means of impacting prescribing patterns are needed. We report a successful method which may generalize to other situations. In 1988, tobramycin (T, $7.26/120 mg) accounted for >30% of aminoglycoside doses dispensed at our facility, despite gentamicin's (G, $0.15/120 mg) similar in vitro susceptibility profile. A feedback intervention was instituted, aimed at reducing T to ≤ 15% of aminoglycoside doses. For each new T order, a pharmacist placed an adhesive label in the patient's medical record comparing costs of each agent and encouraging use of G if appropriate. Pharmacists did not review microbiologic data or personally contact physicians. The proportion of T/(T+G) doses dispensed was assessed quarterly, and the intervention was reapplied for 2- to 3-month periods when T/(T+G) was found to be >15%. Results are shown below (*= intervention periods). We have seen similar efficacy in similar interventions Period Months T/(T+G) Doses % Δ% (95% CI) 6/1/88-8/15/88 2.5 942 / 3037 31.0 -*8/16/88-10/31/88 2.5 1051 / 3895 26.9 -4.1 (-6.3, -1.9) 11/1/88-1/15/89 2.5 684 / 4257 16.1 -14.9(-16.9, -12.9) 4/1/89-6/30/89 3 757 / 3437 22.0 -*7/1/89-9/30/89 3 254 / 3333 7.6 - 4.4 (-16.1,-12.7) 10/1/89-12/31/89 3 80 / 2188 3.7 -18.3 (-19.9,-16.7) 4/1/93-6/30/93 3 366 / 2356 15.5 -*8/1/93-9/30/93 2 299/3110 9.6 -5.9 (-7.7,-4.1) designed to encourage q8 hour vs. q6 hour metronidazole dosing. We conclude that a simple chart label intervention which educates physicians as to the cost of their antibiotic choices can be effective in altering prescribing patterns.