Antibiotic use and cost indicators at a rural hospital: A pilot project

Background: Recently, simple antibiotic use and cost indicators were develo ped for use in long-term care facilities. It was hypothesized that these in dicators also may be applicable to the acute hospital setting. Methods: For a 24-month period, data were collected quarterly on antibiotic use and cost indicators for 11 primary care physicians in a 40-bed rural h ospital. Indicators included antimicrobial use ratio (AUR, ratio of the num ber of antibiotic days to the number of patient care days), cost per antibi otic day, and cost of antibiotics per patient care day. One-way analysis of variance and simple linear regression were used to analyze data. Results: Quinolones (oral plus parenteral) accounted for 26% of the total a ntibiotic days (N = 6020) followed by ceftriaxone (19%) and cefuroxime (11. 8%; oral plus parenteral). Overall trends in antibiotic use and cost includ ed a significant increase in quarterly AUR (R-2 = 0.78, P = .004) and cost per patient care day (R-2 = 0.82, P = .002) but no significant change in qu arterly total antibiotic costs or cost per antibiotic day. Among physicians there was a significant difference in mean quarterly AUR (P < .001) and me an quarterly cost per patient care day (P < .001) but no significant differ ence in mean quarterly cost per antibiotic day. Variation in physician-spec ific cost per patient care day was best explained by variation in AUR (R-2 = 0.75, P < .001). Conclusions: Significant variation in simple antibiotic use and cost indica tors was identified at a rural hospital from both the facility and physicia n perspective. Standardized methods for antibiotic use and cost monitoring, like the one described in this article, are required before the relationsh ip between antibiotic use ana resistance can be fully understood.