Joseph M Mylotte1, Melva Neff. 1. Departments of Medicine and Microbiology, School of Medicine and Biomedical Sciences, State University of New York at Buffalo, NY 14215, USA.
Abstract
BACKGROUND: Antibiotic use and cost indicators have been developed specifically for the long-term care facility (LTCF) setting. Approximately 50% of the variation in these indicators was explained by the variation in infection rate. The objectives of this study were to further assess the utility of the antibiotic use and cost indicators in a different LTCF and to determine the correlation of case-mix and these indicators. METHODS: Antibiotic use and cost indicators were collected monthly by unit (N = 10) and by physician (N = 6) at a 433-bed LTCF in Syracuse, New York, from February 1999 to September 2001. Indicators included incidence (number of antibiotic courses per 1000 resident care-days) of antibiotic use, antibiotic utilization ratio ([AUR]; ratio of the number of antibiotic-days to the number of resident care-days), cost per antibiotic-day, and cost per resident care-day. Case-mix variation was measured with the case-mix index (CMI) of the Resource Utilization Group II system. Simple linear and multilinear regression analyses were used to evaluate correlations of continuous variables. RESULTS: Among the 10 units or 6 physicians, there was a significant difference in the average values for all indicators. Correlation between unit- or physician-specific CMI and antibiotic use or cost indicators or infection rate was poor. However, there was a significant positive correlation between unit- or physician-specific infection rate and incidence of antibiotic use, AUR, and cost per resident care-day but not cost per antibiotic-day. With use of multilinear regression analysis to control for CMI and cost per antibiotic-day, infection rate was a significant predictor of incidence of antibiotic use (R2 = 0.65; P <.001) and AUR (R2 = 0.78; P <.001). CONCLUSIONS: This study provides further evidence that the antibiotic use and cost indicators developed specifically for LTCFs can detect significant variation among units within a facility and among physicians. However, there was no correlation between CMI and antibiotic use or cost indicators. After controlling for case-mix variation and cost per antibiotic-day, variation in infection rate explained most of the variation in incidence of antibiotic use and AUR.
BACKGROUND: Antibiotic use and cost indicators have been developed specifically for the long-term care facility (LTCF) setting. Approximately 50% of the variation in these indicators was explained by the variation in infection rate. The objectives of this study were to further assess the utility of the antibiotic use and cost indicators in a different LTCF and to determine the correlation of case-mix and these indicators. METHODS: Antibiotic use and cost indicators were collected monthly by unit (N = 10) and by physician (N = 6) at a 433-bed LTCF in Syracuse, New York, from February 1999 to September 2001. Indicators included incidence (number of antibiotic courses per 1000 resident care-days) of antibiotic use, antibiotic utilization ratio ([AUR]; ratio of the number of antibiotic-days to the number of resident care-days), cost per antibiotic-day, and cost per resident care-day. Case-mix variation was measured with the case-mix index (CMI) of the Resource Utilization Group II system. Simple linear and multilinear regression analyses were used to evaluate correlations of continuous variables. RESULTS: Among the 10 units or 6 physicians, there was a significant difference in the average values for all indicators. Correlation between unit- or physician-specific CMI and antibiotic use or cost indicators or infection rate was poor. However, there was a significant positive correlation between unit- or physician-specific infection rate and incidence of antibiotic use, AUR, and cost per resident care-day but not cost per antibiotic-day. With use of multilinear regression analysis to control for CMI and cost per antibiotic-day, infection rate was a significant predictor of incidence of antibiotic use (R2 = 0.65; P <.001) and AUR (R2 = 0.78; P <.001). CONCLUSIONS: This study provides further evidence that the antibiotic use and cost indicators developed specifically for LTCFs can detect significant variation among units within a facility and among physicians. However, there was no correlation between CMI and antibiotic use or cost indicators. After controlling for case-mix variation and cost per antibiotic-day, variation in infection rate explained most of the variation in incidence of antibiotic use and AUR.
Authors: H J Appaneal; A R Caffrey; V V Lopes; V Mor; D M Dosa; K L LaPlante; T I Shireman Journal: J Hosp Infect Date: 2021-02-05 Impact factor: 3.926
Authors: Hoa Q Nguyen; Declan T Bradley; Michael M Tunney; Carmel M Hughes Journal: Antimicrob Resist Infect Control Date: 2021-03-09 Impact factor: 4.887
Authors: Charles D Phillips; Omolola Adepoju; Nimalie Stone; Darcy K McMaughan Moudouni; Obioma Nwaiwu; Hongwei Zhao; Elizabeth Frentzel; David Mehr; Steven Garfinkel Journal: BMC Geriatr Date: 2012-11-23 Impact factor: 3.921