Organizational Factors Associated with Antibacterial Use Among Academic Medical Centers
Methods: Data concerning antibacterials were obtained from administrative data from 89 University HealthSystem Consortium hospitals for adult patients (age ≥ 18) for 2011. Antibacterials were measured in Days of Therapy per 100 admissions. The following data were also obtained: case mix index (CMI); bedsize (expressed as < or > the median); the % of intensive care unit days; average length of stay (LOS); a marker for market competition, the Herfindahl Index (HHI); region [Northeast (NE), South, Midwest, West]. The number of admissions per 35 UHC Clinical Service Lines (CSLs) was collapsed into 4: Surgery; Medicine; Transplant; Other [expressed as < or > the median]. Three forward stepwise regression models were conducted to identify factors associated with total antibacterial, total antipseudomonal [antipseudomonal fluoroquinolones, cephalosporins, β-lactam/β-lactamase inhibitors, and carbapenems], and anti-methicillin resistant S. aureus (MRSA) [vancomycin + linezolid + daptomycin + tigecycline + ceftaroline + quinupristin/dalfopristin] drug use.
Results: For all models, the following variables were statistically significant (p < 0.05) and positively associated with antibacterial use: LOS; region South as compared to NE; while HHI and Other CSL were significant and negatively associated with antibacterial use. In addition, Surgery and Transplant CSLs were significant and positively associated with antipseudomonals [p = 0.03 and 0.01, respectively] and the region West as compared to NE was significant and positively associated with anti-MRSA agents [p = 0.04].
Conclusion: Several factors were associated with total antibacterial, antipseudomonal, and anti-MRSA drug use, including HHI, LOS, region, and CSL type. Incorporation of these factors into interhospital analyses can aid in more meaningful antibacterial comparisons.
H. Wang, None
S. Hohmann, None