Drugs of last resort? The use of polymyxins and tigecycline at US Veterans Affairs medical centers, 2005-2010

• Published in: PloS one 2012-05, Vol.7 (5), p.e36649-e36649
• Main Authors: Huttner, Benedikt, Jones, Makoto, Rubin, Michael A, Neuhauser, Melinda M, Gundlapalli, Adi, Samore, Matthew
• Format: Article
• Language: English
• Subjects: Algorithms; Analysis; Antibiotics; Antimicrobial agents; Bacteria; Bar codes; Biology; Correlation coefficient; Correlation coefficients; Drug dosages; Drug resistance; Drug Utilization Review; Drugs; Enzymes; Epidemics; Health aspects; Health care facilities; Hospitals, Veterans - organization & administration; Infectious diseases; Klebsiella pneumoniae; Medical centers; Medicine; Microbial drug resistance; Minocycline - analogs & derivatives; Minocycline - therapeutic use; Multidrug resistance; Natural language processing; Pathogens; Pharmacy benefit management; Polymyxins; Polymyxins - therapeutic use; Staphylococcus aureus; Staphylococcus infections; Tigecycline; Trends; United States; Veterans; Veterans hospitals
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0036649

Multidrug-resistant (MDR) and carbapenem-resistant (CR) Gram-negative pathogens are becoming increasingly prevalent around the globe. Polymyxins and tigecycline are among the few antibiotics available to treat infections with these bacteria but little is known about the frequency of their use. We therefore aimed to estimate the parenteral use of these two drugs in Veterans Affairs medical centers (VAMCs) and to describe the pathogens associated with their administration. For this purpose we retrospectively analyzed barcode medication administration data of parenteral administrations of polymyxins and tigecycline in 127 acute-care VAMCs between October 2005 and September 2010. Overall, polymyxin and tigecycline use were relatively low at 0.8 days of therapy (DOT)/1000 patient days (PD) and 1.6 DOT/1000PD, respectively. Use varied widely across facilities, but increased overall during the study period. Eight facilities accounted for three-quarters of all polymyxin use. The same statistic for tigecycline use was twenty-six VAMCs. There were 1,081 MDR or CR isolates during 747 hospitalizations associated with polymyxin use (1.4/hospitalization). For tigecycline these number were slightly lower: 671 MDR or CR isolates during 500 hospitalizations (1.3/hospitalization) (p = 0.06). An ecological correlation between the two antibiotics and combined CR and MDR Gram-negative isolates per 1000PD during the study period was also observed (Pearson's correlation coefficient r = 0.55 polymyxin, r = 0.19 tigecycline). In summary, while polymyxin and tigecycline use is low in most VAMCs, there has been an increase over the study period. Polymyxin use in particular is associated with the presence of MDR Gram-negative pathogens and may be useful as a surveillance measure in the future.