Effects of targeting higher vancomycin trough levels on clinical outcomes and costs in a matched patient cohort

Study objective: To compare clinical outcomes and costs in patients treated with the new vancomycin guidelines recommending goal serum trough concentrations of 15-20 mg/L versus patients treated with vancomycin doses targeting trough concentrations 5-20 mg/L prior to the new guidelines.

Design: Retrospective quasi-experimental study.

Setting: Urban level I trauma center.

Patients: A total of 200 patients treated with vancomycin for at least 72 hours for confirmed, complicated methicillin-resistant Staphylococcus aureus (MRSA) bacteremia during one of two study phases relative to the implementation of the vancomycin dosing guidelines targeting serum trough concentrations of 15-20 mg/L: 2005-2007 (preperiod phase) and 2008-2010 (postperiod phase). One hundred patients in each phase were matched in a 1:1 ratio according to diagnosis, any concomitant nephrotoxic agents (e.g., aminoglycosides, colistin, acyclovir), and age ± 5 years.

Measurements and main results: Patients in the preperiod had significantly lower success rates with vancomycin than those in the postperiod (45% vs 60%, p=0.034). Median length of stay (LOS) was not significantly higher in patients in the preperiod versus postperiod (15 days vs 13.5 days; p=0.28), and patients in the preperiod received a longer median duration of vancomycin versus those in the postperiod (13 days vs 8.5 days; p<0.001). No statistically significant difference was noted in total hospital costs for patients treated with vancomycin during the preperiod versus the postperiod ($32,754 vs $27,709, p=0.147). However, total drug and monitoring costs of vancomycin were significantly higher for patients in the postperiod. Initial vancomycin trough levels were significantly lower in patients in the preperiod versus postperiod (12.3 mg/L vs 15.8 mg/L, p=0.02). Patients in the preperiod had lower rates of nephrotoxicity than those in the postperiod, although this difference was not statistically significant (15% vs 18%; p=0.85). Median (interquartile range) LOS was significantly longer in patients who developed nephrotoxicity compared with patients who did not develop nephrotoxicity (17 days [11.5-36.5 days] vs 14 days [9-24 days], p=0.017). Costs associated with measurement of serum creatinine concentrations and vancomycin trough levels as well as labor were significantly higher in patients who developed nephrotoxicity.

Conclusion: Higher vancomycin trough concentrations improved outcomes in patients with complicated MRSA bacteremia. In addition, more aggressive dosing was shown to significantly decrease overall duration of vancomycin therapy, which may affect total hospital LOS and cost. Patients who experienced nephrotoxicity had a significantly longer hospital LOS. Additional studies evaluating optimal therapy for MRSA bacteremia in a larger cohort of matched patients are warranted.