Serum trough level as a postmarketing quality measure of generic vancomycin products

Background: The vancomycin trough level (VTL) is the most widely used pharmacokinetic parameter for monitoring its clinical efficacy. Whether the VTL is affected in patients receiving different vancomycin products has not previously been determined. Methods: From 2005 to 2015, five vancomycin products, including the innovator (designated as VAN-Lilly) and four generic versions (designated as VAN-A, VAN-B, VAN-C and VAN-D), were sequentially used in a teaching hospital. The initial VTLs were compared between patients who received different vancomycin products after propensity score (PS) weighting and matching for clinical covariates. Results: Among 8735 patients with initial VTL levels available for analysis, a significant association was identified between the VTL and different vancomycin products in children aged 1 month to 12 years (P < 0.0001). The PS weighting analysis in the paediatric group disclosed children on VAN-C had higher VTL compared to those on other four products (P = 0.0008). PS matching analysis revealed that children who received VAN-C had significantly higher VTLs than those who received VAN-Lily (P = 0.0001), VAN-A (P = 0.0008), VAN-B (P = 0.0002) or VAN-D (P = 0.0015). Furthermore, the coefficient of variation of the VTL was much greater in patients who received VAN-C than in those who received the other four versions, suggesting an unstable quality of this product. Conclusion: A generic version of vancomycin generated significantly higher concentrations and greater variation of VTLs than the innovator and other generic vancomycin products in children. The VTL can serve as an indicator to monitor the quality of vancomycin products after marketing.