Description:
<jats:p><jats:bold>Summary:</jats:bold><jats:bold> Purpose: </jats:bold> To assess interlaboratory variability in the determination of serum levels of new antiepileptic drugs (AEDs).</jats:p><jats:p><jats:bold>Methods:</jats:bold> Lyophilised serum samples containing clinically relevant concentrations of felbamate (FBM), gabapentin (GBP), lamotrigine (LTG), the monohydroxy derivative of oxcarbazepine (OCBZ; MHD), tiagabine (TGB), topiramate (TPM), and vigabatrin (VGB) were distributed monthly among 70 laboratories participating in the international Heathcontrol External Quality Assessment Scheme (EQAS). Assay results returned over a 15‐month period were evaluated for precision and accuracy.</jats:p><jats:p><jats:bold>Results:</jats:bold> The most frequently measured compound was LTG (65), followed by MHD (39), GBP (19), TPM (18), VGB <jats:ext-link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="#b15">(15)</jats:ext-link>, FBM <jats:ext-link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="#b16">(16)</jats:ext-link>, and TGB <jats:ext-link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="#b8">(8)</jats:ext-link>. High‐performance liquid chromatography was the most commonly used assay technique for all drugs except for TPM, for which two thirds of laboratories used a commercial immunoassay. For all assay methods combined, precision was <11% for MHD, FBM, TPM, and LTG, close to 15% for GBP and VGB, and as high as 54% for TGB (p < 0.001). Mean accuracy values were <10% for all drugs other than TGB, for which measured values were on average 13.9% higher than spiked values, with a high variability around the mean (45%). No differences in precision and accuracy were found between methods, except for TPM, for which gas chromatography showed poorer accuracy compared with immunoassay and gas chromatography–mass spectrometry.</jats:p><jats:p><jats:bold>Conclusions:</jats:bold> With the notable exception of TGB, interlaboratory variability in the determination of new AEDs was comparable to that reported with older‐generation agents. Poor assay performance is related more to individual operators than to the intrinsic characteristics of the method applied. Participation in an EQAS scheme is recommended to ensure adequate control of assay variability in therapeutic drug monitoring.</jats:p>