Abstract
BACKGROUND: Monitoring free valproate concentrations, as with other highly protein-bound anticonvulsants, is essential in clinical situations where protein binding may be disrupted. Conversion of measured total concentrations to approximate free concentrations offers a cost-effective alternative. This study evaluated the relationship between total and free valproate concentrations for discordance and the impact of key determinants. A novel formula was devised that incorporates significant variables. METHODS: A multicentre, cross-sectional observational analytical study included 101 subjects 18 years and older using valproate for 6 months or longer. Participants were recruited from private and public sector healthcare settings from primary to tertiary level in, South Africa, during 2017-2019. RESULTS: Free valproate concentrations could be measured for 84 subjects. Discordance for concomitant total and free valproate concentrations was 79.1%. Among 19 participants with elevated free concentrations, 15 (78.9%) had total valproate concentrations within the recommended reference range. Calculations based on the study-derived formula were more accurate in predicting free valproate concentration than previously proposed methods. CONCLUSION: This study proposes that the novel formula for calculating free valproate enables more accurate prediction.