Nomogram for the prediction of valproic acid induced platelet decline: a nested case–control study

Abstract

Abstract Platelet decline is a frequent side effect of valproic acid, a medication commonly prescribed to prevent seizures in neurosurgical patients. However, the risk factors for valproic acid-associated platelet decline remain poorly understood, and it remains unknown whether linezolid or levetiracetam in combination with valproic acid is associated with thrombocytopenia, as both drugs could lead to the decrease in platelet count. This three center, retrospective nested case–control study aimed to develop a predictive model for the prediction of valproic acid-induced platelet decline in a cohort of 356 participants. Multivariate analyses identified advanced age (OR: 1.05; 95% CI 1.03–1.08; P = 0.030) and combination therapy between valproic acid and levetiracetam (OR: 3.03; 95% CI 1.43–6.65; P = 0.005) as independent risk factors, while a trough concentration of valproic acid below 100 μg/mL (OR: 0.41; 95% CI 0.24–0.69; P = 0.010) was an independent protective factor for platelet decline. A nomogram was developed based on these factors, demonstrating robust performance with an area under the curve value of 0.85 in the training cohort and 0.81 in the validation cohort. Calibration plots showed strong agreement between predicted and observed outcomes. This model provides a valuable tool for assessing the risk of platelet decline in valproic acid-treated neurosurgical patients.

Keywords

• Platelet decline
• Valproic acid
• Neurosurgery
• Predictive model
• Nomogram
• Thrombocytopenia