gms | German Medical Science

Introduction/background: Generic substitution of antiepileptic medicines, especially of modified release products, is discussed controversially. All the more in Germany, since rebate contracts between pharmaceutical manufacturers and statutory health insurance (SHI) funds associated with compulsory substitution led to increased generic substitution. The aim of this study was to explore whether generic substitution in patients treated with extended release valproic acid (valproate, VPA) products is associated with a therapeutic modification of antiepileptic drug therapy, i.e. prescribing of further antiepileptic drugs as a proxy for treatment failure.

Methods: Data were extracted from the DAPI database, containing ambulatory drug claims data at the expense of the SHI from more than 80 % of German community pharmacies. Patients were included if they initiated VPA therapy with extended release (ER) formulations between 2003 and 2006 (first prescription) and had a subsequent prescription within no more than 180 days (second prescription = index). Exposure was classified as generic substitution when patients received a different ER VPA product compared to the previous prescription. Patients were followed from the index date until the occurrence of the event (prescription of further antiepileptic drugs) or censoring (due to interruption or discontinuation of ER VPA treatment, or 2009/12/31 at the latest). Hazard ratios (HR) for the event were estimated using extended Cox regression including time-varying exposure to generic substitution, specialty of prescribing physician and pharmaceutical category of ER formulation (monolithic versus multipellet), as well as time-fixed covariates determined at the date of the index prescription such as pretreatment with further antiepileptic drugs, antipsychotics for treatment of bipolar disorders, antidepressants and total number of ATC codes as measure of polypharmacy.

Results: 78,427 medication profiles of patients initiating VPA therapy were identified of which 15,065 (19.2 %) experienced an event (prescription of an additional antiepileptic drug) during follow-up. Total follow-up time amounted to 106,150 years of observation. Of these, 6,924 years were classified as periods with generic substitution. Times with generic substitution were associated with a 10 % increased risk of the outcome (prescription of further antiepileptic drugs). This effect was preserved after adjustment for covariates (HR = 1.12 [99 % CI: 1.03-1.22]).

Discussion/conclusion: Generic substitution of extended release VPA products was associated with an increased risk of antiepileptic drug regimen modification. Bias from residual confounding due to the limitations of claims data can not be excluded. However, these results suggest that generic substitution of ER VPA products may be disadvantageous.