Neither dosage nor serum levels of antiepileptic drugs are predictive for efficacy and adverse effects.

In order to assess whether doses or serum levels are predictive for the efficacy and adverse effects of antiepileptic drugs (AEDs), measures for exposure to drug combinations have to be used. For doses, the ratio of the observed prescribed daily dose (PDD) and the average defined daily dose (DDD) considered effective for the main indication of the drug was used. In analogy for serum levels, the OSL/ATL ratio, i.e. the ratio of the observed serum level and the average therapeutic level was used. In polypharmacy these ratios can be summed as the are normalized measures of strength. The correlations of these ratios with outcome measures were studied in 200 patients attending out-patient clinics of special centres for epilepsy; half of these patients were treated with monopharmacy and half with polypharmacy. As outcome measures the following indices were used: the index of seizures, which quantifies seizure type and frequency, the seizure activity index, the neurotoxicity score, the systemic toxicity score, and the composite index of impairments, which is the sum of the seizure activity index and the neurotoxicity score and the systemic toxicity score. When all data were pooled, the correlation coefficient between the PDD/DDD ratio and the OSL/ATL ratio was 0.77. However, when the data were examined separately for the monopharmacy and polypharmacy groups, the correlation was 0.31 for the monopharmacy group and 0.50 for the polypharmacy group. Neither the PDD/DDD ratio nor the OSL/ATL ratio correlated with the composite index of impairments or with any of the individual indices. Factors such as the difficulty of titrating the endpoint of seizure suppression and the development of tolerance to adverse drug effects may perhaps be responsible for these findings. This observational study signals the problem.