Comparison

Undetectable anticonvulsant blood levels indicate sustained noncompliance (several consecutive doses missed). We com-pared 91 consecutive outpatients with epilepsy and undetectable anticonvulsant blood levels to 100 patients seen during the same time period, verified as compliant by acceptable serum levels. We hypothesized that pay status, application for Supplemental Security Income, patient age, history of missed appointments, and functional status would differ between compliant and noncompliant patients. We were surprised to find large differences between clinic and insurance patients and between Caucasian and non-Caucasian patients. The 100 compliant patients included 44 Caucasian and 56 non-Caucasian patients, whereas only 9 of 91 noncompliant patients were Caucasian, and only 9 had insurance, compared to 32 compliant patients. Applications for Supplemental Security Income and history of missed appointments were signifi-cantly associated with noncompliance, but patient age, seizure type, and seizure control were not. Uninsured Caucasians were more often compliant than non-Caucasians were. Many noncompliant patients had mild epilepsy, which was report-edly doing well. Race and pay status were closely correlated. Several noncompliant females became pregnant, whereas no compliant patients did. Compliant patients were much more likely to be accompanied by a parent or caretaker on clinic visits than noncompliant patients. Noncompliant patients had at least one acceptable subsequent serum level, although 2 patients with intractable epilepsy had undetectable serum levels on three or more occasions. Noncompliance may respon