Use of random-sequence riboflavin as a marker of medication compliance in chronic schizophrenics

Adherence to prescribed medications is believed to be a major problem in medicine. However, actual medication taking behaviour is rarely measured as no reliable objective measures of adherence are available. 50 mg of riboflavin administered at night could be reliably detected in urine, under UV light, for the next 18-24 h. The ability of 20 schizophrenic outpatients to adhere to a once-a-night dosage was studied by dispensing riboflavin or placebo in a random sequence and testing the urine for UV fluorescence the next day. The majority of patients (80%) had errors between 40-80% of the times tested. Only age and sex predicted the level of adherence. Clinicians were very poor predictors of the error rates in their patients.