A Prospective Study of Risk Factors for Nonadherence With Antipsychotic Medication in the Treatment of Schizophrenia

Objectives: This study aimed to prospectively identify the best single predictor and the best set of predictors of risk for nonadherence with antipsychotic medication in the treatment of patients with schizophrenia.

Method: We used data from 1579 patients in a 3-year, prospective, naturalistic, nonrandomized, multisite study of schizophrenia patients conducted from July 1997 to September 2003 (U.S. Schizophrenia Care and Assessment Program). Adherence with any oral antipsychotic medication was assessed using patient-reported medication adherence and an indirect adherence measure based on medical record prescription information. Patients who reported poor medication adherence or had a medication possession ratio < = 80% (percentage of days with prescriptions for any oral antipsychotic) during the first year after enrollment were defined as nonadherent (N = 296, 18.8%). Thirty-nine previously reported potential risk factors of nonadherence with antipsychotic medication were assessed at enrollment with valid and reliable measures. Risk factors represented patient-, environment-, and treatment-related domains, including sociodemographics, symptom severity, substance use, threat to safety of self and others, other illness-related factors, need for supervision, medication-related adverse events, and prior medication-utilization patterns.

Results: The best single predictor of future nonadherence was nonadherence during the 6 months prior to enrollment (odds ratio = 4.1, 95% confidence interval = 3.1 to 5.6, p < .001). The best set of predictors of nonadherence, ordered by strength of association, included prior nonadherence, recent illicit drug use, recent alcohol use, prior treatment with antidepressants, and greater patient-reported, medication-related cognitive impairment.

Conclusion: Nonadherence with antipsychotic medication is associated with a well-defined set of risk factors that can be used to identify patients who are predisposed to poor adherence.