Evaluating nonadherence and its risk factors

Assessing a patient’s adherence to an oral antipsychotic regimen when considering a change to a long-acting injectable preparation can be challenging. Psychiatrists frequently underestimate nonadherence,5 self-reports by patients tend to overestimate adherence,6 and patient and physician evaluations of adherence are not always in sync.7 Clinicians should also bear in mind that nonadherence may not be the sole reason for relapse; other factors, including stressful life events, substance abuse, and the natural course of the disease, should also be considered.1

To evaluate medication adherence and its determinants for oral treatment, Bayle and colleagues conducted a cross-sectional survey of 1,887 patients with psychotic disorders, 61.6% of whom had schizophrenia. The patients were enrolled by 399 psychiatrists, age 28-64, chosen by randomization to participate in the study. Patients ranged in age 18-90 years, had experienced a recent psychotic episode, and were about to switch from oral antipsychotic medication to long-acting injectable risperidone. In 90% of the cases, the change in medication regimen was prompted by nonadherence to oral therapy and the desire to improve treatment efficacy.2

Patients assessed their adherence to oral therapy using the Medication Adherence Questionnaire (MAQ), and therapeutic alliance using the 4-Point ordinal Alliance Scale. Psychiatrist rated treatment acceptance, disease severity, and patient insight using the Compliance Rating Scale, the Clinical Global Impression Scale, and the Positive and Negative Syndrome Scale, respectively.2 

In the 1,610 patients who completed the MAQ, adherence was low in 53%, medium in 29.5%, and high in 17.3%. Age younger than 40 years was associated with low adherence, as was poor insight into one’s illness. The study found a strong positive association between patient-reported adherence and physician assessment of patients’ treatment acceptance; active acceptance was associated with medium or high adherence.  Medication adherence was significantly associated with therapeutic alliance; scores for therapeutic alliance were lowest among patients with severe illness and those with lack of insight. Low insight was also associated with poor adherence, as was severe disease and a diagnosis of schizophrenia.2