Objective: To examine the hypothesis that poorer social and family support, identifiable at the onset of treatment, is associated with nonadherence in the first 6 months of treatment of patients with first-episode psychosis (FEP), independent of other patient-related factors.Method: Consecutive patients (n = 100) admitted to a specialized early intervention service for FEP who met the Diagnostic and Statistical Manual of Mental Disorders, fourth edition, criteria for either a schizophrenia spectrum disorder or an affective psychosis were evaluated monthly for 6 months regarding their adherence to medications. Using sociodemographic and illness-related factors, including social and family support, as independent variables and adherence as the dependent variable, univariate analyses were followed by logistic regression.Results: Fifty-six patients (54.9%) were adherent (76% to 100% of doses taken) and 46 (45.1%) nonadherent (less than 76% of doses taken). Nonadherent patients were less likely to have received a good level of social support (c 2 = 5.89, df = 1, P = 0.02), as rated by their respective case manager, and more likely to be single (Fisher exact test, P = 0.019) and to have refused medication at the first offer of treatment (c 2 = 19.70, df = 1, P = 0.001). Using logistic regression, both the level of social support (OR = 3.552, P = 0.03) and early medication acceptance (OR = 11.092, P < 0.001) were significant as predictors of adherence. Conclusion Clinical Implications· Social and family support may be an important determinant of adherence to medication very early in the course of treatment of FEP. · Refusal to be treated at the first offer of treatment remains the most likely reason for nonadherence to medication, at least in the first 6 months of treatment of FEP. · Interventions may need to be designed to influence adherence to medication very early in the course of treatment. Limitations· The measure of social and family support used may have been too brief and not able to tap into different aspects of support from family and others in the social environment. · The absence of influence of insight on adherence to medication might have resulted from using the one item from the Positive and Negative Symptom Scale for assessing insight and not using a more detailed measure of insight. · The relatively high level of nonadherence may be a reflection of a very conservative measure of adherence to medication (taking greater than 75% of doses of medication prescribed).

This study evaluates how much agreement there is between subjective reports of adherence to antipsychotic medication and objective or derived measures of adherence in first-episode psychosis (FEP) and asks if any adherence measure could approximate a gold standard based on correlation to symptom improvement in the early phase of treatment. Adherence was assessed in 81 FEP subjects on a monthly basis by reports from patients, clinicians, family, and pill counting. A consensus measure of adherence was derived from all available sources of adherence data. Symptoms were measured using the Positive and Negative Syndrome Scale at study entry and 3 months subsequently. Adherence as measured by patient report, pill count, and clinician report were in good agreement with each other (intraclass correlation coefficient = 0.84), and all of these measures were highly correlated to consensus adherence (r values between 0.86 and 0.98). Mean adherence was slightly higher as rated by patients (83% full doses taken per month) and family members (91%) than by clinicians (76%), pill counting (73%), or consensus value (74%). Early in treatment, each measure of adherence (except family report) was significantly associated with positive symptom reduction, although the order of magnitude of this correlation was greater for pill count and consensus adherence (P < 0.01) compared with patient- or clinician-reported adherence (P < 0.05). Patient or clinician reports provide a reasonable estimate of medication adherence in FEP, but introducing pill counting or a derived measure of adherence may allow more accurate measurement.