Agreement between clinically assessed and self-rated medication adherence with schizophrenia in Hong Kong

Abstract

Adherence to antipsychotic treatment is an important aspect of long term management of schizophrenia and other related psychotic disorders. Since medication non-adherence is related to relapse, the correct identification of patients who are at risk of non-adherence problems is important. This study evaluates the sensitivity and specificity of the clinicians screening evaluations and explores factors that are related to non-adherence. Clinicians’ rated questionnaires and matching patients’ self rated questionnaires were administered to 511 consecutive outpatients (mean age 37.8 years and mean duration of illness 11.1 years) and their clinicians in four hospital clinic sites in Hong Kong. Items in the questionnaires addressed adherence behaviors and attitudes towards medication. In addition, clinicians also estimated the presence of adverse factors such as substance abuse, cognitive impairments etc. Patients’ self report and clinicians’ estimation suggested similar proportion (about 20%) of patients to have non-adherence behaviors (i.e. forgot to take the medication and decided to stop medication on their own). Clinicians’ detection of non-adherence behaviors achieved relatively high specificity (0.84) but low sensitivity (0.33). A logistic regression analysis on patients’ decision to stop medication revealed that significant predictors include clinicians’ perception of patients’ unawareness of illness, patients’ reported lack of perception from the benefits of medication, patients’ reported negative feelings associated with regular medication and younger age. Non-adherence is a widespread phenomenon affecting a significant proportion of outpatients under treatment. Clinicians are relatively more accurate in judging patients without non-adherence problems but are less effective in identifying patients with non-adherence problems. Non-adherence behavior is predicted by younger age, patients’ negative attitudes towards medication, patients’ lack of perceived medication benefit and poor illness awareness. We acknowledge partial support for this study from an unrestricted education grant from Janssen Pharmaceuticals.