Treatment adherence and its association with insight, quality of life and other clinical variables in patients with bipolar disorder and schizophrenia in remission

Objective: The present study aimed to examine the association of treatment adherence with insight, quality of life and other clinical variables among subjects with bipolar disorder and schizophrenia who were in remission. This paper also aims to identify predictors of treatment adherence.

Methods: The sample consisted of 300 consecutive admissions to the outpatient clinic at a tertiary care psychiatric research and education hospital during a 13-month study period (from May 2012 through June 2013). One hundred and fifty patients diagnosed with schizophrenia and 150 patients diagnosed with bipolar disorder were involved in the trial. The schizophrenia patient group and the bipolar disorder group were both administered the following battery of tests: Mini International Neuropsychiatric Interview, Brief Psychiatric Rating Scale (BPRS), the Schedule for Assessing the Three Components of Insight (SAI), World Health Organization Questionnaire on Quality of Life: Short Form (WHOQOL-BREF), Beck Cognitive Insight Scale (BCIS), UKU (Udvalg for Kliniske Undersøgelser) Side Effect Rating Scale, MARS (Medication Adherence Rating Scale), McEvoy (McEvoy treatment observation form). Logistic regression was used to determine significant variables associated with treatment adherence to medication.

Results: The results demonstrated that subjects with bipolar disorder in remission had similar levels of quality of life in all four domains compared to those with schizophrenia in remission. No group differences for BPRS scores, treatment adherence and overall insight were found. Correlation analyses indicated that re-labeling of psychotic phenomena domain of insight scale was negatively correlated to quality of life in all four domains. To explore which clinical variables lead to a greater risk for treatment non-adherence in patients with schizophrenia and bipolar disorder, a logistic regression equation was computed. Dependent variable was MARS (treatment adherent vs. non-adherent), while independent variables involved diagnosis (schizophrenia vs. bipolar disorder), gender, duration of illness, BPRS scores, UKU scores, BCIS scores. The results of this logistic regression analysis indicated that only the total score of insight scale is a predictor of treatment adherence even when other potential factors were controlled for. The secondary analysis was a logistic regression with forward selection of BCIS subscores.

Conclusion: The treatment compliance domain of insight scale seems to be a key variable in the prediction of treatment adherence in patients with schizophrenia and bipolar disorder. Keywords: bipolar disorder, insight, schizophrenia