Send to

Choose Destination
Encephale. 2003 Sep-Oct;29(5):430-7.

[Relationships between insight and medication adherence in subjects with psychosis].

[Article in French]

Author information

Université Victor Segalen-Bordeaux 2, CH Charles-Perrens, Centre Carreire, 121, rue de la Béchade, 33076 Bordeaux.



Poor medication adherence in subjects with psychosis has a high prevalence and a negative impact on clinical outcome. Several studies have reported that a poor level of insight was a strong predictor of poor medi-cation adherence. However, few studies have investigated whether insight was associated with medication adherence, independently from other clinical and treatment characteristics.


To explore the link between insight and medi-cation adherence in subjects with psychosis, and to assess the impact of potential confounding factors on this association.


Subjects included in the study were patients aged 60 or less, consecutively admitted in a psychiatric ward, and presenting with at least one psychotic symptom (delusion or hallucination). Medication adherence was assessed using: 1) history of total discontinuation of treatment against medical advice over the 2 weeks before admission; 2) the 7-point rating scale developed by Kemp et al.; 3) the self-report questionnaire Drug Attitude Inventory (DAI). The Scale to assess Unawareness of Mental Disorder (SUMD) was used to measure level of insight. Assessment of symptoms was performed using the Scale for the Assessment of Positive Symptoms (SAPS), the Scale for the Assessment of Negative Symptoms (SANS), and the Calgary Depression Scale (CDS). DSM IV diagnoses were assessed using the Diagnostic Interview for Psychosis (DIP). The associations between level of insight (SUMD scores) and the three measures of medication adherence were explored using the non-parametric Mann-Whitney and Spearman's tests. Logistic regression models giving Odds Ratios (ORs) and 95% confidence intervals (95% CI) were used to examine the impact of potential confounding variables on the associations between level of insight and medication adherence.


42 patients presenting with schizophrenia broadly defined (n=25) or psychotic mood disorder (n=17) were assessed. Significant associations were found between higher SUMD scores (ie poorer insight) and discontinuation of treatment before admission (z=- 2.6, p=0.009), poor medication adherence rated using the Kemp et al.'s scale (r=- 0.64; p=0.0001), and negative perception of treatment assessed using the DAI (r=- 0.405; p=0.009). The Kemp'scale score and the DAI score were categorised into poor vs. good according to the median for logistic regression analyses. Subjects were 1.7 times more likely (OR=1.7, 95% CI 1.1-2.5, p=0.01) to have discontinued their treatment, 1.9 times more likely (OR=1.9, 95% CI 1.3-2.8), p=0.0003) to have poor medication adherence rated with the Kemp's scale, and 1.8 times (OR=1.8, 95% CI 1.2-2.6, p=0.005) more likely to have a negative perception of the treatment for one point increase at the SUMD score (ie lower level of insight). The associations between SUMD score and the three measures of medication adherence were not modified after adjustment for demographic characteristics (age, gender, educational level, occupational status, marital status) and categorical diagnosis (schizophrenia broadly defined vs. psychotic mood disorder), severity of symptoms (SANS, SAPS, CDS scores), characteristics of the psychotropic treatment, diagnosis of substance or alcohol use disorder, age at onset, and number of previous admissions.


The study demonstrates that medication adherence is associated with the level of insight, independently from other patient's demographic and clinical characteristics. The association between low level of insight and poor medication adherence should be confirmed using prospective studies carried out in ambulatory patients. These findings suggest that psycho-educational programs aimed at improving insight should be developed in order to improve medication adherence.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Masson (France)
Loading ...
Support Center