# Antipsychotic Medication Adherence Among Outpatients Attending the Centre Neuro Psycho Pathologique in Kinshasa: A Cross-Sectional Pilot Study of Treatment Attitudes, Insight, and Side-Effect Burden

## TL;DR

This study found that nearly half of patients in Kinshasa did not properly take their antipsychotic medications, with treatment attitude and insight being key factors.

## Contribution

The study provides new insights into antipsychotic adherence in the DRC, identifying modifiable factors like treatment attitude and insight.

## Key findings

- Poor adherence was reported in 45.2% of patients.
- Treatment attitude and insight were independently associated with lower odds of poor adherence.
- Adverse effects were strongly correlated with adherence but not independently associated after adjustment.

## Abstract

Background and objective

Incomplete adherence to antipsychotics commonly undermines relapse prevention and functional stabilization in severe mental disorders. Published data from the Democratic Republic of the Congo (DRC) are scarce. The objective of this pilot study was to estimate antipsychotic medication adherence among outpatients attending the Centre Neuro Psycho Pathologique (CNPP) in Kinshasa, DRC, and examine associations with treatment attitude, insight, and adverse effects using brief standardized instruments.

Methods

We conducted a cross-sectional pilot study from April to May 2025 at the outpatient service of the CNPP in Kinshasa, DRC. Adult patients (≥18 years) who had been prescribed at least one antipsychotic for a minimum duration of five weeks were included. Adherence was measured with the 10-item Medication Adherence Rating Scale (MARS). Adherence was dichotomized as good adherence (MARS ≥ 7) and poor adherence (MARS < 7). The Drug Attitude Inventory (DAI) was used to assess treatment attitude. Insight was assessed with the 8-item Birchwood Insight Scale (BIS-8). The Glasgow Antipsychotic Side-Effect Scale (GASS) measured the adverse effects of antipsychotics. Correlations between scales were examined by Spearman’s rank correlation coefficient rho (ρ). Modelled associations were examined using logistic regression (exploratory, parsimonious models). Associations were explored using logistic regression. Significance was set at α = 0.05 with 95% confidence intervals (CI).

Results

A total of 62 outpatients were included: mean age 37.5 years, standard deviation (SD) 14.4; 37 (59.7%) were men. Poor adherence occurred in 28 patients (45.2%; 95% CI 33.4-57.5). Mean scores were 5.89 (SD 3.04) for MARS, 2.52 (SD 5.37) for DAI, 5.80 (SD 2.86) for BIS, and 20.63 (SD 20.84) for GASS. Of the enrolled patients, 39 (62.9%) had a positive treatment attitude, nine (14.5%) had good insight, and 12 (19.4%) had a severe side-effect burden. The MARS correlated with DAI (ρ=0.787; p<0.001) and BIS-8 (ρ=0.616; p<0.001) and inversely with GASS (ρ=-0.660; p<0.001). In the multivariable extended model, higher DAI (adjusted odds ratio (aOR) 0.54 per 1-point increase; 95% CI 0.32-0.93; p=0.025) and higher BIS-8 (aOR 0.39 per 1-point increase; 95% CI 0.17-0.87; p=0.022) were independently associated with lower odds of poor adherence; GASS was not independently associated (aOR 1.30; 95% CI 0.96-1.78; p=0.092).

Conclusion

Nearly half of the outpatients receiving antipsychotics in this Kinshasa pilot study reported poor adherence. Treatment attitude and insight, both modifiable, showed independent associations with poor adherence, while adverse effects were strongly correlated with adherence but were not independently associated after adjustment. These exploratory findings suggest that structured side-effect monitoring and shared decision-making may be relevant components to evaluate in future adherence-oriented interventions in constrained settings.