Evaluation of mhGAP training for primary healthcare workers in Mulanje, Malawi: a quasi-experimental and time series study

There has been a growing global movement championed by the World Health Organization (WHO) to integrate mental health into primary health care as the most effective way of reducing the mental health treatment gap. This study aimed to investigate the impact of WHO Mental Health Gap Action Programme (mhGAP) training and supervision on primary health workers’ knowledge, confidence, attitudes and detection rate of major mental disorders in Mulanje, Malawi.

The study used a quasi-experimental method (single cohort pre- and post-measures) with an interrupted time-series design. A 2 day mhGAP training was delivered to 43 primary healthcare workers (PHWs) working in 18 primary care clinics serving the entire population of Mulanje, Malawi (population 684,107). Modules covered were psychosis, moderate-severe depression, and alcohol & substance use disorders. The PHWs completed pre and post-tests to assess knowledge, confidence and attitudes. Number of diagnosed cases was obtained from clinic registers for 5 months prior to and 7 months following training. Data was analyzed using mean scores, t-test, one-way analysis of variance and linear regression.

The mean knowledge score increased significantly from 11.8 (SD: 0.33) before training to 15.1 (SD: 0.38) immediately after training; t (42) = 7.79, p < 0.01. Similarly, mean knowledge score was significantly higher 6 months post training at 13.9 (SD: 2.52) compared to before; t (42) = 4.57, p < 0.01. The mean confidence score also increased significantly from 39.9 (SD: 7.68) before training to 49.6 (SD: 06.14) immediately after training; t (84) = 8.43, p < 0.01. It was also significantly higher 6 months post training 46.8, (SD: 6.03) compared to before; t (84) = 6.60, p < 0.01. One-way analysis of variance showed no significant difference in mean scores on all four components of the scale used to measure attitudes. A significant positive change in the trend in mental health service utilization after the intervention was demonstrated using a segmented linear regression (β = 2.43 (95% CI 1.02; 3.83) as compared to before (β = − 0.22 (95% CI − 2.67; 2.23) and immediately after (β = 1.63 (95% CI − 7.31; 10.57).

The findings of this study add to the growing evidence for policy makers of the effectiveness of mhGAP training and supervision in a resource-constrained country.