Validity, Reliability, and Diagnostic Cut-off of the Kinyarwandan Version of the Hamilton Depression Rating Scale in Rwanda

Introduction: In Rwanda, major depressive disorder affects 11.9% of the population and up to 35% of genocide survivors. Mental health services remain underutilized due to stigma and lack of awareness. Increasing the ability and capacity to diagnose and treat mental disorders is considered important to close this gap. We describe the translation, validity, and reliability assessment of the Hamilton Depression Rating Scale (HDRS) as a diagnostic tool for moderate to severe depression in Rwanda. Methods: The HDRS-21 was translated by a multi-group taskforce. We validated the translation against expert assessment in a comparative study on a sample of patients living with depression and of healthy volunteers. Psychometric properties, namely internal structure, reliability, and external validity were assessed using confirmatory factor analysis, three reliability calculations, and correlation analysis, respectively. Maximized Youden’s index was used for determining diagnostic cut-off. Results: The translated version demonstrated a kappa of 0.93. We enrolled 105 healthy volunteers and 105 patients with confirmed mild to severe depression. In the confirmatory factor analysis, HDRS had good factor loadings of 0.32–0.80. Reliability coefficients above 0.92 indicated strong internal consistency. External validity was shown by good sensitivity (0.95) and specificity (0.94) to differentiate depression from absence of depression. At a cut-off point of 17 for the diagnosis of depression, sensitivity and specificity were both 0.95 relative to gold standard. Conclusion: The validated HDRS in Kinyarwanda with diagnostic cut-off provides mental healthcare staff with an accurate tool to diagnose moderate to severe depression, enabling closure of the diagnosis and treatment gap.