Psychometric Evaluation of a Kinyarwanda Intimate Partner Violence Instrument among Rwandan Women

In the East African country of Rwanda, 1 in 3 women experience a form of gender-based violence throughout their life span. Currently, there is no established clinical screening tool, nor standardized screening process in place within the healthcare system. Although many valid clinical GBV screening tools exist for use within different settings, the Assessment Screen to Identify Survivors Toolkit for Gender Based Violence (ASIST-GBV) was recently created for use in low resource and humanitarian areas, including post-conflict areas associated with refugee and internally displaced peoples. The ASIST-GBV has been previously tested and reported as a valid tool in similar contexts to Rwanda but had not been translated to Kinyarwanda (the indigenous language) or tested there. Thus, the objectives of this doctoral dissertation were to translate and adapt the ASIST-GBV to the Rwandan context, establish the psychometric properties of the new Kinyarwanda ASIST-GBV, and to identify factors associated with GBV from among women screened using the Kinyarwanda ASIST-GBV. Cognitive interviewing of 20 Kinyarwanda speaking Rwandan women was conducted to establish contextual validation of the translated version. This modified Kinyarwanda version of the ASIST-GBV was administered to a sample of two hundred women at a large public hospital in central Rwanda. Item response theory (IRT) testing provided evidence of the tool’s reliability and validity. Multivariate logistic regression modeling was fit to the data and five risk factors associated with GBV in the best fit model are reported, along with the probabilities of experiencing GBV for each factor among these women. These findings reveal that the Kinyarwanda ASIST-GBV is both valid and functional in screening women in a healthcare setting, while also confirming a high rate of GBV occurring within the sample. The risk factors are useful in informing collaboration with the Ministry of Health to create policies for using the screening tool appropriately to fill the existing gap between GBV survivors and available resources. Further modifications towards a shortened version of the Kinyarwanda ASIST-GBV should be tested for efficient use in the clinical setting