Effects of Post-Election Violence on HIV patients care and Treatment in Selected Districts in Central, Nyanza and Rift Valley Provinces, Kenya, 2008.

Abstract

Kenya, like other African countries, has been affected significantly by the HIV/AIDS epidemic. Kenya has been a leader in rapidly expanding access to HIV prevention, care and treatment, and great effort has been invested in developing national to facility-level systems to deliver prevention, care, and treatment services for HIV patients. In December 30, 2007; civil unrest broke out in several parts of the country after the presidential election results were announced. People fled their homes and health care services were disrupted. The effects of the post-election violence on HIV infected individuals and those on antiretroviral therapy (ART) are not clearly understood. A cross-sectional study was carried out to determine the effects of post-election violence on HIV patients care and treatment in selected districts in Central, Nyanza, and Rift Valley provinces, Kenya. A total of 35 health facilities from 18 districts were selected using person proportion to size method. HIV patients were interviewed and facility records reviewed to asses the effects of the post-election violence. A standardized questionnaire was administered to the study participants. Odds Ratio (OR) was used to assess the strength of association between various risk factors and missing ART and chi square to test for statistical significance.A total of 947 participants were interviewed. On analysis of factors that could have lead to patients missing treatment, there was significant increased risk of missing ART therapy among those who reported to have moved into different districts since December 30, 2007 (OR=2.6, p=0.002), unable to access health XVIcare during the post-election violence period (OR=9.5, p<0.0001), having experienced physical violence personally (OR=2.7, p=0.0055) and having witnessed acts of physical violence (OR=2.0, p=0.02). For participants who had remained at the hospital of normalcare they were less likely to miss ART (OR=0.3, p=0.002). A Logistic regression model of factors that lead to patients missing treatment showed two factors that remained independent. There was an increased risk of missing ART among those participants who were unable to access routine health care (OR=5.25, p<0.0001) and care when ill (OR=3.81 p<0.0001). The post-election violence had an effect on patients who were on ART. Inability to access health care and experiencing physical violence were important factors leading to participants missing ART treatment. This may have implications on ARV drug resistance and further research is needed.