Assessment of progress towards universal health coverage for people with disabilities in Afghanistan : a multilevel analysis of repeated cross-sectional surveys

Abstract: Background Since 2002, Afghanistan has made much effort to achieve universal health coverage. According to the UN Sustainable Development Goal 3, target eight, the provision of quality care to all must include usually underserved groups, including people with disabilities. We investigated whether a decade of international investment in the Afghan health system has brought quality health care to this group. Methods We used data from two representative household surveys, one done in 2005 and one in 2013, in 13 provinces of Afghanistan, that included questions about activity limitations and functioning difficulties, socioeconomic factors, perceived availability of health care, and experience with coverage of health-care needs. We used multilevel modelling and tests for interaction to investigate factors associated with differences in perception between timepoints and whether village remoteness affected changes in perception. Findings The 2005 survey included 334 people, and the 2013 survey included 961 people. Mean age, employment, and asset levels of participants with disabilities increased slightly between 2005 and 2013, but the level of education decreased. Formal education and higher asset level were associated with improved availability of health care and positive experience with coverage of health-care needs, whereas being employed was only associated with the latter. Perceived availability of health care and positive experience with coverage of health-care needs significantly worsened in 2013 compared with in 2005 (227 [69%] perceived that services were available in 2005 vs 405 [44%] in 2013, p<0·0001; 255 [78%] perceived a positive experience in 2005 vs 410 [45%] in 2013, p<0·0001). Village remoteness increased in 2013 (no connectivity by paved road 186 [57%] in 2005 vs 797 [87%] in 2013, p<0·0001; mean time to reach health-care facility 64·3 min [SD 167·7] vs 84·4 min [107·7], p