Measuring health system responsiveness at facility level in Ethiopia: performance, correlates and implications

BACKGROUND: Health system responsiveness measures (HSR) the non-health aspect of care relating to the environment and the way healthcare is provided to clients. The study measured the HSR performance and correlates of HIV/AIDS treatment and care services in the Wolaita Zone of Ethiopia. METHODS: A cross-sectional survey across seven responsiveness domains (attention, autonomy, amenities of care, choice, communication, confidentiality and respect) was conducted on 492 people using pre-ART and ART care. The Likert scale categories were allocated percentages for analysis, being classified as unacceptable (Fail) and acceptable (Good and Very Good) performance. RESULTS: Of the 452 (91.9%) participants, 205 (45.4%) and 247 (54.6%) were from health centers and a hospital respectively. 375 (83.0%) and 77 (17.0%) were on ART and pre-ART care respectively. A range of response classifications was reported for each domain, with Fail performance being higher for choice (48.4%), attention (45.5%) and autonomy (22.7%) domains. Communication (64.2%), amenities (61.4%), attention (51.4%) and confidentiality (50.1%) domains had higher scores in the ‘Good’ performance category. On the other hand, ‘only respect (54.0%) domain had higher score in the ‘Very Good’ performance category while attention (3.1%), amenities (4.7%) and choice (12.4%) domains had very low scores. Respect (5.1%), confidentiality (7.6%) and communication (14.7%) showed low proportion in the Fail performance. 10.4 and 6.9% of the responsiveness percent score (RPS) were in ‘Fail’ and Very Good categories respectively while the rest (82.7%) were in Good performance category. In the multivariate analysis, a unit increase in the perceived quality of care, satisfaction with the services and financial fairness scores respectively resulted in 0.27% (p < 0.001), 0.48% (p < 0.001) and 0.48% (p < 0.001) increase in the RPS. On the contrary, visiting traditional medicine practitioner before formal HIV care was associated with 2.1% decrease in the RPS. CONCLUSION: The health facilities performed low on the autonomy, choice, attention and amenities domains while the overall RPS masked the weaknesses and strengths and showed an overall good performance. The domain specific responsiveness scores are better ways of measuring responsiveness. Improving quality of care, client satisfaction and financial fairness will be important interventions to improve responsiveness performance. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12913-017-2224-1) contains supplementary material, which is available to authorized users.