Analytical Study of Predictive Analytics Adoption Barriers in Nigerian Public Hospitals: A Multi-Level Empirical Investigation

Predictive analytics in hospital settings can improve patient outcome prediction, resource allocation, and disease surveillance, but adoption barriers in Nigerian public hospitals operate at individual, organisational, and technological levels that have not been simultaneously examined in an empirical multi-level study. This study analytically investigates multi-level barriers to predictive analytics adoption in Nigerian public hospitals across three geopolitical zones. A multi-level cross-sectional survey was applied to 360 clinical and administrative staff across nine public hospitals in Abuja, Ibadan, and Enugu, selected through stratified cluster sampling. A multilevel model testing individual-level barriers (digital skills, attitude, awareness), organisational-level barriers (leadership support, data governance, IT infrastructure), and technological barriers (interoperability, EHR data quality) was estimated using hierarchical linear modelling in R. Available predictive analytics adoption literature from Sub-Saharan African hospital contexts identifies EHR data incompleteness and absence of data science talent as the most constraining adoption barriers. The Multi-Level Theory of Innovation Adoption and the Technology-Organisation-Environment Framework provide the analytical basis. Findings confirm that organisational-level barriers (specifically leadership digital literacy and data governance quality) explain 58 percent of the variance in adoption intention after controlling for individual and technical factors. This study fills a gap in multi-level, empirical predictive analytics adoption research from Nigerian public hospital contexts. Recommendations address health ministry digital leadership development and national EHR data quality standards. Keywords: predictive analytics, adoption barriers, Nigerian public hospitals, multi-level analysis, health informatics.