Growing international concern about the need for improved health systems in Africa has catalyzed an expansion of the health systems literature. This review applies a bibliometric procedure to analyze the acceleration of scientific writing on this theme. We focus on research published during the Millennium Development Goal (MDG) era between 1990 and 2014, reporting findings from a systematic review of a database comprised of 17,655 articles about health systems themes from sub-Saharan African countries or subregions. Using bibliometric tools for co-word textual analysis, we analyzed the incidence and associations of keywords and phrases to generate and visualize topical foci on health systems as clusters of themes, much in the manner that astronomers represent groupings of stars as galaxies of celestial entities. The association of keywords defines their relative position, with the size of images weighted by the relative frequency of terms. Sets of associated keywords are arrayed as stars that cluster as "galaxies" of concepts in the knowledge universe represented by health systems research from sub-Saharan Africa. Results show that health systems research is dominated by literature on diseases and categorical systems research topics, rather than on systems science that cuts across diseases or specific systemic themes. Systems research is highly developed in South Africa but relatively uncommon elsewhere in the region. "Black holes" are identified by searching for terms in our keyword library related to terms in widely cited reviews of health systems. Results identify several themes that are unexpectedly uncommon in the country-specific health systems literature. This includes research on the processes of achieving systems change, the health impact of systems strengthening, processes that explain the systems determinants of health outcomes, or systematic study of organizational dysfunction and ways to improve system performance. Research quantifying the relationship of governance indicators to health systems strengthening is nearly absent from the literature. Long-term experimental studies and statistically rigorous research on cross-cutting themes of health systems strengthening are rare. Studies of organizational malaise or corruption are virtually absent. Trend analysis shows the emergence of organizational research on specific priority diseases, such as on HIV/AIDS, malaria, and tuberculosis, but portrays a lack of focus on integrated systems research on the general burden of disease. If health systems in Africa are to be strengthened, then organizational change research must be a more concerted focus in the future than has been the case in the past.
© Phillips et al.