Reviewed by Josefien Van Olmen
This commentary in the American Journal of Public Health is not wrong, but not complete. Perhaps it reflects a different phase in the discussions of public health and health systems at either side of the Atlantic. In this commentary, the authors call for the appreciation of other methods than the RCT in public health research, such as surveys, surveillance and mathematical modelling. Indeed, these are methods to show associations and possibly say something about the effect / impact of interventions and in that sense, they answer a need of policy makers. But the authors bypass the discussion about the complexity of most public health interventions and the fact that impact is thus highly dependant on the context. The Health Systems Research series in Plos widens the scope of questions and methods in public health research much more.