Double-counting of populations in evidence synthesis in public health: a call for awareness and future methodological development