highly significantp < 0.000
Between 1981 and 2015 there was a highly significant upwards trend in the number of studies analysing NCD media debates published per year (Poisson regression coefficient 0.113, p < 0.000) (Fig. 1 ).
Between 1981 and 2015 there was a highly significant upwards trend in the number of studies analysing NCD media debates published per year (Poisson regression coefficient 0.113, p < 0.000) (Fig. 1 ).