Table 1 Vote-counting rules
From: The effectiveness of champions in implementing innovations in health care: a systematic review
(1)To make conclusions pertaining to champions’ effectiveness at increasing innovation use or outcomes (patient, provider, or system/facility) four or more studies must have evaluated a relationship or correlation between champions and innovation use or the outcome of innovation use |
(2)Champions’ effectiveness at increasing innovation use or outcomes of innovation use were coded as follows:  a.Champions are effective if 60% or more of the studies demonstrated a positive significant relationship between exposure to champions and either innovation use or outcome of innovation use  b.Champions are ineffective if 60% or more of the studies demonstrated a non-significant or significant negative relationship between exposure to champions and either innovation use or outcome of innovation use  c.Champions’ effectiveness is mixed if less than 60% of the studies reported a non-significant/significant relationship between exposure champions and either innovation use or outcome of innovation use |
(3)We applied the same rules as above (rule number 2) to determine whether individual studies demonstrated a significant, non-significant, or mixed relationship between exposure to champions and either innovation use or outcome of innovation use. The analysis was based on percentage of statistical results reported in a study. We performed these evaluations to counteract double counting articles with multiple study outcomes |
(4)When both bivariate and multivariate statistics are reported in a study, we used the more robust multivariate findings in our synthesis |
(5)We assessed categories examined in three or less studies to determine trends in champion effectiveness using the same rules detailed above |