“…These findings are therefore vulnerable to response bias as well as recall bias, when participants were required to make a post-intervention assessment of their pre-intervention skills, knowledge or confidence [65,66]. Several studies were limited by their small sample size [11,16,29,30,34,36,59,61] and only four studies [9,32,43,57] reported calculation of sufficient power to detect a statistically significant difference. Twelve studies (21.4%) had a poor response rate (less than 60%) [9,10,12,14,18,19,22,39,40,46 ,60,62] and four (7.1%) studies [26,29,37,53] did not report response rate.…”