Text this: The impact of adjustment for covariates on meta-analysis of randomised intervention studies for binary outcome