Fig 1.
Flow chart of study selection.
Fig 2.
Summary of risk of bias for the included randomized controlled trials.
Table 1.
Comparisons of characteristics between highly and poorly cited RCTs.
Fig 3.
Scatter plots of the total citations against the journal impact factor, stratified by publication year.
Fig 4.
Scatterplot matrix describing the correlations among continuous variables.
Participating centers and sample sizes appeared to be closely correlated with each other, thus we only retained sample size in the model.
Fig 5.
Diagnostic plots of multivariable linear regression.
The dependent variable, total citation, was not transformed. The scale-location plot shows a correlation trend between fitted values and square root of the standardized residual, indicating the violation of homoscedasticity. The Q-Q plot also shows the violation of normality.
Table 2.
Linear model built by a stepwise approach, with a square root transformation of the total citation outcome variable.
Fig 6.
Diagnostic plots of multivariable linear regression with the dependent variable total citation transformed by square root.
The relationship between the fitted values and square root of standardized residual was weakened. The Q-Q plot showed that the normal assumption of the dependent variable was well satisfied.
Table 3.
Sensitivity analysis by excluding RCTs with total citations greater than 500 without transforming the outcome variable.