Meta-analysis of hazard ratio (HR) is available on Revman 5 by using generic inverse variance method.
However, if some of the studies presented HR using treatment A as reference while some using treatment B as reference, then how could I deal with it ?
HR could be theoretically inverted to change the different reference group, but what about the 95% confidence interval ?
Thanks.
However, if some of the studies presented HR using treatment A as reference while some using treatment B as reference, then how could I deal with it ?
HR could be theoretically inverted to change the different reference group, but what about the 95% confidence interval ?
Thanks.