Faulty risk-of-bias assessment in a meta-analysis of hydroxyethyl starch for non-septic ICU patients: a rebuttal