Mixed colloids and acute kidney injury: a case of selection bias?

In a previous issue of Critical Care, Frenette and colleagues [1] described a retrospective study of 984 cardiac surgery patients receiving mixed colloids, including hydroxyethyl starch (HES) 130/0.4 in 82% or HES 200/ 0.5 in 43%. In 25% or more of the patients, both HES solutions were used. A small minority (16%) received 5% or 25% albumin. An association was observed between albumin exposure and acute kidney injury (AKI). Much essential information is not provided, including what colloids were administered, why, and when. How many patients received 5% albumin? 25% albumin? How often was albumin co-administered with HES 130/0.4? With HES 200/0.5? With both? In what co-administered doses? How much of each colloid was used for extracorporeal circuit priming? For volume expansion? How much albumin was used to correct hypoalbuminemia? What was the temporal relationship between albumin infusion and AKI development?