Effects of hydroxyethyl starch in subgroups of patients with severe sepsis: exploratory post-hoc analyses of a randomised trial

It has been speculated that certain subgroups of sepsis patients may benefit from treatment with hydroxyethyl starch (HES) 130/0.42, specifically in the earlier resuscitation of patients with more severely impaired circulation. This was a post-hoc, subgroup analysis of all 798 patients with severe sepsis randomised in the 6S trial according to time from ICU admission to randomisation, surgery and fluids given prior to randomisation and markers of shock at randomisation. Intervention effects estimated as risk ratios were analysed between the HES versus Ringer's acetate groups to detect subgroup heterogeneity of the effects on 90-day mortality. Multiple logistic regression was used to adjust for risk factors. Most baseline characteristics were comparable between the HES and Ringer's acetate groups in the different subgroups. There was no heterogeneity in the intervention effect on 90-day mortality in the following subgroups: randomisation earlier than 4 h after ICU admission versus later (test of interaction P = 0.85), surgery versus no surgery (P = 0.42), colloids given versus not given (P = 0.57), < 2 l of crystalloids given prior to randomisation vs. > 2 l (P = 0.88) or plasma lactate > 4 mmol/l versus < 4 mmol/l (P = 0.54), hypotension versus no hypotension (P = 0.32) or use of vasopressor or inotropic agents at randomisation versus no use (P = 0.10). The increased 90-day mortality observed in patients with severe sepsis resuscitated with HES 130/0.42 did not appear to depend on time course, surgery or fluids given prior to randomisation or on markers of shock at randomisation. As the analyses were planned post hoc and their power is reduced, the results should be interpreted with caution.