Letters to the Editor

In a recent commentary in the Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine Chappell and Jacob provide their view of the three largest, randomised clinical trials of hydroxyethyl starch in critically ill patients [1]. The authors’ interpretation of these trials and their view of the safety with HES are far from those of independent authors conducting high-quality systematic reviews [2–7], the US Food and Drug Administration [8] and the European Medicines Agency’s Pharmacovigilance Risk Assessment Committee [9].

Unfortunately, several claims regarding the Scandinavian Starch for Severe Sepsis/Septic Shock (6S) trial rely on misreading of the main publication of this trial [10] and are not supported by the trial data.

The 6S trial aimed at testing the effect of HES versus Ringer’s acetate as used in clinical practice in a high number of different intensive care units. The 24 hour window from diagnosis of severe sepsis resembles clinical practice, where fluid resuscitation in septic patients is initiated without waiting for the results of new blood samples to confirm the diagnosis of severe sepsis.

Forty-nine percent, not 60%, received colloids in the 24 hours prior to randomisation, but the clinician judged that fluid resuscitation was still needed as this was an inclusion criterion. Such judgement is complex, but may have been based on poor peripheral perfusion, oliguria, increased vasopressor need etc. Thus, the claim that the patients were hemodynamically stabilized at randomisation is unjustified. Also, CVP and ScvO2 were only registered in a minority of patients and either value may have been abnormal or considered abnormal in combination with other parameters by the clinicians assessing these patients. Neither we nor Chappell and Jacob can know how individual clinicians judged patients.

Most trial fluid was given in the first 38 hours (day 1 lasted median 14 hours) and the cumulative total dose of HES was less than the labelled maximum daily dose of HES (Table 1). Thus, the statement that HES was given in high amounts and over a prolonged period of time is not true for the majority of patients.

Table 1 Doses of trial fluid
Full size table

During the trial, 38 of 400 patients (10%) in the Ringer’s group received synthetic colloid (almost exclusively HES). Sixty-five patients (16%) in the Ringer’s group received albumin for other indications than volume expansion. Overall, 92 (23%) patients in the Ringer’s group received a colloid during the trial, and not 32% as stated by Chappell and Jacob. As the use of albumin was similar in both intervention groups, this unlikely affected the trial results.

Most patients discontinued trial fluid due to severe bleeding or renal replacement therapy (RRT) as stipulated in the protocol [11]. To exclude these patients from the primary analysis would mask the harmful side effects of HES. However, per-protocol analyses excluding patients, who discontinued trial fluid for other reasons, showed comparable estimates of increased risk of death supporting the primary results.

Kidney failure without RRT was not a ‘clear’ or ‘absolute’ contraindication for HES according to the steering committee and scientific advisors of the trial, B Braun Melsungen AG and the Medicines Agencies in Denmark, Norway, Finland and Iceland, who all approved the protocol and the inclusion of these patients. It is important to note that the increased risk of death with HES was independent of kidney failure at inclusion in the pre-planned subgroup analysis [10].

Data on mechanical ventilation and hospital length of stay can be found in the main publication of the 6S trial as days alive without ventilation and days alive and out of hospital, respectively. We find these endpoints more valid as they are less affected by survival bias.

Taken together, the 6S trial showed that HES was harmful compared to Ringer’s acetate in patients with severe sepsis when used according to clinical practice and in alignment with the recommendations by the manufactures and authorities. Several hypotheses may be made regarding less harmful ways of using HES in critically ill patients, but clinicians, guideline committee members and authorities need to acknowledge that such safer ways have not yet been identified.