The author concluded that hydroxyethyl starch increases the risk of acute renal failure in patients with sepsis and may also increase mortality, especially at higher doses. The review was well-conducted in many respects and the data were consistent and based on randomised evidence. However, a degree of caution is required with interpretation as study quality was not assessed systematically.

To assess the effectiveness and safety of hydroxyethyl starch (HES) for fluid management in patients with sepsis.

MEDLINE, EMBASE and the Cochrane Library were searched. Search terms were provided. Speciality journals and reference lists of relevant articles were handsearched. The search was not limited by language or publication status.

Randomised controlled trials (RCTs) reporting any outcomes related to the use of any type of HES among patients with sepsis were eligible for inclusion.

The studies in the review included patients with severe sepsis, septic shock, postoperative sepsis and sepsis with hypovolaemia. Patients in some studies were mechanically ventilated and/or haemodynamically controlled. HES with a molecular weight of 200 kilodaltons (kDa) and molar substitution of 0.5 (HES 200/0.5) was evaluated in most of the studies. Other types of HES used were 200/0.62, 130/0.4 and 450/0.7. The median daily dose, cumulative dose and duration of HES were 12.6 millilitres per kilogram (mL/kg) (interquartile range (IQR) 11.0 to 13.7 mL/kg), 49.8 mL/kg (IQR 22.6 to 63.0 mL/kg) and five days (IQR one to five days). Control fluids were 5 per cent or 20 per cent albumin, gelatin or crystalloid. Most RCTs had follow up of one hour to five days (acute studies), but one reported outcomes at 34 and another at 90 days in patients with severe sepsis or septic shock. The outcomes most commonly reported were cardiorespiratory, haemodynamic and coagulation measures. Other outcomes in acute studies were extravascular lung water, gastric mucosal acidosis and circulating soluble adhesion molecules. The two longer-term studies reported acute renal failure (ARF), morbidity and mortality.

The author did not state how the papers were selected for the review or how many reviewers performed the selection.

The author did not state that he assessed validity.

Descriptive data were extracted, with odds ratios and/or p values for significant changes from baseline or significant differences between the groups. Additional data were requested from study investigators. The author did not state how many reviewers performed the data extraction.

Studies were combined in a narrative synthesis accompanied by summary tables. Results were grouped by outcome and length of follow up.

Twelve RCTs (n=1,062) were included, all published.

Acute effects
HES 130/0.4 and HES 200/0.5 significantly improved haemodynamic and cardiorespiratory variables (p<0.05); 20 per cent albumin did not (three RCTs). HES 200/0.5 significantly improved Acute Physiology and Chronic Health Evaluation II score (one RCT), improved gastric intramucosal pH (pHi) (one RCT) and avoided a decline in pHi (one RCT) (all p<0.05); 20 per cent albumin did not.

No differences were detected in haemodynamic and cardiorespiratory variables when HES 200/0.5 and gelatin were compared (one RCT). Gelatin significantly decreased gastric mucosal arterial gradient (p<0.0005) and raised pHi (p<0.001); HES 200.0/62 did not (one RCT).

HES 450/0.7 significantly impaired coagulation (prolonged partial thromboplastin time and decreased platelet count, p=0.01); 5 per cent albumin did not (one RCT). One RCT reported no difference in coagulation and platelet count between HES 200/0.5 and 20 per cent albumin. HES 200/0.5 was significantly associated with diminished factor VIII levels compared with 5 per cent albumin (p=0.05, one RCT) and impaired coagulation compared to crystalloid (p<0.001, one RCT).

Longer term effects
HES 200/0.62 was associated with a significantly higher incidence of renal failure over 34-day follow up than gelatin (OR 2.57, 95% CI: 1.13, 5.83, p=0.018, one RCT, n=129). In a second RCT (n=537), over 90-day follow up HES 200/0.5 was significantly associated with impaired renal function (p=0.02) and a higher incidence of ARF (OR 1.81, 95% CI: 1.22, 2.71, p=0.002) and use of renal replacement therapy (RRT) (OR 1.95, 95% CI: 1.23, 2.98, p=0.001) than crystalloid. RRT use significantly correlated with cumulative HES dose (p<0.001). There was a trend to higher 90-day mortality in the HES group (OR 1.35, 95% CI: 0.94, 1.95, p=0.11). Mortality was significantly higher among patients receiving a higher dose of HES (more than 22 mLs/kg for at least one day) than among those receiving a lower dose (p<0.001).

HES infusion increases the risk of ARF in patients with sepsis and may increase the risk of mortality, especially at higher doses.

The review objectives and inclusion criteria were clear. Relevant sources were searched without language restriction. It was unclear whether specific efforts were made to identify unpublished studies and potential publication bias was not discussed. Search dates were not stated. It was unclear whether steps were taken to minimise bias and error by having more than one reviewer make decisions on study selection and data extraction. It does not appear that study validity was systematically assessed. The authors highlighted the clinical and methodological heterogeneity between the studies and appropriately decided not to pool them statistically. The narrative synthesis was concise and clear, except that it was hard to determine in some cases whether reported results represented differences between treatments or changes from baseline for individual treatment arms. Precedence was given to the larger long-term studies. The text was supported by summary tables with detailed data. The review was well-conducted in many respects and the data were consistent and based on randomised evidence. However, a degree of caution is required with interpretation, mainly because study quality was not assessed systematically and study methods were not reported fully.

Practice: the author stated that HES 200/0.5 and 200/0.62 should not be used for patients with sepsis.

Research: the author stated that clinical trials are needed to test the safety and effectiveness of HES solutions other than 200/0.5 and 200/0.62 in patients with sepsis. HES and albumin should be compared directly in a well-powered study.

Not stated.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.