[Colloids vs. crystalloids in fluid resuscitation for septic shock: a meta-analysis]

To systematically review the effects of colloids compared with crystalloids in fluid resuscitation for septic shock.

The PubMed, Embase, Cochrane Library, CBM and CNKI were searched. All randomized controlled trials (RCTs) of adult patients requiring fluid resuscitation comparing colloids vs. crystalloids were retrieved. Study selection and assessment, data collection and analyses were undertaken. Meta-analysis was done using the Cochrane Collaboration RevMan 5.0 software.

Nine RCTs, involving a total of 1 998 patients were included in the review. No differences were found between colloid and crystalloid resuscitation with respect to 28-day mortality [relative risk (RR) was 0.92, 95% confidence interval (95% CI) was 0.80-1.05, P=0.21], mean arterial pressure [MAP, standardized mean difference (SMD) was 1.04, 95% CI was -0.44-2.51, P=0.17] and heart rate (HR, SMD was 0.51, 95% CI was -5.84-6.86, P=0.87). But a significant difference was observed in cardiac index (CI, SMD was 0.39, 95% CI was 0.17-0.62, P=0.000 5) between two groups. Subgroup analysis suggested a statistically significant difference in CI when albumin was used in resuscitation (SMD was 0.46, 95% CI was 0.16-0.75, P=0.002). But synthetic colloids did not have such effect (P=0.07).

Colloids have been shown to improve CI significantly, and no differences were found between two groups with respect to 28-day mortality, MAP, HR.