Colloids or crystalloids for fluid replacement in critically ill

Using colloids compared to crystalloids for fluid replacement probably makes little or no difference to the number of critically ill people who die, according to a study published in Cochrane. Whether gelatin or crystalloids are used for the fluid replacement makes little or no difference to the mortality rate.

Trauma, burns or serious conditions or infections such as sepsis may cause heavy loss of blood and such critically ill patients require additional fluids urgently to prevent dehydration or kidney failure.

Often intravenously (via a tube straight into the blood) fluid replacement such as colloids and crystalloids are used. Crystalloids are low-cost salt solutions (saline) with small molecules, which can move around easily when injected into the body. On the other hand, colloids can be artificial (e.g. starches, dextrans, or gelatins), or naturally occurring (e.g. albumin or fresh frozen plasma (FFP)), and have bigger molecules, which stays in the blood for a longer duration before passing to other parts of the body. Colloids are more expensive than crystalloids.

Lewis SR and his associates conducted a study to ascertain whether colloids are better than crystalloids at reducing death, need for blood transfusion or need for renal replacement therapy (filtering the blood, with or without dialysis machines, if kidneys fail) when given to critically ill people who need fluid replacement.

The investigators searched the medical literature and identified 69 relevant studies with 30,020 critically ill participants who were given fluid replacement in a hospital or in an emergency out-of-hospital setting. Studies compared colloids (starches; dextrans; gelatins; or albumin or FFP) with crystalloids.

Key study findings:

• Moderate-certainty evidence was found which showed that using colloids (starches; dextrans; or albumin or FFP) compared to crystalloids for fluid replacement probably makes little or no difference to the number of critically ill people who die within 30 or 90 days, or by the end of study follow-up. There was low-certainty evidence that using gelatins or crystalloids may make little or no difference to the number of deaths within each of these time points.


• Moderate-certainty evidence was found which showed that using starches probably slightly increases the need for blood transfusion. However, the authors are uncertain whether using other types of colloids, compared to crystalloids, makes a difference to whether people need a blood transfusion because the certainty of the evidence is very low.


• Moderate-certainty evidence was found which showed that using starches for fluid replacement probably slightly increases the need for renal replacement therapy. Using albumin or FFP compared to crystalloids may make little or no difference to the need for renal replacement therapy. One study comparing gelatins did not report results for renal replacement therapy according to the type of fluid given, and no studies comparing dextrans assessed renal replacement therapy.

The authors concluded that using starches, dextrans, albumin or FFP (moderate-certainty evidence), or gelatins (low-certainty evidence), versus crystalloids probably made little or no difference to death of the people. Starches probably slightly increased the need for blood transfusion and RRT (moderate-certainty evidence), and albumin or FFP may make little or no difference to the need for renal replacement therapy (low-certainty evidence).

For reference log on to https://www.cochrane.org/CD000567/INJ_colloids-or-crystalloids-fluid-replacement-critically-people