Is chloride in intravenous fluids nephrotoxic?

JAMA2012;308:1566-72OpenUrlCrossRefPubMedWeb of Science

When one intensive care unit in Australia reduced the chloride content of permitted intravenous fluids, patients’ renal function improved significantly. The incidence of acute kidney injury and use of renal replacement therapy fell. In addition, serum creatinine concentrations rose less sharply than they had during a control period before the switch (mean increase 14.8 μmol/L (95% CI 9.8 to 19.9) v 22.6 μmol/L (17.5 to 27.7); P=0.03).

The lower concentration of chloride in intravenous fluids could have been responsible, say the authors, although the new fluids also contained more anion buffers such as lactate, more albumin, more potassium, and less sodium. Staff also stopped using a 4% gelatine solution during the six months of the intervention.

It is impossible to isolate the impact of lower chloride concentrations from all the other changes that went with it, says a linked editorial (p 1583). Before and after studies can’t establish cause and effect either. But this carefully done pilot does pave the way for proper trials of intravenous fluids, which are long overdue.

Even ubiquitous fluids such as normal saline contain supraphysiological doses of chloride ion. We simply don’t know what these fluids do to the human kidney. Animal studies hint that chloride is a renal vasoconstrictor that reduces perfusion and impairs glomerular filtration.

Trials of chloride sparing strategies would be difficult but not impossible, says the editorial. A cluster design that randomises intensive care units, not patients, may be the best way forward.