The risk of venous thromboembolism after a red blood cell transfusion is 80% higher if administered through a multi-lumen peripherally-inserted central catheter compared to a peripheral intravenous line, a large study reports.

“Consideration of a peripheral intravenous catheter for red blood cell delivery when possible, or careful monitoring for thrombosis, in patients transfused through multi-lumen PICCs is warranted,” the authors wrote in Lancet Haematology.

About one in 13 patients (7.7%) receiving red cells through a multi-lumen (PICC) had a DVT (of which two-thirds were in the arms) or pulmonary embolism.

The hazard ratio for multi-lumen peripherally-inserted central catheter PICCs was 1.96 compared to patients with a PICC receiving no transfusions and 1.79 for those transfused through a peripheral intravenous line.

VTE risk was similar in those transfused through a single-lumen PICC or central venous catheter as for a peripheral intravenous line.

The findings were based on 10,604 patients with PICCs inserted for any reason in 47 hospitals. Overall, 5% had a VTE regardless of transfusion status.

“When gauge, lumens, and catheter length were simultaneously assessed, only the number of lumens was a significant predictor for thrombosis,” the researchers said.

“These findings suggest that the PICC itself might not impart venous thromboembolism risk or red blood cells alone; rather, red blood cell transfusion through a multi-lumen PICC seems to confer this risk.”

Co-administration of other treatments around the time of transfusion might have accounted for some of the risk.

Fluids and medications that could interact with red cells to increase the risk of VTE, when given through a multi-lumen PICC, included parenteral nutrition containing calcium, dextrose solutions that caused hyperglycaemia, and vancomycin.

“Whether the inflammatory response during an active infection or a specific antibiotic delivered in proximity to transfusion act as independent thrombotic contributors is an important issue that needs additional study,” the researchers said.

They emphasised that the study was not randomised. Although the analysis accounted for differences in the severity of disease, other differences might have remained. For example, some patients might not have received red cells through a peripheral intravenous catheter because of difficult venous access or poor physiological status.