Rogers MAM, Blumberg N, Bernstein SJ, Flanders SA, Chopra V. Association between delivery methods for red blood cell transfusion and the risk of venous thromboembolism: a longitudinal study.
LANCET HAEMATOLOGY 2016;
3:e563-e571. [PMID:
27818170 DOI:
10.1016/s2352-3026(16)30132-6]
[Citation(s) in RCA: 4] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Received: 08/23/2016] [Revised: 09/09/2016] [Accepted: 09/13/2016] [Indexed: 10/20/2022]
Abstract
BACKGROUND
Mechanisms of red blood cell delivery and their contribution to the incidence of venous thromboembolism are not well understood in the clinical setting. We assessed whether red blood cell transfusion through peripherally inserted central catheters (PICCs) affects the risk of venous thromboembolism compared with transfusion through non-PICC devices.
METHODS
We implemented a prospective study between Jan 1, 2013, and Sept 12, 2015, in patients (age ≥18 years) admitted to a general medicine ward or intensive care unit who received a PICC for any reason during clinical care in 47 hospitals in Michigan, USA, with a maximum follow-up of 70 days. The exposure of interest was route of red blood cell transfusion. The primary outcome was symptomatic, radiographically confirmed, deep-vein thrombosis in the arm or leg or pulmonary embolism. We used Cox proportional hazards regression for analyses.
FINDINGS
Venous thromboembolism developed in 482 (5%) of 10 604 patients with PICCs. Of 788 patients who received a red blood cell transfusion through a multi-lumen PICC, 61 had venous thromboembolism. The adjusted hazard ratio (HR) for venous thromboembolism in all patients whose transfusions were administered through a multi-lumen PICC was 1·96 (95% CI 1·47-2·61; p<0·0001) compared with patients not receiving a transfusion, and was 1·79 (1·09-2·95; p=0·022) compared with patients transfused through a peripheral intravenous line. Compared with delivery through a peripheral intravenous line, venous thromboembolism risk was not elevated if transfusions were delivered through a single-lumen PICC (HR 0·98, 95% CI 0·44-2·14; p=0·95) or central venous catheter (1·50, 0·77-2·91; p=0·23). For every red blood cell unit transfused through a PICC, there was a significantly increased risk of venous thromboembolism (adjusted HR 1·24, 95% CI 1·01-1·52; p=0·037). Patients who received a transfusion through a PICC in the left arm were significantly more likely to develop a deep-vein thrombosis in the ipsilateral arm compared with the contralateral side (HR 23·44, 95% CI 2·96-185·83; p=0·0028). Similarly, patients transfused through a right-sided PICC were more likely to develop deep-vein thrombosis in the ipsilateral arm (HR 3·37, 95% CI 1·02-11·14; p=0·047).
INTERPRETATION
Red blood cell delivery through a multi-lumen PICC is associated with a greater risk of thrombosis than transfusion through a peripheral intravenous catheter. Careful monitoring for venous thromboembolism when transfusing red blood cells through multi-lumen PICCs seems necessary.
FUNDING
Blue Cross Blue Shield of Michigan and Blue Care Network, as part of the BCBSM Value Partnerships program.
Collapse