Warfarin, Needle Aspiration, and Injection: What’s the Risk?


Source: Thumboo J, O’Duffy JD. Arthritis Rheum 1998;41: 736-739.

Following the dictum, "first do no harm," clinicians have generally hesitated to do needle aspiration and joint or soft tissue injections in patients on warfarin. Intra-articular hemorrhage or bleeding into the soft tissues which could complicate the procedure was assumed to be a significant, though unquantified risk.

Thumboo and O’Duffy have tried to provide an estimate of the risk of local hemorrhage following needle insertion into a joint or the soft tissues by assessing, in a prospective fashion, the incidence of bleeding in a group of 25 patients on warfarin, all of whom had INRs less than 4.5, were not on heparin, and did not have evidence to suggest overlying skin infection. Nonsteroidal anti-inflammatory drug (NSAID) use was not a contraindication, nor was mild thrombocytopenia. No special precautions were taken, most aspirations were performed after local injection of 2 mL of lidocaine. Eighteen gauge needles were used for aspiration of large joints, such as the knee, 20 gauge needles for aspiration and injection of other sites, except the great toe, where 25 gauge needles were used. The primary outcome, local hemorrhage, was assessed by telephone interview at a mean of five weeks after the procedure. The results were encouraging, with no patient-reported hemorrhages in the 32 procedures performed. The aspirations were diagnostically useful in 53% of instances, with crystal arthritis being confirmed by microscopic examination of the joint fluid, or infection being ruled out. The authors calculate that the absence of any reported hemorrhage in 32 procedures indicates an incidence of local bleeding of 0-10% with 95% confidence.

Comment by Jerry M. Greene, MD

Anticoagulation is a relative contraindication to aspiration or injection of joints and soft tissues. The results above are reassuring. When needle aspiration is necessary to help exclude septic arthritis or bursitis, or when more conservative methods for treating joint or soft tissue inflammation have failed or are contraindicated, injection of the joints or soft tissues can be done with a small risk. Whether the risk is acceptable is a decision that must be made by the patient after being counseled. With Thumboo and O’Duffy’s data, it is possible to give a much better estimate of the risk of local hemorrhage. (Dr. Greene is Instructor in Medicine, Harvard Medical School, Chief, Rheumatology Section, Brockton/W. Roxbury VA Hospital.)