Background/Purpose: We hypothesized that mechanical compression of the knee in rheumatoid arthritis (RA) and osteoarthritis (OA) would mobilize occult extractable fluid and improve arthrocentesis success.

Methods: 186 consecutive knees with grade II-III OA and 67 knees with RA were included. Conventional arthrocentesis was performed and success and volume (milliliters) determined; the needle was left intraarticularly, and mechanical compression was applied with an elastomeric knee brace. Arthrocentesis was then resumed until fluid return ceased. Fluid was characterized as to volume and cell counts.

Results: In the RA knee mechanical compression decreased failed diagnostic arthrocentesis from 56.7% (38/67) to 26.9% (18/67) (-47.4%, p =0.003), and increased absolute arthrocentesis yield from 4.7±10.3 ml to 9.8±9.8 ml (108% increase, 95% CI’ -8.5 <-5.1< -1.7 p=0.0038).  Total extractable fluid yield was 96% greater in RA (9.8±9.8 ml) than OA (5.0±9.4 ml, p = 0.0008), and occult extractable fluid was 77% greater in RA than OA (RA: 5.3±8.7 ml, OA 3.0±5.5 ml, p=0.046).  Large effusions (≥ 5 ml) in RA were associated with an increase in neutrophils in synovial fluid p=0.04) but not radiologic arthritis grade (p=0.87). In contrast, in the OA knee large effusions were associated with more severe arthritis grade (p=0.0001) but not synovial fluid neutrophil count (p=0.96). 

Conclusion: Mechanical compression improves the success of diagnostic and therapeutic knee arthrocentesis in both RA and OA.  Large effusions in RA are associated with inflammatory synovial fluid but not arthritis grade; in contrast, large effusions in OA are associated with more severe arthritis grades.