Survey of arthroscopy performed by rheumatologists

To determine the international distribution and practice of arthroscopy performed by rheumatologists and to evaluate proposed guidelines on minimum standards for training in arthroscopy in the context of current clinical practice.

A questionnaire was sent to all rheumatology centres identified as practising arthroscopy, by (i) searching Medline from 1966 to 1999, (ii) searching the abstract books of the annual general meetings of ACR, BSR and EULAR from 1980 to 1999, and (iii) correspondence with all the centres identified.

Thirty-six rheumatology centres were confirmed as performing arthroscopy (24 in Europe, 10 in USA and two in Australia) and 33 (92%) centres completed the questionnaire. Twenty-five (76%) of the 33 centres performing arthroscopy had started to perform it since 1990 and 72 rheumatologists are now trained in arthroscopy. A total of 16532 arthroscopies had been performed (median=220 arthroscopies/centre, range 20-5000); 50.5% of the arthroscopies had a primary clinical indication and 49.5% had a primary research indication. Most centres fulfilled the minimum standards for arthroscopic facilities and the proposed minimum standards in training were acceptable to 76% of respondents. Complication rates were calculated for 15682 arthroscopies where routine follow-up data were available [joint infection, 16 (0.1%); wound infection, 17 (0.1%); haemarthrosis, 141 (0.9%); deep venous thrombosis, 31 (0.2%); neurological damage, 3 (0.02%), thrombophlebitis, 12 (0.08%), other, 8 (0.06%)]. Irrigation volume correlated with wound infection rate (r=0.41, P=0.03) and centres performing cartilage biopsy had a higher rate of haemarthrosis (P=0.007).

The last decade has seen rapid growth in arthroscopy performed by rheumatologists in an out-patient setting under local and regional anaesthesia. Proposed minimum standards for training in rheumatological arthroscopy reflect current practice accurately and are acceptable to the majority of arthroscopists. Complication rates of rheumatological arthroscopy are similar to those reported in the orthopaedic literature.