Objectives: Painful osteoarthritis (OA) of the hand is common and a validated US scoring system would be valuable for epidemiologic and therapeutic outcome studies. Ultrasound (US) is increasingly used to assess peripheral joints, though most of the US focus in rheumatic diseases has been on rheumatoid arthritis. We aimed to develop a preliminary US hand OA scoring system, initially focussing on relevant pathological features with potentially high reliability.
Methods: A group of experts in the fields of OA, US and novel tool development agreed on domains and suggested scaling of the items to be used in an US hand OA scoring system. A multi-observer reliability exercise was then performed to evaluate the draft items.
Results: Synovitis (grey-scale and Power Doppler) and osteophytes (representing activity and damage domains) were included and evaluated as the initial components of the scoring system. All 3 features were evaluated for their presence/absence and if present were scored using a 1-3 scale. The reliability exercise demonstrated intra-reader kappa values of 0.444 -1.0, 0.211-1.0 and 0.087-1.0 for grey scale synovitis, power Doppler and osteophytes respectively. Inter-reader reliability kappa values were 0.398, 0.327 and 0.530 grey scale synovitis, power Doppler and osteophytes respectively. Without extensive standardisation, both intra- and inter-reader reliability were moderately good.
Conclusions: The draft scoring system demonstrated substantive to almost perfect percentage exact agreement on the presence/absence of the selected OA features and moderate to substantive percentage exact agreement on semi-quantitative grading. This preliminary process provides a good basis from which to further develop an US outcome tool for hand OA that has potential to be utilised in multi-centre clinical trials.