The National Institute for Health and Clinical Excellence (NICE) found evidence from 13 studies (one cohort and 12 case series) investigating the role of clinical indicators in the diagnosis of rheumatoid arthritis (RA). All studies were methodologically sound. Some of the studies had an element of prognostic design; they assessed which clinical features were able to predict people who went on to develop RA after at least 1 year of follow up [National Collaborating Centre for Chronic Conditions, 2009].

• The cohort study (n = 474) investigated the clinical features of people with arthritic symptoms who attended an early arthritis clinic, compared with a routine clinic, over a 1 year follow-up period.
• The 12 case series investigated the clinical features of people with early inflammatory arthritis, and in some studies, people were followed up to look at the features of those that went on to develop RA. Studies differed with respect to sample size (ranging from n = 41 to n = 903) and study length (1–8 years).
• NICE concluded that the following clinical features help identify people who are likely to have persistent synovitis and go on to develop RA:
• Ever having had prolonged morning stiffness is more helpful than currently having morning stiffness.
• In a case series of people with suspected RA, the presence of morning stiffness increased the odds of developing RA, compared with not developing RA, nearly 10-fold (odds ratio [OR] 9.4, 95% CI 3 to 28.7, p < 0.001).
• In another case series, doctors in primary care found a history of morning stiffness or stiffness at rest increased the odds of inflammatory arthritis, compared with non-inflammatory arthritis, nearly 13-fold (OR 12.7, 95% CI 3.6 to 45.8, p = 0.0001).
• Both swelling and tenderness in affected joints, particularly small joints.
• In a case series of people with suspected RA, the presence of more than 10 tender and swollen joints increased the odds of developing RA, compared with not developing RA, nearly 3-fold (OR 2.8, 95% CI 1.1 to 7.6, p = 0.038).
• Involvement of proximal interphalangeal joints and metacarpophalangeal joints.
• In a case series, significantly more people who went on to develop RA had swelling or pain of their wrist or finger joints compared with people who did not go onto to develop RA (89.4% compared with 60%, p = 0.0006).
• Symmetrical joints affected.
• In one case series, symmetrical synovitis of the hands and feet was present in 100% of people who went on to develop RA (no data were reported on statistical significance).
• A greater number of joints affected.
• In a case series, a greater number of tender joints (mean 9.8, compared with 6.0) and swollen (mean 7.9, compared with 4.4) was a useful predictor for distinguishing between people with or without RA (no data were reported on statistical significance).
• An inability to make a fist or flex finger.
• In a case series, the likelihood of developing RA was 10-fold higher in people who had some limitation when clenching their fist, compared with people without any limitation (likelihood ratio 10.3, p = 0.001). No data were reported for limitations on finger flexion.
• A positive metacarpophalangeal (MCP) squeeze test.
• In a case series of people with very early persistent synovitis, 68% had a positive squeeze test of their MCP joints (no data was reported on statistical significance).