A four-item lupus arthritis score anchored to ultrasound

Design

derivation and external validation cohort study, anchored to hand and wrist ultrasound

Population

133 adults with SLE receiving intramuscular glucocorticoid for lupus arthritis (USEFUL, seven English hospitals), plus 44 in a Leeds validation cohort

Primary outcome

construct validity and responsiveness against grey-scale and power Doppler synovitis scores

Effect

four-item score retained; effect size 0.40 after glucocorticoid (p<0.0001); thresholds outperformed swollen joint count alone

Lupus arthritis trials keep producing flat results, and one reason may be that the instruments used are not sensitive to change. The LAMDA score was derived by regressing the American College of Rheumatology inflammatory arthritis core set against combined grey-scale and power Doppler ultrasound of both hands and wrists in the USEFUL cohort — 133 people with SLE receiving intramuscular glucocorticoid for lupus arthritis at seven English hospitals.

Four variables survived: swollen joint count in 66 joints, ESR, physician's musculoskeletal disease activity visual analogue scale, and the patient's own musculoskeletal pain scale. The 28-joint count substituted for the 66-joint count almost perfectly, which makes the score usable in a clinic rather than only in a study visit. It moved after glucocorticoid with a medium-to-large effect size of 0.40, and in a separate 44-patient Leeds cohort the proposed thresholds discriminated treatment intention and acceptable symptom state better than swollen joint counts alone.

This is a measurement study, not a treatment study, and the derivation cohort was small and predominantly white. Nothing here tells you to change a drug. It does say that when a lupus patient's joints are the problem, a swollen joint count on its own is a poor instrument, and that ESR and both the patient's and clinician's global assessment of the joints add real information for very little extra time.

• Record a 28-joint swollen count rather than a tender count when lupus arthritis is the question
• Add ESR and separate patient and physician musculoskeletal scales — these carried the score
• Do not read a flat swollen joint count as absence of synovitis
• Thresholds are provisional and still being validated in multinational trials
• Derivation was in a mostly white English cohort; performance in Indian patients is untested

The statistics, in plain English

An effect size of 0.40 means the average change after treatment was about four-tenths of the spread of scores between patients — enough to detect in a trial, not enough to judge one patient. The very high intraclass correlation above 0.99 between the 66- and 28-joint counts is why the shortcut is safe. External validation in 44 patients is small: it supports the score but does not settle the thresholds.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.