. Author manuscript; available in PMC: 2022 Oct 1.

Published in final edited form as: J Rheumatol. 2021 Apr 15;48(10):1547–1551. doi: 10.3899/jrheum.210175

Table 2.

Report of Studies of Test-Retest Reliability for HAQ-DI and SF-36 PF in PsA with OMERACT Filter 2.1

Author, year	Study description		Results					Judgement
Author, year	Characteristics of sample	Characteristics of testing situation	Sample recruited and sample considered stable for analysis	Scores at baseline and retest	Statistic used	Results	Minimal detectable change (95%CI) SEM=SD_baseline x √ (1-ICC) MDC=1.96 × SEM × √2	Interpretation of adequacy (+, +/−, −)
HAQ-DI
Study 1 Tillett 2019	140 consecutive patients with PsA fulfilled CASPAR, recruited for validation of composite measures	• 1 week apart • Assumed no change in condition	• 31 patients (77% men) who required no medication change had data for test-retest reliability • Mean (SD) age 54 (11) years • Mean (SD) duration of PsA 5.7 (4.7) years	Mean (SD) T1: 0.54 (0.62) T2: 0.52 (0.69) Mean (SD) difference = −0.02 (0.30), p=0.77 95% CI: −0.13 to 0.09	• ICC • Spearman’s rho (r)	• ICC=0.90 (95% CI: 0.79–0.95) • r=0.94 (p<0.01)	SEM=0.62 × √ (1–0.90) =0.20 MDC=1.96 × 0.20 × √2 = 0.54	(+) Good ICC and correlation between scores in a situation where changes in scores were not expected. Bland-Altman plot provided supportive evidence.
Study 2 Leung 2016	• 98 consecutive patients with PsA fulfilled CASPAR recruited for validation of PsAQoL study • Mean (SD) duration of illness 5.5 (8.4) years	• 2 weeks apart • Stability between timepoints assumed, given that PsA is a chronic illness and no change in medicine between T1 and T2	• 38 patients (44.7% men) who required no medication change had data for test-retest reliability • Mean (SD) age 53.9 (11.5) years	Mean (SD) T1: 0.38 (0.55) T2: 0.35 (0.56) Mean difference=0 (SD=0.19), p=1.00 95% CI: −0.373 to 0.373	• ICC • Spearman’s rho (r)	• ICC=0.94 (95% CI: 0.89–0.97) • r=0.83 (p<0.001)	SEM=0.55 × √ (1–0.94) =0.13 MDC=1.96 × 0.13 × √2 = 0.36	(+) Excellent ICC and correlation between scores in a situation where changes in scores were not expected. Bland-Altman plot provided supportive evidence.
SF-36 PF
Study 1 Tillett 2019	140 consecutive patients with PsA fulfilled CASPAR, recruited for validation of composite measures	• 1 week apart • Assumed no change in condition	• 31 patients who required no medication change had data for test-retest reliability • Mean (SD) age 54 (11) years • Mean (SD) duration of PsA 5.7 (4.7) years	Mean (SD) T1: 63.6 (29.6) T2: 66.7 (29.2) Mean (SD) difference = 3.10 (8.06), p = 0.05 95% CI: −0.04 to 6.17	• ICC • Spearman’s rho (r)	ICC=0.96 (95% CI: 0.92 to 0.98) r=0.95 (p<0.01)	SEM=29.6 × √ (1–0.96) = 5.92 MDC=1.96 × 5.92 × √2 = 8.37	(+) Excellent ICC and correlation between scores in a situation where changes in scores were not expected. Bland-Altman plot provided supportive evidence.

(+) adequate performance; (+/−) equivocal performance; (−): inadequate performance.

CASPAR: Classification criteria for Psoriatic Arthritis; ICC: intraclass correlation coefficient; HAQ-DI: Health Assessment Questionnaire-Disability Index; MDC: minimally detectable change; OMERACT: Outcome Measures in Rheumatology; PsA: psoriatic arthritis; PsAQoL: Psoriatic Arthritis Quality of Life index; r: Spearman’s rho correlation; SEM: standardized error of mean; SD: standard deviation; SF-36 PF: Medical Outcome Survey Short-Form - 36 item physical functioning domain; T1: time point 1; T2: time point 2.