Psychometric properties of the RAID combinations for 570 RA patients in the cross-sectional study
| Correlation | |||||||
|---|---|---|---|---|---|---|---|
| Combination | Patients with missing data (N (%)) | Mean (SD) values of the combination | Patient global VAS | Global (SF-36)* | PCS (SF-36) | MCS (SF-36) | DAS28-ESR |
| Seven domains by NRS (final RAID) | 27 (5.7) | 4.4 (2.3) | 0.76 | 0.55 | −0.60 | −0.55 | 0.69 |
| Six domains by NRS (without coping) | 26 (4.5) | 4.3 (2.3) | 0.75 | 0.54 | −0.57 | −0.57 | 0.65 |
| Seven domains, pain by SF-36 pain | 24 (4.2) | 4.4 (1.4) | 0.70 | 0.49 | −0.50 | −0.56 | 0.58 |
| Seven domains, sleep by MOS sleep | 54 (9.5) | 4.4 (2.2) | 0.75 | 0.58 | −0.61 | −0.56 | 0.69 |
| Seven domains, coping by questionnaire | 146 (25.6) | 4.7 (2.1) | 0.74 | 0.53 | −0.59 | −0.54 | 0.68 |
| Seven domains, function by HAQ | 27 (5.7) | 4.2 (2.2) | 0.75 | 0.56 | −0.62 | −0.55 | 0.69 |
| Seven domains, function by mHAQ | 61 (10.7) | 3.9 (2.1) | 0.77 | 0.55 | −0.62 | −0.56 | 0.71 |
The table presents the psychometric properties of some of the combinations of questions/questionnaires assessed. Missing data percentages are presented without any imputation technique for missing data, except for the HAQ in which the validated method to impute missing data was applied.27 Correlations are Spearman's correlation. For all correlations, p values were less than 0.0001.
↵* Global assessment of status by the first question in the SF-36 questionnaire.
DAS28, disease activity score; ESR, erythrocyte sedimentation rate; HAQ, health assessment questionnaire; mHAQ, modified HAQ; MOS, medical outcome study subscale; NRS, numeric rating scale; SF-36, short form 36 (summary values: physical, PCS and mental, MCS); RAID, rheumatoid arthritis impact of disease; VAS, visual analogue scale.