Table 3.
Longitudinal associations of outcomes with different SLE unmet need definitions
| Outcomes | LLDAS-never | AMS > 4 | HDAS-ever |
|---|---|---|---|
| HR1 (95% CI), p-value | HR1 (95% CI), p-value | HR1 (95% CI), p-value | |
| Damage accrual | |||
| Unadjusted | 1.52 (1.31, 1.76), p < 0.001 | 1.38 (1.18, 1.61), p < 0.001 | 1.85 (1.47, 2.31), p < 0.001 |
| Adjusteda | 1.46 (1.26, 1.69), p < 0.001 | 1.36 (1.16, 1.59), p < 0.001 | 1.81 (1.43, 2.30), p < 0.001 |
| Mortality | |||
| Unadjusted | 6.64 (2.83, 15.6), p < 0.001 | 2.99 (1.68, 5.3), p < 0.001 | 6.97 (3.82, 12.7), p < 0.001 |
| Adjustedb | 4.98 (2.07, 12.0), p < 0.001 | 2.36 (1.29, 4.33), p = 0.006 | 5.45 (2.75, 10.80), p < 0.001 |
| RC2 (95% CI), p-value | RC2 (95% CI), p-value | RC2 (95% CI), p-value | |
| Cumulative prednisolone (PNL) | |||
| Unadjusted | 5.61 (5.34, 5.88), p < 0.001 | 4.08 (3.66, 4.51), p < 0.001 | 8.96 (8.02, 9.91), p < 0.001 |
| Adjustedc | 5.71 (5.38, 6.03), p < 0.001 | 3.39 (2.95, 3.83), p < 0.001 | 9.04 (7.80, 10.3), p < 0.001 |
| TAM-PNL at visit | |||
| Unadjusted | 1.25 (1.08, 1.41), p < 0.001 | 2.33 (1.90, 2.76), p < 0.001 | 1.18 (0.88, 1.47), p < 0.001 |
| Adjustedc | 1.35 (1.17, 1.52), p < 0.001 | 2.52 (2.20, 2.85), p < 0.001 | 1.41 (0.88, 1.94), p < 0.001 |
| PCS | |||
| Unadjusted | −1.59 (−1.90, −1.28), p < 0.001 | −1.04 (−1.58,−0.50), p < 0.001 | −2.49 (−3.07, −1.90), p < 0.001 |
| Adjustedd | −1.40 (−1.71, −1.09), p < 0.001 | −0.96 (−1.50,−0.43), p < 0.001 | −2.17 (−2.78, −1.57), p < 0.001 |
| MCS | |||
| Unadjusted | −1.22 (−1.58, −0.85), p < 0.001 | −0.84 (−1.42, −0.27), p < 0.001 | −1.37 (−2.03, −0.70), p < 0.001 |
| Adjustede | −1.20 (−1.57, −0.84), p < 0.001 | −0.97 (−1.56, −0.38), p = 0.001 | −1.29 (−1.96, −0.63), p < 0.001 |
1Hazard ratios (HR) derived using Cox regression analyses. 2Regression coefficients (RCs) derived using generalised estimating equations (GEE). RC indicates the mean difference between the unmet need definition and the corresponding comparator (e.g. not in LLDAS vs. in LLDAS). TAM-PNL at visit = time-adjusted mean prednisolone since the baseline visit to each routine visit
aHRs adjusted for age, disease duration, Asian ethnicity, tertiary education, and cumulative PNL. Full multivariable models are presented in Supplementary table S1
bHRs adjusted for cumulative PNL and ACR/SLICC SDI score. Full multivariable models are presented in Supplementary table S2
cRCs adjusted for age, disease duration, Asian ethnicity, presence of flare, and ACR/SLIC SDI score. Full multivariable models for prednisolone are presented in Supplementary table S3
dRCs adjusted for age, disease duration, Asian ethnicity, tertiary education, cumulative PNL, presence of flare, and organ damage. Full multivariable models are presented in Supplementary table S4
eRCs adjusted for Asian ethnicity, tertiary education, and cumulative PNL. Full multivariable models are presented in Supplementary table S5