Table 3.
Univariate Cox regression for time to infection post V3 in 157 patients (30 infected, 127 uninfected) who had no change in immunotherapy group from V1 through to V3
| Variable | Univariate Cox regression (outcome: time to infection post V3) |
|||
|---|---|---|---|---|
| Event vs. reference level | Hazard ratio (95% CI) | p value | Omnibus p value |
|
| Age | Year | 0.978 (0.952–1.004) | 0.103 | |
| Female | Female vs. male | 2.198 (0.589–8.198) | 0.241 | |
| Diagnosis | – | – | – | 0.642 |
| AQP4-Ab NMOSD vs. MS | 0.745 (0.330–1.682) | 0.479 | ||
| MOGAD vs. MS | 0.541 (0.099–2.949) | 0.477 | ||
| EDSS | EDSS ≥ 6 vs. EDSS < 6 | 1.025 (0.422–2.490) | 0.956 | |
| Immunotherapy group | – | – | – | 0.059 |
| Anti-CD20s, S1PRMs vs. Nil | 3.201 (1.150–8.907) | 0.026 | ||
| DMARDs vs. Nil | 0.853 (0.249–2.920) | 0.800 | ||
| IRTs vs. Nil | 1.419 (0.299–6.738) | 0.660 | ||
| Other DMTs vs. Nil | 1.956 (0.530–7.226) | 0.314 | ||
| Detectable NAbs response within 2 – 16 weeks post V3 (i.e. ≥ 30% inhibition)* | Detectable NAbs response vs. negative response | 0.464 (0.182– 1.183) | 0.108 | |
*Data available for 48 patients: 5 infected and 43 uninfected. For the 5 infected patients, there was no correlation between NAbs levels and time to infection (r = – 0.226, p = 0.715)
The adequacy of the Cox regression model was assessed via proportional hazard assumption by Kolmogorov-type supremum test of 1000 simulated patterns. Only ‘Detectable NAbs response within 2–16 weeks post V3’ variable could not satisfy the underlying assumption which was due to the small sample size