Table 2.

Annualised relapse rate on and off tocilizumab treatment

	On TCZ (N=65)	Off TCZ (N=65)	Overall (N=65)
Follow-up time, median (IQR), years	1.7 (1.0–3.0)	1.0 (0.4–1.6)	2.8 (2.1–3.9)
Rate of flares per year*
Rate (95% CI)	0.14 (0.09 to 0.22)	0.42 (0.27 to 0.65)
Rate ratio (95% CI)†	0.33 (0.18 to 0.60)
P value	0.0004
Patients with ≥1 flare, n (%)	14 (21.5)	21 (32.3)	31 (47.7)
1 flare	8 (12.3)	16 (24.6)	20 (30.8)
2 flares	4 (6.2)	5 (7.7)	6 (9.2)
3 flares	2 (3.1)	0	4 (6.2)
≥4 flares	0	0	1 (1.5)

Log of the time at risk is used as an offset variable. All follow-up time is included in the time at risk for a new event. The TCZ analysis population is defined as the participants whose initial TCZ course was ≥9 months.

* Rates are estimated from a Poisson regression model, with ongoing TCZ treatment, age, relapse status, sex, and prednisone dose at TCZ intiation as covariates, and random patient effect. For both on TCZ and after TCZ discontinuation, the time period considered is aggregated for all periods on and off TCZ, respectively, after TCZ initiation.

† Rate ratio was computed as the ratio of the rate on TCZ by the rate off TCZ.

TCZ, tocilizumab.