Skip to main content
. Author manuscript; available in PMC: 2016 Jun 24.
Published in final edited form as: Lancet. 2015 Oct 1;386(10012):2489–2498. doi: 10.1016/S0140-6736(15)00347-5

Table 2.

Multivariable logistic regression analysis for the effect of treatment on the primary endpoint, ACR20 at 48 weeks post-randomisation (ITT population with multiple imputation)

Parameter OR & 95% CI p-value
Treatment arm: Tight control vs Standard care 1·91 (1·03, 3·55) 0·0392
Arthritis type: Oligoarthritis vs Polyarthritis 0·62 (0·31, 1·24) 0·1733
Centre: Other sites vs Chapel Allerton Hospital 2·33 (0·87, 6·27) 0·0929
Centre: St Luke’s Hospital Bradford vs Chapel Allerton Hospital 0·93 (0·41, 2·09) 0·8607
Centre: York District Hospital vs Chapel Allerton Hospital 0·50 (0·17, 1·52) 0·2223

Other sites included: Harrogate District Hospital, Manchester Royal Infirmary, St Bartholomew’s Hospital, North Tyneside General Hospital and Royal National Hospital for Rheumatic Diseases. These smaller recruiting sites were combined to prevent model convergence problems.

OR, Odds Ratio; CI, Confidence Interval