Table 3

Predicted yearly changes in outcomes and survival rates according to initial protocol, with and without adjusting (95% CI)

Embedded Image
  • Significance p: Fisher's test for equality of change rates between protocols, for each outcome variable.

  • *Results are reported in terms of changes after 12 months. However, all study data (from baseline to the 24-month end point) were used in estimation. To obtain 24-month changes, multiply results above by 2.

  • †For the subanalysis involving the subset of patients with pulmonary fibrosis at baseline, patients with definite bibasal pulmonary fibrosis confirmed on HRCT were included, irrespective of FVC value. If no HRCT scan was performed at baseline, an FVC<55%, DLCO<55% predicted or definite bibasal shadowing on X-ray was also a basis for inclusion.

  • ‡Changes expressed in units for the Cochin regression are an approximation derived from the 95% CI of percentage changes between baseline and 12 months (on a scale shifted by one unit), applied to the predicted baseline values for each group in the original scale.

  • DLCO, carbon monoxide diffusing capacity; FVC, forced vital capacity; HAQ-DI, Health Assessment Questionnaire Disability Index; HRCT, high-resolution CT; mRSS, modified Rodnan skin score (17 sites); PF, pulmonary fibrosis.