Table 2.

Descriptive statistics for each patient reported outcome measure used in the 6 month follow up post randomisation in the OSCAR trial

	Range	N	Mean	SD	Response rates	Floor effects	Ceiling effects
EQ-5D-3 L (Utility)	−0.594 to 1	234	0.60	0.36	95.29 %	0.00 %	19.73 %
SF-6D (Utility)	0.345 to 1	218	0.66	0.16	93.16 %	0.92 %	0.92 %
SF-12 PCS	0 to 100	209	36.88	12.32	89.32 %	0.00 %	0.00 %
SF-12 MCS	0 to 100	209	43.43	13.30	89.32 %	0.00 %	0.00 %
SGRQ Symptoms Score	0 to 100	221	33.71	27.89	94.44 %	1.36 %	1.36 %
SGRQ Activity Score	0 to 100	193	46.04	31.80	82.47 %	7.77 %	7.77 %
SGRQ Impact Score	0 to 100	194	22.13	22.42	82.90 %	0.00 %	0.00 %
SGRQ Total Score	0 to 100	175	32.08	24.18	74.79 %	0.00 %	0.00 %

Scores or Utility Values of participants which could not be calculated due to missing items were not included within the analysis

N stands for the sample size. SD denotes the standard deviation. CI is abbreviated for confidence intervals

Floor effects is the proportion of patients in the lowest health state possible and ceiling effects is the proportion of patients in the highest health state poss

EQ-5D-3 L is anchored at -0.594 for worst possible health and 1 for best possible health

SF-6D is anchored at 0.345 for worst possible health and 1 for best possible health

SF-12 PCS & MCS scores are anchored at 0 for worst possible health and 1 for best possible health

SGRQ scores are anchored at 100 for worst possible health and 0 for best possible health

The EQ-5D-3 L and SF-12 was assessed in terms of complete responses for all questions

The SF-6D was regarded complete if there was sufficient items to calculate a score

The SGRQ was assessed by calculating if more than 24 % of the items were missing or not answered, then a total score could not be calculated