Table 2– COPD assessment test (CAT) nonresponse rate, floor and ceiling effect, internal consistency and test–retest reliability

Completed CATs and missing items	0 missing items %	1 missing item %	Subjects n
Agusti [9]	99.7	0.3	377
Ghobadi [17]	100	0.0	105
Jones [21]^#	99.3	0.7	6469
Mackay [26]	98.3	1.7	161
Ringbaek [41]	98.9	1.1	90
Tsuda [15]	100	0.0	301
Floor^¶ and ceiling⁺ effect	Floor effect %	Ceiling effect %	Subjects n
Agusti [9]	0.5	0.3	377
Nishimura [36]	0.0	7.6	145^##
Internal consistency^§	Cronbach’s α	Subjects n	p-value
Agusti [9]	0.86	NR	NR
da Silva [10]	0.98	50	0.001
Horita [18]	0.85	85	NR
Hwang [19]	0.85	100	NR
Jones [5]	0.88	1490	NR
Jones [21] Arabic version	0.85	4807	NR
Jones [21] Turkish version	0.86	1590	NR
Tsiligianni [14]	0.86	90	NR
Tsuda [15]	0.89	301	NR
Test–retest^ƒ	ICC	n	Administration duration weeks			p-value
			1	2	3
Agusti [9]	0.83	NR	NR	NR	NR	NR
Al Moamary [30]	0.90	45	0	2	NA	0.00008
da Silva [10]	0.96	50	0	1	NA	0.93–0.97^¶¶
Jones [5]	0.80	53	0	1	NA	NR
Tsiligianni [14]	0.94	90	0	2	6	0.92–0.96^¶¶

COPD: chronic obstructive pulmonary disease; ICC: intra-class correlation coefficient; NR: not reported; NA: not applicable. ^#: missing defined as more than one item not completed so the questionnaire was considered unusable; if the score was missing for only one item, this was replaced by the mean value of the remaining seven items. ^¶: percentage of patients with the minimum possible CAT total score (0/40). ⁺: percentage of patients with the maximum possible (40/40). ^§: interrelatedness of the items within the CAT. ^ƒ: ability of the CAT to provide consistent scores over time when administered repeatedly under stable disease condition. ^##: floor and ceiling effect of the CAT for COPD participants defined by fixed ratio. ^¶¶: 95% CI for ICC estimate.