Randomized controlled phase 2 trial of hydroxychloroquine in childhood interstitial lung disease

Table 4 Changes observed from START and those from STOP treatment groups were combined to explore maximum number of treatment effects

	“HCQ-effect”	“Placebo-effect”	P	Odds ratio/Kappa coefficient /Effect size (95% CI)
Groups combined	B, F, C, G	A, E
Responder (protocol definition) [yes, no]	4, 23	0, 16	0.106	Not appl^a,b
Responder (minimum important difference) [yes, no]	7, 19	2, 14	0.269	2.6 (0.5–14.4)^b
O2 saturation (%) in room air	1.6 (3.7) 25	− 0.1 (2.2) 16	0.062	0.5 (− 0.1–1.2)
Respiratory rate (/min) in room air	− 0.4 (6.7) 26	1.4 (4.9) 16	0.306	− 0.3 (− 0.9 − 0.3)
QoL chILD specific	− 3.2 (4.3) 12	6.6 (9.1) 8	0.019	− 1.5 (− 2.5–− 0.5)
QoL total score	4.3 (5.1) 12	− 2.0 (7.6) 8	0.063	1.0 (0.1–2.0)
BMI percentile	− 3.0 (10.0) 28	6.7 (10.9) 16	0.007	− 0.9 (− 1.6–− 0.3)
FVC absolute change (% predicted)	5.8 (15.1) 17	− 0.6 (4.7) 11	0.119	0.5 (− 0.2–1.3)

Data are number of subjects for responder/non-responder and means (SD) n for the other variables. Changes in Stop were multiplied by -1 before combining. After Bonferroni correction for multiple (four) comparisons, a P value < 0.0125 was considered significant
^anot appl = cannot be calculated if in any group no responder exists (i.e. no odds exists for independent groups)
^bOdds ratios and Kappa coefficients were calculated for responder evaluation, see also Table 2

ISSN: 1750-1172