Skip to main content

Table 2 Number of responders to treatment

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

Variable

Start

Stop

Group (treatment)

Yes (n)

No (n)

Comparison to A (placebo 4wks); P

Odds ratio/Kappa coefficient (95% CI)

Group (treatment)

Yes (n)

No (n)

Comparison to E (HCQ 4wks); P

Odds ratio/ Kappa coefficient (95% CI)

Responders to treatment (defined per protocol)

A (placebo 4wks)

0

13

  

E (HCQ 12wks)

0

3

  

B (HCQ 4wks)

0

7

1.000a

Not applc

F (placebo 12wks)

0

5

1.000a

Not applc

C (HCQ 4wks)

3

9

Not applb,c

Not applc

G (no med. 12wks)

1

2

Not applb,c

Not applc

B + C (combined)

3

16

0.253a

Not applc

F + G combined

1

7

1.000a

Not applc

Responders to treatment (defined by MID)

A (placebo 4wks)

2

11

  

E (HCQ 12wks)

0

3

  

B (HCQ 4wks)

1

5

1.000a

1.100 (0.079–15.16)

F (placebo 12wks)

3

2

0.196a

Not applc

C (HCQ 4wks)

2

10

1.000b

− 0.2 (− 0.340–0.008)

G (no med. 12wks)

1

2

Not applb,c

Not applc

B + C (combined)

3

15

1.000a

1.100 (0.156–7.740)

F + G (combined)

4

4

0.236a

Not applc

  1. Data are given as number (n) of patients responding or not. MID = minimum important difference (defined as 3% change in O2-saturation)
  2. Remark: in group B “Responders to treatment (defined by MID)” has one subject less than “Responders to treatment (defined per protocol)”, because in that case the O2 saturation in room air was missing. no med. = no medication, means withdrawal of HCQ
  3. aFisher exact test
  4. bMcNemar test for paired samples and compares A and C or E and G
  5. cnot appl = cannot be calculated if in any group no responder exists (i.e. no odds exists for independent groups or no disconcordant pairs could be determined for dependent groups). P – values are given and a P < 0.05 was considered significant
  6. Using the per protocol definition, the 4 patients who responded to HCQ or HCQ withdrawal had the diagnosis: 2 × ABCA3 deficiency, 1 × SP-C deficiency and 1 × Filamin A deficiency; no patients responded to placebo. Using the MID definition, the 7 patients who responded to HCQ (groups B, C) or HCQ withdrawal (groups F, G) had the following diagnoses: 4 × ABCA3 deficiency, 1 × NSIP, 1 × SP-C deficiency and 1 × Filamin A deficiency. The 2 responders to placebo (group A) had the following diagnoses: 1 × COPA-syndrome, 1 × cNSIP