CIBIC+ — Categorical Analysis — LOCF

CIBIC+ as categories rather than a mean — how many subjects fell in each impression band at three visits, with a stratified Cochran-Mantel-Haenszel test across treatment groups.

Slug: t-cibic-categorical
Regulatory ID: EFF05
Type: Table
E3 position: §14.2
Status: evidenced
Current iteration: v002
Datasets: adqscibc f6461f4, adsl 70c7278
Evidence: 22/22 passing
Requirements: DSP-ALL-001 DSP-ALL-002 DSP-CIBIC-001 DSP-CIBIC-002 DSP-CIBIC-003 DSP-CIBIC-004 DSP-CIBIC-005 DSP-INTXT-001 DSP-INTXT-002 DSP-INTXT-003 STD-GATE-001 STD-GATE-002 STD-GATE-003 STD-GATE-004 STD-GATE-005 STD-MODEL-001 STD-MODEL-002 STD-MODEL-003 STD-MODEL-004 STD-MODEL-005 STD-SRC-001 STD-SRC-002
Used in: Clinical Study Report Table 14.2.13
Post-Text Display Package Table 14.2.13
Abbreviated Clinical Study Report (ICH E3) Table 14.2.13
Study model: Efficacy Analysis Set EFFFL · grouped by TRT01A
Specification: 2 iterations
Environment: R 4.3.3 Darwin 23.6.0
Bound by: No text block binds it
Values: No value is sourced from it

Inputs

adsl254 rows · 70c7278
adqscibc730 rows · f6461f4
analysis.yamlwhat to compute
display.yamlhow to show it
custom.Rthe display's own method
study modelEfficacy Analysis Set

Function

prepare_data()regenerate('t-cibic-categorical')
build_ard()one row per statistic
render_display()every variant
render_rtf()the same cells

Outputs

ard.jsonv002 · 147 rows · 4847131
table.htmlthe rendered display
table.rtfsha256 413133a
manifest.jsonwho, why, every hash

RTFFull display413133a

Generated by the same pipeline run that wrote the ARD and the HTML above, from the same rendered cells; the short hash is the manifest's record of the file.

t-cibic-categorical · post_text variant · study CDISCPILOT01 · data cut-off 2014-07-01

	Placebo (N=79)	Xanomeline Low Dose (N=81)	Xanomeline High Dose (N=74)
CIBIC+ — Categorical Analysis — LOCF
Study CDISCPILOT01 — Efficacy Analysis Set
Week 8
n	77	81	73
Marked improvement	0 (0%)	0 (0%)	0 (0%)
Moderate improvement	1 (1%)	2 (2%)	1 (1%)
Minimal improvement	19 (25%)	16 (20%)	13 (18%)
No change	45 (58%)	48 (59%)	38 (52%)
Minimal worsening	10 (13%)	14 (17%)	20 (27%)
Moderate worsening	2 (3%)	1 (1%)	1 (1%)
Marked worsening	0 (0%)	0 (0%)	0 (0%)
p-value			0.2727
Week 16
n	79	81	74
Marked improvement	0 (0%)	0 (0%)	0 (0%)
Moderate improvement	0 (0%)	3 (4%)	2 (3%)
Minimal improvement	12 (15%)	12 (15%)	13 (18%)
No change	41 (52%)	46 (57%)	39 (53%)
Minimal worsening	25 (32%)	19 (23%)	20 (27%)
Moderate worsening	1 (1%)	1 (1%)	0 (0%)
Marked worsening	0 (0%)	0 (0%)	0 (0%)
p-value			0.4003
Week 24
n	79	81	74
Marked improvement	0 (0%)	0 (0%)	0 (0%)
Moderate improvement	1 (1%)	1 (1%)	0 (0%)
Minimal improvement	9 (11%)	14 (17%)	11 (15%)
No change	38 (48%)	37 (46%)	33 (45%)
Minimal worsening	28 (35%)	27 (33%)	25 (34%)
Moderate worsening	3 (4%)	2 (2%)	5 (7%)
Marked worsening	0 (0%)	0 (0%)	0 (0%)
p-value			0.6180
An analysis not specified in the protocol, treating the CIBIC+ score as a categorical rather than a continuous variable.
The seven categories are the CIBIC+ scale as the study's statistical analysis plan defines it (Appendix 1, section 14.1.2): 1 = marked improvement through 7 = marked worsening, with 4 = no change. Every category is reported at every visit, including those no subject was scored in.
Values are last observation carried forward within the analysis window (ADQSCIBC ANL01FL = 'Y').
Column headings are the size of the efficacy analysis set; n is the number of subjects with a value at the visit, and percentages are of that n.
The p-value is an overall comparison of the three treatment groups by the Cochran-Mantel-Haenszel row-mean-scores statistic on 2 degrees of freedom, stratified by site group. It is a single test per visit, printed once.
Source: adqscibc, adsl (CDISCPILOT01 ADaM package, phuse-org/phuse-scripts). Data cut-off: 2014-07-01.
open.csr display t-cibic-categorical (post_text variant); generated from the committed ARD.

Rendered from outputs/t-cibic-categorical/v002/table.html.

analysis	group1	group1_level	group2	group2_level	variable	variable_level	context	stat_name	stat_label	warning	error	stat
population	TRT01A	Placebo	—	—	POPULATION	—	population	N	Subjects in the analysis set	—	—	79
population	TRT01A	Xanomeline Low Dose	—	—	POPULATION	—	population	N	Subjects in the analysis set	—	—	81
population	TRT01A	Xanomeline High Dose	—	—	POPULATION	—	population	N	Subjects in the analysis set	—	—	74
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	77
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	1
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0.012987012987013
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	19
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.246753246753247
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	45
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.584415584415584
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	10
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.12987012987013
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	2
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0.025974025974026
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week8	TRT01A	Placebo	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	81
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	2
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0.0246913580246914
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	16
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.197530864197531
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	48
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.592592592592593
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	14
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.172839506172839
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	1
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0.0123456790123457
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week8	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	73
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	1
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0.0136986301369863
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	13
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.178082191780822
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	38
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.520547945205479
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	20
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.273972602739726
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	1
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0.0136986301369863
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week8	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cmh_week8	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	pval	p-value (CMH, controlling for site group)	—	—	0.272728404037124
cmh_week8	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	chisq	CMH statistic	—	—	2.59855767200549
cmh_week8	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	df	Degrees of freedom	—	—	2
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	79
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	0
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	12
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.151898734177215
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	41
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.518987341772152
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	25
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.316455696202532
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	1
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0.0126582278481013
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week16	TRT01A	Placebo	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	81
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	3
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0.037037037037037
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	12
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.148148148148148
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	46
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.567901234567901
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	19
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.234567901234568
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	1
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0.0123456790123457
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week16	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	74
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	2
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0.027027027027027
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	13
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.175675675675676
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	39
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.527027027027027
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	20
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.27027027027027
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	0
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week16	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cmh_week16	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	pval	p-value (CMH, controlling for site group)	—	—	0.400314714491458
cmh_week16	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	chisq	CMH statistic	—	—	1.83100850999908
cmh_week16	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	df	Degrees of freedom	—	—	2
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	79
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	1
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0.0126582278481013
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	9
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.113924050632911
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	38
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.481012658227848
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	28
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.354430379746835
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	3
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0.0379746835443038
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week24	TRT01A	Placebo	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	81
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	1
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0.0123456790123457
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	14
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.172839506172839
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	37
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.45679012345679
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	27
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.333333333333333
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	2
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0.0246913580246914
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week24	TRT01A	Xanomeline Low Dose	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	—	categorical	N	Subjects with a value at this visit	—	—	74
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked improvement	categorical	n	Subjects	—	—	0
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate improvement	categorical	n	Subjects	—	—	0
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate improvement	categorical	p	Percent of subjects at this visit	—	—	0
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal improvement	categorical	n	Subjects	—	—	11
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal improvement	categorical	p	Percent of subjects at this visit	—	—	0.148648648648649
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	No change	categorical	n	Subjects	—	—	33
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	No change	categorical	p	Percent of subjects at this visit	—	—	0.445945945945946
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal worsening	categorical	n	Subjects	—	—	25
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Minimal worsening	categorical	p	Percent of subjects at this visit	—	—	0.337837837837838
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate worsening	categorical	n	Subjects	—	—	5
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Moderate worsening	categorical	p	Percent of subjects at this visit	—	—	0.0675675675675676
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked worsening	categorical	n	Subjects	—	—	0
cat_week24	TRT01A	Xanomeline High Dose	—	—	CIBICCAT	Marked worsening	categorical	p	Percent of subjects at this visit	—	—	0
cmh_week24	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	pval	p-value (CMH, controlling for site group)	—	—	0.61804393101048
cmh_week24	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	chisq	CMH statistic	—	—	0.962391476546889
cmh_week24	TRT01A	Xanomeline High Dose	—	—	CMH	—	cmh	df	Degrees of freedom	—	—	2

147 rows, one per computed statistic.

Schema: opencsr/ard/v1
Generated: 2026-09-02T15:23:21Z
Spec hash: sha256:f6675d7fa980619760f57857f5cba5bf0e75b6aa92d486c7664b19f7f216ba7b
Display hash: sha256:3e86f7c60fc952738119de57fc95e6056240dd3ef152d16e56966d5853a3ed35
ARD file: outputs/t-cibic-categorical/v002/ard.json
ARD hash: sha256:4847131cf5f3a20db3ebb89fe76d1f674b9f21f57cf0bae1e6c04f907a52cb2c

Source datasets

Dataset	Rows	Package	Hash
adsl	254	phuse-org/phuse-scripts:data/adam 398a6d33ced9359ffb58c46650a6d488811176b1	sha256:70c7278e83e571a7d4a94fd864c98bf8ca243c7c2cc9eec8b99a40fa7bc98e35
adqscibc	730	phuse-org/phuse-scripts:data/adam 398a6d33ced9359ffb58c46650a6d488811176b1	sha256:f6461f436ff8d39a434fe5ecb7e1bcaf1b6da2913c4cbf2bf2c09d4c71d77873

Environment

R 4.3.3 · Darwin 23.6.0 · cards 0.6.1 · cardx 0.2.5 · gtsummary 2.3.0 · gt 1.0.0 · dplyr 1.1.4 · pharmaverseadam 1.1.0 · jsonlite 2.0.0 · yaml 2.3.10 · digest 0.6.37

library/tfl/t-cibic-categorical/analysis.yaml

id: t-cibic-categorical
title: "CIBIC+ — Categorical Analysis — LOCF"
regulatory_id: EFF05
type: table
dataset: adqscibc
analysis_set: efficacy
denominator: adsl
sources: phuse
group:
  - TRT01A
total: false
analyses:
  - name: population
    method: subject_count
    custom: ard_population_n
    label: "Subjects in the efficacy analysis set"
  - name: cat_week8
    method: categorical
    custom: ard_cibic_categories
    label: "CIBIC+ category at Week 8 (LOCF)"
    variables: [AVAL]
    filter: "PARAMCD == 'CIBICVAL' & AVISIT == 'Week 8' & ANL01FL == 'Y'"
  - name: cmh_week8
    method: categorical
    custom: ard_cibic_cmh
    label: "CMH test of the Week 8 distributions, controlling for site group"
    variables: [AVAL]
    filter: "PARAMCD == 'CIBICVAL' & AVISIT == 'Week 8' & ANL01FL == 'Y'"
  - name: cat_week16
    method: categorical
    custom: ard_cibic_categories
    label: "CIBIC+ category at Week 16 (LOCF)"
    variables: [AVAL]
    filter: "PARAMCD == 'CIBICVAL' & AVISIT == 'Week 16' & ANL01FL == 'Y'"
  - name: cmh_week16
    method: categorical
    custom: ard_cibic_cmh
    label: "CMH test of the Week 16 distributions, controlling for site group"
    variables: [AVAL]
    filter: "PARAMCD == 'CIBICVAL' & AVISIT == 'Week 16' & ANL01FL == 'Y'"
  - name: cat_week24
    method: categorical
    custom: ard_cibic_categories
    label: "CIBIC+ category at Week 24 (LOCF)"
    variables: [AVAL]
    filter: "PARAMCD == 'CIBICVAL' & AVISIT == 'Week 24' & ANL01FL == 'Y'"
  - name: cmh_week24
    method: categorical
    custom: ard_cibic_cmh
    label: "CMH test of the Week 24 distributions, controlling for site group"
    variables: [AVAL]
    filter: "PARAMCD == 'CIBICVAL' & AVISIT == 'Week 24' & ANL01FL == 'Y'"

library/tfl/t-cibic-categorical/display.yaml

id: t-cibic-categorical
title: "CIBIC+ — Categorical Analysis — LOCF"
study: CDISCPILOT01
population_label: "Efficacy Analysis Set"
cutoff: "2014-07-01"
source: "Source: adqscibc, adsl (CDISCPILOT01 ADaM package, phuse-org/phuse-scripts). Data cut-off: 2014-07-01."
footnotes:
  - "An analysis not specified in the protocol, treating the CIBIC+ score as a categorical rather than a continuous variable."
  - "The seven categories are the CIBIC+ scale as the study's statistical analysis plan defines it (Appendix 1, section 14.1.2): 1 = marked improvement through 7 = marked worsening, with 4 = no change. Every category is reported at every visit, including those no subject was scored in."
  - "Values are last observation carried forward within the analysis window (ADQSCIBC ANL01FL = 'Y')."
  - "Column headings are the size of the efficacy analysis set; n is the number of subjects with a value at the visit, and percentages are of that n."
  - "The p-value is an overall comparison of the three treatment groups by the Cochran-Mantel-Haenszel row-mean-scores statistic on 2 degrees of freedom, stratified by site group. It is a single test per visit, printed once."
columns:
  order: [Placebo, "Xanomeline Low Dose", "Xanomeline High Dose"]
format:
  digits:
    "N": 0
    "n": 0
    p: 0
    pval: 4
rows:
  - { label: "Week 8", section: true }
  - { analysis: cat_week8, variable: CIBICCAT, pattern: "{N}", label: "n", indent: 1 }
  - { analysis: cat_week8, variable: CIBICCAT, levels: all, pattern: n_pct, indent: 1,
      level_order: ["Marked improvement", "Moderate improvement", "Minimal improvement",
                    "No change", "Minimal worsening", "Moderate worsening", "Marked worsening"] }
  - { analysis: cmh_week8, variable: CMH, pattern: "{pval}", label: "p-value", indent: 1 }
  - { label: "Week 16", section: true }
  - { analysis: cat_week16, variable: CIBICCAT, pattern: "{N}", label: "n", indent: 1 }
  - { analysis: cat_week16, variable: CIBICCAT, levels: all, pattern: n_pct, indent: 1,
      level_order: ["Marked improvement", "Moderate improvement", "Minimal improvement",
                    "No change", "Minimal worsening", "Moderate worsening", "Marked worsening"] }
  - { analysis: cmh_week16, variable: CMH, pattern: "{pval}", label: "p-value", indent: 1 }
  - { label: "Week 24", section: true }
  - { analysis: cat_week24, variable: CIBICCAT, pattern: "{N}", label: "n", indent: 1 }
  - { analysis: cat_week24, variable: CIBICCAT, levels: all, pattern: n_pct, indent: 1,
      level_order: ["Marked improvement", "Moderate improvement", "Minimal improvement",
                    "No change", "Minimal worsening", "Moderate worsening", "Marked worsening"] }
  - { analysis: cmh_week24, variable: CMH, pattern: "{pval}", label: "p-value", indent: 1 }
variants:
  post_text: {}

library/tfl/t-cibic-categorical/custom.R

# Display-specific statistics for the CIBIC+ categorical analysis.
#
# Two things the built-in vocabulary cannot do, both needed by this display.
#
# 1. ADQSCIBC stores CIBIC+ as a bare number in AVAL. It carries no AVALC and the
#    study's define.xml declares no code list for it, so `method: categorical`
#    would produce rows labelled "2", "3", "4". The seven category names are
#    stated in the study's statistical analysis plan (Appendix 1, §14.1.2):
#
#      1 = Marked improvement   2 = Moderate improvement   3 = Minimal improvement
#      4 = No change            5 = Minimal worsening      6 = Moderate worsening
#      7 = Marked worsening
#
#    They are transcribed here, once, and every category is reported at every
#    visit — including the ones nobody was scored in, which is information: at no
#    visit, in any arm, was a subject rated markedly improved or markedly worsened.
#    `method: categorical` would silently drop those rows.
#
# 2. The published table's p-value is a stratified Cochran-Mantel-Haenszel test.
#    No built-in method fits a model.
#
# Contract: each function receives (data, spec, denominator) and returns a
# {cards}-shaped ARD. `spec$group` holds the grouping variable in force.

#' The CIBIC+ seven-point scale, as the study's analysis plan defines it
#'
#' Ordered from best to worst outcome, which is the order the published table
#' prints and the order the display's `level_order` repeats.
#' @return A named character vector, score (as text) to category name.
cibic_categories <- function() {
  c(
    "1" = "Marked improvement",
    "2" = "Moderate improvement",
    "3" = "Minimal improvement",
    "4" = "No change",
    "5" = "Minimal worsening",
    "6" = "Moderate worsening",
    "7" = "Marked worsening"
  )
}

#' Category frequencies for one visit
#'
#' Percentages are of the subjects with a value AT THIS VISIT, not of the
#' efficacy analysis set — which is what the published table does, and why `N`
#' is reported per visit as well as per population.
#'
#' @param data Analysis dataset, already restricted to the analysis set and to
#'   the entry's `filter` (which carries the visit).
#' @param spec The analysis entry, including `spec$group`.
#' @param denominator Subject-level denominator (unused: the denominator of this
#'   table is the visit, not the population).
#' @return A data frame carrying the `{cards}` ARD columns.
ard_cibic_categories <- function(data, spec, denominator) {
  group <- spec$group
  if (!length(group)) {
    stop("The CIBIC+ categorical analysis needs a grouping variable.", call. = FALSE)
  }
  cats <- cibic_categories()
  score <- as.character(as.integer(data$AVAL))
  unknown <- setdiff(unique(score[!is.na(score)]), names(cats))
  if (length(unknown)) {
    stop(
      "CIBIC+ scores outside the seven-point scale the analysis plan defines: ",
      paste(sort(unknown), collapse = ", "), ".",
      call. = FALSE
    )
  }
  levels <- levels(data[[group]])
  if (is.null(levels)) levels <- sort(unique(as.character(data[[group]])))

  rows <- list()
  for (lv in levels) {
    in_group <- !is.na(data[[group]]) & as.character(data[[group]]) == lv
    n_visit <- sum(in_group)
    rows[[length(rows) + 1]] <- cat_row(
      group, lv, NA_character_, "N", n_visit, "Subjects with a value at this visit"
    )
    for (code in names(cats)) {
      n <- sum(in_group & !is.na(score) & score == code)
      rows[[length(rows) + 1]] <- cat_row(group, lv, cats[[code]], "n", n, "Subjects")
      rows[[length(rows) + 1]] <- cat_row(
        group, lv, cats[[code]], "p",
        if (n_visit > 0) n / n_visit else NA_real_, "Percent of subjects at this visit"
      )
    }
  }
  do.call(rbind, rows)
}

#' Stratified Cochran-Mantel-Haenszel test of the category distributions
#'
#' The published table footnotes this as an "overall comparison of treatments
#' using CMH test ... controlling for site group". Three CMH statistics answer to
#' that description and they do not agree; the one that reproduces the published
#' p-values at all three visits (0.2727, 0.4003, 0.6180) is the ROW MEAN SCORES
#' DIFFER statistic — the mean-score (ANOVA) statistic on 2 degrees of freedom,
#' with the CIBIC+ score itself as the score. The general-association statistic
#' on 8 degrees of freedom, which the footnote's parenthetical "(Pearson
#' Chi-Square)" would suggest, does not. Which of the two SAS printed is settled
#' by the numbers, and the qualification record carries all three so the choice
#' is inspectable rather than asserted.
#'
#' Implemented from the generalised Mantel-Haenszel definition rather than by
#' calling a test function, because [stats::mantelhaen.test()] computes only the
#' general-association statistic.
#'
#' @param data Analysis dataset, already restricted to the analysis set and the
#'   entry's `filter`.
#' @param spec The analysis entry, including `spec$group`.
#' @param denominator Subject-level denominator (unused).
#' @return A data frame carrying the `{cards}` ARD columns.
ard_cibic_cmh <- function(data, spec, denominator) {
  group <- spec$group
  if (!"SITEGR1" %in% names(data)) {
    stop("The CIBIC+ CMH test stratifies on SITEGR1, which the analysis dataset does not carry.", call. = FALSE)
  }
  levels <- levels(data[[group]])
  if (is.null(levels)) levels <- sort(unique(as.character(data[[group]])))
  scores <- sort(unique(as.integer(data$AVAL[!is.na(data$AVAL)])))

  tab <- table(
    factor(as.integer(data$AVAL), levels = scores),
    factor(as.character(data[[group]]), levels = levels),
    factor(as.character(data$SITEGR1))
  )
  res <- cmh_row_mean_scores(tab, scores)

  # One model-level statistic, printed once, in the last treatment column —
  # which is where the published table prints it.
  col <- levels[length(levels)]
  out <- rbind(
    cat_row(group, col, NA_character_, "pval", res[["p"]], "p-value (CMH, controlling for site group)"),
    cat_row(group, col, NA_character_, "chisq", res[["Q"]], "CMH statistic"),
    cat_row(group, col, NA_character_, "df", res[["df"]], "Degrees of freedom")
  )
  out$variable <- "CMH"
  out$context <- "cmh"
  out
}

#' Generalised Mantel-Haenszel "row mean scores differ" statistic
#'
#' @param tab A scores x groups x strata contingency table.
#' @param scores Numeric scores for the first dimension.
#' @return A named numeric vector with `Q`, `df` and `p`.
#' @noRd
cmh_row_mean_scores <- function(tab, scores) {
  R <- dim(tab)[1]
  C <- dim(tab)[2]
  K <- dim(tab)[3]
  if (C < 2) stop("The CMH test needs at least two treatment groups.", call. = FALSE)
  B <- matrix(as.numeric(scores), nrow = 1)
  Cm <- diag(C)[-C, , drop = FALSE]

  A <- 0
  E <- 0
  V <- 0
  for (k in seq_len(K)) {
    n <- matrix(tab[, , k], nrow = R, ncol = C)
    N <- sum(n)
    if (N < 2) next
    r <- rowSums(n)
    cc <- colSums(n)
    A <- A + kronecker(Cm, B) %*% as.vector(n)
    E <- E + kronecker(Cm %*% cc, B %*% r) / N
    VR <- B %*% (diag(r, nrow = R) - outer(r, r) / N) %*% t(B)
    VC <- Cm %*% (diag(cc, nrow = C) - outer(cc, cc) / N) %*% t(Cm)
    V <- V + kronecker(VC, VR) / (N - 1)
  }
  d <- A - E
  rank <- qr(V)$rank
  if (rank < ncol(V)) {
    stop(
      "The CMH covariance matrix is rank ", rank, " of ", ncol(V),
      "; the treatment groups are not separately estimable within site groups.",
      call. = FALSE
    )
  }
  Q <- drop(t(d) %*% solve(V, d))
  c(Q = Q, df = rank, p = stats::pchisq(Q, rank, lower.tail = FALSE))
}

#' Size of the analysis set in each treatment column
#'
#' Identical in intent to the function of the same name in the CIBIC+ summary
#' displays: the column heading is the efficacy analysis set, which is larger
#' than the number of subjects assessed at any one visit.
#'
#' @param data Analysis dataset (unused).
#' @param spec The analysis entry, including `spec$group`.
#' @param denominator Subject-level denominator, already restricted to the
#'   display's analysis set.
#' @return A data frame carrying the `{cards}` ARD columns.
ard_population_n <- function(data, spec, denominator) {
  group <- spec$group
  levels <- levels(denominator[[group]])
  if (is.null(levels)) levels <- sort(unique(as.character(denominator[[group]])))
  rows <- lapply(levels, function(lv) {
    sub <- denominator[as.character(denominator[[group]]) == lv, , drop = FALSE]
    cat_row(group, lv, NA_character_, "N", length(unique(sub$USUBJID)), "Subjects in the analysis set")
  })
  out <- do.call(rbind, rows)
  out$variable <- "POPULATION"
  out$context <- "population"
  out
}

#' One ARD row in the `{cards}` shape
#' @noRd
cat_row <- function(group, group_level, variable_level, stat_name, value, label) {
  data.frame(
    group1 = group,
    group1_level = group_level,
    variable = "CIBICCAT",
    variable_level = variable_level,
    context = "categorical",
    stat_name = stat_name,
    stat_label = label,
    stat = as.numeric(value),
    warning = NA_character_,
    error = NA_character_,
    stringsAsFactors = FALSE
  )
}

Every iteration is kept: outputs/t-cibic-categorical/vNNN/ holds the spec snapshot, the ARD, the rendered output, and a manifest recording who regenerated it, why, and from which commit.