Chronic conditions

[caitlink12]

Included ICES survey cycles to existing chronic condition variables CCC_071, CCC_091, CCC_101, CCC_111, CCC_121, CCC_151, CCC_280 to variable and variable_details.

[rafdoodle]

CCC_ Variable Metadata Review — databaseStart / variableStart Updates

✅ Variables Updated (variables.csv + variable_details.csv)

The following CCC_ variables had their databaseStart and variableStart extended to include the full range of master file (_m) databases (where applicable), mirroring their existing public file (_p) coverage. Changes were applied to both inst/extdata/variables.csv and inst/extdata/variable_details.csv:

Variable	Description
CCC_031	Asthma
CCC_035	Asthma — had symptoms/attacks
CCC_036	Asthma — took medication
CCC_051	Arthritis/Rheumatism
CCC_061	Back problems
CCC_071	Hypertension
CCC_081	Migraine headache
CCC_091	COPD/Emphysema/Bronchitis
CCC_101	Diabetes
CCC_105	Diabetes — currently takes insulin
CCC_10A	Diabetes diagnosed when pregnant
CCC_10B	Diabetes diagnosed — other than when pregnant
CCC_10C	Diabetes diagnosed — when started with insulin
CCC_111	Epilepsy (already complete)
CCC_121	Heart disease
CCC_131	Active cancer
CCC_151	Stroke
CCC_181	Alzheimer's disease or other dementia
CCC_280	Mood disorder
number_conditions	Number of chronic conditions (minor formatting fix only)

---

⚠️ Variables Needing Review (variables.csv + variable_details.csv)

The following variables have not yet been updated. Their databaseStart and variableStart in both variables.csv and variable_details.csv may be incomplete with respect to master file (_m) database coverage and inconsistent with respect to PUMF file (_p) database coverage. They should thus be verified against the CCHS data dictionaries.

Variable	Description	Notes
CCC_041	Fibromyalgia	Only cchs2010_m listed
CCC_072	Hypertension diagnosis	Starts at cchs2005_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_073	Hypertension medication	Starts at cchs2005_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_073A	Hypertension — pregnant at first diagnosis	Starts at cchs2009_2010_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_073B	Hypertension — diagnosed outside pregnancy	Starts at cchs2009_2010_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_075	Blood cholesterol	No _m entries
CCC_102_A	Age of diabetes diagnosis (categorical, 2003)	Only cchs2009_m, cchs2010_m, cchs2012_m
CCC_102_B	Age of diabetes diagnosis (categorical, 2005+)	Starts at cchs2005_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_102_cont	Age of diabetes diagnosis (continuous)	Starts at cchs2003_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_106	Diabetes — pills to control blood sugar	Starts at cchs2005_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_141	Stomach or intestinal ulcers	No cchs2015_2016_p/cchs2017_2018_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_161	Urinary incontinence	No cchs2015_2016_p/cchs2017_2018_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_171	Bowel disorder	No cchs2015_2016_p/cchs2017_2018_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_17A	Type of bowel disorder	Starts at cchs2005_p; no cchs2015_2016_p/cchs2017_2018_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_251	Chronic fatigue syndrome	Only cchs2010_m
CCC_261	Multiple chemical sensitivities	Only cchs2010_m
CCC_290	Anxiety disorder	Starts at cchs2003_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_31A	Cancer diagnosis	Starts at cchs2005_p; only cchs2009_m, cchs2010_m, cchs2012_m
CCC_91A	Chronic bronchitis	No _m entries
CCC_91E	Emphysema	No _m entries
CCC_91F	COPD	No _m entries
COPD_Emph_der	COPD/Emphysema (derived)	Only cchs2009_m, cchs2010_m, cchs2012_m; see note below
resp_condition	Respiratory condition (derived)	Only cchs2009_m, cchs2010_m, cchs2012_m; see note below

Note on COPD_Emph_der and resp_condition

The functions in R/respiratory-condition.R that produce COPD_Emph_der and resp_condition do not currently account for master file (_m) databases as they use DHHGAGE_cont as a parameter. The databaseStart/variableStart metadata for these derived variables should be updated in both variables.csv and variable_details.csv once the underlying source variables (CCC_91E, CCC_91F, CCC_091) have been reviewed, and the R functions should be updated to handle _m data accordingly.

[DougManuel]

Review: PR #167 (Chronic conditions)

Reviewed 19 CCC variables (expanded from original 7 after Rafi's update) plus number_conditions derived variable. L6 integration testing passed for all PUMF cycles. Source variable names verified against MCP metadata database.

Full review details in CEP-012 (ceps/cep-012-chronic-conditions/).

Fixes applied (commit bd90746)

1. _s → _m in databaseStart (114 rows across all CCC_* variables): Converted cchs2009_s, cchs2010_s, cchs2012_s to single-year master equivalents.
2. _NA::a → _NAa / _NA::b → _NAb in dummyVariable (102 rows): Colons are invalid in identifiers.

L6 integration results

All 19 variables pass rec_with_table(). Expected gaps:

• CCC_091 (COPD): MISS for 2005 and 2007-2008 (question split into 91A/91E/91F those cycles)
• CCC_111 (epilepsy): MISS from 2007+ (dropped from CCHS)
• CCC_181 (dementia): MISS all cycles (master-only, never in PUMF)
• CCC_280 (mood disorder): MISS for 2001 (introduced 2003)

No step changes at era boundaries for CCC_091 prevalence.

Respiratory conditions (CCC_091 family)

Three questionnaire eras documented:

• 2001-2003: 91A (bronchitis, standalone) + 91B (COPD/emphysema combined)
• 2005-2008: 91A + 91E (emphysema) + 91F (COPD) as three separate questions
• 2009+: Recombined as CCC_091 "Do you have COPD (eg bronchitis, emphysema)?" → CCC_030 in 2015+

Harmonized CCC_091 correctly maps to 91B (2001-2003), 91F (2005-2008 master), CCC_091 (2009-2014), CCC_030 (2015+).

CCC_181 (Alzheimer's/dementia)

Confirmed master-only (suppressed from PUMF for disclosure control). Universe changed: 18+ (2001-2008) → 35+ (2009-2014) → 41+ (2015+). 2015+ rename CCC_181 → CCC_145 correctly mapped.

Pre-existing issues (not introduced by this PR)

• CCC_151 ccsh2009_2010_m typo (swapped s/h in databaseStart)
• number_conditions requires full feeder chain including DHHGAGE_cont (documented in roxygen)

Recommendation

PR is good to merge. Close and delete the branch after merge.

[StaceyFisher]

For COPD/emphysema/chronic bronchitis (CCC_091):

As mentioned above, there are three 'eras' for this concept:
2001-2003: 91A (bronchitis, standalone) + 91B (COPD/emphysema combined)
2005-2008: 91A + 91E (emphysema) + 91F (COPD) as three separate questions
2009+: Recombined as CCC_091 "Do you have COPD (eg bronchitis, emphysema)?" → CCC_030 in 2015+

However, in 2009/10/11, CCC_091 is labelled 'Has COPD' however, the question actually asks "Do you have chronic bronchitis, emphysema or chronic obstructive pulmonary disease or COPD?', therefore all three of these conditions need to be combined in the cycles in which they are separated.

1.1 (2000/01): CCCA_91A (CB) or CCC_91B (E + COPD)
2.1 (2003) : CCCC_91A (CB) or CCC_91B (E + COPD)
3.1 (2005) : CCCE_91A (CB) or CCCE_91E (E) or CCCE_91F (COPD)
4.1 (2007/08): CCC_91A (CB) or CCC_91E (E) or CCC_91F (COPD)
2009 - 2014: CCC_091 (CB or E or COPD)
2015 - 2023: CCC_030 (CB or E or COPD)
Abbreviations: Chronic bronchitis (CB); Emphysema (E); COPD

Of note, the documentation also talks of CCCDCPD as a derived variable in 4.1 that would also work, but I don't see this variable in the ICES data dictionary

[rafdoodle]

Hello Stacey,

Your summary of the three eras is exactly right. Here's what I implemented in cchsflow to harmonize across them:

CCC_091 (harmonized "any chronic respiratory condition" flag):

• 2001/2003: CCC_091_fun1(CCC_91A, CCC_91B) — returns 1 if either chronic bronchitis (91A) or COPD/emphysema (91B) is positive
• 2005/2007-08: CCC_091_fun2(CCC_91A, CCC_91E, CCC_91F) — returns 1 if any of CB (91A), emphysema (91E), or COPD (91F) is positive
• 2009–2014: direct recode of CCC_091 (already combined in the survey)
• 2015–2018: maps CCC_030 → CCC_091 (already handled via cycle-specific variableStart)

COPD_Emph_der (age-stratified derived variable):

This variable classifies respondents as: (1) age ≥35 with condition, (2) age <35 with condition, or (3) no condition. We updated it to properly cover all cycles including the annual/master file split, using the era-appropriate source variables:

• 2001/2003: COPD_Emph_der_fun2(age, CCC_91B) — uses CCC_91B (COPD/emphysema combined)
• 2005/2007-08: COPD_Emph_der_fun1(age, CCC_91E, CCC_91F) — uses emphysema + COPD separately
• 2009+: COPD_Emph_der_fun2(age, CCC_091) — uses the harmonized CCC_091

Note that COPD_Emph_der intentionally excludes standalone chronic bronchitis (CCC_91A) for the 2001–2008 eras, consistent with its focus on COPD/emphysema specifically. CCC_091 is the appropriate variable if you want all three conditions combined.

Didn't see CCCDCPD in the 613 data dictionary anywhere.

Please let me what you think. More chronic condition variables still need fixing.

[DougManuel]

Note from v3-smoking branch review (2026-03-13):

check_recode_blocks() — a new validation function added on the v3-smoking branch — flags COPD_Emph_der as having a recStart collision across two DerivedVar:: blocks. After the Func:: router rows are excluded, two recode blocks remain with recStart = N/A across different variableStart values (DerivedVar::[DHHGAGE_cont, CCC_091] and DerivedVar::[DHHGAGE_cont, CCC_91E, CCC_91F, CCC_091]). This is likely a pre-existing structural issue but should be reviewed before merging.

[rafdoodle]

Changes manually merged to v3 via commit 96f627b. Will close this PR now.