Skip to content

Dataset Comparison Guide

This guide helps you understand the similarities and differences between Orca datasets to choose the right data for your analysis.

Quick Comparison Table

Feature GI-DAMPs MUSIC Mini-MUSIC
Study ID Prefix GID- MID- MINI-
Population Adults only Adults only Pediatric only
Data Structure Sampling visits Fixed timepoints Fixed timepoints
Timepoints Variable 1-5 1-3
Columns 227 369 423
Primary Focus Biomarkers, drug monitoring Mucosal healing Pediatric outcomes, EEN
Disease Activity Scores HBI, SCCAI HBI, SCCAI, Mayo PCDAI, PUCAI

Variable Availability

Common to All Datasets

All datasets share standardized variables across demographics, laboratory values, medications, and phenotyping. For complete variable definitions and documentation, see the Unified Data Dictionary.

Dataset-Specific Variables

GI-DAMPs Unique Features

  • sampling_date, sampling_setting (inpatient/outpatient/endoscopy)
  • redcap_repeat_instance for multiple sampling visits per participant
  • comment, medication_comments (free text fields)
  • More detailed medication history with start/stop dates

MUSIC Unique Features

  • endoscopic_mucosal_healing, complete_mucosal_healing
  • endoscopic_mucosal_healing_at_3_6_months, endoscopic_mucosal_healing_at_12_months
  • cd_pro2_raw, cd_pro2_weighted, uc_pro2
  • ibdresponse_criteria_cd_met, ibdresponse_criteria_uc_met
  • saliva_sample, saliva_setting, saliva-specific variables

Mini-MUSIC Unique Features

  • pucai_score, pcdai_score (pediatric disease activity)
  • cdparis_*, ucparis_* (pediatric classification)
  • impact3_score, promis_fatigue_score (pediatric PROs)
  • een_use, een_formula_type, een_start_date, een_end_date
  • patient_age_group (6-10, 10-13, 14-18)
  • ibd_drug_1_* through ibd_drug_5_* (detailed current medication tracking)

Disease Activity Scores

Comparison of Scoring Systems

Score GI-DAMPs MUSIC Mini-MUSIC Notes
HBI (Harvey-Bradshaw) ✓ ✓ ✗ Adults with CD only
SCCAI (Simple Clinical Colitis) ✓ ✓ ✗ Adults with UC only
Mayo Score ✓ ✓ ✗ Adults with UC only
SES-CD (Crohn's endoscopy) ✓ ✓ ✓ All studies
UCEIS (UC endoscopy) ✓ ✓ ✓ All studies
PCDAI (Pediatric CD) ✗ ✗ ✓ Pediatrics only
PUCAI (Pediatric UC) ✗ ✗ ✓ Pediatrics only

Important: PCDAI and PUCAI are not comparable to HBI and SCCAI. Use pediatric scores only for Mini-MUSIC analyses.

Classification Systems

Montreal Classification (Adults)

  • Used in: GI-DAMPs, MUSIC
  • Variables: montreal_cd_location, montreal_cd_behaviour, montreal_uc_extent, montreal_uc_severity

Paris Classification (Pediatrics)

  • Used in: Mini-MUSIC
  • Variables: cdparis_location, cdparis_behaviour, cdparis_upper_gi, cdparis_growth, cdparis_perianal, ucparis_extent, ucparis_severity

Note: These systems are designed for different age groups and should not be directly compared.

Disease Activity Variables

Disease activity classifications vary significantly across studies. Each study uses different variable names and value sets for representing disease activity. For a detailed comparison table, standardization suggestions, and example implementation code, see Known Issues - Disease Activity Definitions.

Longitudinal Structure

GI-DAMPs

  • Structure: Sampling visits (not fixed intervals)
  • Key Variable: redcap_repeat_instance (instance number)
  • Visit Date: sampling_date
  • Considerations: Variable intervals between visits, based on clinical events

MUSIC

  • Structure: Fixed timepoints
  • Key Variable: redcap_event_name (timepoint_1 through timepoint_5)
  • Visit Date: visit_date
  • Considerations: Fixed intervals (Baseline, 3 months, 6 months, 9 months, 12 months)

Mini-MUSIC

  • Structure: Fixed timepoints
  • Key Variable: redcap_event_name (timepoint_1, timepoint_2, timepoint_3)
  • Visit Date: Available in dataset
  • Considerations: Fixed intervals (Baseline, 3 months, 6 months)

Medication Variables

Sampling Status (All Studies)

All studies use sampling_* prefix to indicate medications at time of visit/sampling:

  • sampling_asa, sampling_ifx, sampling_ada, sampling_vedo, sampling_uste, etc.
  • Values: 1 = yes, 0 = no

Historical Medication (Study-Specific)

  • GI-DAMPs: ifx, ada, vedo, etc. with *_start, *_stop dates
  • MUSIC: baseline_* prefix (e.g., baseline_ifx, baseline_ada)
  • Mini-MUSIC: baseline_ibd_drug_1_* through baseline_ibd_drug_5_* (structured format)

Combining Datasets

  1. Focus on Common Variables: Use variables documented in the Unified Data Dictionary

  2. Standardize Disease Activity: Consider creating a standardized variable based on:

    • has_active_symptoms
    • crp (threshold >5 mg/L)
    • calprotectin (threshold >250 μg/g)

    See Known Issues for example implementation.

  3. Respect Study-Specific Differences:

    • Don't compare pediatric scores (PCDAI/PUCAI) with adult scores
    • Don't mix Montreal and Paris classifications
    • Account for different timepoint structures

  4. Use Combined Dataset When Available:

Example: Cross-Study Analysis

Python
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
# Load datasets
gidamps = pd.read_csv('gidamps_sampling_2025-10-30.csv')
music = pd.read_csv('music_main_2025-10-30.csv')

# Select common variables
common_vars = [
    'study_id', 'study_group', 'age', 'sex', 'bmi',
    'crp', 'calprotectin', 'has_active_symptoms',
    'sampling_ifx', 'sampling_ada'
]

# Filter and combine
gidamps_subset = gidamps[common_vars].assign(source='GI-DAMPs')
music_subset = music[music['redcap_event_name'] == 'timepoint_1'][common_vars].assign(source='MUSIC')

combined = pd.concat([gidamps_subset, music_subset])

# Create standardized disease activity
def standardize_activity(row):
    if row['has_active_symptoms'] == 'yes' or row['has_active_symptoms'] == 1:
        if row['crp'] > 5 and pd.notna(row['calprotectin']) and float(str(row['calprotectin']).replace('<', '').replace('>', '')) > 250:
            return 'biochem_active'
        else:
            return 'active'
    else:
        if row['crp'] > 5 or (pd.notna(row['calprotectin']) and float(str(row['calprotectin']).replace('<', '').replace('>', '')) > 250):
            return 'remission'
        else:
            return 'biochem_remission'

combined['standardized_activity'] = combined.apply(standardize_activity, axis=1)

Choosing the Right Dataset

Use GI-DAMPs if you need:

  • ✅ Sampling-based data collection
  • ✅ Diverse recruitment settings
  • ✅ Rich biomarker data
  • ⚠️ Not suitable for fixed-interval longitudinal analyses
  • ⚠️ Adults only (no paediatric data)

Use MUSIC if you need:

  • ✅ Adult longitudinal data
  • ✅ Mucosal healing outcomes
  • ✅ Fixed timepoint structure
  • ⚠️ Adults only (no paediatric data)

Use Mini-MUSIC if you need:

  • ✅ Paediatric-specific data
  • ✅ EEN (exclusive enteral nutrition) information
  • ✅ Pediatric disease activity scores (PCDAI, PUCAI)
  • ✅ Age-appropriate classifications (Paris)
  • ⚠️ Cannot combine with adult scores

Combine Multiple Datasets if you need:

  • ✅ Cross-study comparisons
  • ✅ Larger sample sizes
  • ✅ Validation across populations
  • ⚠️ Must standardize variables first (see above)