From: Assessing clusters of comorbidities in rheumatoid arthritis: a machine learning approach
No comorbidities | + Individual comorbidities | + Clustering K-mode | + Clustering regression | + Clustering DIANA | |
---|---|---|---|---|---|
Beta coefficient (95% confidence interval) | |||||
Race/ethnicity | |||||
Asian | − 0.31 (− 1.34, 0.73) | 0.23 (− 0.78, 1.25) | − 0.05 (− 1.08, 0.97) | − 0.09 (− 1.11, 0.94) | − 0.29 (− 1.33, 0.74) |
Black | 0.70 (0.14, 1.26) | 0.72 (0.16, 1.29) | 0.68 (0.12, 1.25) | 0.68 (0.12, 1.24) | 0.69 (0.13, 1.26) |
Hispanic | − 0.08 (− 0.64, 0.48) | − 0.01 (− 0.56, 0.53) | − 0.04 (− 059, 0.51) | − 0.03 (− 0.58, 0.52) | − 0.08 (− 0.63, 0.45) |
Other | 0.43 (− 0.40, 1.26) | 0.54 (− 0.28, 1.36) | 0.47 (− 0.35, 1.30) | 0.56 (− 0.27, 1.38) | 0.42 (− 0.41, 1.26) |
White | reference | reference | reference | reference | reference |
Duration of RA, years | 0.03 (0.01, 0.04) | 0.02 (0.01, 0.04) | 0.02 (0.01, 0.04) | 0.03 (0.01, 0.04) | 0.03 (0.01, 0.04) |
Erosions | 0.50 (0.21, 0.79) | 0.61 (0.32, 0.89) | 0.48 (0.19, 0.76) | 0.57 (0.29, 0.86) | 0.49 (0.21, 0.78) |
Serologic status, positive | − 0.88 (− 1.18, − 0.58) | − 0.72 (− 1.02, − 0.43) | − 0.84 (− 1.14, − 0.54) | − 0.76 (− 1.06, − 0.46) | − 0.88 (− 1.18, − 0.58) |
CDAI, baseline | 0.30 (0.29, 0.32) | 0.29 (0.28, 0.30) | 0.30 (0.29, 0.31) | 0.30 (0.28, 0.31) | 0.30 (0.29, 0.32) |
HAQ-DI, baseline | 3.43 (3.08, 3.79) | 2.60 (2.24, 2.96) | 3.11 (2.76, 3.47) | 3.00 (2.64, 3.35) | 3.43 (3.08, 3.79) |
Individual comorbidities | |||||
Coronary artery disease | --- | 0.77 (0.33, 1.21) | --- | --- | --- |
Heart failure | --- | 1.78 (0.78, 2.78) | --- | --- | --- |
Hypertension | --- | 0.58 (0.29, 0.85) | --- | --- | --- |
DVT/PE | --- | 0.72 (− 0.15, 1.58) | --- | --- | --- |
Stroke/TIA | --- | 0.68 (− 0.03, 1.40) | --- | --- | --- |
Arrhythmia | --- | − 0.64 (− 2.36, 1.08) | --- | --- | --- |
Gastrointesinal bleed | --- | 0.72 (0.05, 1.39) | --- | --- | --- |
Liver disease | --- | 0.50 (− 0.15, 1.16) | --- | --- | --- |
Solid tumor | --- | − 0.19 (− 0.96, 0.57) | --- | --- | --- |
NMSC | --- | 0.11 (− 0.32, 0.54) | --- | --- | --- |
Lymphoma | --- | 0.67 (− 0.89, 2.24) | --- | --- | --- |
Other cancer | --- | − 0.01 (− 0.53, 0.51) | --- | --- | --- |
Melanoma | --- | 0.57 (− 0.30, 1.44) | --- | --- | --- |
Diabetes | --- | 0.54 (0.17, 0.92) | --- | --- | --- |
Hyperlipidemia | --- | 0.57 (0.29, 0.85) | --- | --- | --- |
Osteoporosis | --- | 0.11 (− 0.22, 0.44) | --- | --- | --- |
Demyelinating | --- | − 1.01 (− 2.79, 0.78) | --- | --- | --- |
Mental health | --- | 1.42 (1.15, 1.68) | --- | --- | --- |
Fibromyalgia | --- | 3.01 (2.54, 3.48) | --- | --- | --- |
Psoriasis | --- | 0.21 (− 0.35, 0.76) | --- | --- | --- |
Asthma/COPD | --- | 0.20 (− 0.33, 0.74) | --- | --- | --- |
RA Lung | --- | 0.41 (− 0.34, 1.15) | --- | --- | --- |
Acute kidney injury | --- | − 0.06 (− 0.47, 0.34) | --- | --- | --- |
Clusters from ML | |||||
1 | --- | --- | reference | reference | reference |
2 | --- | --- | 0.03 (− 0.48, 0.54) | 2.84 (2.22, 3.47) | − 0.18 (− 2.15, 1.78) |
3 | --- | --- | 1.48 (1.00, 1.97) | 0.29 (− 0.25, 0.83) | − 0.53 (− 2.73, 1.67) |
4 | --- | --- | 1.32 (0.78, 1.87) | − 0.70 (− 1.19, − 0.21) | 0.24 (− 2.33, 2.82) |
5 | --- | --- | − 0.41 (− 0.91, 0.09) | 0.60 (0.03, 1.16) | − 2.98 (− 11.2, 5.27) |
Degrees of freedom | 22 | 46 | 26 | 26 | 26 |
Model fit statistics | |||||
Adjusted R2 | 0.30 | 0.33 | 0.31 | 0.31 | 0.30 |
Root mean square error | 7.07 | 6.95 | 7.03 | 7.00 | 7.07 |