Skip to main content

Table 1 Baseline patient characteristics comparing patients who progressed to RA with those who did not in KURAMA training cohort

From: Predicting rheumatoid arthritis progression from seronegative undifferentiated arthritis using machine learning: a deep learning model trained on the KURAMA cohort and externally validated with the ANSWER cohort

Baseline characteristics

Overall (n = 210)

RA (n = 57)

Non-RA (n = 153)

p-value

age (median, [Q1-Q3])

54 [44–66]

60 [49–70]

52 [42-63.5]

0.0094

sex (female%)

72.9%

63.2%

76.5%

0.079

BMI (median, [Q1-Q3])

21.6 [19.6–24]

22.8 [20.4–25]

21.2 [19.3–23.6]

0.010

Family history of RA (positive %)

28.1%

29.8%

27.5%

0.87

Smoking (current or previous %)

31.9%

42.1%

28.1%

0.077

CRP, mg/L (median, [Q1-Q3])

1 [1–6]

7 [1-19.5]

1 [0–1]

< 0.001

ESR_1h (median, [Q1-Q3])

14 [6–26]

26 [8–48]

11 [6–19]

< 0.001

RF (median, [Q1-Q3])

8 [8–8]

8 [8–8]

8 [8–8]

 

ACPA (median, [Q1-Q3])

0.6 [0.6–0.6]

0.6 [0.6–0.6]

0.6 [0.6–0.6]

 

MMP-3 (median, [Q1-Q3])

49.8 [32.5–96.8]

95.9 [50.8-175.25]

43.7 [30.4-66.225]

< 0.001

PhGA (median, [Q1-Q3])

13 [3–26]

25 [15.5–40.5]

8 [2–20]

< 0.001

PtGA (median, [Q1-Q3])

48.5 [21–60]

56 [40.75-75]

45.5 [18–54]

< 0.001

TJC28 (median, [Q1-Q3])

2 [0–4]

3 [1–6]

1 [0–3]

< 0.001

SJC28 (median, [Q1-Q3])

1 [0–2]

2 [1–4]

0 [0–1]

< 0.001

CDAI (median, [Q1-Q3])

9.05 [5.375–14.025]

13.75 [10.1-21.125]

7.55 [4.375–11.6]

< 0.001

HAQ-DI (median, [Q1-Q3])

0.38 [0-0.75]

0.63 [0.38-1]

0.25 [0-0.5]

< 0.001

Days to diagnosis (median, [Q1-Q3])

 

37 [28–76]

105 [35–427]

 
  1. Data showing descriptive statistics before imputation
  2. Abbreviations
  3. BMI: body mass index, CRP: C-reactive protein, ESR: erythrocyte sedimentation rate, RF: rheumatoid factor, ACPA: anti-citrullinated protein antibody, MMP-3: matrix metalloproteinase 3, PhGA: physician global assessment, PtGA: patient global assessment, TJC: 28 tender joint counts, SJC: 28 swollen joint counts, CDAI: clinical disease activity index, HAQ-DI: Health Assessment Questionnaire Disability Index