de Punder et al developed 2 models for predicting joint damage progression in a patient with Rheumatoid Arthritis (RA). This can help to identify a patient who may benefit from more aggressive management. The authors are from Radboud Univeristy in Nijmegen, the Netherlands.
Patient selection: rheumatoid arthritis >= 18 years old not on DMARD
Outcome: progression of joint damage based on the Ratingen score
Parameters:
(1) subscore for risk factors
(1a) anti-CCP
(1b) rheumatoid factor (RF)
(2) number of erosions
(3) erythrocyte sedimentation rate (ESR) in mm/h
(4) age in years
(5) gender
Parameter |
Finding |
Points |
anti-CCP |
<= 25 U/mL |
0 |
|
> 25 U/mL |
1 |
rheumatoid factor |
<= 10 U/mL (negative) |
0 |
|
> 10 U/mL (positive) |
1 |
subscore =
= SUM(points for the 2 risk factors)
Parameter |
Finding |
beta coefficient |
subscore |
0 |
0 |
|
1 |
0.19 |
|
2 |
1.34 |
number of erosions |
0 |
0 |
|
1 to 5 |
0.95 |
|
6 to 10 |
1.35 |
|
> 10 |
2.61 |
ESR |
< 25 mm/h |
0 |
|
25 to 50 mm/h |
0.99 |
|
> 50 mm/h |
1.61 |
Age |
< 45 years of age |
0 |
|
45 to 64 years |
-0.18 |
|
> 64 years |
-0.89 |
gender |
male |
0 |
|
female |
-0.45 |
where:
• In the scoring of erosions the presence of 5 erosions is listed twice. The table above is one way of resolving this.
X =
= SUM(beta coefficients) – 1.43
probability of joint damage progression =
= 1 / (1 + EXP((-1)*X))
Performance:
• The area under the ROC curve is 0.77.
Specialty: Immunology/Rheumatology
ICD-10: ,