Model May Predict Rheumatoid Arthritis
New Assessment Could Identify Which Arthritis Patients Need Aggressive Treatment
WebMD News Archive
Jan. 30, 2007 -- A new prediction model developed in the Netherlands may
make it easier for doctors to identify which arthritis patients need early,
aggressive treatment and which do not.
It is now clear that the best strategy for preventing potentially crippling
joint damage in patients with rheumatoid arthritis is very early, aggressive
treatment with a potentially toxic combination of drugs.
But not all patients with arthritis have the progressive form of the
Studies suggest that pain and stiffness symptoms resolve on their own in
time in as many as half of newly diagnosed patients with undifferentiated
arthritis. Undifferentiated arthritis is arthritis that doesn't meet criteria
for a more specific type.
But about a third of undifferentiated arthritis patients end up with a
diagnosis of rheumatoid arthritis, an autoimmune disease that affects joints
and other parts of the body.
In an effort to help guide treatment decisions, researchers in the
Netherlands have developed a model for predicting a patient’s rheumatoid
arthritis risk. The research is published in the February issue of Arthritis
“This model would be very easy to adapt to clinical practice, because it is
based on assessments rheumatologists already make,” researcher Annette van der
Helm-van Mil, MD, PhD, tells WebMD.
Model Identified Rheumatoid Arthritis Early
The model was developed using data from 570 newly diagnosed patients with
undifferentiated arthritis who were followed for a year.
During that time, 177 were diagnosed with rheumatoid arthritis, while the
remaining 393 either achieved remission, did not progress, or were diagnosed
with other rheumatologic diseases.
Using a combination of questionnaires, physical examinations, and blood
samples, van der Helm-van Mil and her colleagues from the Leiden University
Medical Center developed their nine-point model.
Rheumatoid Arthritis Danger Signs
Important predictive variables included a patient’s age, sex (most
rheumatoid arthritis patients are women), number of tender joints and swollen
joints, and certain symptoms characteristic of rheumatoid arthritis -- such as
morning stiffness and location of affected joints.
Other tests, including blood tests for C-reactive protein level and
rheumatoid factor, were also included in the model.
Based on the assessments, the researchers came up with a 14-point predictive
score, with 0 being the lowest likelihood of progression to rheumatoid
arthritis and 14 representing the highest likelihood.
None of the study's patients with a score of 3 or less ended up with a
diagnosis of rheumatoid arthritis; all of those with a score of 11 or greater
The likelihood of progression to rheumatoid arthritis increased in tandem
with the scores for those between 4 and 10.
Van der Helm-van Mil says the findings must be confirmed in other patient
populations. But she says she’s confident the model can be useful in hospitals
and doctor's offices.
“This model has very good predictive ability,” she says. “It is very
Clinical Value Uncertain
Dallas rheumatologist Scott J. Zashin, MD, tells WebMD the model may prove
to be a useful tool for predicting rheumatoid arthritis.
One major unanswered question, he adds, is whether its use will lead to
different treatment decisions. “I’m not sure that it will, but it would be
worthwhile to find out.”
Zashin is a clinical assistant professor at the University of Texas
Southwestern Medical Center in Dallas.
“For years we have been using our own clinical judgments, based on the
measurements used in this model, to make decisions about treatment,” says
“Formalizing these measurements may help us better identify the patients who
will benefit from early treatment, but I think that remains to be seen,” he