Description

Fries et al developed a score for predicting the risk of an adverse gastrointestinal event in a patient with Rheumatoid Arthritis taking NSAIDS (non-steroidal anti-inflammatory agents). This is based on risk factors identified using stepwise logistic regression. The authors are from the Stanford University School of Medicine.


 

NSAIDS included (available in 1991): aspirin, naproxen, ibuprofen, piroxicam, indomethacin, sulindac, meclofenamate, tolmetin, fenoprofen, ketoprofen, nonacetylated salicylates, salsalate, diflunisal, diclofenac.

 

Parameters associated with GI complications from NSAID use:

(1) age

(2) history of a previous NSAID GI side effect

(3) disability index or American Rheumatism Association (ARA) class

(4) NSAID dose as a fraction of maximum recommended dose

(5) current prednisone use (even low daily doses)

 

Parameter

Finding

Points

history of previous NSAID GI side effect

absent

0

 

present

1

disability index or ARA class

disability 0, or ARA 1

0

 

disability 1, or ARA 2

1

 

disability 2, or ARA 3

2

 

disability 3, or ARA 4

3

current prednisone use

absent

0

 

present

1

 

 

RA Function

ARA class

normal

1

adequate

2

limited

3

unable

4

 

GI event score =

= (2 * (age in years)) + (50 * (points for history of previous GI side effect)) + (10 * (points for disability index or ARA class)) + (15 * (NSAID dose as a fraction of maximum recommended daily dose)) + (40 * (points for current prednisone use))

 

risk per year of a GI event in a rheumatoid arthritis patient while on NSAIDs =

= ((GI event score) – 100) / 40

 

Performance:

• The actual risk vs predicted risk from the equation is shown in Figure 3, page 220. This shows a linear relation with a strong correlation.

 

Limitations:

• The predicted risk may be too high in a person taking a newer NSAID formulated for reduced GI toxicity.

 


To read more or access our algorithms and calculators, please log in or register.