Baldini et al developed a simple clinical prediction rule for identifying a patient with primary Sjogren's syndrome who may be at increased risk for malignant lymphoma. The authors are from the University of Pisa.
Patient selection: primary Sjogren's syndrome
Parameter
(1) salivary gland size
(2) serum C4 concentration
(3) disease duration in years
|
Parameter
|
Finding
|
Points
|
|
salivary gland size
|
not enlarged
|
0
|
|
|
enlarged
|
1
|
|
serum C4 concentration
|
normal or elevated
|
0
|
|
|
low (< 20 mg/dL)
|
1
|
|
disease duration
|
short
|
0
|
|
|
long
|
1
|
number of risk factors for malignant lymphoma =
= SUM(points for all 3 parameters)
|
Parameter
|
Odds Ratio
|
LN(odds ratio)
|
|
salivary gland enlargement
|
10.20
|
2.32
|
|
low serum C4
|
11.53
|
2.446
|
|
disease duration
|
1.14
|
0.132
|
linear predictor =
= (2.32 * (points for salivary gland size)) + (2.446 * (points for serum C4)) + (0.132 * (duration in years)
Interpretation:
• minimum score: 0
• maximum score in study: 6.5, increases with disease duration
• The higher the score the greater the risk for malignant lymphoma.
• The optimum cutoff point for lymphoma was 4.26. Scores for patients with malignant lymphoma ranged from 3.4 to 5.4.
Performance:
• The area under the ROC curve was 0.905.
• The sensitivity was 78% and specificity 95% at cutoff 4.26.
