Ho et al reported a prediction rule for identifying patients with severe acute respiratory syndrome (SARS). The authors are from the University of Hong Kong and Queen Mary Hospital in Hong Kong.
Patient selection: acute respiratory symptoms
Parameters:
(1) myalgia at presentation
(2) radiographic appearance between days 1 and 3
(3) lymphopenia between days 1 and 3 (< 1,500 per µL absolute count)
(4) elevated serum ALT between days 1 and 3 (> 53 IU/L)
(5) epidemiologic link (high risk exposure)
Parameter
|
Finding
|
Points
|
myalgia at presentation
|
absent
|
0
|
|
present
|
3
|
radiographic appearance between days 1 and 3
|
deterioration
|
10
|
|
unchanged
|
1
|
|
improvement
|
0
|
lymphopenia between days 1 and 3
|
absent
|
0
|
|
present
|
3
|
elevated serum ALT between days 1 and 3
|
absent
|
0
|
|
present
|
3
|
epidemiologic link
|
absent
|
0
|
|
present
|
11
|
total score =
= SUM(points for all 5 parameters)
Interpretation:
• minimum score: 0
• maximum score: 30
• The higher the score the greater the likelihood of SARS.
• A score >=11 was considered high risk for SARS.
Performance:
• The sensitivity was 98% and specificity 81%.
• The negative predictive value was 99.5% and positive predictive value 81%.
• The area under the ROC curve is 0.96.