Tattevin et al developed a model for predicting pulmonary tuberculosis in a patient. This can help to target a patient who may benefit from additional testing and pulmonary isolation. The authors are from multiple hospitals in France.
Parameters:
(1) chest X-ray appearance
(2) HIV infection
(3) clinical symptoms
(4) history of BCG immunization
(5) immigrant status
(6) homeless living status
|
Parameter
|
Finding
|
Points
|
|
chest X-ray findings
|
typical
|
14
|
|
|
compatible
|
7
|
|
|
neither
|
0
|
|
HIV infection
|
yes
|
0
|
|
|
no
|
6
|
|
clinical symptoms
|
typical
|
12
|
|
|
compatible
|
5
|
|
|
neither
|
0
|
|
BCG immunization
|
never
|
2
|
|
|
> 10 years ago
|
1
|
|
|
<= 10 years ago
|
0
|
|
immigrant status
|
high risk country
|
2
|
|
|
lower risk country
|
1
|
|
|
other or nonimmigrant
|
0
|
|
homeless living status
|
yes
|
1
|
|
|
no
|
0
|
where:
• High risk countries (3): sub-Saharan Africa, Southeast Asia, Haiti
• Lower risk countries (5): North Africa, Central America, South America, Southern Europe, Eastern Europe
• There are several discrepancies between the text on page 1250 and Table 6 on page 1251.
total score =
= SUM(points for all 6 parameters)
Interpretation:
• minimum score: 0
• maximum score: 37
• A score > 18 identified patients at risk for positive sputum smears.
Performance:
• The performance varied in different populations.
• The sensitivity ranged from 82 to 100%. The specificity was around 50%, The positive predictive value was 25%.
