Lee et al developed a model for predicting the risk of a prolonged air leak following lung resective surgery. These can help to identify a patient who may benefit from more aggressive management. The authors are from McGill University in Montreal.
Parameters:
(1) type of lung resection
(2) pleural adhesions
(3) observed FEV1 as percent of predicted (from 0 to 100)
(4) observed DLco as percent of predicted (from 0 to 100)
observed FEV1 as percent of predicted =
= (observed FEV1) / (predicted) * 100%
observed DLco as percent of predicted =
= (observed DLco) / (predicted) * 100%
If the lung resection was limited to a wedge resection, then the risk of a prolonged air leak was 4.8%. If a larger resection was performed then a logistic regression model can be used.
X =
= (-0.483 if pleural adhesions present) - (0.026 * (100 - (FEV1 as percent of predicted)) - (0.013 * (100 - (DLco as percent of predicted)) + 2.42
probability of not having a prolonged air leak =
= 1/ (1 + EXP((-1) * X)
probability of having a prolonged air leak =
= 1 - (1/ (1 + EXP((-1) * X))
This can also be represented as a score:
|
Parameter
|
Finding
|
Points
|
|
pleural adhesion
|
absent
|
0
|
|
|
present
|
2
|
points for FEV1 =
= (100 - (FEV1 as percent of predicted)) / 10
points for DLco =
= (100 - (DLco as percent of predicted)) / 20
total score =
= (points ofr pleural adhesions) + (points for FEV1) + (points for DLco)
|
Total Score
|
Percent with Air Leak
|
|
0
|
8.2%
|
|
1
|
10.2%
|
|
2
|
12.8%
|
|
3
|
15.8%
|
|
4
|
19.5%
|
|
5
|
23.7%
|
|
6
|
28.5%
|
|
7
|
33.8%
|
|
8
|
39.7%
|
|
9
|
45.8%
|
|
10
|
52%
|
|
11
|
58.2%
|
The equation for this is:
risk in percent =
= (0.2371 * ((score)^2)) + (2.023 * (score)) + 7.895
