Zainab et al reported 2 models for predicting respiratory failure in a patient who has undergone cardiac surgery. One model includes the ASA status while the second does not. This can help to identify a patient who may benefit from more aggressive management. The authors are from Houston Methodist Hospital.
Patient selection: cardiac surgery
Outcome: respiratory failure
Parameters:
(1) body mass index (BMI)
(2) diabetes
(3) chronic lung disease
(4) home oxygen
(5) recent pneumonia
(6) eGFR in mL per min per 1.73 square meter
(7) previous cardiac intervention
(8) NYHA classification
(9) cardiopulmonary bypass
(10) cardiogenic shock
(11) mechanical support
(12) total number of blood product units used
|
Parameter |
Finding |
Beta-Coefficient |
Points |
|
body mass index |
< 30 kg per sq m |
0 |
0 |
|
|
>= 30 |
0.26 |
3 |
|
diabetes |
no |
0 |
0 |
|
|
yes |
0.15 |
2 |
|
chronic lung disease |
no |
0 |
0 |
|
|
yes |
0.29 |
4 |
|
home oxygen |
no |
0 |
0 |
|
|
yes |
0.49 |
6 |
|
recent pneumonia |
no |
0 |
0 |
|
|
yes |
0.35 |
3 |
|
eGFR |
>= 60 |
0 |
0 |
|
|
< 60 |
0.32 |
4 |
|
cardiac intervention |
no |
0 |
0 |
|
|
yes |
0.34 |
4 |
|
NYHA class |
1 or 2 |
0 |
0 |
|
|
3 or 4 |
0.45 |
6 |
|
CP bypass |
no |
0 |
0 |
|
|
yes |
0.36 |
4 |
|
cardiogenic shock |
no |
0 |
0 |
|
|
yes |
1.28 |
15 |
|
mechanical support |
no |
0 |
0 |
|
|
yes |
0.54 |
7 |
|
blood products |
none |
0 |
0 |
|
|
|
0.08 * (number) |
(number) |
value of X =
= SUM(beta-coefficients for all of the parameters) - 2.59955
probability of respiratory failure =
= 1 / (1 + EXP((-1) * X))
total score =
= SUM(points for all of the parameters)
Interpretation:
• minimum score: 0
• maximum score: around 70 (varies with the number of blood products received)
|
Score |
Risk Group |
Rate Respiratory Failure |
|
0 to 9 |
low |
11% |
|
10 to 16 |
moderate |
19% |
|
>= 17 |
high |
36% |
Performance:
• The area under the ROC curve is 0.70.
Specialty: Pulmonology
