Mehta et al developed 2 versions of score to predict mortality in a patient with a Type A aortic dissection. One version is a simple bedside score based on a logistic regression equation. The authors are investigators with the International Registry of Acute Aortic Dissection (IRAD).
Type A aortic dissection (Stanford classification): any dissection involving the ascending aorta (DeBakey Types I and II)
Parameters for score:
(1) age
(2) gender
(3) onset of chest pain
(4) ECG findings on presentation
(5) pulse deficit on presentation
(6) renal failure
(7) hypotension, shock or tamponade
|
Parameter |
Finding |
Points for LR |
Points for Bedside |
|
age |
< 70 years |
0 |
0 |
|
|
>= 70 |
1 |
0.5 |
|
gender |
male |
0 |
0 |
|
|
female |
1 |
0.3 |
|
onset of chest pain |
not abrupt |
0 |
0 |
|
|
abrupt |
1 |
1.0 |
|
ECG on presentation |
normal |
0 |
0 |
|
|
abnormal |
1 |
0.6 |
|
pulse deficit on presentation |
absent |
0 |
0 |
|
|
present |
1 |
0.7 |
|
renal failure prior to surgery |
absent |
0 |
0 |
|
|
present |
1 |
1.6 |
|
hypotension, shock or tamponade |
absent |
0 |
0 |
|
|
present |
1 |
1.1 |
where:
• The points assigned for the bedside score are the coefficients for the logistic regression equation rounded to 1 decimal place (see equation below).
|
Finding |
Odds Ratio for Death |
95% CI |
p value |
|
age >= 70 |
1.7 |
1.05 to 2.77 |
0.03 |
|
female gender |
1.38 |
0.85 to 2.27 |
0.20 |
|
abrupt onset |
2.6 |
1.22 to 5.54 |
0.01 |
|
abnormal ECG |
1.77 |
1.06 to 2.95 |
0.03 |
|
pulse deficit |
2.03 |
1.25 to 3.29 |
0.004 |
|
renal failure |
4.77 |
1.80 to 12.6 |
0.002 |
|
hypotension, etc. |
2.97 |
1.83 to 4.81 |
< 0.0001 |
from Table 4, page 203
X =
= (0.5331 * (points for age)) + (0.3224 * (points for sex)) + (0.9569 * (points for pain onset)) + (0.7089 * (points for pulse deficit)) + (0.5714 * (points for ECG)) + (1.5616 * (points for renal failure)) + (1.0876 * (points for hypotension/shock/tamponade)) – 2.94
probability of death =
= 1 / (1 + EXP((-1)*X)
bedside score =
= SUM(points for all 7 parameters)
Interpretation:
• minimum bedside score: 0
• maximum bedside score: 5.8
• The higher the score, the greater the mortality rate.
|
Bedside Score |
Mortality Rate |
|
0 |
< 5% |
|
0.5 |
16% |
|
1.0 |
8% |
|
1.5 or 2.0 |
20% |
|
2.5 or 3.0 |
42% |
|
3.5 |
62% |
|
4.0 |
69% |
|
4.5 |
80% |
approximated from Figure 2, page 204
where:
• The mortality with a score of 0.5 is higher than that of 1.0. In the implementation I have set the probability for both at 10%. I also did some rounding of percentages.
Performance:
• The area under ROC curve 0.74
• Hosmer-Lemeshow statistic 0.75
Specialty: Cardiology
