O'Connor et al developed a model for predicting the risk of mortality after discharge for a patient hospitalized with heart failure. This can help to identify a patient who may benefit from more aggressive management. The authors are Duke University and other medical centers around the United States.
Parameters:
(1) age in years
(2) body weight in kilograms
(3) systolic blood pressure in mm Hg
(4) serum sodium in mmol/L
(5) serum creatinine in mg/dL
(6) history of liver disease
(7) history of depression
(8) history of reactive airway disease
points for age over a range of 25 to 95 years =
= (0.2476 * (age in years)) - 6.357
points for body weight over a range of 60 to 140 kg =
= (0.000357 * ((weight)^2)) - (0.1614 * (weight)) + 17.49
points for systolic blood pressure over a range of 80 to 280 mm Hg =
= (0.000280 * ((pressure)^2)) - (0.2134 * (pressure)) + 38.60
points for serum sodium over a range of 110 to 140 mmol/L =
= (-0.4 * (sodium)) + 56
points for serum creatinine over a range of 0 to 4 mg/dL =
= (4.7 * (creatinine))
|
Parameter
|
Finding
|
Points
|
|
history of liver disease
|
no
|
0
|
|
|
yes
|
8
|
|
history of depression
|
no
|
0
|
|
|
yes
|
4
|
|
history of reactive airway disease
|
no
|
0
|
|
|
yes
|
4
|
total points =
= SUM(points for all 8 parameters)
Interpretation:
• minimum score: 0
• maximum score: 97
• The higher the score the greater the mortality rate.
If the data in Figure 1 is modeled in Minitab:
probability of mortality =
= 1/ (1 + EXP((-1) * X))
X =
= (0.00094 * ((total points)^2)) + (0.00498 * (total points)) - 5.409
