Description

Philbin and DiSalvo developed a risk score for predicting readmission for a patient hospitalized with heart failure. This can help identify patients who may benefit from more aggressive management. The authors are from Henry Ford Hospital (Detroit) and Massachusetts General Hospital.


Parameters:

(1) race

(2) health coverage

(3) health services on discharge

(4) ischemic heart disease

(5) valvular heart disease

(6) diabetes mellitus

(7) renal disease

(8) chronic lung disease

(9) idiopathic cardiomyopathy

(10) history of cardiac surgery

(11) use of telemetry during the index hospitalization

(12) treatment hospital

(13) echocardiogram performed during the index hospitalization

(14) cardiac catheterization performed during the index hospitalization

 

Parameter

Finding

Points

race

Black

1

 

other

0

health coverage

Medicare

1

 

Medicaid

1

 

other

0

health services on discharge

none

0

 

home health care services

1

 

skilled nursing facility

(-1)

ischemic heart disease

absent

0

 

present

1

valvular heart disease

absent

0

 

present

1

diabetes mellitus

absent

0

 

present

1

renal disease

absent

0

 

present

1

chronic lung disease

absent

0

 

present

1

idiopathic cardiomyopathy

absent

0

 

present

1

history of cardiac surgery

absent

0

 

present

1

use of telemetry

absent

0

 

present

1

treatment hospital

urban

0

 

rural

(-1)

echocardiography

not performed

0

 

performed

(-1)

cardiac catheterization

not performed

0

 

performed

(-1)

 

where:

• A rural hospital is any not urban. An urban hospital is one located in a county which is part of a Federal Metropolitan Statistical Area.

 

total risk score =

= SUM(points for all 14 parameters) + 4

 

Interpretation:

• minimum score: 0 (adding 4 prevents a negative score)

• maximum score: 15

• The higher the score the more likely the patient will be readmitted.

 

Total Score

Probability of Readmission

0 to 3

10%

4

14%

5

16%

6

20%

7

23%

8

27%

9

29%

10

34%

11 to 15

46%

from Figure 1, page 1564

 

Performance:

• The area under the ROC curve is 0.60.


To read more or access our algorithms and calculators, please log in or register.