Levine et al developed a model for predicting 1-year mortality for an older adult in the hospital. This can help to identify a patient who may benefit from more aggressive management. The authors are from the University of Chicago.
Patient selection: older adult (>- 65 years of age) in the hospital
Outcome: 1-year mortality after discharge
Parameters:
(1) age in years
(2) length of hospital stay in days
(3) discharge location
(4) diagnosis of metastatic cancer
(5) congestive heart failure
(6) peripheral vascular disease
(7) renal disease
(8) hematologic or solid nonmetastatic cancer
(9) dementia
|
Parameter
|
Finding
|
Points
|
|
age in years
|
65 to 69 years
|
0
|
|
|
70 to 74 years
|
1
|
|
|
>= 75 years
|
2
|
|
length of stay
|
< 5 days
|
0
|
|
|
>= 5 days
|
1
|
|
discharge location
|
nursing home
|
1
|
|
|
other
|
0
|
|
metastatic cancer
|
absent
|
0
|
|
|
present
|
2
|
|
congestive heart failure
|
absent
|
0
|
|
|
present
|
1
|
|
peripheral vascular disease
|
absent
|
0
|
|
|
present
|
1
|
|
renal disease
|
absent
|
0
|
|
|
present
|
1
|
|
hematologic or nonmetastatic cancer
|
absent
|
0
|
|
|
present
|
1
|
|
dementia
|
absent
|
0
|
|
|
present
|
1
|
where:
• The handling of malignancy could cause problems. Metastatic cancer would supersede nonmetastatic solid cancer, but a patient could have both a metastatic cancer and a hematologic malignancy.
total score =
= SUM(points for all 9 parameters)
Interpretation:
• minimum score: 0
• maximum score: 11
• The higher the score the greater the 1-year mortality.
• For a patient with a score of 0 or 1 the 1-year mortality was 11%.
• For a patient with a score >=4 (high risk group) the mortality was 48%.
Performance:
• The area under the ROC curve was 0.70 (derivation cohort) vs 0.68 (validation cohort).
