Owusuaa et al reported a model for predicting 1-year survival for a patient with advanced cancer. This can help to identify a patient who may benefit from more aggressive management or palliative care. The authors are from Erasmus University and affiliated hospitals.
Patient selection: advanced cancer, age 20 to 90 years
Outcome: 1-year survival
Parameters:
(1) Surprise question to the physician
(2) age of the patient in years from 20 to 90 years
(3) cancer prognostic type
(4) visceral metastases
(5) brain metastases
(6) ECOG performance status
(7) weight loss in previous 6 months
(8) VAS pain score from 0 to 10
(9) dyspnea (CTCAE grade)
(10) hemoglobin in mmol/L (conversion: multiple g/dL by 0.6206
(11) serum CRP in mg/L
(12) serum albumin in g/L
points for age =
= (0.593 * (age)) - 11.857
points for pain score
= (2.78 * (VAS score))
points for hemoglobin
= 152 - (19 * (hemoglobin))
points for CRP
= (32.63 * LOG10(CRP)) + 9.806
points for serum albumin
= (99.73 - (1.81 * (albumin))
|
Parameter
|
Finding
|
Points
|
|
surprise question
|
yes
|
0
|
|
|
no
|
100
|
|
cancer prognostic type
|
good (breast, prostate, thyroid)
|
0
|
|
|
poor
|
25.3
|
|
visceral metastases
|
no
|
0
|
|
|
yes
|
20.2
|
|
brain metastases
|
no
|
0
|
|
|
yes
|
34.1
|
|
ECOG performance
|
0
|
2
|
|
|
1
|
0
|
|
|
2 to 4
|
28.6
|
|
weight loss
|
none
|
0
|
|
|
any
|
7.7
|
|
dyspnea
|
Grade 0
|
0
|
|
|
Grade 1
|
16
|
|
|
Grade 2 to 4
|
22.4
|
total score =
= SUM(points for all of the parameters)
Interpretation:
• minimum score: 0
• maximum score: 513.4
value of X =
= (0.0000295 * ((score)^2)) + (0.0049148 * (score)) - 2.7259
probability of 1-year mortality =
= 1 / (1 + EXP((-1) * X))
Performance:
• The area under the ROC curve is 0.76.
