Patient selection: renal replacement therapy for acute renal failure in a postoperative patient
Outcome: in-hospital mortality
Parameters:
(1) serum lactate in mmol/L
(2) SOFA score at start of renal replacement therapy
(3) age in years (score based on decade)
(4) mechanical circulatory support (MCS) use (ECMO, balloon pump, etc)
(5) total parenteral nutrition (TPN)
(6) post-CPR
(7) sepsis
points for age =
= 3 * INT((age)/10)
|
Parameter
|
Finding
|
Points
|
|
MCS use
|
no
|
0
|
|
|
yes
|
8
|
|
TPN
|
no
|
0
|
|
|
yes
|
10
|
|
post-CPR
|
no
|
0
|
|
|
yes
|
11
|
|
sepsis
|
no
|
0
|
|
|
yes
|
13
|
modified SOFA score =
= (serum lactate) + (SOFA score) + (points for age) + SUM(points for the remaining 4 parameters)
Interpretation:
• minimum score: 0
• maximum score: perhaps around 100
• The higher the modified SOFA score the greater the in-hospital mortality.
X =
= (0.099 * (modified SOFA score)) - 4.935
probability of in-hospital mortality =
= 1 / (1 + EXP((-1) * X))
Performance:
• The reported area under the ROC curve was 0.84.
