Description

Del Carpio et al reported a model for predicting severe acute kidney injury (AKI) acquired by a hospitalized patient on a general hospital ward. This can help to identify a patient who may benefit from more aggressive management. The authors are from multiple institutions in Spain.


Patient selection: hospitalized patient not in intensive care unit (ICU), AKI stage 0 to 3

 

Outcome: severe AKI (increase in serum creatinine >= 3 times over baseline OR >= 4 mg/dL) with onset after the first 24 hours of admission

 

Parameters:

(1) age in years

(2) hypertension

(3) diabetes

(4) peripheral vascular disease

(5) anemia

(6) congestive heart failure

(7) ischemic heart disease

(8) chronic obstructive pulmonary disease (COPD)

(9) chronic liver disease

(10) chronic urologic disease

(11) CKD stage on admission

(12) SIRS vs shock

(13) acute heart failure

(14) major surgery

(15) acute respiratory failure

(16) nephrotoxic drugs

(17) exposure to contrast dyes

(18) urgent admission (probably includes emergency admission)

 

Parameter

Finding

Points

age in years

 

0.024 * (age)

hypertension

no

0

 

yes

0.539

diabetes

no

0

 

yes

1.184

peripheral vascular disease

no

0

 

yes

0.597

anemia

no

0

 

yes

0.664

congestive heart failure

no

0

 

yes

0.405

ischemic heart disease

no

0

 

yes

0.653

COPD

no

0

 

yes

0.469

chronic liver disease

no

0

 

yes

1.013

chronic urologic disease

no

0

 

yes

1.309

CKD stage

none

0

 

stage 1

0.582

 

stage 2

1.425

 

stage 3

2.187

SIRS or shock

no

0

 

SIRS

0.698

 

shock

2.055

acute heart failure

no

0

 

yes

0.801

major surgery

no

0

 

yes

1.213

acute respiratory failure

no

0

 

yes

1.283

nephrotoxic drugs

no

0

 

yes

0.345

exposure to contrast dyes

no

0

 

yes

0.931

urgent admission

no

0

 

yes

1.899

 

value of X =

= SUM(points for all of the parameters) - 11.211

 

probability of developing severe AKI =

= 1 / (1 + EXP((-1) * X))

 

Performance:

• The area under the ROC curve is 0.91.


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