Description

Bell et al developed a model for predicting acute kidney injury (AKI) in patients undergoing orthopedic surgery. This can help to identify a patient who may benefit from more aggressive management. The authors are from Ninewells Hospital, Leiden University Medical Centre, and the University of Dundee.


 

Patient selection: Scottish adults undergoing orthopedic surgery

 

Outcome: any degree of acute kidney injury during first postoperative week (increase in serum creatinine concentration >= 26.4 µmol/L from preoperative OR postoperative creatinine > 1.5 times preoperative baseline)

 

Parameters:

(1) gender

(2) age in years at time of operation

(3) diabetes mellitus

(4) number of prescribed drugs

(5) therapy with ACE inhibitor or angiotensin receptor blocker

(6) baseline eGFR in mL per minute based on the CKD-EPI equation

(7) ASA grade for surgery

 

Parameter

Finding

Beta Coefficient

gender

male

0

 

female

-0.708

age in years

 

0.022 * (age)

diabetes

no

0

 

yes

0.427

prescribed drugs

0

0

 

1 or 2

0.130

 

>= 3

0.347

ACE inhibitor or angiotensin receptor blocker

no

0

 

yes

0.534

eGFR

> 60 mL per min

-1.417

 

45 to 59 mL per min

-1.108

 

30 to 44 mL per min

-0.676

 

< 29 mL per min

0

ASA

ASA 1

-1.037

 

ASA 2

-0.141

 

ASA 3

0.014

 

ASA 4

0

 

X =

= SUM(beta coefficient) – 2.385

 

probability of AKI =

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

 

Performance:

• The area under the ROC sucrve was 0.74 in the development and 0.70 in the validation cohorts.

 

Limitatons:

• The study is based on an observational cohort study.

 


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