• Konlawij Trongtrakul, Jayanton Patumanond, Piyarat Phairatwet, Chaiwut Sawawiboon, Anusang Chitsomkasem, Sathit Kurathong, Surasee Prommoon, Thananda Trakarnvanich, and Phichayut Phinyo.
• Department of Internal Medicine, Faculty of Medicine, Chiang Mai University, Chiang Mai 50200, Thailand.
• Healthcare (Basel). 2021 Feb 15; 9 (2).

BackgroundAcute kidney injury (AKI) is a common complication encountered in an intensive care unit (ICU). In 2020, the AKI prediction score was developed specifically for critically ill surgical patients who underwent major non-cardiothoracic surgeries. This study aimed to externally validate the AKI prediction score in terms of performance and clinical utility.MethodsExternal validation was carried out in a prospective cohort of patients admitted to the ICU of the Faculty of Medicine Vajira Hospital between September 2014 and September 2015. The endpoint was AKI within seven days following ICU admission. Discriminative ability was based on the area under the receiver operating characteristic curves (AuROC). Calibration and clinical usefulness were evaluated.ResultsA total of 201 patients were included in the analysis. AKI occurred in 37 (18.4%) patients. The discriminative ability dropped from good in the derivation cohort, to acceptable in the validation cohort (0.839 (95%CI 0.825-0.852) vs. 0.745 (95%CI 0.652-0.838)). No evidence of lack-of-fit was identified (p = 0.754). The score had potential clinical usefulness across the range of threshold probability from 10 to 50%.ConclusionsThe AKI prediction score showed an acceptable discriminative performance and calibration with potential clinical usefulness for predicting AKI risk in surgical patients who underwent major non-cardiothoracic surgery.

5 keywords...

hide…