Early detection of patients at high risk for acute kidney injury during disasters: development of a scoring system based on the Bam earthquake experience

Abstract

Background: Early prophylactic hydration therapy in patients prone to acute kidney injury (AKI) can reduce its incidence in disasters. As sufficient availability of fluids might be problematic, it is important to discriminate those at risk versus not at risk for AKI. The present study uses biochemical findings from victims of the Bam earthquake to design decision rules for early detection of patients at risk for AKI.

Methods: Data were collected in hospitals admitting Bam earthquake victims. Biochemical factors on day 1 (creatinine, CPK, LDH, SGOT, uric acid, calcium, phosphorus, sodium, potassium, white blood cell count, platelet count) were introduced into multivariate models for prediction of AKI (serum creatinine >1.6 mg/dL) at a later stage.

Results: A rule of thumb to assess the absence of risk for AKI was developed, using ROC analysis: if on day 1, serum creatinine was <2 mg/dL, LDH was <2,000 IU and serum uric acid was <6 mg%, the risk for development of AKI was virtually nil. Using multiple regression analysis (0.45 CPK + 2.5 LDH + 2,700 K + 2,000 uric acid - 14,000)/10,000, was most predictive for serum creatinine on day 3. Dichotomizing this value at 2.0 yielded a sensitivity and specificity for prediction of AKI of 96.6% and 95.7%, respectively.

Conclusions: We propose 2 decision rules to predict development of AKI in earthquake victims. These scores can easily be calculated already at an early stage of a disaster, enabling rationalization of prophylactic hydration therapy in crushed earthquake victims.