BACKGROUND

High-volume fluid resuscitation and the administration of sodium bicarbonate and diuretics have a theoretical renoprotective role in patients at high risk of acute kidney injury (AKI) following rhabdomyolysis. Abnormally elevated creatine kinase has previously been used as a biological marker for the identification of patients at high risk of AKI following rhabdomyolysis.

OBJECTIVE

To assess the sensitivity and specificity of plasma creatine kinase (admission and peak values) for the prediction of AKI requiring renal replacement therapy (RRT) or of death in patients with confirmed rhabdomyolysis. To compare the diagnostic performance of creatine kinase with the McMahon score.

DESIGN

Retrospective observational study. Data collection included McMahon and Acute Physiology and Chronic Health Evaluation II (APACHE II) scores; daily creatine kinase; daily creatinine and electrolytes; ICU length of stay and mortality.

SETTING

Neurosciences and Trauma Critical Care Unit (Cambridge, UK).

PATIENTS

In total, 232 adults with confirmed rhabdomyolysis (creatine kinase > 1000 Ul) admitted to Neurosciences and Trauma Critical Care Unit between 2002 and 2012.

MAIN OUTCOME MEASURES

AKI, RRT and mortality.

RESULTS

Forty-five (19%) patients developed AKI and 29 (12.5%) patients required RRT. Mortality was significantly higher in patients who developed AKI (62 vs. 18%, P < 0.001). Average creatine kinase on admission was 5009 (range 69-157 860) Ul. Creatine kinase peaked between the day of admission and day 3 in 91% of cases. PEAK creatine kinase of at least 5000 Ul is 55% specific and 83% sensitive for the prediction of AKI requiring RRT. A McMahon Score of at least 6 calculated on admission is 68% specific and 86% sensitive for RRT.

CONCLUSIONS

Creatine kinase is not a specific or early predictor of AKI in patients with rhabdomyolysis. Although a PEAK creatine kinase of at least 5000 Ul has sensitivity acceptable for screening purposes, this is often a delayed finding. A McMahon score of at least 6 calculated on admission allows for a more sensitive, specific and timely identification of patients who may benefit from high-volume fluid resuscitation.