Abstract
Background: Few parameters are available for diagnosing renal potassium loss in emergency patients or patients receiving treatment. This study aimed to investigate the ratio of random urine potassium‒creatinine to the synchronous serum potassium concentration squared ([UK/UCr]/SK(2)) and compare it with other parameters in the diagnostic ability of renal potassium loss. Methods: This single-center study enrolled 380 subjects, including 218 hypokalemia patients (91 with nonrenal potassium loss and 127 with renal potassium loss) and 162 normal potassium controls. The values of serum and urine were based on laboratory data. Groups were compared in pairs, and the ROC curve analysis was used to evaluate the predictive ability of parameters related to renal potassium loss. Results: (UK/UCr)/SK(2) was significantly elevated in potassium loss patients, especially in females. Moreover, a greater (UK/UCr)/SK(2) ratio was observed in those with nonrenal potassium loss, which demonstrated a trend toward increases in the normal potassium, nonrenal potassium loss, and renal potassium loss groups among males and females. Ultimately, the AUC of (UK/UCr)/SK(2) was the highest at 0.880 (95% CI: 0.822-0.938) in males and 0.878 (95% CI: 0.831-0.924) in females for potassium loss diagnosis. Conclusion: Random (UK/UCr)/SK(2) has good diagnostic value for renal potassium loss in patients with hypokalemia. Given that these serum and random urine parameters are easily obtainable from patients during treatment, regularly observing (UK/UCr)/SK(2) may prove to be an effective indicator.