Objective: To predict severe hyperkalemia from single electrocardiogram (ECG) tracings.
Methods: Ten conditioned dogs each underwent this protocol three times: Under isoflurane anesthesia, 2 mEq/kg/hr of potassium chloride was given intravenously until P-waves were absent from the ECG and ventricular rates decreased > or =20% in < or =5 minutes. Serum potassium levels (K(+)) were measured at regular intervals with concurrent digital storage of lead II of the surface ECG. A three-layer artificial neural network with four hidden nodes was trained to predict K(+) from 15 separate elements of corresponding ECG data. Data were divided into a training set and a test set. Sensitivity, specificity, and diagnostic accuracy for recognizing hyperkalemia were calculated for the test set based on a prospectively defined K(+) = 7.5.
Results: The model produced data for 189 events; 139 were placed in the training set and 50 in the test set. The test set had 37 potassium levels at or above 7.5 mmol/L. The neural network had a sensitivity of 89% (95% CI = 75% to 97%) and a specificity of 77% (95% CI = 46% to 95%) in recognizing these. The positive likelihood ratio was 3.87. Overall accuracy of this model was 86% (95% CI = 73% to 94%). Mean (+/-SD) difference between predicted and actual K(+) values was 0.4 +/- 2.0 (95% CI = -0.2 to 1.0).
Conclusions: An artificial neural network can accurately diagnose experimental hyperkalemia using ECG parameters. Further work could potentially demonstrate its usefulness in bedside diagnosis of human subjects.