Variability in reported midpoints of (in)activation of cardiac INa

Electrically active cells like cardiomyocytes show variability in their size, shape, and electrical activity. But should we expect variability in the properties of their ionic currents? In this meta-analysis, we gather and visualize measurements of two important electrophysiological parameters: the midpoints of activation and inactivation of the cardiac fast sodium current, INa. We find a considerable variation in reported mean values between experiments, with a smaller cell-to-cell variation within experiments. We show how the between-experiment variability can be decomposed into a correlated component, affecting both midpoints almost equally, and an uncorrelated component, affecting the midpoints independently, and we find that the correlated component is much larger than the uncorrelated one. We then review biological and methodological issues that might explain the observed variability and attempt to classify each as a within-experiment or a correlated or uncorrelated between-experiment effect. Although the existence of some variability in measurements of ionic currents is well-known, we believe that this is the first work to systematically review it and that the scale of the observed variability is much larger than commonly appreciated, which has implications for modelling and machine-learning as well as experimental design, interpretation, and reporting.