Comparison of the results of the direct swab–to–RT-LAMP assay with the RT-LAMP assay using isolated RNA revealed a much broader distribution of the ΔOD measurements in negative samples (Fig. 5A versus Fig. 3A). This was likely due to a sample-specific variability that influenced the starting pH in the LAMP reaction. This might have affected the interpretability of the measurement at 30 min (ΔOD30min). We investigated how this pH shift influenced the RT-LAMP assay. For three plates, the data acquired for the RT-LAMP assay also included measurements for the 10-min time point (ΔOD10min) (Fig. 6A). We plotted the change of the ΔOD between the 10- and 30-min time points (i.e., the difference ΔOD30min – ΔOD10min, corresponding to the slope of the lines) versus ΔOD30min (Fig. 6B). This removed the variability of the values for samples that did not change their color (negative samples) and permitted a better separation of the positive from the negative samples.