(a,b) Relative feature importance in predicting COVID-19 according to the study of Soltan et al. [19]. Two models were built, one with data based on emergency presentation (a) and one with data based on hospital admission. MetHb was a relevant parameter in both models (indicated in red). (c) COVID-19 non-survivors (n = 22) show a steeper and higher increase of COHb during the curse of the disease compared to COVID-19 survivors (n = 41), according to the study of Paccaudi et al. [24]. The two COHb time-series are statistically significantly different. The yellow bar refers to the COHb reference ranges for healthy non-smokers [16]. (d) Time-course of MetHb of a COVID-19 patient reported by Palmer et al. [22]. The yellow bar refers to the MetHb reference range for healthy non-smokers [15]. The figures show data extracted from the original figures of the respective publications.