Figure 2.

Flow diagram. A total of 1167 patients were screened. Exclusion criteria included unremovable fingernail polish, admission for a vascular complication (e.g., grafting or stenting), amputation, and large areas of skin discoloration where the accuracy of skin tone measurements could be affected due to arterial insufficiency or cytopenias. Pairs containing either a Sao₂ or a Spo₂ measurement of the 70–100% range were excluded. Of these, 301 patients qualified for this prospective study and were approached. Among the 134 patients who signed consent forms, one patient later withdrew, one patient did not have complete skin measurement data, and four patients did not have skin measurements. For patients who had pulse oximetry measurements done on the finger, we used the average of four palm locations (left ventral, right ventral, left dorsal, right dorsal). For patients who did not have pulse oximetry locations specified, we presumed the measurement was done on the finger and imputed it using the four palm locations as well.