Fig. 4. Evaluation of differences (bias) over time in 360 paired pulse hemoglobin (SpHb) to arterial hemoglobin measurements. No systematic relationship was found between bias and elapsed time since first arterial hemoglobin sample (R 2less than 0.01). Bias was the same for pairs in each hourly interval. However, individual variability was present (A ). Bias was higher in the sixth paired sample compared with the first or second samples (P = 0.003) but not other samples. There was no systematic relationship between increasing sample number and bias (R 2= 0.04) (B ). Plot of mean and range of SpHb to arterial hemoglobin bias in the 88 patients with two or more paired samples. Results are shown for individuals with increasing numbers of paired measurements on the x-axis. In 19 patients, bias was within ±1 g/dl for all sample pairs. The remaining patients had bias outside this range, including 32 patients with at least one paired measurement with positive and one paired measurement with negative bias or a range of bias more than 2 g/dl (C ).