The appearance of the blood-collection tube after the first low-speed spin and representative P-PRP and L-PRP preparations after the second high-speed spin and subsequent re-suspension are shown in Fig. 2. Although low-speed spinning did not result in the formation of a clear buffy coat in the interface between the plasma and RBC fractions, the buffy coat corresponding to the plasma was not included in the second spin for P-PRP preparation. Therefore, the resulting P-PRP was light yellow in color, not reddish. In contrast, for the L-PRP preparation, the buffy coat and the surface of the RBC fraction just below the interface were included in the second spin. The inclusion of significant amounts of RBC turned the L-PRP red. The strength of this color was variable depending on the operators’ pipetting skills; however, L-PRP preparations were more or less reddish when the maximum amount of platelets was recovered.

To characterize both the P-PRP and L-PRP preparations used for the calibration curves, blood cells were counted using an AHA (Fig. 3). For platelet counts, there was no significant difference between the two types of PRP. For WBC and RBC counts, in contrast, L-PRP contained significantly more WBCs and RBCs than P-PRP.

The samples were serially diluted, and platelets in individual dilutions were counted using the AHA. In parallel, the absorbance of each sample was measured with the SPM. The resulting calibration curves for P-PRP and L-PRP are shown in Fig. 4. Compared with P-PRP, the calibration curves for L-PRP varied with the samples and appeared generally inappropriate for linear regression. The calibration curve for P-PRP was expressed as “y = 0.00308x − 0.0157,” while that of L-PRP was “y = 0.00852x − 0.638.” The SD values for both the slope and intercept values were much higher in L-PRP. In addition, the R² value (coefficient of determination) for the linear regression of P-PRP was 0.995, while that of L-PRP was a little lower than that of P-PRP, 0.975, with almost 6.5-times higher SD values.