Use of Peripheral Perfusion Index Derived From the Pulse Oximetry Signal as a Noninvasive Indicator of Perfusion

Alexandre Pinto Lima, MD, Peter Beelen, RN, Jan Bakker, MD, PhD


Crit Care Med. 2002;30(6) 

In This Article


Group 1. One hundred and eight healthy volunteers were included, and a total of 216 measurements were made. The distribution of age in the healthy volunteers was normal: skewness, 0.06; median, 36 yrs (inner quartile range, 30-45 yrs). The distribution of PFI was skewed (Fig. 1, Table 1 ).

Figure 1.

Frequency distribution of all 216 peripheral perfusion index values in the normal volunteers. Line represents normal distribution.

Descriptive analysis showed no significant difference between variance and skewness of the measurements before and after the meal ( Table 1 ). Also, no significant differences were found between smokers (n = 26) and nonsmokers (n = 82), as well as between volunteers with (n = 11) or without (n = 97) vascular disease (diabetes, hypertension). All volunteers had a normal capillary refill time and arterial oxygen saturation (96% to 100%).

Group 2. A total of 74 measurements were carried out in the 37 patients studied. Descriptive statistics revealed a mean PFI of 2.2 ± 0.22 with a median of 1.8 (inner quartile range, 0.5-3.2). Table 2 summarizes hemodynamic data during abnormal peripheral perfusion and normal peripheral perfusion, as well as the mean doses of vasoactive drugs. No significant relationship between core temperature and PFI or core-to-toe temperature difference was found. A significant exponential relationship between PFI and the core-to-toe temperature difference was found (R2= .52, p < .001; Fig. 2).

Figure 2.

Relationship between peripheral perfusion index (PFI) and core-to-toe temperature difference in all 74 measurements in the 37 patients studied. Displayed is the best fit curve (logarithmic) R2= .52, p < .001. Reference lines are the median PFI of healthy volunteers and the reference for an abnormal core-to-toe temperature difference.

We found a significant linear correlation between changes in PFI and changes in the core-to-toe temperature difference (R2= .52, p < .001; Fig. 3).

Figure 3.

Relationship between changes in peripheral perfusion index (PFI) and changes in core-to-toe temperature difference. Displayed are the linear regression and the correlation coefficient. dPerfusion index, PFI during abnormal perfusion - PFI during normal perfusion; dCore-Toe temperature, temperature difference during abnormal perfusion - temperature difference during normal perfusion.

In all cases, a concordant change in PFI and core-to-toe temperature difference was found. No significant relationship was found between mean arterial pressures, dose of vasoactive agents, and PFI or between changes in these variables and changes in PFI.

In 16 patients, cardiac output was measured. No significant relationship was found between changes in cardiac output and changes in either core-to-toe temperature difference or PFI.

We assessed the ability of the PFI to indicate an abnormal peripheral perfusion, as reflected by an abnormal core-to-toe temperature difference by constructing a receiver operating characteristic curve. A PFI of 1.4 discriminated best between a normal and abnormal core-to-toe temperature difference in these critically ill patients (area under the curve, 0.91; 95% confidence interval, 0.84-0.98). Table 3 reports the corresponding sensitivity, specificity, and likelihood ratios.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: