Comparison of two methods based on photoplethysmography for the diagnosis of peripheral arterial disease

To examine the interchangeability of two methods for distal pressure measurement based on photoplethysmography using a truncated or full display of the arterial inflow curve, respectively.

Toe and ankle pressures were obtained from 69 patients suspected of peripheral arterial disease (PAD). Observer reproducibility of the curve readings was examined by blinded reassessment of the pressure curves in a randomly selected subgroup (60 limbs).

There were no significant differences in mean pressures between the two methods (p for all > .455). The limits of agreement for the differences were -15.0-15.4 mmHg for right toe pressures, -16.3-16.2 mmHg for left toe pressures, -14.2-15.7 mmHg for right ankle pressures, and -18.3-17.7 mmHg for left ankle pressures. Correlation analysis revealed intraclass correlation coefficients ≥0.960 for all measuring sites. Cohen's Kappa showed excellent agreement in diagnostic classification, with κ = 0.930 for the diagnosis of PAD and perfect agreement in the diagnosis of critical limb ischemia (κ = 1.000). The analysis of intra-observer variation for curve reading showed limits of agreement of -3.9-4.0 for toe pressures and -7.6-7.7 for ankle pressures for the method involving truncated display and -3.1-3.2 for toe pressures and -6.3-8.6 for ankle pressures for the method involving full display of the signal.

The present study shows minimal differences in diagnostic classification, as well as in ankle and toe pressures, between the full display and the truncated display of the photoplethysmographic pulse signal. Furthermore, the inter-observer variation was low for both of the photoplethysmographic methods investigated.