Accuracy of pulse oximetry in detecting lower extremity arterial disease

Marie Simonette Ganzon, Neva Jean Reloj, Adelina Paule

Objective: To evaluate the accuracy of pulse oximetry vs ankle-brachial index (ABI) in detecting lower extremity arterial disease (LEAD) using arterial duplex scan as the reference test.

Design:

A prospective validation study.

Participants:

Purposively selected adult patients referred to SLMC for evaluation of LEAD using ABI and arterial duplex scan were included. !

Interventions:

Oxygen saturation (SaO2) of index fingers and toes were measured both in horizontal and at 12-inch elevation of the leg. The averaged measurement of toes in each extremity was compared with the higher SaO2 in the upper extremity. A finger-to-toe SaO2 gradient of > 2% was considered abnormal. ABI was performed with abnormal value defined as < 0.9. LEAD was defined as > 50% stenosis with monophasic waveform pattern on arterial duplex study.

Results:

The 52 patients involved in this study had an averaged age of 63.42 years old (SD + 12.85), where there was equal distribution of both males and females. Majority of the patients had intermittent claudication (36.5%) and resting leg pain (25%), with 67.3% of the patients being hypertensive. Of the 104 legs, 35.6% had LEAD. Abnormal pulse oximetry at horizontal (p=0.002) and at 12-inch elevation of the leg (p<0.01) were significantly associated with LEAD However, accuracy indices indicated that pulse oximetry at horizontal position of the leg yielded sensitivity of only 16.2% (95% confidence interval CI, 4-28.1), specificity of 100% and negative likelihood ratio of 0.84 (95% CI, 0.73-0.97); while at 12-inch elevation, sensitivity improved to 23.4% (95% CI, 10.5-38.1), specificity of 100%, and negative likelihood ratio of 0.76 (95% CI 0.63-0.91). On the other hand, ABI measures also demonstrated a lower sensitivity of 29.7% (95% CI, 15-44.5), specificity of 100%, and negative likelihood ratio of 0.7 (95% CI, 0.57-0.87).

Conclusion

Pulse oximetry did not perform clinically acceptable accuracy in detecting LEAD.