Limitations of Pulse Oximetry:
Effect of Venous Pulsation
and Congestion on Pulse Oximetry
Erroneous SpO2 readings have been reported due to venous pulsations caused by tricuspid valve regurgitation, obstructed venous return , measurement in a dependent limb, constrictive tourniquets or bandages in the monitored limb, and wide swings of intrathoracic pressure (for example, during cyclic high pressure ventilation). Many of these reports have focused on ear or forehead sensors as being particularly sensitive to the phenomenon. Sensors placed on the extremities seemed to be less affected.
In a 1994 report in the Lancet, Bucher and coworkers demonstrated that the effect of venous congestion from tight fixation of the neonatal sensor caused under reading of the actual SaO2.1 They report an incident of unexplained hypoxemia (SpO2 in the 80-82% range) being resolved by readjustment of the sensor. They then performed a series of experiments with 20 infants studying the effect of venous congestion (impaired venous return) at the sensor site. Each infant had a sensor on placed on both feet. One foot was used as the control the other was the test site. A blood pressure cuff was wrapped around the test sensor site. As the pressure was slowly raised the difference in the control SpO2 versus test SpO2 was recorded. They found that at 50 mm Hg pressure the test difference from the control exceeded 3% in 5 infants. With increasing pressure between 25 to 30% of the infants had a difference >3% up to total failure due to inadequate pulsatile signal. In a second set of experiments the blood pressure cuff was placed around the lower leg (to study the effects of venous congestion). At an occlusion pressure of 30 mm Hg, 3 infants had a DSpO2 >3%, and at 40 and 50 mm Hg, 6 and 7 infants respectively, had DSpO2 >3%. In a separate study they measured the amount of pressure that 26 nurses used to apply sensors (to a device with built-in pressure transducer). Forty-two (42)% of the nurses exceeded 50 mm Hg. They conclude that when the tissue pressure increases or venous return is impeded, the venous blood may become pulsatile and thus, become included in the calculation and cause the SpO2 to be erroneously low. This can occur without affecting accurate pulse detection.
Kim and coworkers demonstrated that dependent extremities can exhibit lower SpO2 than elevated ones.2 They found that the fingertip was a better placed to measure than the base of the finger (less of a spurious drop in SpO2). They postulated that the pulse of blood measured by the pulse oximeter, although arterial in content, is actually present in veins, having been shunted there by the systolic pressure wave through the arteriovenous anastomoses in the cutaneous circulation.
In 1989, Mark reported two cases where the ear sensor (Ohmeda Biox 3700) gave erroneously low readings due to venous pulsations.3 One patient had tricuspid regurgitation and exhibited SpO2 < 80% despite an arterial PaO2 > 400 mm Hg. She had an elevated CVP with large systolic v-waves present. A second patient presented with severe coronary artery disease, biventricular dysfunction and a history of heart failure. . His ECG showed junctional rhythm and left bundle branch block. His CVP = 20 mm Hg with large early systolic waves. The ear sensor consistently under read SpO2 (60 - 80%). When a finger sensor was placed it read 100%.
Stewart and Rowbottom studied a series of 22 patients with severe tricuspid value regurgitation.4 They measured both finger and ear SpO2 versus actual SaO2. They were able to obtain CVP readings from 19 patients. The peak systolic venous pressure ranges from 8 to 40 mm Hg. They found the finger to be a more accurate site than ear in this group of patient. They attribute this possibly due to greater attenuation of the retrograde pulsations at the more peripheral sites due to the distance from the central veins and the presence of venous values. However, they cautioned that there were 6 instances where the finger read lower than the ear. They recommend caution in interpreting the information obtained by pulse oximetry in this group of patients.
Clinicians need to be aware of the effects of venous pulsations and venous congestion on pulse oximetry. Cautions as simple as maintaining the sensor at heart level and not placing a tourniquet on the same arm as the pulse oximeter sensor are simple but often forgotten. One should consider venous congestion or venous pulsation as a possible cause when the pulse oximeter is under reading the actual SaO2.
1) Bucher HU, Keel M, Wolf M, von Siebenthal K, Duc G. Artifactual pulse-oximetry estimation in neonates. Lancet 1994;343:1135-1136
2) Kim JM, Arakawa K, Benson KT, Fox DK. Pulse Oximetry and Circulatory Kinetics Associated with Pulse Volume Amplitude Measured by Photoelectric Plethysmography. Anesth Analg 1986;65:1333-1339
3) Marks JB. Systolic Venous Waves Cause Spurious Signs of Arterial Hemoglobin Desaturation. Anesthesiology 1989;71:158-160
4) Stewart KG, Rowbottom SJ. Inaccuracy of pulse oximetry in patients with severe tricuspid regurgitation. Anaesthesia 1991;46:668-670