Improving pulse oximetry
Pulse oximetry has come a long way. Yet it can go a lot further. Barbara Hein’s article [“Taking the pulse of pulse oximeters,” August 24x7] wonderfully explains how the existing technology of pulse oximeter testers work. But let me put a more clinical perspective on it. When a clinician puts a SPO2 probe on a patient he, or she does not inform the patient what model and manufacture the probe is. My point? Neither should we have to input the pulse oximeter probe model into the tester. We need realistic tests, not idealized preprogrammed comparison curves to ensure the tester checks a particular model probe as good.

I use and like Nonin’s finger phantom and commend them on the simplicity of their device. However it would be nice if their tester incorporated a means (mechanical cam(s) to modulate various arterial pressure waves. A wider range of real patient hemodynamics could be simulated if this was developed and marketed.

More importantly how can we improve pulse oximetry itself? The basic detection formula for pulse oximetry hasn’t changed since two wavelengths were used to quantify SPO2 using the plethysmograph waveform (IEEE Biomedical Engineering Handbook, pages 1349-1352). For nearly a decade, pulse oximetry companies have experimented with multiwavelength pulse oximetry. NASA has a patent on a certain three-wavelength technique (#6,083,158). Companies instead have chosen to add algorithms and look up tables to improve and decrease artifact effects. I commend all these companies in improving the technology. However, I do contend we (biomedical engineers/technicians) need to improve the detection process. Using additional wavelengths, you may detect SPO2 values more accuracy and with a wider range of varying patient arterial pressure hemodynamics (and with less reliance on plethysomography). Not to mention removing the assumption by the clinician that only oxyhemoglobin saturation is the displayed SPO2 value (detection of carboxyhemoglobin, methemoglobin and other blood gas data may be possible). Let’s face it, most of our patients have less than ideal hemodynamics — or they wouldn’t be in our hospitals.

If you’re wondering how I test SPO2 probes, I do it the same way you do. However, it serves no purpose for me to explain to a doctor intraoperatively that the pulse oximeter is fine and it just doesn’t have an R curve programmed in it for your current patient. They just want it to provide accurate data to take care of the patient. Not too much to ask is it?

Kenneth N. Hicks CETsr, CBET, GROL
Manager, biomedical engineering anesthesia/surgery Baylor University Medical Center