I work in a sleep research lab. We have a couple of Masimo pulse oximeters that report pulse oximetry values in terms of integers (e.g. 90, 91, 92, etc.).

Now our principle investigators used to use an Ohmeda Biox 3700 that reported data via analog output in a nice (noisy) curve (e.g. after filtering, 90, 90.1, 90.3, 90.4, etc.).

There's some discussion that because of the lack of decimal, the Masimo is useless for data reporting. The Ohmeda though is from the 90s and I can't find any documentation that indicates if taking the decimal is even acceptable for the device's resolution. Any thoughts?

  • $\begingroup$ Possible cross-site duplicate of biology.stackexchange.com/questions/82301/… $\endgroup$
    – tripleee
    Mar 27, 2019 at 19:12
  • $\begingroup$ This really is a medical question not an engineering question. $\endgroup$
    – Eric S
    Mar 28, 2019 at 14:25
  • $\begingroup$ This is not a medical question - it is an engineering question. It's really about making a practicable device, and how the manufacturer has swept all the fudges under the carpet. It's a question about noise and how small changes in a measurement propagate. What's the variance in the the photodiode outputs at the wavelengths used for the pulse oximetry? How to propagate forward to the calculation of SpO2 and how this is related to the absorption and scattering coefficients of blood and tissue? It's specmanship - my device has an extra decimal point, so it must be better. $\endgroup$
    – D Duck
    Mar 29, 2019 at 9:12
  • $\begingroup$ @DDuck There is a difference between accuracy and precision. If the extra decimal place doesn’t make any medical treatment difference then it is unimportant. $\endgroup$
    – Eric S
    Mar 31, 2019 at 2:34
  • $\begingroup$ It is important is for the sales people. The implication is that this device is better because it is more accurate and/or precise than the competitors product. $\endgroup$
    – D Duck
    Mar 31, 2019 at 17:22

1 Answer 1


The accepted error in pulse oximtery isn't that good. It's about $\pm$2 % so your additional significant figures are useless. A good review is Aoyagi, T. J Anesth (2003) 17: 259. https://doi.org/10.1007/s00540-003-0192-6


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.