The first thing to know is that pulse oximeters lie. A lot. The second thing to know is that for any value less than 80%, it's a garbage number. The algorithm is based on a sample of real human beings. It is not ethical or legal to make people extremely hypoxic, so the cutoff for healthy volunteers is 80%. Below 80%, the machine is making a mathematical best guess. It is not valid measurement data. The lower the number, the wider the margin of error.
In a normal person with a normal hemoglobin level, you would not be conscious with a saturation of 65%. You would be quite blue and in significant respiratory distress. With lower-than-normal hemoglobin, you wouldn't turn blue but you'd pass out before it ever got down to 65% (you need at least 5g/dL of deoxygenated hemoglobin before cyanosis is apparent).
If you have chronic lung disease and have a higher than normal hemoglobin (polycythemia), you can tolerate a much lower saturation than average, but 65% is still too prone to measurement error on a finger probe. That has to be confirmed with a blood gas.
Far and away the most common cause of a reading that low is an equipment problem, not a physiology problem.
This is also why I am not a big fan of pulse oximeters at home. Without a hemoglobin value, it's not a very meaningful number. And either way, if it doesn't help decision-making, why have it? If it says 65% and you feel totally fine, would you scream and call 911? No. If it says 100% and you can't breathe are you going to sit there and do nothing? No. So why have the number? It doesn't help you make decisions. It's just another number to fuss over -- like you're doing right now.
The best advice is frankly to ditch the pulse oximeter. Unless you REALLY have to have it because your doc wants you keeping track of trends, you don't need it. You can manage your oxygen needs without it.
Respiratory therapist (B.S., RRT, RPFT)
Working on my master's in nursing