Sten Osis
Well-known member
You guys are arguing classic "old school" quality control methods. In manufacturing, you have control charts with the average process output (~your INR) and control limits based on statistical variation of historical results (usually 3-sigma). Then you have to make sure that those limits are within your drawing specifications (INR range from the doctor). The key in either scenario is to target the optimal value, but not make adjustments unless you fall outside of the allowable range.
You can see by my individuals control chart that my process is not yet stable.