Following comments on valvereplacement.org, here I have posted the following further information:

Thanks for taking the time to read and comment.

I knew it would be controversial, particularly the frequency of testing.

The reason I chose to test so frequently was really to challenge the incumbent beliefs about stability and to try and uncover the truth of what the INR level does on a more regular basis.

I agree that if the typical dosage calculators are followed then it would cause yo-yo effect, and that's because the calculators only ever look at a single dose.  My strategy is different - i look at the range of doses and INRs over the last few days and look for a trend.  I then make a dose change to arrest that change…. or often just don't bother.

My time on the drug so far has seem some very successful dosing and small dose adjustments, and it's also seen some mistakes.  I will keep refining this.

I'm a software developer by trade and i'm working on a new algorithm calculator tool that will support my dosing.  It's a very complex algorithm so I haven't managed to create it yet, but it's definitely on the way.  It will look at a whole week's results and doses before making a dose suggestion.

I believe that the "perfect" system would be to test every single day, and to change the dosage daily making small adjustments to arrest any trend up and down.  I this were followed I believe warfarin would be in range for an even higher percentage of the time.

I will continue to test very regularly and keep refining my method.  Certainly regular testing doesn't do any harm, it gives a great insight into what's actually going on.

Remember that all of this is based on my statistical interest, and on the fact I don't believe warfarin to be a stable drug.  Not a mainline opinion, but certainly mine.