applaud the decision to change projects- this project really fits with your research interests.
I might add to the negative reinforcement- many of the patients may have issues with innumeracy and will need alot more education/explanation from the surgeon about risks, even before using the tool. this might cause frustration on both patient and surgeon perspectives.
Also- with tracking/compliance (or is it intention?)- if I were a surgeon, i would want access to tools that might help me characterize risk better, but also leave "wiggle room" for clinical gestault/impression. My take is that there are probably cases where the patients "numbers" or risk seems high but the surgeon feels the surgery is important and/or the surgeon feels it actually will be lower risk; and vice versa with patients who seem low risk on calculators but perhaps the surgeon knows something that may make the particular pt higher risk than advertised-- addressing these scenarios might be important to consider.